Cytochemical Study of Estrogen Receptor in Human Mammary Cancer SIN HANG LEE, M.D., F.R.C.P.(C) Lee, Sin Hang: Cytochemical study of estrogen receptor in human mammary cancer. Am J Clin Pathol 70: 197-203, 1978. A 17/3-estradiol-6-carboxymethyl-oxime-bovine serum albumin-fluorescein isothiocyanate conjugate is prepared by attaching on the average 11 moles of the fluorescein dye and 24 moles of the steroid hormone to each mole of the protein carrier. Thisfluorescentestradiol conjugate is used as a tracer to detect estrogen receptor of human mammary cancer cells in frozen sections. The cytochemical findings indicate that mammary carcinomas are composed of heterogeneous populations of receptor-positive and receptor-negative cancer cells in varying proportions and probably should be classified according to the percentages of receptor-positive cells in the cancer cell populations for better correlation with endocrine therapies. (Key words: 17/3-estradiol-6-carboxymethyl-oxime-bovine serum albumin-fluorescein isothiocyanate; Fluorescent estradiol conjugate; Estrogen receptor; Cytochemical study; Human mammary cancer.)
Department of Pathology, Hospital of St. Raphael, New Haven, Connecticut
IN RECENT YEARS, assay of estrogen receptor in cytosol protein of tumor tissue homogenates has proved to be useful in selecting patients with advanced mammary carcinoma for hormonal therapy. 5 Patients selected by estrogen receptor assay have been reported to show an objective remission rate of 32-60%, 4 , 6 1 2 in contrast to a 20-30% 3 1 5 response in those selected by clinical criteria alone. However, failure to obtain higher response rates among patients whose tumors were interpreted as positive for estrogen receptor and the fact that about 10% of those patients who had mammary cancers with no estrogen receptor value demonstrable still responded favorably to a variety of endocrine therapies indicate the need for further improvement of the existent methodology of discriminating potential hormone-responsive cases. Since the current biochemical assay measures estradiol-binding capacity of the proteins in the cytosol fraction of tissue homogenates, its results are influenced by the actual size of the epithelial component of the cancer in the tissue submitted for assay, the percentage of cancer receptor protein extractable into the cytosol, and the proteins extractable from other noncancerous tissue components. Consequently, variations in these factors alone may
Materials and Methods Synthesis of 17(i-estradiol-6-(0-carboxymethyl)
Oxime
According to the method of Longwell and Wintersteiner, 10 10 g 17y3-estradiol diacetate* were dissolved in 34.5 ml glacial acetic acid, and a solution of 8.5 g chromium trioxide in 6.9 ml water and 51 ml of glacial acetic acid added. The mixture was stirred at room temperature for 24 hr, diluted to 500 ml with water, and extracted four times each with 500 ml ether. Most of the acetic acid was removed from the combined ether extract by washing it with saturated N a H C 0 3 solution until a faint pink color appeared in the aqueous phase. The ether phase was further exhaustively washed with a 3:1 mixture of 5% Na 2 C0 3 and saturated N a H C 0 3 solutions, then with water, and evaporated to dryness. The residue was dissolved in a 1:4 ethyl acetate-petroleum ether (boiling point 6 0 100 C)t mixture, and passed through a 22 x 4.5-cm column of silica gel. The column was eluted with 3 liters 1:4 ethyl acetate-hexane, and 200-ml fractions were collected. The peak of 6-keto-17/3-estradiol diacetate preceded the appearance of the yellowish overoxidized byproducts, probably a keto acid and its derivatives. After evaporation of the solvent, the crude products of 6-keto-17/3-estradiol diacetate were com-
Received January 6, 1978; received revised manuscript March 3, 1978; accepted for publication March 3, 1978. Presented at the 62nd Annual Meeting of the Federation of American Societies for Experimental Biology, Atlantic City, N.J., April 9-14, 1978. Address reprint requests to Dr. Lee: Department of Pathology, Hospital of St. Raphael, New Haven, Connecticut 06511.
* Sigma Chemical Co., St. Louis, Mo. t J. T. Baker Chemical Co., Phillipsburg, N.J.
0002-9173/78/0800/0197 $00.85 © Ai
can Society of Clinical Pathologists
197
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 5, 2016
contribute to wide fluctuations of the final deduced estrogen receptor values, not always reflecting true physiologic states of the cancer cell population. The purpose of this study was to explore the possibility of developing a cytochemical method that can be used to investigate estrogen binding of individual cancer cells, and potentially as an alternative or supplementary predictor in the selection of patients for hormonal manipulation. Since the key chemical compounds needed for preparation of the receptor tracer are not readily available from commercial sources, technical procedure will be described in detail.
198
Preparation of Fluorescein Isothiocyanate (FITC)Bovine Serum Albumin (BSA) Complex (FITC-BSA) Standard procedures14 were followed except that a high FITC:BSA ratio was used. In detail, 1 g BSA§ (crystallized and lyophilized) was dissolved in 10 ml 0.5 M carbonate-bicarbonate buffer, pH 9.5. One gram of FITC on celite (10%)§ was added to the solution and stirred for 4 hr at room temperature. After removal of the celite by centrifugation, the FITC-BSA complex was passed through a Sephadex G-25 column which previously equilibrated with 0.05 M phosphate buffer, pH 7.8, and dialyzed against the same buffer for additional 48 hr to assure complete removal of any loosely bound FITC. Its protein content was determined by the biuret reaction. The average molecular fluorescein/ protein ratio was 11.2 (±0.5), calculated from the absorbancy of pure FITC§ in 0.1 N NaOH at 495 nm.
A portion of FITC-BSA complex in phosphate buffer containing 200 mg protein, 200 mg l-ethyl-3-(3-diX Aldrich Chemical Co., Milwaukee, Wis. § Sigma Chemical Co., St. Louis, Mo.
methylaminopropyl) carbodiimide hydrochloride,11 and 200 mg 17)3-estradiol-6-(0-carboxymethyl) oxime dissolved in 6 ml dioxane, 5 ml water and adequate 0.05 M phosphate buffer, pH 7.8, to make a total volume of 21 ml were stirred for 20 hr at room temperature. The conjugate was exhaustively dialyzed first against water, then against phosphate-buffered saline solution (PBS), pH 7.4, for 48 hr, and centrifuged (28,000 x g, 15 min). The supernatant was collected, and sodium azide added to a final concentration of 1 mg/ml as preservative. It was further diluted to the desired concentration with PBS before use (see results). Cytochemical Staining of Estrogen Receptors in Tissue Sections Unfixed frozen sections of tissues containing primary or metastatic mammary carcinoma removed at operation were cut about 14 p.m thick, mounted on microscopic slides, and dried in a refrigerator (2-5 C) for 1 hr. They were then rehydrated with a brief rinse in PBS, covered with fluorescent estradiol conjugate, and incubated in a humid chamber at room temperature for 2 hr. After removal of the conjugate, the sections were rinsed gently in PBS, immersed in the buffer for 1 hr with one change of fresh washing solution, covered with phosphate-buffered glycerin and coverslip, and examined with a fluorescence microscope. Positive and negative blocking of specific staining of the cellular estrogen receptor was performed by preincubation of the dried frozen sections in a 17/3-estradiol-6-(0-carboxymethyl) oxime-BSA compound without FITC and in a plain buffered BSA solution, respectively, for 1 hr before serial dilutions of fluorescent estradiol conjugate were applied. The sections were incubated, washed and examined as described above. Results Identification of 17/3-estradiol-6-CMO and its precursors was carried out by melting points and thin-layer chromatography on silica gel plates (solvent system, ethyl acetate:N-hexane:ethanol:acetic acid, 72:13.5: 4.5:10). The values were found to be those expected for these chemicals as reported in the literature.1-9 The sequence of chemical reaction involved in the preparation of the 17/3-estradiol-6-CMO-BSA-FITC conjugate is illustrated in Figure 1. For routine staining, the dialyzed fluorescent estradiol conjugate was diluted with PBS, pH 7.4, to a final concentration equivalent to 0.5 p-mol (± 10%) FITC per ml, and should show a 17/3-estradiol-6-CMO:FITC 1 Story Chemical Corporation, Muskegon, Mich.
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 5, 2016
bined and recrystallized from absolute ethanol, yielding 2.5 g of needle-like crystals (melting point 173175 C). To prepare 6-keto-17/3-estradiol, 2.5 g 6-keto-17/3estradiol diacetate were hydrolyzed in 75 ml 20% KOH in methyl alcohol under nitrogen for 24 hr at room temperature. The yellow solution was then diluted with water to about 400 ml, acidified with hydrochloric acid topH 2, and extracted three times with 500 ml of ether each. The ether layer was washed with 5% Na 2 C0 3 solution saturated with NaHCO,-,, and with water. After evaporation to dryness, the crude product was recrystallized twice from absolute ethanol as crystal plates (melting point 280-282 C; yield 1.5 g). The general method outlined by Erlanger and associates2 was followed to prepare 17/3-estradiol-6-(0-carboxymethyl) oxime. In 75 ml ethanol and 6.8 ml 2 N KOH, 1.5 g 6-keto-17/3-estradiol was dissolved; 1.5 g carboxymethoxylamine hemihydrochloride (94%)$ was added. The mixture was refluxed for 3 hours, and the alcohol evaporated. After addition of 150 ml water to dissolve the residue, the solution was adjusted to pH 8.5, filtered to remove insoluble particles, if any, extracted twice with 30 ml ethyl acetate each, and acidified to pH 2 with HC1, giving a white precipitate. The latter was collected by filtration, dried in a 37 C incubator, and recrystallized twice from acetone, yielding 1.4 g needles (melting point 197-199 C).
Preparation of 17p-Estradiol-6-CMO-BSA-FITC Conjugate (Fluorescent Estradiol Conjugate)
A J C.P. . Augusl 1978
LEE
CYTOCHEMISTRY OF ESTROGEN RECEPTOR
Vol. 70 • No. 2
OAc
199
OAc
NH r 0-CH r C00H AcO-^
BSA +
FITC
CH2C0CIH
BSA-FITC (carbodiimide)
Fiu. I. Sequence of chemical reaction for preparation of 17/3-estradiol-6-carboxymethyl-oxime-bovine serum albumin-fluorescein isothiocyanate (fluorescent estradiol conjugate).
molecular ratio of 2.2 (±10%), estimated by absorbancy in 0.1 N NaOH at 340 nm and 495 nm, respectively, using the unconjugated oxime and pure FITC (Sigma Chemical Co.) in 0.1 N NaOH as standards. No free hormone or fluorescein dye was detected on thin-layer chromatography. After incubation with the fluorescent estradiol conjugate, mammary cancer cells in the unfixed frozen sections, when examined with a fluorescence microscope, were readily separated into two major categories, i.e., cancer cells manifesting coarse brilliant green, primarily cytoplasmic, fluorescence, and those that did not fluoresce. The cellular fluorescence was interpreted as a result of specific binding of estradiol residues of the conjugate and accepted as evidence indicating the presence of estrogen receptor, since this fluorescent staining was successfully blocked by preincubation of the tissue sections in nonfluorescent 17/3estradiol-6-CMO-BSA, but not in plain BSA solution. In order to obtain complete specific blocking, it was found necessary to dilute the fluorescent estradiol conjugate or to use conjugates with low steroid/protein ratios so that the concentration of estradiol residues in the stain would not exceed 1/10 of that in the blocking solutions, which had a steroid/protein ratio of 20 to 25. Attempts to use unconjugated 17/3-estradiol or 17/3estradiol-6-CMO as blocking agent were unsuccessful, because in the absence of a protein carrier the solubilities of these chemicals in aqueous solutions of physiologic pH were not high enough to saturate the binding sites of estrogen receptor-positive cancer cells.
The fluorescent estradiol conjugate did not stain tissue sections of the vermiform appendices, tonsils, or carcinomas of the colon and stomach, which were prepared in the same manner as negative controls. FITC-BSA complex diluted to a concentration equivalent to 0.5 ixmo\ FITC/ml was also used as a negative nonspecific staining control, and failed to stain estrogen receptor-positive cancer cells. Fixation of the tissue sections containing estrogen receptor-positive cancer cells in cold acetone (—20 C), 3.7% buffered formaldehyde, and 1% glutaraldehyde resulted in complete abolition of the cellular estrogenbinding capacity. Application of this cytochemical method to clinical biopsy materials at the time of this reporting was largely confined to mammary lesions, including 17 primary infiltrating ductal carcinomas, two metastatic mammary carcinomas in axillary lymph nodes, three fibroadenomas, ten cases of benign "fibrocystic mastopathy," and one case of male gynecomastia, all from different patients. The preliminary findings were as follows. Among all 19 breast cancer cases studied only two carcinomas (one primary and one metastatic) consisted of almost homogeneous populations of strongly estrogen receptor-positive cancer cells, which were characterized by coarse heavy deposits of the fluorescent estradiol conjugate in the cytoplasm, and to a lesser extent in the nuclei (Fig. 2). The concomitant lymphocytes, polymorphonuclear leukocytes, histiocytes, plasma cells, and blood vessels, and the intervening
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 5, 2016
NH-BSA-FITC
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 5, 2016
^S.*
A.J.C.P. • August 1978
LEE
200
8
10
FIGS. 2-10 (legends on facing page).
Vol. 70 . No. 2
CYTOCHEMISTRY OF ESTROGEN RECEPTOR
201
FIG. 2. Frozen section of a metastatic mammary carcinoma in an axillary lymph node stained with fluorescent estradiol conjugate. Almost all cancer cells in this tumor were positive for estrogen receptor, characterized by green cytoplasmic fluorescence. The nuclei showed less binding activity than the cytoplasm. The lymphoid tissue between the cancer cells was not stained. x416. FIGS. 3 and 4. Two consecutive frozen sections of a primary cancer of the breast, showing the same area stained with fluorescent estradiol conjugate and with hematoxylin and eosin, respectively. The conjugate was deposited selectively in the cancer cells, but not in the inflammatory cells and fibroblasts in the vicinity. Originally x260. FIG. 5. A primary cancer of the breast composed of about 50% estrogen receptor-positive and 50% receptor-negative cells. This photograph was purposely overexposed to outline the weakly autofluorescent stroma surrounding the estrogen receptor-negative cancer cells. xl66. FIG. 6. Primary cancer of the breast with mostly ER-positive cancer cells infiltrating the stroma around a duct filled with hyperplastic epithelium. The ductal epithelial cells showed weak estrogen binding, x 166. FIG. 7. Benign epithelial cells in a female breast with "fibrocystic mastopathy," showing uneven weak estrogen receptivity among cells lining the ducts and lobules, x 166. FIG. 8. Consecutive section adjacent to that shown in Figure 7. Hematoxylin and eosin. x 104.
FIG. 10. Male breast with gynecomastia, showing pattern of estrogen binding similar to that in Figure 9. x260.
Department of Pathology, Hospital of St. Raphael, New Haven, Connecticut
IN RECENT YEARS, assay of estrogen receptor in cytosol protein of tumor tissue homogenates has proved to be useful in selecting patients with advanced mammary carcinoma for hormonal therapy. 5 Patients selected by estrogen receptor assay have been reported to show an objective remission rate of 32-60%, 4 , 6 1 2 in contrast to a 20-30% 3 1 5 response in those selected by clinical criteria alone. However, failure to obtain higher response rates among patients whose tumors were interpreted as positive for estrogen receptor and the fact that about 10% of those patients who had mammary cancers with no estrogen receptor value demonstrable still responded favorably to a variety of endocrine therapies indicate the need for further improvement of the existent methodology of discriminating potential hormone-responsive cases. Since the current biochemical assay measures estradiol-binding capacity of the proteins in the cytosol fraction of tissue homogenates, its results are influenced by the actual size of the epithelial component of the cancer in the tissue submitted for assay, the percentage of cancer receptor protein extractable into the cytosol, and the proteins extractable from other noncancerous tissue components. Consequently, variations in these factors alone may
Materials and Methods Synthesis of 17(i-estradiol-6-(0-carboxymethyl)
Oxime
According to the method of Longwell and Wintersteiner, 10 10 g 17y3-estradiol diacetate* were dissolved in 34.5 ml glacial acetic acid, and a solution of 8.5 g chromium trioxide in 6.9 ml water and 51 ml of glacial acetic acid added. The mixture was stirred at room temperature for 24 hr, diluted to 500 ml with water, and extracted four times each with 500 ml ether. Most of the acetic acid was removed from the combined ether extract by washing it with saturated N a H C 0 3 solution until a faint pink color appeared in the aqueous phase. The ether phase was further exhaustively washed with a 3:1 mixture of 5% Na 2 C0 3 and saturated N a H C 0 3 solutions, then with water, and evaporated to dryness. The residue was dissolved in a 1:4 ethyl acetate-petroleum ether (boiling point 6 0 100 C)t mixture, and passed through a 22 x 4.5-cm column of silica gel. The column was eluted with 3 liters 1:4 ethyl acetate-hexane, and 200-ml fractions were collected. The peak of 6-keto-17/3-estradiol diacetate preceded the appearance of the yellowish overoxidized byproducts, probably a keto acid and its derivatives. After evaporation of the solvent, the crude products of 6-keto-17/3-estradiol diacetate were com-
Received January 6, 1978; received revised manuscript March 3, 1978; accepted for publication March 3, 1978. Presented at the 62nd Annual Meeting of the Federation of American Societies for Experimental Biology, Atlantic City, N.J., April 9-14, 1978. Address reprint requests to Dr. Lee: Department of Pathology, Hospital of St. Raphael, New Haven, Connecticut 06511.
* Sigma Chemical Co., St. Louis, Mo. t J. T. Baker Chemical Co., Phillipsburg, N.J.
0002-9173/78/0800/0197 $00.85 © Ai
can Society of Clinical Pathologists
197
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 5, 2016
contribute to wide fluctuations of the final deduced estrogen receptor values, not always reflecting true physiologic states of the cancer cell population. The purpose of this study was to explore the possibility of developing a cytochemical method that can be used to investigate estrogen binding of individual cancer cells, and potentially as an alternative or supplementary predictor in the selection of patients for hormonal manipulation. Since the key chemical compounds needed for preparation of the receptor tracer are not readily available from commercial sources, technical procedure will be described in detail.
198
Preparation of Fluorescein Isothiocyanate (FITC)Bovine Serum Albumin (BSA) Complex (FITC-BSA) Standard procedures14 were followed except that a high FITC:BSA ratio was used. In detail, 1 g BSA§ (crystallized and lyophilized) was dissolved in 10 ml 0.5 M carbonate-bicarbonate buffer, pH 9.5. One gram of FITC on celite (10%)§ was added to the solution and stirred for 4 hr at room temperature. After removal of the celite by centrifugation, the FITC-BSA complex was passed through a Sephadex G-25 column which previously equilibrated with 0.05 M phosphate buffer, pH 7.8, and dialyzed against the same buffer for additional 48 hr to assure complete removal of any loosely bound FITC. Its protein content was determined by the biuret reaction. The average molecular fluorescein/ protein ratio was 11.2 (±0.5), calculated from the absorbancy of pure FITC§ in 0.1 N NaOH at 495 nm.
A portion of FITC-BSA complex in phosphate buffer containing 200 mg protein, 200 mg l-ethyl-3-(3-diX Aldrich Chemical Co., Milwaukee, Wis. § Sigma Chemical Co., St. Louis, Mo.
methylaminopropyl) carbodiimide hydrochloride,11 and 200 mg 17)3-estradiol-6-(0-carboxymethyl) oxime dissolved in 6 ml dioxane, 5 ml water and adequate 0.05 M phosphate buffer, pH 7.8, to make a total volume of 21 ml were stirred for 20 hr at room temperature. The conjugate was exhaustively dialyzed first against water, then against phosphate-buffered saline solution (PBS), pH 7.4, for 48 hr, and centrifuged (28,000 x g, 15 min). The supernatant was collected, and sodium azide added to a final concentration of 1 mg/ml as preservative. It was further diluted to the desired concentration with PBS before use (see results). Cytochemical Staining of Estrogen Receptors in Tissue Sections Unfixed frozen sections of tissues containing primary or metastatic mammary carcinoma removed at operation were cut about 14 p.m thick, mounted on microscopic slides, and dried in a refrigerator (2-5 C) for 1 hr. They were then rehydrated with a brief rinse in PBS, covered with fluorescent estradiol conjugate, and incubated in a humid chamber at room temperature for 2 hr. After removal of the conjugate, the sections were rinsed gently in PBS, immersed in the buffer for 1 hr with one change of fresh washing solution, covered with phosphate-buffered glycerin and coverslip, and examined with a fluorescence microscope. Positive and negative blocking of specific staining of the cellular estrogen receptor was performed by preincubation of the dried frozen sections in a 17/3-estradiol-6-(0-carboxymethyl) oxime-BSA compound without FITC and in a plain buffered BSA solution, respectively, for 1 hr before serial dilutions of fluorescent estradiol conjugate were applied. The sections were incubated, washed and examined as described above. Results Identification of 17/3-estradiol-6-CMO and its precursors was carried out by melting points and thin-layer chromatography on silica gel plates (solvent system, ethyl acetate:N-hexane:ethanol:acetic acid, 72:13.5: 4.5:10). The values were found to be those expected for these chemicals as reported in the literature.1-9 The sequence of chemical reaction involved in the preparation of the 17/3-estradiol-6-CMO-BSA-FITC conjugate is illustrated in Figure 1. For routine staining, the dialyzed fluorescent estradiol conjugate was diluted with PBS, pH 7.4, to a final concentration equivalent to 0.5 p-mol (± 10%) FITC per ml, and should show a 17/3-estradiol-6-CMO:FITC 1 Story Chemical Corporation, Muskegon, Mich.
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 5, 2016
bined and recrystallized from absolute ethanol, yielding 2.5 g of needle-like crystals (melting point 173175 C). To prepare 6-keto-17/3-estradiol, 2.5 g 6-keto-17/3estradiol diacetate were hydrolyzed in 75 ml 20% KOH in methyl alcohol under nitrogen for 24 hr at room temperature. The yellow solution was then diluted with water to about 400 ml, acidified with hydrochloric acid topH 2, and extracted three times with 500 ml of ether each. The ether layer was washed with 5% Na 2 C0 3 solution saturated with NaHCO,-,, and with water. After evaporation to dryness, the crude product was recrystallized twice from absolute ethanol as crystal plates (melting point 280-282 C; yield 1.5 g). The general method outlined by Erlanger and associates2 was followed to prepare 17/3-estradiol-6-(0-carboxymethyl) oxime. In 75 ml ethanol and 6.8 ml 2 N KOH, 1.5 g 6-keto-17/3-estradiol was dissolved; 1.5 g carboxymethoxylamine hemihydrochloride (94%)$ was added. The mixture was refluxed for 3 hours, and the alcohol evaporated. After addition of 150 ml water to dissolve the residue, the solution was adjusted to pH 8.5, filtered to remove insoluble particles, if any, extracted twice with 30 ml ethyl acetate each, and acidified to pH 2 with HC1, giving a white precipitate. The latter was collected by filtration, dried in a 37 C incubator, and recrystallized twice from acetone, yielding 1.4 g needles (melting point 197-199 C).
Preparation of 17p-Estradiol-6-CMO-BSA-FITC Conjugate (Fluorescent Estradiol Conjugate)
A J C.P. . Augusl 1978
LEE
CYTOCHEMISTRY OF ESTROGEN RECEPTOR
Vol. 70 • No. 2
OAc
199
OAc
NH r 0-CH r C00H AcO-^
BSA +
FITC
CH2C0CIH
BSA-FITC (carbodiimide)
Fiu. I. Sequence of chemical reaction for preparation of 17/3-estradiol-6-carboxymethyl-oxime-bovine serum albumin-fluorescein isothiocyanate (fluorescent estradiol conjugate).
molecular ratio of 2.2 (±10%), estimated by absorbancy in 0.1 N NaOH at 340 nm and 495 nm, respectively, using the unconjugated oxime and pure FITC (Sigma Chemical Co.) in 0.1 N NaOH as standards. No free hormone or fluorescein dye was detected on thin-layer chromatography. After incubation with the fluorescent estradiol conjugate, mammary cancer cells in the unfixed frozen sections, when examined with a fluorescence microscope, were readily separated into two major categories, i.e., cancer cells manifesting coarse brilliant green, primarily cytoplasmic, fluorescence, and those that did not fluoresce. The cellular fluorescence was interpreted as a result of specific binding of estradiol residues of the conjugate and accepted as evidence indicating the presence of estrogen receptor, since this fluorescent staining was successfully blocked by preincubation of the tissue sections in nonfluorescent 17/3estradiol-6-CMO-BSA, but not in plain BSA solution. In order to obtain complete specific blocking, it was found necessary to dilute the fluorescent estradiol conjugate or to use conjugates with low steroid/protein ratios so that the concentration of estradiol residues in the stain would not exceed 1/10 of that in the blocking solutions, which had a steroid/protein ratio of 20 to 25. Attempts to use unconjugated 17/3-estradiol or 17/3estradiol-6-CMO as blocking agent were unsuccessful, because in the absence of a protein carrier the solubilities of these chemicals in aqueous solutions of physiologic pH were not high enough to saturate the binding sites of estrogen receptor-positive cancer cells.
The fluorescent estradiol conjugate did not stain tissue sections of the vermiform appendices, tonsils, or carcinomas of the colon and stomach, which were prepared in the same manner as negative controls. FITC-BSA complex diluted to a concentration equivalent to 0.5 ixmo\ FITC/ml was also used as a negative nonspecific staining control, and failed to stain estrogen receptor-positive cancer cells. Fixation of the tissue sections containing estrogen receptor-positive cancer cells in cold acetone (—20 C), 3.7% buffered formaldehyde, and 1% glutaraldehyde resulted in complete abolition of the cellular estrogenbinding capacity. Application of this cytochemical method to clinical biopsy materials at the time of this reporting was largely confined to mammary lesions, including 17 primary infiltrating ductal carcinomas, two metastatic mammary carcinomas in axillary lymph nodes, three fibroadenomas, ten cases of benign "fibrocystic mastopathy," and one case of male gynecomastia, all from different patients. The preliminary findings were as follows. Among all 19 breast cancer cases studied only two carcinomas (one primary and one metastatic) consisted of almost homogeneous populations of strongly estrogen receptor-positive cancer cells, which were characterized by coarse heavy deposits of the fluorescent estradiol conjugate in the cytoplasm, and to a lesser extent in the nuclei (Fig. 2). The concomitant lymphocytes, polymorphonuclear leukocytes, histiocytes, plasma cells, and blood vessels, and the intervening
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 5, 2016
NH-BSA-FITC
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 5, 2016
^S.*
A.J.C.P. • August 1978
LEE
200
8
10
FIGS. 2-10 (legends on facing page).
Vol. 70 . No. 2
CYTOCHEMISTRY OF ESTROGEN RECEPTOR
201
FIG. 2. Frozen section of a metastatic mammary carcinoma in an axillary lymph node stained with fluorescent estradiol conjugate. Almost all cancer cells in this tumor were positive for estrogen receptor, characterized by green cytoplasmic fluorescence. The nuclei showed less binding activity than the cytoplasm. The lymphoid tissue between the cancer cells was not stained. x416. FIGS. 3 and 4. Two consecutive frozen sections of a primary cancer of the breast, showing the same area stained with fluorescent estradiol conjugate and with hematoxylin and eosin, respectively. The conjugate was deposited selectively in the cancer cells, but not in the inflammatory cells and fibroblasts in the vicinity. Originally x260. FIG. 5. A primary cancer of the breast composed of about 50% estrogen receptor-positive and 50% receptor-negative cells. This photograph was purposely overexposed to outline the weakly autofluorescent stroma surrounding the estrogen receptor-negative cancer cells. xl66. FIG. 6. Primary cancer of the breast with mostly ER-positive cancer cells infiltrating the stroma around a duct filled with hyperplastic epithelium. The ductal epithelial cells showed weak estrogen binding, x 166. FIG. 7. Benign epithelial cells in a female breast with "fibrocystic mastopathy," showing uneven weak estrogen receptivity among cells lining the ducts and lobules, x 166. FIG. 8. Consecutive section adjacent to that shown in Figure 7. Hematoxylin and eosin. x 104.
FIG. 10. Male breast with gynecomastia, showing pattern of estrogen binding similar to that in Figure 9. x260.
