Estradiol and progesterone receptors in cultured normal human breast epithelial cells and fibroblasts: immunocytochemical studies.

0021-972X/91/7301-0008/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1991 by The Endocrine Society

Vol. 73, No. 1 Printed in U.S.A.

Estradiol and Progesterone Receptors in Cultured Normal Human Breast Epithelial Cells and Fibroblasts: Immunocytochemical Studies* C. MALET, A. GOMPEL, H. YANEVA, H. CREN, N. FIDJI, I. MOWSZOWICZ, F. KUTTENN, AND P. MAUVAIS-JARVIS Department of Endocrinology and Reproductive Medicine (CM., A.G., H.C., N.F., I.M., F.K., P.M-JJ, Department of Pathology (H. Y.), Department of Biochemistry (H.C., I.M.), Hopital Necker, 75743 Paris Cedex 15, France

ABSTRACT. The estradiol (E2) and progesterone (P) receptors (ER and PR) were studied in normal human breast epithelial (HBE) cells and fibroblasts cultured separately in our laboratory from surgical reductive mammoplasty samples. Immunocytochemical studies were performed on cytospun cells using the anti-ER antibody H222 SPY and the anti-PR antibodies JZB39 and KD68. A specific immunostaining was observed for ER and PR in HBE cells. This immunostaining was nuclear, varying from cell to cell in positivity and intensity of staining. Moreover, ER and

S

TUDIES of estradiol (E2) and progesterone receptors (ER and PR) in normal human breast tissue remain rare (1-3). This is due to the weak sensitivity of classical biochemical receptor assays to the low ER and PR levels of normal breast tissue. These studies have also been discouraged by the heterogeneity of this tissue, which includes epithelial, stromal, adipose, and vascular elements, contrasting with epithelial homogeneity and high receptor levels of breast cancer. The availability of monoclonal antibodies against ER and PR made possible immunohistochemical studies and demonstration of ER and PR pressure in normal breast tissue (4-7). However, hormone regulation of ER and PR in normal breast cells can only be observed in breast cells cultured under varying hormonal conditions. Normal human breast epithelial (HBE) cells and fibroblasts are routinely cultured in our laboratory from surgical reductive mammoplasty samples. These cells remain hormone dependent in culture (8, 9) and have Received August 17, 1990. Address all correspondence and requests for reprints to: Frederique Kuttenn, Department of Endocrinology and Reproductive Medicine, Hopital Necker, 149, rue de Sevres, 75743 Pairs, Cedex 15, France. * This work was supported by grants from INSERM (to F.K.) and the Conseil Scientifique de la Faculte Necker. These data were presented in part at the 71st Meeting of the Endocrine Society (Seattle, WA, June 1989, Abstract 561).

PR immunostaining was hormone-modulated: it increased in E2treated cells and decreased after addition of the progestin R5020. In fibroblasts, a weak ER immunostaining and a stronger PR immunostaining could be observed; however it was not modified by either E2 or progestogen treatment. Thus, in normal breast epithelial cells, E2 stimulates both its own receptor and PR, whereas the progestin R5020 lowers ER and PR content. In contrast, ER and PR content in normal breast fibroblasts seem to be independent of E2 or P action. (J Clin Endocrinol Metab 73: 8-17, 1991)

proved to be well-adapted for the study of hormone/ antihormone actions (8-11). The presence of ER has been previously demonstrated in HBE cells by the immunocytochemical method (11). Parallel immunocytochemical studies of ER and PR were carried out in HBE cells as well as in fibroblasts. The percentage of positive cells and the intensity of positivity were evaluated as a function of the various hormonal conditions, in the presence or absence of E2 and/or the progestin promegestone (R5020) in the cell culture medium, in order to determine the presence and hormonal modulation of these steroid receptors in normal human breast cells.

Materials and Methods Materials Hank's balanced salt solution, trypsine, and fetal calf serum were obtained from GIBCO (Cergy-Pontoise, France). Human serum was provided by the Centre National de Transfusion Sanguine (Les Ulis, France). Ham F12 (HAM) and collagenase were purchased from Boehringer Mannheim Chemical Corporation (Meylan, France), hyaluronidase, cholera toxin, transferrin, bovine pancreas crystalline insulin, T3, Ilj8,17a,21-trihydroxy-4-pregnene-3,20-dione, E2, from Sigma Chemical Corp. (St-Louis, MO) and epidermal growth factor from Clin-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 June 2015. at 20:04 For personal use only. No other uses without permission. . All rights reserved.

ER AND PR IN HUMAN BREAST CELLS isciences (Paris, France). R5020 was purchased from New England Nuclear Corp. (Dupontde Nemours, Paris, France). ER-immunocytochemical assay (ICA) H222 Sp7 monoclonal kit was obtained from Abbott Laboratories (West Germany). Human PR antibodies JZB39 and KD68 were generously provided by G. L. Greene (Chicago, IL). Goat antirat immunoglobulin (IgG) and monoclonal rat peroxidase-antiperoxidase (PAP) complex were obtained from Jackson Immunoresearch Lab. (West Grove, IL). Tissue collection Specimens of normal breast tissue were obtained from women (aged 15-27 yr) who underwent reduction mammoplasty. These patients had no history of benign breast disease and in all patients, pathological study of the tissue revealed only normal breast tissue. Enzymatic digestion The breast tissue enzymatic digestion procedure has been previously described (8). Briefly, the tissue was enzymatically digested with collagenase (0.15%) and hyaluronidase (0.05%) in HAM, and then filtered consecutively through 500, 300, and 150-jum sieves in order to retain any undigested tissue. Cell material retained on the final 60-jum sieve was used for epithelial cell culture, whereas the filtrate was used for fibroblast culture. Culture procedure The cells were pelleted, distributed into T25 plastic flasks, and maintained at 37 C in a humidified atmosphere with 5% CO2. The basal culture medium consisted of HAM without phenol red, containing NaHCO3 (0.24%), kanamycin (0.04%) F (5 ng/ml), T3 (6.5 ng/ml). For HBE cells, culture conditions were specifically chosen to provide epithelial cell growth and in order to inhibit fibroblast development; the medium (medium A) consists of basal medium to which cholera toxin (10 ng/ml), transferrin (5 jug/ml), insulin (0.12 U/ml), epidermal growth factor (10 ng/ml), and 5% compatible human serum were added. Fibroblast cultures were carried out in the basal culture medium supplemented only with 10% fetal calf serum (medium B). In case of accidental contamination of epithelial cell cultures by fibroblasts, a brief washing with trypsin (0.25% 1 ml) was used to eliminate the fibroblasts, which are more rapidly detached from the surface of the plastic flask. Steroid treatments In order to study the estrogen and progestogen dependency of ER and PR in HBE cells and fibroblasts, one-third confluent primary cell cultures were carried out using medium A for HBE cells with a reduced concentration of serum (1% instead of 5%) in order to lower endogenous steroid concentration, and medium B for fibroblasts, with 1% serum instead of 10%, and treated under various hormonal conditions as follows: 1) medium without any steroid addition, 2) medium with E2 (10~9 to 10~7 M), or 3) with R5020 (10~7 or 10"6 M), or 4) with E2 (10~8 M + R5020 (10~7 M). The medium was changed every 2 days and the steroids were added on the alternate days in

amounts permitting maintenance of a constant final concentration. ER and PR-ICA ER and PR immunocytochemical studies were carried out on 1) primary cultured epithelial cells or fibroblasts at varying times and with various concentrations during 8-day steroid treatments, and also in some cases 2) tissue sections of original breast samples, and 3) dissociated epithelial cells obtained after enzymatic digestion of the original breast tissue. Preparation of the slides Samples of breast tissue obtained at surgery were immediately frozen in isopentane and kept in liquid nitrogen until further tissue slicing with a cryostat (Slee) and immunohistochemical study of the original tissue. After breast tissue enzymatic digestion, the dissociated isolated cells and/or organoids were washed three times with PBS before being pelleted, resuspended in 3 ml PBS, and distributed in 200-^1 aliquots of cytospin. Cultured cells (epithelial cells or fibroblasts) were harvested with trypsin, washed twice with PBS, resuspended in PBS at a final concentration of 75,000 cells/ml, and 200-/A aliquots of the cell suspensions were spun in a cytospin 2 device (Shandon, Eragny, France) for 8 min at 400 rpm. In some experiments, E2 (10~8 M) or R5020 (10~7 M) were added to the culture medium before harvesting in order to rule out biochemical effects of these steroids on receptor immunodetection. The slides were immediately fixed in 3.7% formaldehyde 0.1 M PS (pH 7.4) for 15 min, washed in PBS for 10 min, and successively soaked in cold methanol and cold acetone at -20 C for 3-4 min each. After being rinsed in PBS, the slides were stored at -20 C in a preservative solution (0.33 g anhydrous MgCl2,42.5 g saccharose, and 250 ml glycerol, qsp 500 ml PBS). ER-ICA After being rinsed in PBS, the slides were incubated with normal goat serum (in PBS) for 15 min to reduce nonspecific staining. The slides were then successively incubated with the monoclonal anti-ER antibody (0.1 iig/m\) or with normal IgG (0.1 fig/ml) as a control antibody for 24 h at 4 C, and then with the bridging goat antibody antirat IgG (10 Mg/ml) and with rat PAP (0.5 /xg/ml) for 30 min each at room temperature in a humidified chamber. Each incubation was followed by two 5min washes in PBS. After final rinsing in PBS, the cell areas were covered with 0.5% diaminobenzidine 0.4 HC1 substrate solution for 6 min in the dark and counterstained with lightgreen (Biolyon Laboratories, France) except for the cryostat tissue sections, which were counterstained with Mayer hematoxylin (Sigma, St. Louis, MO). Abbott kit control and slides of normal endometrium tissue sections were used as positive controls for ER. PR-ICA The successive steps of the immunocytochemical study of PR were identical to those of the ER study. After a 15-min


MALET ET AL.

10

incubation with normal goat serum (2% in PBS), three antibodies were sequentially used: 1) a rat monoclonal antiPRantibody (JZB39, 1 jug/ml) for 24 h at 4 C or a normal rat IgG as a negative control, 2) a goat antirat IgG as a bridging antibody (1.8 ng/ml, 1:80 dilution; 30 min), 3) a rat PAP complex (1:300 dilution; 30 min). Each antibody incubation was followed by two 5-min washes in PBS. Slides prepared with cytospun cells of the PR-rich T47D breast cancer cell line were used as positive control. In some studies, the KD68 antiPR-antibody was used in parallel with the JZB antibody.

Results

JCE&M«1991 Vol 73 • No 1

the nucleus. Nevertheless, this staining was heterogeneous: not all the cells were positively stained (ER+) and among ER+ cells the intensity of staining varied from cell to cell. Specificity was demonstrated by the absence of staining with a control antibody (normal IgG). A staining intensity scale was established: +++ for strong, ++ for moderate, and + for weak staining. Intraassay variations were less than 10%. Interassay variations did not exceed 15%. ER immunostaining as a function of various hormone culture conditions: ±E2, ±R5020

Results of the immunocytochemical study of human breast epithelial cells cultured in the presence or absence of hormones for 8 days are presented in Fig. 2 and in Immunocytochemical localization of ER in breast epithelial cells Table 1. Under all hormone conditions, ER staining was The study was carried out on cells which had been nuclear and varied in positivity and intensity from one cytospun. This explains why, contrary to the original cell to another. Moreover, the percentage of positive cells cultured cells (Fig. 1, a and b), they appear dispersed and and staining intensity varied as a function of the horflattened, and with a spread out cytoplasm. mone culture condition and duration of treatment. The We observed a specific immunostaining for ER. It was percentage of ER+ cells was greater in E2-treated cells exclusively nuclear, and diffusely distributed throughout (10~8 M) than in cells cultured in the absence of E2, and the intensity of staining was higher. When the progestin R5020 (10~7 M) was added to E2, the percentage of ER+ cells and the intensity of staining were considerably lower, comparable to those of untreated cells and sometimes even lower. When R5020 was added alone, staining and percentage of ER+ cells were even lower than when E2 was added simultaneously (Table 1). ER study

Attempt to detect time- and dose-dependent effects of steroid hormones on ER immunostaining of epithelial cells

It was principally with E2 that a time- and dosedependent stimulation of ER immunostaining was investigated. After 48 h E2 treatment, the ER immunostaining was weak, as in untreated cells. Staining was greater after 4-5 days under both hormonal conditions, but with an E2-stimulative effect already evident. While the ER level remains similar in untreated cells at both 4-5 and 8 days of culture, the stimulatory effect of E2 appeared to be maximal at 8 days of treatment. However, it was not possible to demonstrate a dose-dependent effect of E2 on ER immunostaining. Identical results were obtained at 10~9, 10"8, and 10"7 M concentrations. A 10"8 M E2 concentration was used in further studies, particularly in association with R5020. With R5020, no difference could be found between 10~7 and 10~6 M concentrations in their reductive effect on ER immunostaining. ER immunostaining in breast fibroblasts in culture

FIG. 1. Morphological aspect of normal HBE and fibroblast cells in primary culture, a) Epithelial cells; b) fibroblasts. Magnification, X200.

After cytocentrifugation, cultured fibroblasts do not retain their characteristic fusiform shape. However, they can still be distinguished from epithelial cells by the


ER AND PR IN HUMAN BREAST CELLS

11

FIG. 2. Stimulation effect of E2 and inhibitory effect of R5020 on ER immunostaining in normal HBE cells. HBE cells were cultured 8 days: a) in the absence of steroids, or b) in the presence of E2 (10~8 M), or c) E2 (10~8 M) + R5020 (10~7 M). The cells were cytospun and immunocytochemical study of ER was carried out using the monoclonal anti-ER antibody H222 Sp7; d) negative control: the primary antibody was replaced by normal IgG. Counterstaining: light green. Magnification, X400. TABLE 1. Variations ofER immunostaining in human breast epithelial (HBE) cells in culture as a function of the steroid treatments ER immunostaining {% of stained cells) Control E2 E2 + R5020 R5020

Positive

Negative

55 73 44 35

45 27 56 65

Intensity of staining

HBE cells were cultured 8 days 1) in the absence of steroids: control cells, or 2) in the presence of E2 (10~8 M), or 3) E2 (1(T8 M) + R5020 (10-7 M), or 4) R5020 (1(T7 M) alone.

ovoid form of their nuclei. In breast fibroblasts, we observed weak nuclear ER immunostaining (Fig. 3), which was positive in 60% of the cells (Table 2). This staining was specific since it was absent when the antiER-antibody was replaced by the control antibody. Variations in the positivity (=percentage of ER+ cells,

staining intensity) under varying hormonal conditions are less clear-cut in fibroblasts than in epithelial cells. No significant difference was observed in percentage and intensity of ER immunostaining between E2- and/or R5020-treated or untreated fibroblasts. ER in isolated epithelial cells or organoids obtained after enzymatic digestion of breast tissue

Percentage of ER+ cells varied from 30-75% according to the patient and to the field examined on the slide. On the organoid presented in fig. 4, 73% of the cells are ER+ (62% with high and 11% with moderate intensity). ER immunohistochemical study of breast tissue sections

In sections of the original breast tissue, ER immunostaining was clearly observed in the epithelial cells. However, it was localized only at the inner layer of cells and varied from one sample to another, involving 7-45% of the cells (25% on the galactophore fragment presented


MALET ET AL.

12

JCE & M • 1991 Vol 73 • No 1

FIG. 4. Immunocytochemical staining of ER in isolated cells and/or organoids obtained after partial enzymatic digestion of the surgical breast tissue samples. Counterstaining: light green. Magnification, X200.

FIG. 3. ER immunostaining in normal breast fibroblasts. Fibroblasts were cytospun after an 8-day period of culture and immunocytochemical study of ER was carried out using the anti-ER antibody H222 Sp7. a) Fibroblasts after culture in the absence of steroids; b) negative control: the primary antibody was replaced by normal IgG. Counterstaining: light green. Magnification, X400. TABLE 2. ER immunostaining in normal human breast fibroblasts in culture ER immunostaining

Positive

Negative

% of stained cells

60%

40%


on Fig. 5 with about two thirds highly positive and one third moderately positive).

FIG. 5. ER immunostaining on histological sections of normal human breast tissue. Samples of breast tissue were frozen and immunohistochemical study of ER was carried out on 6-jrni cryostat tissue sections. Counterstaining: hematoxylin. Magnification, X200.

it was rather granular, with a crescent or clod-like shape (Fig. 6). In general, it was also more intense. Specific staining was also observed when the JZB39 antibody was replaced by another human PR antibody (KD68). However, at a dilution of KD68 adequate to avoid nonspecific staining, specific staining was lower than that observed with JZB39. Effects of steroid hormone treatments

When E2 (10~8 M) was added to the culture medium, the number of PR positive cells and the intensity of Immunocytochemical localization of PR in breast epithelial cells staining were higher than in nontreated cells (Fig. 6, Studies using the anti-PR JZB39 antibody directed Table 3). For PR as for ER, no dose-dependent stimuagainst the PR of human T47D breast cancer cells also lation could be clearly demonstrated. showed nuclear staining of breast epithelial cells. HowWhen R5020 (10~7 M) was added in association with ever, PR staining was often less diffuse than ER staining; E2 (10~8 M), the staining was weaker than with E2 alone. PR study



13

FIG. 6. Stimulatory effect of E2 and inhibitory effect of R5020 on PR immunostaining in HBE cells. HBE cells were cultured 8 days: a) In the absence of steroids, or b) In the presence of E2 (10~8 M), or c) E2 (10~8 m) + R5020 (10~7 M). They were cytospun and immunocytochemical study of PR was carried out using the monoclonal anti-PR antibody JZB39. d) Negative control: the primary antibody was replaced by normal IgG. Counterstaining: light green. Magnification, X400.

TABLE 3. Variations of PR immunostaining in cultured HBE cells as a function of the steroid treatments (for hormone conditions see the legend to Table 1) PR immunostaining (% of stained cells)

Positive

Negative

Control E2 E2 + R5020 R5020

68 81 60 42

32 19 40 58


There were few positive cells and less intense staining (Fig. 6, Table 3). When the cells were treated with R5020 (10~7 M) alone, staining was weaker in percentage of PR+ cells as well as in intensity (Table 3). PR immunostaining in breast fibroblasts

In cytospun fibroblasts (Fig. 7), PR immunostaining was also exclusively nuclear. The percentage of stained

fibroblasts was very high: 84% of the cells were positive with high staining intensity in 57% (Table 4). This percentage and the intensity of staining did not seem to change under the various steroid conditions. Specific staining was also observed when using the KD68 anti-PR antibody, but it was of lower intensity. In isolated epithelial cells or organoids obtained after enzymatic digestion of breast tissue

As observed for ER, the percentage of PR+ cells varied according to the individual and from one slide to another (Fig. 8). PR immunohistochemical study of breast tissue sections

In sections of breast tissue samples obtained during reductive mammoplasty prepared for immunohistochemical study and counterstained with hematoxylin, nuclear staining of epithelial cells was evident. However, the


MALET ET AL.

14

JCE&M«1991 Vol73«Nol

FiG. 8. Immunocytochemical staining on PR in isolated cells and/or organoids obtained after partial enzymatic digestion of the original breast tissue. Counterstaining: light green. Magnification, X400. FlG. 9. PR immunostaining on histological sections of normal human breast tissue. Samples of breast tissue were frozen and immunohistochemical study of PR was carried out on 6-jtm cryostat tissue sections. Counterstaninig: hematoxylin. Magnification, X200.

FIG. 7. PR immunostaining in normal breast fibroblasts. Fibroblasts were cytospun after an 8-day period of culture and immunocytochemical study of PR was carried out using the antiPR-antibody JZB30; a) fibroblasts after culture in the absence of steroids; b) negative control: the primary antibody was replaced by normal IgG. Counterstaining: light green. Magnification, X400.

TABLE 4. PR immunostaining in normal human breast fibroblasts in culture PR immunostaining

Positive

Negative

% of stained cells

84%

16%


percentage of PR+ cells varied from 5-50%, according to tissue origin. In the breast duct section presented in Fig. 9, 48% of the cells are PR + , and of these, four fifths present high intensity staining. All positive cells are localized in the inner cell layer. In loose connective tissue, the brown PR stain is difficult to distinguish from the hematoxylin counterstain. However, some fibroblast nuclei seem to stain weakly. When the tissue section was not counterstained, it was possible to observe some nuclear staining in the connective tissue in addition to the epithelial cell stain-

FIG.

9.

ing. This could correspond to the immunostaining observed for PR in the isolated cytospun fibroblasts.

Discussion Hormone dependence of the breast has essentially been determined from experimental study on animals (12, 13) and biochemical investigations on breast cancer cell lines (14-16) as well as observations of physiological development of the human breast (12, 17). Studies of hormonal risk factors for benign breast disease (18), and, in the long term, for breast cancer (17), have also provided information on the hormone dependence of breast tissue. The presence of ER and PR (19-22) was first demonstrated in endometrium, an easily biopsied target tissue,



and then in breast cancer. In cancer tissue, the presence of these receptors is considered to be a marker of hormone dependency for E2 and therefore an indication of its possible responsiveness to adjuvant hormone therapy (23). Unlike endometrium, normal breast tissue is difficult to obtain. Moreover, it is histologically heterogeneous, and its cyclic variations are less well known than those of endometrium. There are few studies of steroid receptor content of normal breast tissue (1-3); receptor distribution remains imprecise and its regulation an unresolved question. We have developed a culture system of human breast cells derived from breast tissue samples obtained from reductive mammoplasty. This permits, after enzymatic digestion and with appropriate culture media, obtention of separated cultures of epithelial cells and fibroblasts (8). In culture, these cells remain hormone dependent for E2 and P, as has been shown by studies of cell growth, ultrastructure, and enzyme activity under various hormonal conditions (8, 9, 11, 24). The existence of steroid receptors within these cells and their possible hormonal regulation had yet to be demonstrated. Classical biochemical methods for studying steroid receptors have rather low sensitivity and are difficult to apply to normal breast cells because of relatively low receptor levels in this tissue, which require a large number of cells for this assay. In addition, classical biochemical methods do not allow identification of receptor distribution among the different types of cells, nor intracellular localization of receptors. The recent availability of monoclonal anti-ER and anti-PR antibodies renders feasible the immunocytochemical study of these receptors. It resolves the problem of limited quantities of cell material as well as allowing determination of the intracellular localization and hormone regulation of these receptors. Using the anti-ER H222 Sp7 Abbott antibody and the anti-PR JZB39 and KD68 antibodies (25), we observed an immunostaining specific for ER and PR in normal breast epithelial cells and also in fibroblasts. This immunostaining is exclusively nuclear, indicating a nuclear localization of these steroid receptors, as has been recently shown with this method in breast cancer cells and endometrium (26, 27). This immunostaining is also heterogeneous; not all cells are stained, and staining intensity varies from cell to cell. Such staining heterogeneity, even in cells of the same type with the same tissue, had previously been reported in ER and PR immunohistochemical studies carried out on uterus and normal breast tissue sections. This difference in receptor levels from one cell to another suggests differences in age, degree of maturation, and function, and reflects the turnover of cells indispensible to tissue survival.

15

It is essential to point out certain other characteristics of ER and PR in these normal cells: In cultured epithelial cells, either untreated or treated with E2, the percentage of ER-I- cells is greater than observed in immunohistochemical studies of breast tissue sections, as reported in the present study as well as in other authors' publications, since a percentage range of 7-45% ER-I- cells was observed in previous immunohistochemical studies on normal breast tissue (4, 5). Several explanations can be proposed 1) in immunohistochemical studies, the nucleus may have been left out of the section, leaving only cytoplasm; since staining is exclusively nuclear, the cells may appear artefactually receptor negative. 2) In addition, it is highly probable that the tissue enzymatic digestion process preceding culture is aggressive to the cells, thereby exercising a selective effect, which may explain why only young, active, receptor-rich cells survive. Initiation of culture is certainly an additional selective step, which explains that young active cells, which may still be highly disposed to become strongly hormone dependent at a given time of their evolution, prove to be particularly well adapted to adhesion and survival in culture. Indeed, in our immunocytochemical study the small epithelial cells seem to stain more intensely than the larger, sprawled-out cells; however it remains difficult to determine whether such small and younger cells are actually richer in steroid receptors, or if it is the packed character of their nuclei that accentuates staining. The same phenomenon was also observed with PR in the present study. PR immunostaining tends to be more intense than ER immunostaining in epithelial cells as well as in fibroblasts. Staining intensity depends upon two main factors: antibody-receptor affinity and amount of receptor. In our experience, PR immunostaining was first studied using the Transbio antibody directed against PR of rabbit endometrium, which provided only weak staining. The anti-PR antibodies JZB39 and KD68 provided by G. Greene were elected for further studies. The better results obtained with these two antibodies are most probably the consequence of greater specificity of JZB39 and KD68 which are directed against PR of human breast cancer cells. On the other hand, the more intense immunostaining of PR compared to that of ER, whereas using antibodies of similar origins, agrees with results obtained using the classical biochemical method in endometrium, breast EP+ PR+ cancers and fibroadenomas, in which PR has been found to be 2 to 3 times higher than ER (28, 29). The specific ER and PR immunostaining observed in fibroblasts constitutes new data, since it was not observed in most previous immunohistochemical studies on cancer, benign disease and normal breast tissue (4, 5) with the exception of Markopoulos et al. (30). These


16

MALET ET AL.

reports contrasted with observations made on the stroma of uterus (endometrium, cervix), Fallopian duct, vagina, ovary in which ER and PR are immunostained (4, 5, 7, 31, 32). These discrepancies among different target tissues, as well as between the present study and others, raise several questions. First, are these antibodies specific for ER and PR in fibroblasts? A possible hypothesis is that the epitopes against which the anti-ER H222 SPY and anti-PR JZB39 and K68 antibodies are directed exist in fibroblasts but are unrelated to E2 and P receptors. However, positive immunostaining found in variable degrees with all these antibodies should exclude nonspecific immunostaining. Other hypotheses can be proposed: the connective tissue appears to be very loose in the immunohistochemical sections and the nuclei are less visible than epithelial cell nuclei even though some of the epithelial nuclei are also missed by the histological section. On the contrary, in immunocytochemical studies of ER and PR carried out on cytospun fibroblasts, cells and nuclei appear distinctly and in their totality and the antireceptor antibody is exclusively available for the cytospun cells from these pure fibroblast cultures. Moreover, in sections of frozen breast tissue which we studied by immunohistochemistry, some nuclei dispersed in the connective tissue appeared immunostained for ER and PR, even though such nuclei are smaller and the staining weaker than in epithelial cells. Finally, and above all, if receptor content is lower in fibroblasts than in epithelial cells, the long incubation time with the first antibody (24 h) in our study certainly allows better visualization than do the time periods usually employed by other authors in their studies of breast cancer cells or endometrium (1 or 2 h). Even though it is heterogeneous, ER and PR immunostaining is nevertheless hormone dependent at least in epithelial cells, since in these cells it is stimulated by E2 and reduced by R5020. Indeed, despite the semi-quantitative value of the immunocytochemical method, hormone dependence of normal breast cells is now clear. Hormone regulation of ER and PR had been demonstrated by histochemical studies conducted on endometrium throughout the menstrual cycle (25, 31, 33, 34) but no immunohistochemical study of normal breast tissue had addressed this question. To study E2 effect on ER and PR immunostaining, a control medium containing a minimal amount of E2 had to be defined, as shown in our previous studies of E2 effect on cell growth (9,11). Charcoal-treated serum does not sustain HBE cell growth. We therefore used untreated serum after checking that the 1% serum medium used in our immunocytochemical studies never exceeded 1CT11 M E2. Furthermore, these studies were carried in phenol red free medium, since this pH indicator has been

JCE & M • 1991 Vol 73 • No 1

reported as having a weak estrogenic activity (35). Under these conditions, the stimulatory effect of E2 on ER and PR immunostaining was correlated to the duration of treatment (8 days > 4-5 days > 2 days); however, the absence of a clear dose-dependent effect might be related to the merely semi-quantitative value of immunocytochemical assay. For the study of P action, the progestin R5020 was preferred to P itself, because of its greater stability in vitro, P being rapidly metabolized in culture. R5020 treatment of the epithelial cells led to a decrease in percentage and intensity of immunostaining for both ER and PR. However, while ER and PR hormone regulation is evident in epithelial cells, with stimulation by E2 and repression by the progestin R5020 and a time-dependent effect, hormone regulation is far more difficult to demonstrate in fibroblasts. It is possible, as shown in studies of endometrium stroma, that the receptor content of fibroblasts in normal breast tissue is constitutive and less sensitive to hormone regulation than it is in epithelial cells (25, 33, 34). E2-stimulatory and P-repressive effects on both ER and PR had been observed in the endometrium (21, 22), in breast cancer cells (15), as well as in benign breast diseases (29). These effects have also been suggested for ER by biochemical assay carried out on samples of normal breast tissue throughout normal ovulatory cycles (1). However, to our knowledge, this is the first time that E2stimulatory and P-repressive effects have been demonstrated on both the ER and PR contents of epithelial cells. In these normal epithelial human breast cells with limited steroid receptor content, whether cultured in the absence or in the presence of infinitesimal quantities of E2, the most probable mechanism for the increase in ER and PR under E2 treatment is the stimulation of their synthesis, even if recent studies on breast cancer cell lines have shown (at least in the short-term and only under certain hormonal conditions) down-regulation of ER by E2 (36). In contrast, PR and mRNA of PR are usually stimulated by E2, and R5020 has a repressive effect on both ER and PR (36, 37). By stimulating ER and PR in normal breast epithelial cells as in other target tissues, E2 tends to amplify its own action and to promote the complementary action of progesterone, whereas P tends to limit both E2's and its own actions. This regulation of steroid receptors is one of the mechanisms by which E2 induces cell multiplication and P controls cell differentiation in their target tissues (9,10). In conclusion, in normal breast epithelial cells, E2 stimulates both its own receptor and PR, whereas the progestin R5020 lowers ER and PR content. In contrast,



ER and PR content in normal breast fibroblasts seem to be independent of E2 or P action, and could be constitutive. This culture system of normal breast cells has proven itself to be especially well adapted to the study of the actions of various hormone and antihormone molecules on cell multiplication and on markers of differentiation. The action of antiestrogens, which have a cytoplasmic effect on both breast cancer and normal cells (11), should now be studied on ER and PR in these normal cells.

15. 16. 17. 18. 19. 20.

Acknowledgment The authors wish to thank Dr. Geoffrey L. Greene for the kind gift of the anti-PR antibodies.

21. 22.

References 1. Silva JS, Georgiade GS, Dilley WG, McCarty KS, Wells SA. Menstrual cycle-dependent variations of breast cyst fluid proteins and sex steroid receptors in the normal human breast. Cancer. 1983;51:1297-1302. 2. Leung B, Manaugh LC, Wood DC. Estradiol receptors in benign and malignant disease of the breast. Clin Chim Acta. 1973;46:6976. 3. Menandez-Botet CJ, Nisselbaum JS, Fleisher M, et al. Correlation between estrogen receptor protein and carcinoembryonic antigen in normal and carcinomatous human breast tissue. Clin Chem. 1976;22:1366-71. 4. Jacquemier J, Gandilhon PH, Charpin C, Pourreau-Schneider N, Hassoun J, Martin PM. Tissu mammaire: recepteurs estrogeniques. Analyse immunohistochimique et biochimique du tissu normal, hyperplasique et du carcinome in situ. A propos de 86 cas. Bull Cancer (Paris). 1987;74:129-49. 5. Petersen OW, Hoyer PE, Van Deurs B. Frequency and distribution of estrogen receptor-positive cells in normal nonlactating human

6.

7. 8. 9.

10.

11.

12. 13. 14.

breast tissue. Cancer Res. 1987;47:5748-51. Marchetti E, Querzoli P, Moncharmont B, et al. Immunocytochemical demonstration of estrogen receptors by monoclonal antibodies in human breast cancer. Correlation with estrogen receptor assay by dextran-coated charcoal method. Cancer Res. 187;47:2508-13. Press MP, Greene GL. Localization of progesterone receptor with monoclonal antibodies to the human progestin receptor. Endocrinology. 1988;122:1165-75. Prudhomme JF, Malet C, Gompel A, et al. 17/3-hydroxysteroid dehydrogenase activity in human breast cell and fibroblast cultures. Endocrinology. 1984;114:1483-9. Gompel A, Malet C, Spritzer P, Lalardrie JP, Kuttenn F, MauvaisJarvis P. Progestin effect on cell proliferation and 17/3-hydroxysteroid dehydrogenase activity in normal breast cell in culture. J Clin Endocrinol Metab. 1986;63:1174-80. Malet C, Gompel A, Spritzer P, Kuttenn F, Mauvais-Jarvis P. Effect of estradiol and the synthetic progestin promegestone (R5020) on the proliferation of human mammary cells in culture. In: Angeli A, Bradlow HL, Dogliotti L, (eds). Endocrinology of the breast. Basic and clinical aspects. Ann NY Acad Sci. 1986;464:48092. Malet C, Gompel A, Spritzer P, Bricout N, Kuttenn F, MauvaisJarvis P. Tamoxifen and hydroxytamoxifen isomers versus estradiol effects on normal human breast cells in culture. Cancer Res. 1988;48:7193-9. Bassler R. The morphology of hormone-induced structural changes in the female breast. Curr Top Pathol. 1970;53:7-89. Eisen MJ. The occurrence of benign and malignant mammary lesions in rats treated with crystalline estrogen. Cancer Res. 1932;2:1029-38. Lippman ME, Bolan G, Huff K. The effects of estrogens and

23.

24. 25.

26. 27. 28. 29. 30.

31. 32. 33.

34. 35.

36.

37.

17

antiestrogens on hormone-responsive human breast cancer in longterm tissue culture. Cancer Res. 1976;36:4595-601. Horowitz KB, Koseky Y, McGuire WL. Estrogen control of progesterone receptor in human breast cancer: role of estradiol and antiestrogen. Endocrinology. 1978;103:1742-51. Chalbos D, Vignon F, Keydar I, Rochefort H. Estrogens stimulate cell proliferation and induce secretory proteins in a human breast cancer line (T47D). J Clin Endocrinol Metab. 1982;55:276-83. Mauvais-Jarvis P, Gompel A. Hormones et sein. En amont du cancer. Paris: Flammarion Medecine-Sciences; 1989. Sitruk-Ware LR, Sterkers N, Mauvais-Jarvis P. Benign breast diseases: hormonal investigation. Obstet Gynecol. 1979;53:457-60. Jensen EV, Jacobson HI. Basic guides to the mechanism of estrogen action. Rec Prog Horm Res. 1962;18:387-414. Milgrom E, Thi L, Atger M, Baulieu EE. Mechanisms of regulating the conformation of progesterone receptor(s) in the uterus. J Biol Chem. 1973;248:6366-74. Tseng L, Gurpide E, Gusberg SB. Estradiol receptor and 17/3hydroxysteroid dehydrogenase in normal and abnormal human endometrium. Ann NY Acad Sci. 1977;286:190-8. Bayard F, Damilano S, Robel P, Baulieu EE. Cytoplasmic and nuclear estradiol and progesterone receptors in human endometrium. J Clin Endocrinol Metab. 1978;46:635-48. McGuire WL, Pearson OH, Segaloff A. Predicting hormone responsiveness in human breast cancer. In: McGuire WL, Carbone PP, Vollmer EP (eds). Estrogen receptors in human breast cancer. New York: Raven Press; 1975;17-30. Gompel A, Chomette G, Malet C, et al. Estradiol-progestin interaction in normal human breast cells in culture: ultrastructural studies. Breast disease-Senologia. 1985;34:149-56. Press MF, Udove JA, Greene GL. Progesterone receptor distribution in the human endometrium: analysis using monoclonal antibodies to the human progesterone receptor. Am J Pathol. 1988;131:112-24. King WJ, Greene GL. Monoclonal antibodies localize oestrogen receptor in the nuclei of target cells. Nature. 1984;307:745-7. Perrot-Applanat M, Logeat F, Groyer-Picard MT, Milgrom E. Immunocytochemical study of mammalian progesterone receptor using monoclonal antibodies. Endocrinology. 1985;116:1473-84. Martin P, Kuttenn F, Serment H, Mauvais-Jarvis P. Progesterone receptors in fibroadenomas. J Steroid Biochem. 1979;ll:1295-98. Kuttenn F, Fournier S, Durand JC, Mauvais-Jarvis P. Estradiol and progesterone receptors in human breast fibroadenomas. J Clin Endocrinol Metab. 1981;52:1225-9. Markopoulos C, Berger U, Wilson P, Gazet JC, Coombes RC. Oestrogen receptor content of normal breast cells and breast carcinomas throughout the menstrual cycle. Br Med J. 1988;296:134951. Press MF, Nousek-Goebl NA, Bur M, Greene GL. Estrogen receptor localization in the female genital tract. Am J Pathol. 1986;123:289-92. Perrot-Applanat M, Groyer-Picard MT, Lorenzo F, et al. Immunocytochemical study with monoclonal antibodies to progesterone receptor in human breast tumors. Cancer Res. 1987:47:2652-61. Lessey BA, Killam AP, Metzger DA, Haney AF, Greene GL, McCarty KS. Immunohistochemical analysis of human uterine estrogen and progesterone receptors throughout the menstrual cycle. J Clin Endocrinol Metab. 1988;67:334-40. Garcia E, Bouchard P, DeBrux J, et al. Use of immunocytochemistry of progesterone and estrogen receptors for endometrial dating. J Clin Endocrinol Metab. 1988;67:80-7. Berthois Y, Katzenellenbogen JA, Katzenellenbogen BS. Phenol red in tissue culture media is a weak estrogen: implications concerning the study of estrogen-responsive cells in culture. Proc Natl Acad Sci USA. 1986;83:2496-500. Ree AH, Landmark BF, Eskild W, et al. Autologous down regulation of messenger ribonucleic acid and protein levels for estrogen receptors in MCF7 cells: an inverse correlation to progesterone receptor levels. Endocrinology 1989;124:2577-83. Read LD, Srider CE, Miller JS, Greene GL, Katzenellenbogen BS. Ligand-modulated regulation of progesterone receptor messenger ribonucleic acid and protein in human breast cancer cells lines. Mol Endocrinol. 1988;2:263-71.


Studies on steroid hormone receptors (5alpha-dihidrotesterone, estradiol, and dexamethasone) in cultured human fibroblasts and amniotic fluid cells.

Endothelin messenger RNA and receptors are differentially expressed in cultured human breast epithelial and stromal cells.

Characterization of normal breast epithelial cells in primary cultures: differentiation and growth factor receptors studies.

Progesterone stimulates progenitor cells in normal human breast and breast cancer cells.

Immunocytochemical localization of progesterone receptors in endocrine cells of the human pancreas.

Endothelin receptors in cultured renal epithelial cells.

Cytoplasmic receptors for 17 beta-estradiol, 5 alpha-dihydrotestosterone and progesterone in normal and abnormal human uterine tissues.

Expression of estradiol and progesterone receptors by histologically normal endometria of women with postmenopausal bleeding.

Down-Regulation of miR-92 in Breast Epithelial Cells and in Normal but Not Tumour Fibroblasts Contributes to Breast Carcinogenesis.

Changes in the apical microfilaments of rat uterine epithelial cells in response to estradiol and progesterone.

Estrone and estradiol content in human breast tumors: relationship to estradiol receptors.

Estrogen and progesterone receptors in breast cancer.

Progesterone receptor determined by immunocytochemical and biochemical methods in human breast cancer.

Pharmacological and immunocytochemical characterization of metabotropic glutamate receptors in cultured Purkinje cells.

Effects of ozone exposure on human epithelial adenocarcinoma and normal fibroblasts cells.

Progesterone receptors in breast fibroadenomas.

Morphological and immunocytochemical characterization of cultured rat incisor cervical epithelial cells.

Estrogen control of progesterone receptor in human breast cancer: role of estradiol and antiestrogen.

Estradiol formation from testosterone by continuously cultured human breast cancer cells.

Estrogen and progesterone receptors in epithelial ovarian tumours.

Surface modification of microparticles causes differential uptake responses in normal and tumoral human breast epithelial cells.

Electrophysiology of cultured human lens epithelial cells.

GPER mediates estrogen-induced signaling and proliferation in human breast epithelial cells and normal and malignant breast.

Sildenafil Effect on Nitric Oxide Secretion by Normal Human Endometrial Epithelial Cells Cultured In vitro.