The distribution of keratin type intermediate filaments in human breast cancer. An immunohistological study

Antibodies to different intermediate filament proteins can be used to distinguish cells of epithelial, mesenchymal, muscle, glial and neuronal origin. Antibodies to prekeratin which characterize cells of epithelial origin, and antibodies to vimentin which recognize cells of mesenchymal origin have been used to study twenty cases of breast carcinoma (sixteen infiltrating ductal carcinomas and four infiltrating intraductal carcinomas), two cases of cystic breast disease, two fibroadenomas and one case of benign cystosarcoma phylloides. The prekeratin and vimentin were detected using specific antibodies to these proteins by immunofluorescence microscopy using alcohol fixed paraffin-embedded tissues. In eighteen out of the twenty carcinomas the tumor cells were strongly and specifically stained by antibodies to prekeratin. DIfferent tumors gave different patterns of prekeratin staining. In contrast, when the same specimens were tested with the vimentin antibody, the tumor cells were unstained, and instead only the usual strong staining to fibroblasts and blood vessels in the stroma was observed. In cystic breast disease, fibroadenomas, and benign cystosarcoma phylloides, cells of epithelial origin were strongly stained by the prekeratin but not by the vimentin antibody.     

Development of (p-O-sulfamoyl)-N-alkanoyl-phenylalkyl amines as non-steroidal estrone sulfatase inhibitors

Estrogen levels in breast tumors of postmenopausal women are as much as 10 times higher than estrogen levels in plasma, presumably due to in situ formation of estrogen. The major source of estrogen in breast cancer cells may be conversion of estrone sulfate to estrone by the enzyme estrone sulfatase. Thus, inhibitors of estrone sulfatase are potential agents for treatment of estrogen-dependent breast cancer. Several steroidal compounds have been developed that are potent estrone sulfatase inhibitors, most notably estrone-3-O-sulfamate. However, these compounds and their metabolites may have undesired effects, including estrogenicity. To avoid the problems associated with a potentially active steroid nucleus, we designed and synthesized a series of nonsteroidal estrone sulfatase inhibitors, the (p-O-sulfamoyl)-N-alkanoyl phenylalkyl amines. The compounds synthesized vary in the length of their alkanoyl chain and in the number of carbons separating the phenyl ring and the carbonyl carbon. The ability of these compounds to inhibit estrone sulfatase activity was tested using human placental microsomes and intact cultured human breast cancer cells. Estrogenicity was also evaluated, using growth of estrogen-dependent human breast cancer cells. All of the test compounds inhibited estrone sulfatase activity of human placental microsomes to some extent, with the most effective compound having an IC50 value of 72 nM. In general, compounds with longer alkanoyl chains (12-14 carbons) were more effective than those with shorter chains. The test compounds also inhibited estrone sulfatase activity in intact cultures of MDA-MB-231 human breast cancer cells. Again, the longer chain compounds were more effective. In both the placental and breast cancer cell sulfatase assays, the optimal distance between the phenyl ring and the carbonyl carbon was 1-2 carbons. The MCF-7 cell proliferation assay revealed that estrone and estrone-3-O-sulfamate were both estrogenic, but the (p-O-sulfamoyl)-N-alkanoyl phenylalkyl amines were not. Our data indicate the utility of (p-O-sulfamoyl)-N-alkanoyl phenyl alkylamines for inhibition of estrone sulfatase activity. Furthermore, our data support the concept that nonsteroidal estrone sulfatase inhibitors may be useful as therapeutic agents for estrogen-dependent breast cancers.     

Steroidal oxathiazine inhibitors of estrone sulfatase

The presence of estrone sulfatase in breast tumors and the high levels of circulating estrone sulfate may contribute the major portion of estrogen synthesized locally in breast tissues through conversion of estrone sulfate to estrone by the enzyme. Using inhibitors of estrone sulfatase for the treatment of estrogen-dependent (estrogen receptor positive, ER(+)) breast cancer could be a very effective therapeutic strategy for the treatment of estrogen-dependent breast tumors in postmenopausal women. Therefore, we designed and synthesized several steroidal 2',3'-oxathiazines that inhibit estrone sulfatase and have greatly reduced estrogenic side effects. Our in vitro studies indicate that the oxathiazine compounds have inhibitory activity on estrone sulfatase in MCF-7 human breast cancer cells. These estrone sulfatase inhibitors (ESIs) also inhibit the growth of MCF-7 cells induced by estrone sulfate. In addition, our in vivo experiments demonstrate that our ESIs have moderate antitumor activity against MCF-7 breast cancer xenografts in Balb/c athymic nude mice. The synthesis and biological activity of a number of these unique steroidal ESIs are described.     

Sulfamoyloxy-substituted 2-phenylindoles: antiestrogen-based inhibitors of the steroid sulfatase in human breast cancer cells

Estrone sulfate (E1S) is an endogenous prodrug that delivers estrone and, subsequently, estradiol to the target cells following the hydrolysis by the enzyme estrone sulfatase which is active in various tissues including hormone dependent breast cancer cells. Blockade of this enzyme should reduce the estrogen level in breast cancer cells and prevent hormonal growth stimulation. Sulfamates of a variety of phenolic compounds have been shown to be inhibitors of estrone sulfatase. Our rational is based on findings that these inhibitors can undergo hydrolysis and the pharmacological effects of the free hydroxy compounds contribute to the bioactivity of the sulfamates. A desirable action of the metabolites would be an estrogen antagonism to block stimulatory effects of residual amounts of estrogens. Thus, we synthesized a number of sulfamoyloxy-substituted 2-phenylindoles with side chains at the indole nitrogen that guarantee antiestrogenic activity. All of the new sulfamates were studied for their inhibitory effects on the enzyme estrone sulfatase from human breast cancer cells and their (anti)hormonal activities in stably transfected human MCF-7/2a mammary carcinoma cells. The hormonal profile of the sulfamates was partly reflected by the properties of the corresponding hydroxy precursors. Some of the sulfamoylated antiestrogens strongly inhibited estrone sulfatase activity with IC₅₀ values in the submicromolar range. They were devoid of agonist activity and suppressed estrone sulfate-stimulated gene expression mainly by blocking the enzyme. Examples are the disulfamates of the indoles ZK 119, 010 and ZK 164, 015. Their IC₅₀s for sulfatase inhibition were 0.3 and 0.2 μM, respectively, and 50 and 80 nM, respectively, for the inhibition of E1S-stimulated luciferase expression in transfected MCF-7 cells. With some of the new sulfamates an additional direct antiestrogenic effect was noticed which might be due to a partial hydrolysis during incubation and would improve the growth inhibitory effect on estrogen-sensitive breast cancer cells.     

Fine needle aspiration cytology, immunocytochemistry and electron microscopy of fibromatosis of the breast. Report of two cases

Two cases of primary fibromatosis of the breast are described. The lesions were suspected to be carcinomas both clinically and mammographically. Fine needle aspiration (FNA) yielded bland-appearing isolated spindle cells associated with small groups of benign ductal cells and lymphocytes. Immunoperoxidase staining performed on the original FNA smears showed positivity for vimentin and muscle-specific actin only in the spindle cells and for cytokeratin only in the epithelial cells. Electron microscopy study of one case demonstrated the ultrastructural characteristics of well-differentiated fibroblasts and myofibroblasts.     

Development of novel steroid sulfatase inhibitors; II. TZS-8478 potently inhibits the growth of breast tumors in postmenopausal breast cancer model rats

In postmenopausal breast cancer tissue, steroid sulfatase (STS) activity is high and much estrone sulfate also exists; these facts reveal that estrone sulfate may be involved in the growth of breast cancer as an estrogen source. Steroid sulfatase is an enzyme, which catalyzes hydrolysis from estrone sulfate to estrone, and the development of steroid sulfatase inhibitors is expected as novel therapeutic drugs for postmenopausal breast cancer. We have developed a novel compound 2',4'-dicyanobiphenyl-4-O-sulfamate (TZS-8478), which has potent steroid sulfatase-inhibitory activity and exhibits no estrogenicity in vitro and in vivo. To elucidate its usefulness as a therapeutic drug for postmenopausal breast cancer, we examined the breast cancer cell proliferation- and breast tumor growth-inhibitory activity of TZS-8478 in postmenopausal breast cancer model rats. TZS-8478 dose-dependently suppressed the estrone sulfate-stimulated proliferation of MCF-7 cells. Regarding nitrosomethylurea (NMU)-induced postmenopausal breast cancer models, furthermore, TZS-8478 (0.5 mg/kg per day) markedly inhibited the estrone sulfate-stimulated growth of breast tumors similarly to estrone sulfate-depletion. TZS-8478 completely inhibited steroid sulfatase activity in tumor, uterus and liver, and also markedly lowered plasma concentrations of estrone and estradiol. The above mentioned results suggested that TZS-8478 may be useful as a therapeutic drug for estrogen-dependent postmenopausal breast cancer.     

Effect of nomegestrol acetate on estrogen biosynthesis and transformation in MCF-7 and T47-D breast cancer cells

Although ovaries serve as the primary source of estrogen for pre-menopausal women, after menopause estrogen biosynthesis from circulating precursors occurs in peripheral tissues by the action of several enzymes, 17beta-hydroxysteroid dehydrogenase 1 (17beta-HSD1), aromatase and estrogen sulfatase. In the breast, both normal and tumoral tissues have been shown to be capable of synthesizing estrogens, and this local estrogen production can be implicated in the development of breast tumors. In these tissues, estradiol (E(2)) can be synthesized by three pathways: (1) estrone sulfatase transforms estrogen sulfates into bioactive estrogens, (2) 17beta-HSD1 converts estrone (E(1)) into E(2), (3) aromatase which converts androgens into estrogens is also present and contributes to the in situ synthesis of active estrogens but to a far lesser extent than estrone sulfatase. Quantitative assessment of E(2) formation in human breast tumors indicates that metabolism of estrone sulfate (E(1)S) via the sulfatase pathway produces 100-500 times more E(2) than androgen aromatization. Breast tissue also possesses the estrogen sulfotransferase involved in the conversion of estrogens into their sulfates that are biologically inactive. In the present review, we summarized the action of the 19-nor-progestin nomegestrol acetate (NOMAC) on the sulfatase, 17beta-HSD1 and sulfotransferase activities in the hormone-dependent MCF-7 and T47-D human breast cancer cell lines. Using physiological doses of substrates NOMAC blocks very significantly the conversion of E(1)S to E(2). It inhibits the transformation of E(1) to E(2). NOMAC has a stimulatory effect on sulfotransferase activity in both cell lines, with a strong stimulating effect at low doses but only a weak effect at high concentrations. The effects on the three enzymes are always stronger in the progesterone-receptor rich T47-D cell line as compared with the MCF-7 cell line. Besides, no effect is found for NOMAC on the transformation of androstenedione to E(1) in the aromatase-rich choriocarcinoma cell line JEG-3. In conclusion, the inhibitory effect provoked by NOMAC on the enzymes involved in the biosynthesis of E(2) (sulfatase and 17HSD pathways) in estrogen-dependent breast cancer, as well as the stimulatory effect on the formation of the inactive E(1)S, can open attractive perspectives for future clinical trials.     

Steroid sulfatase and estrogen sulfotransferase in normal human tissue and breast carcinoma

Steroid sulfatase (STS) hydrolyzes inactive estrone sulfate (E1-S) to estrone (E1), while estrogen sulfotransferase (EST; SULT 1E1 or STE gene) sulfonates estrogens to estrogen sulfates. They are considered to play important roles in the regulation of local estrogenic actions in various human tissues, however, their biological significance remains largely unknown. Therefore, we examined the expression of STS and EST in non-pathologic human tissues and breast carcinomas. STS expression was very weak except for the placenta, while EST expression was markedly detected in various tissues examined. In breast carcinoma tissues, STS and EST immunoreactivity was detected in carcinoma cells in 74 and 44% of cases, respectively, and was significantly associated with their mRNA levels and enzymatic activities. STS immunoreactivity was significantly correlated with the tumor size, and an increased risk of recurrence. EST immunoreactivity was inversely correlated with the tumor size or lymph node status. Moreover, EST immunoreactivity was significantly associated with a decreased risk of recurrence or improved prognosis. Our results suggest that EST is involved in protecting various peripheral tissues from excessive estrogenic effects. In the breast carcinoma, STS and EST are suggested to play important roles in the regulation of in situ estrogen production in the breast carcinomas.     

Immunohistochemical localization of sex hormone-binding globulin in normal and neoplastic breast tissue.

The distribution of sex hormone-binding globulin-like antigens (SHBG-LA) in normal and neoplastic human breast tissues was investigated by immunohistochemistry, employing a monospecific polyclonal antiserum against highly purified human SHBG and an avidin-biotin-peroxidase complex (ABC) method in formalin-fixed paraffin embedded tissue sections. In normal breast tissues the staining of SHBG-LA was present exclusively in the cytoplasm of epithelial cells of ductal and ductular types. Nuclei as well as stromal and lymphatic tissues remained unstained. While the staining was positive in all cases of intraductal carcinoma, only 4 out of 15 infiltrating carcinomas revealed SHBG-LA. The demonstration of a plasma sex steroid binding globulin in the cytoplasm of endocrine target cells is consistent with the hypothesis that steroid-binding globulins are able to enter target cells. The apparent loss of this specific cell function in infiltrating carcinomas may result from dedifferentiation and change of cell membrane properties occurring during the process of neoplastic progression.     

Steroid sulfatase, arylsulfatases A and B, galactose-6-sulfatase, and iduronate sulfatase in mammary cells and effects of sulfated and non-sulfated estrogens on sulfatase activity

Sulfatase enzymes have important roles in metabolism of steroid hormones and of glycosaminoglycans (GAGs). The activity of five sulfatase enzymes, including steroid sulfatase (STS; arylsulfatase C), arylsulfatase A (ASA; cerebroside sulfatase), arylsulfatase B (ASB; N-acetylgalactosamine-4-sulfatase), galactose-6-sulfatase (GALNS), and iduronate-2-sulfatase (IDS), was compared in six different mammary cell lines, including the malignant mammary cell lines MCF7, T47D, and HCC1937, the MCF10A cell line which is associated with fibrocystic disease, and in primary epithelial and myoepithelial cell lines established from reduction mammoplasty. The effects of estrogen hormones, including estrone, estradiol, estrone 3-sulfate, and estradiol sulfate on activity of these sulfatases were determined. The malignant cell lines MCF7 and T47D had markedly less activity of STS, ASB, ASA, and GAL6S, but not IDS. The primary myoepithelial cells had highest activity of STS and ASB, and the normal epithelial cells had highest activity of GALNS and ASA. Greater declines in sulfatase activity occurred in response to estrone and estradiol than sulfated estrogens. The study findings demonstrated marked variation in sulfatase activity and in effects of exogenous estrogens on sulfatase activity among the different mammary cell types.     

Tissue culture and estrogen, to clarify the roles of estrone sulfate

Estrone sulfate (E1-S) has been shown to be quantitatively the most important estrogen in peripheral blood. But, the physiological and/or pathological role of E1-S is not yet clarified. At present, we tried to clarify it using tissue cultures. In tissue cultures of human endometrium, secretory endometrium showed higher activity of estrone sulfatase (E1----E1-S) than proliferative endometrium. Progesterone added in the medium induced an increase of estrone sulfotransferase in the proliferative endometrium. The results suggest a reducing effect of estrogen by progesterone in secretory endometrium in physiological conditions. Estrogen dependent malignant tumors (breast cancer, endometrial cancer) have high estrone sulfatase. It converts E1-S to E1 (----E2) which are abundant in these tumors. Ishikawa cell line increased estrone sulfotransferase activity with progesterone, somewhat like the physiological conditions. From out study in vivo, there is a possibility of some ameliorative effects of E1-S on the central nervous system of patients with senile dementia (Alzheimer's type). Effects of E1-S on central nerves were investigated using tissue cultures.     

STEROID SULFATASE IN BRAIN: COMPARISON OF SULFOHYDROLASE ACTIVITIES FOR VARIOUS STEROID SULFATES IN NORMAL AND PATHOLOGICAL BRAINS, INCLUDING THE VARIOUS FORMS OF METACHROMATIC LEUKODYSTROPHY

Abstract– The enzymatic hydrolysis by brain homogenate of the sulfate esters of estrone, pregnenolone, dehydroepiandrosterone, testosterone, cholesterol and p-nitrophenol was studied. With homogenate of young rat brain, the pH optima of estrone sulfatase ⁴ 4 The term steroid sulfatase is used as a general name for the enzyme(s) which hydrolyzes the sulfate ester of a steroid. Simplified terms, such as estrone sulfatase, instead of the more formal terms, such as estrone sulfate sulfohydrolase, have been used throughout.
and arysulfatase C (p-nitrophenyl sulfate as substrate) were 8.2 and all other steroid sulfatases had pH optima at 6.6. Apparent K_ms for these steroid sulfates were widely different. The highest K_m value was 32.2 μm for estrone sulfate and the lowest was 0.66 μm for testosterone sulfate; the K_m for p-nitrophenyl sulfate was 30 fold higher than for estrone sulfate. Specific activity was also highest with estrone sulfatase and lowest with testosterone sulfatase; specific activity with aryl sulfatase C was over 3 fold higher than with estrone sulfatase. Estrone sulfatase activity was inhibited noncompetitively by sulfate esters of dehydroepiandrosterone, pregnenolone, and cholesterol; on the other hand, other steroid sulfatases were inhibited by these latter three sulfates competitively. Developmental changes of these sulfohydrolase activities in rat brain were almost identical with the exception of testosterone sulfatase activity; the latter sulfatase had a peak activity at 30 days old, while all other sulfatase had a peak at 20 days old. Thermal stability of all these activities was identical. Testosterone sulfatase activity in neurological mouse mutants, jimpy, msd, and quaking mice, was less than one half of littermate controls, while other steroid sulfatase levels in these mutants' brain were normal. All sulfatase activities were diminished in the brain of a metachromatic leukodystrophy patient with multiple sulfatase deficiency. The brains of classical metachromatic leukodystrophy patients contained normal levels of all steroid sulfatases and arylsulfatase C, with the single exception of testosterone sulfatase which level was less than 50% of control.     

2-phenylindole sulfamates: inhibitors of steroid sulfatase with antiproliferative activity in MCF-7 breast cancer cells

A number of 2-phenylindole sulfamates with lipophilic side chains in 1- or 5-position of the indole were synthesized and evaluated as steroid sulfatase (estrone sulfatase) inhibitors. Most of the new sulfamates inhibited the enzymatic hydrolysis of estrone sulfate in MDA-MB 231 breast cancer cells with IC₅₀ values between 2 nM and 1 μM. A favorable position for a long side chain is the nitrogen of a carbamoyl group at C-5 of the indole when the phenyl ring carries the sulfamate function. These derivatives inhibit gene activation in estrogen receptor (ER)-positive MCF-7 breast cancer cells in submicromolar concentrations and reduce cell proliferation with IC₅₀ values of ca. 1 μM. All of the potent inhibitors were devoid of estrogenic activity and have the potential for in vivo application as steroid sulfatase inhibitors.     

Recent data on estrogen sulfatases and sulfotransferases activities in human breast cancer

Of the total number of breast cancers approx. 30-50% are hormone-dependent and estradiol is one of the main factors of cancerization. Consequently, the control of this hormone inside the cancer cell is of capital importance because it is well established that the inhibition of estradiol biosynthesis can have a positive effect on the evolution of the disease. The blockage of estradiol can be obtained by the action of anti-aromatases, anti-sulfatases, the control of the 17 beta-hydroxysteroid dehydrogenase activity or by the stimulation of the sulfotransferase which converted the estrogens in their sulfates. In breast cancer tissue estrone sulfate is quantitatively the most important source of estradiol. In the intact cell, estrone sulfatase activity is very intense in the hormone-dependent cell lines (e.g. MCF-7, T-47D) but very small activity is observed in the hormone-independent (e.g. MDA-MB-231, MDA-MB-436) cell lines. However, this activity became very strong after homogenization in the hormone-independent cells, suggesting the presence of repressive factor(s) for this enzyme or its sequestering in an inactive form, in the intact cells of these cell lines. In a series of previous studies it was found that in hormone-dependent cell lines different anti-estrogens: tamoxifen and derivatives, ICI 164,384, very significantly decrease the estradiol concentration originated from estrone sulfate, and recently it was observed that Decapeptyl (D-Trp6-gonadotropin-releasing hormone) in the presence of heparin can also decrease the conversion of estrone sulfate into estradiol. No significant effect was obtained in the presence of heparin or Decapeptyl alone. The estrone sulfatase activity can be inhibited by progesterone, the progestagen R-5020, and testosterone. In another series of recent studies the presence of very strong estrogen sulfotransferase activity has been shown in one breast cancer cell line, the MDA-MB-468. We can conclude that: (1) the control of estradiol concentration can be carried out in the breast cancer tissue itself; (2) estrone sulfate can play an important role in the bioavailability of estradiol in the breast cancer cell; and (3) as is the case for the aromatase, the control of: the estrogen sulfatase, estrogen sulfotransferase, and 17 beta-hydroxysteroid dehydrogenase can be new targets for therapeutic applications in breast cancer.     

Steroid sulfatase activities in normal and cirrhotic livers and plasma levels of estrone sulfate,estrone and estradiol-17β in men

Estrone and dehydroepiandrosterone (DHEA) sulfatases were studied in livers of normal and cirrhotic men. Their Km were 3.2μM and 1.2μM respectively. The musomal sulfatases were solubilized by Miranol H2M and ultrasound. After gel filtration, the soluble material gave a single peak of activity for both substrates with a molecular weight of approximately 330,000. In terms of pmol of product.min^?1 per mg of fresh tissue, the mean (±SD) values of estrone and DHEA sulfatase activities were lower in cirrhotic livers [(n=7) (4.09±2.90 and 0.38±0.20)] than in normal livers [(n=13)(8.29±4.00 and 0.69±0.20)]. The differences were statistically significant: p<0.03 for estrone sulfatase and p<0.01 for DHEA sulfatase. In cirrhotic men, the mean level of plasma estrone is increased whereas that of estrone sulfate is decreased. The variations may be related to the decrease of serum albumin in cirrhotic subjects.     

Changed expression of 9-O-acetyl GD3 (CDw60) in benign and atypical proliferative lesions and carcinomas of the human breast

 Expression of gangliosides is affected in various ways by malignant cell transformation. In the present study, we investigated the expression of CDw60, a constituent of O-acetylated disialogangliosides, in benign and atypical proliferative breast diseases, and preinvasive and invasive carcinomas by immunohistochemistry and thin-layer chromatography (TLC). In normal ducts, antibodies to CDw60 (mAb M-T21) reacted to membranes of the Golgi apparatus in the juxtaluminal cell compartment. A similar polarized distribution of Golgi cisterns in epithelial cells was observed in several benign lesions, i.e., fibroadenomas, intraductal papillomas, and gynecomastia. In contrast, blunt duct adenosis and duct hyperplasia exhibited an abnormal cytosolic and cell surface staining, whereas atypical duct hyperplasia showed randomly dispersed immunoreactive Golgi cisterns, indicating loss of epithelial polarity. In mammary carcinomas and in two breast carcinoma cell lines (MCF-7 and EFM-19) the neoplastic cells contained CDw60-immunolabelled Golgi complexes, which were distributed in a disorderly fashion throughout the cytoplasm, thus reflecting a loss of epithelial polarity. Additionally, only well differentiated ductal carcinomas in situ or invasive ductal carcinomas disclosed a strong cell surface labelling, which was absent in lower differentiated carcinomas of the same types. In all carcinomas, the intensity of CDw60 immunostaining decreased with progressing loss of differentiation (grade of dedifferentiation), as demonstrated by staining intensity in paraffin sections and by evaluation of the relative amounts of extracted 9-O-acetyl GD3 by TLC. Our results indicate that abnormal CDw60 expression is already detectable in benign proliferative breast lesions with different risk rates to develop into malignant lesions. Downregulation of CDw60 expression in poorly differentiated invasive carcinomas may be the consequence of loss of cell functions usually associated with poor prognosis. Received: 19 February 1998     

Detection by immunohistochemistry of c-erbB2 oncoprotein in breast carcinomas and benign mammary lesions.

The presence of the c-erbB2 oncoprotein was demonstrated by immunohistochemistry in a study involving 173 mammary lesions. The lesions included infiltrating cancers, non-invasive neoplasia, as well as atypical and benign lesions. Our aim was to investigate the correlation between the c-erbB2 oncoprotein overexpression and the morphological features of the different mammary tissues analyzed to obtain a better characterization for the growth potential of certain lesions, with emphasis placed on the non-invasive neoplasia and the atypical lesions. Nearly 30% of infiltrating ductal carcinomas (27/89 cases) and 2 out of 24 infiltrating lobular carcinomas were positive. The comedocarcinomas were mostly stained (83%). In contrast, the intraductal carcinomas of cribriform or papillar patterns were consistently negative. No staining was observed in the atypical epithelial hyperplasia located in the vicinity of positive cancers for anti-oncoprotein c-erbB2 antibody. Furthermore, the only 5 positive cases for c-erbB2 out of 32 cancer-free cases were three fibroadenomas and two fibrocystic diseases with atypical ductal hyperplasia. A close correlation was thus observed between c-erbB2 oncoprotein overexpression and cancerous cell morphology, characterized by a marked nuclear hypertrophy often associated with cellular pleomorphism. However, predictive abnormalities of malignant transformation in non-neoplastic epithelial proliferation was difficult to identify, considering only the c-erbB2 expression. A group of tumors with little nuclear abnormalities were found positive for c-erbB2 immunostaining. These probably corresponded to a particular cellular phenotype. Further studies involving other oncogenes should lead to a better characterization of the different tumor phenotypes and help to clarify breast carcinogenesis.