Pathophysiology of estrogen receptors in mammary tissue by monoclonal antibodies

J. sreroid B&hem. Vol. 27, No. 1-3, pp. 171-176, Printed in Great Britain. All rights reserved

1987

0022-473 l/87 $3.00 + 0.00 Copyright @ 1987 Pergamon Journals Ltd

Proceedings of the VII International Congress on Hormonal Steroids (Madrid, Spain, 19X6)

PATHOPHYSIOLOGY OF ESTROGEN RECEPTORS IN MAMMARY TISSUE BY MONOCLONAL ANTIBODIES G. FABRIS*,E.MARCHETTI,A.MARZOLA,A.BAGNI,P.QUERZOLI andI. NENCI Istituto

di Anatomia

e Istologia

Patologica,

Universita

di Ferrara,

Ferrara,

Italy

Summary--Identification of preneoplastic lesions of the breast has mainly rested on morphological grounds, supported by epidemiological data. These studies assign a definite precancerous potential to a group of atypical hyperplastic lesions and in situ carcinoma. In spite of much effort no criteria are yet available to understand which, among these lesions, is committed to infiltrative growth, in other words, to understand the risk to a single patient. Estrogens are know to play a critical role in the etiology of breast cancer. The hypothesis is investigated that this role is dependent on a modified expression of their receptor. To approach this question estrogen receptor expression was traced by specific monoclonal anti-receptor antibodies and immunocytochemistry, on a spectrum of breast tissue changes, from normal tissue to infiltrating cancer. Estrogen receptor expression is heterogeneous in normal tissue and in infiltrating cancer, and on the contrary is homogeneous in proliferative atypical lesions and in in situ carcinomas. Present results show that receptor expression is enhanced and becomes homogeneous, maybe constitutive, in atypical hyperplasia and in in situ carcinoma and that this phenomenon could subserve important changes of proliferative capacity which are necessary and possibly sufficient for autonomous growth.

INTRODUCTION

A major task for the pathologist who studies neoplastic diseases and in particular breast cancer is to identify the epithelial proliferations associated with an increased risk of developing a carcinoma; in other words to assess which proliferative lesions among others can be considered precancerous on the basis of their biological characteristics and clinical behaviour. However, though the search for precancerous lesions of the breast dates back to the beginning of this century[l, 21, only recently was it possible to identify with reliability a subgroup of proliferative lesions with an increased risk of cancer. These studies were delayed by the difficulty in understanding the biological potential of several hyperplastic lesions, otherwise quite distinct on morphological grounds, which constitute a spectrum from normal tissue to carcinoma. Actually, assessment of dysplastic features of breast tissue is not as readily achievable as for other tissues; this can be mainly due to the variety of proliferative lesions and to particular functional and anatomical characteristics of the breast tissue. However, despite a significant progress in understanding breast cancer, among all proliferative and hyperplastic lesions other than in situ lobular and ductal carcinomas, comprising fibrocystic changes, a significant risk of subsequent breast cancer equal to 5-fold, could be assigned with reliability only to atypical hyperplasia [3]. The identification of precancerous lesions on morphological grounds is

*To whom all correspondence

should

be addressed. 171

supported by epidemiological studies. In fact, the incidence of fibrocystic disease is low among populations with a low incidence of breast cancer [4]. On the other hand, there is a 2-4-fold increased frequency of atypical hyperplastic lesions among benign breast biopsies from populations at high risk than from those at low risk for breast cancer [5]. Other studies have shown a clinically significant increase of subsequent breast cancer risk for atypical hyperplastic lesions of the breast [6]; the risk increasing according to the degree of atypia [6,7]. However, the incidence of atypical hyperplastic lesions among unselected autopsy cases is extremely low; this would support a very low frequency of atypical hyperplastic lesions in a general population as well [LX]. Moreover, other factors other than purely histological ones, such as age of occurrence of epithelial atypia or family history of breast carcinoma, strongly affect the level of risk [3,6,9]. Studies on experimental mammary tumorigenesis in the rat and mouse have met similar difficulties. In fact, some studies show that some hyperplastic lesions, in particular hyperplastic alveolar nodules, are definitely precancerous lesions [ 101, while others suggest that at least in the rat, the tumorigenie process is not a multi-step process, at least at the histological level, and that neoplastic transformation does not occur obligatorily after defined precancerous lesions identifiable on morphological grounds [ 111. Two conclusions can be drawn from these observations: (1) The different degrees of atypical hyperplasia identify groups of women with different levels of risk

112

G. FABRISet al

of subsequent carcinoma, but lack any prognostic significance for the single patient. (2) Histological studies have identified precancerous lesions. but by no means do they tell us the true biological potential of each one. in other words which are definitely committed to invasive growth. The multiple endocrine control of the functional activity of the mammary gland has been taken into consideration increasingly also in the studies of mammary tumorigenesis. Much attention has been devoted to both the constituents of hormonal regulation: hormonal levels and metabolism, and cellular components through which hormones exert their action (the receptors). Many epidemiological studies support the fundamental role of hormones, and in particular of estrogens, in the development of breast cancer [ 121. Though specific hormonal aberrations have been shown to increase the risk of this disease, it is also realized that their etiologic importance is the exception rather than the rule in the occurrence of breast cancer and suggest other mechanisms by which mammary hormones tumoritake part in genesis [ 13. 141. For this reason, much attention has been paid also to the other constituent of hormonal regulations, that is, the receptor protein, which is responsible for the phenotypic expression of the hormone action [IS, 161. The hypothesis is currently investigated that the role of estrogens in the development of breast cancer of their depends on a modified expression receptors [ 171. Several observations seem to substantiate this hypothesis. Neoplastic cells often shown abnormal responsiveness to growth factors [ 18, 191; estrogen receptor levels are significantly higher in breast cancer than in normal breast tissue. One may suggest that enhanced estrogen receptor expression provides cells with proliferative advantage, with enhanced responsiveness to growth-promoting stimuli of estrogen despite a normal endocrine milieu. The question arises whether enhanced receptor expression occurs early in neoplastic transformation or is inherent in later events of carcinogenesis and tumor progression. Several methodological difficulties are met when facing this question. In fact, available biochemical evaluations of estrogen receptor on tissue homogenates are highly reliable when a quantitative assay is needed, but are not the assay of choice when estrogen receptor must be evaluated on small pieces of tissue, as on early known morphological expression of neoplastic transformation and, moreover, on currently identified precancerous lesions. These lesions, which include, as discussed above, atypical hyperplasias, can only be checked under the microscope, are asymptomatic and do not give rise to palpable or clinically identifiable breast lumps. These shortcomings are readily overcome by the immunocytochemical approach: this assay traces

the receptor within a morphologically indentifiable structure, thus allowing the study of estrogen receptor expression in proliferative lesions identifiable only under the microscope. This paper deals with the immunocytochemical tracing of estrogen receptor by means of specific monoclonal anti-estrogen receptor antibodies on breast tissue with special attention to precancerous lesions, in the attempt of identifying a modified receptor expression as a clue to neoplastic transformation. EXPERIMENTAL

The study was performed on over 150 cases of breast cancer and over 150 specimens of benign breast disease, including the whole variety of nonproliferative and proliferative lesions of fibrocystic changes and non-invasive, in situ carcinomas. Tissue samples were obtained either from biopsies or from surgical mastectomies sent to this department for histological diagnosis. The biochemical assay of estrogen receptors has been performed on the majority of cancer specimens. The following monoclonal antibodies against estrogen receptor have been used: JS34/32 [20, 2 I], ER-ICA [22,23], and DS [24.25]. Fresh tissue specimens were processed according to the requirements of each antibody. Therefore, they were either frozen with N, or CO2 to obtain cryostatic sections or were fixed in formaldehyde. F, or Carnoy. Fixed specimens were processed and paraffin-embedded according to standard procedures. Histological assessment was performed on tissue sections stained with H.E. The display systems were either the avidin-biotin complex or the P.A.P. method, depending on the antibody used. The majority of specimens has been studied by the JS34/32 and ER-ICA. Only on IO cases of cancer and proliferative lesions was the DS antibody used. The histochemical assay of receptors was compared with the cytosol receptor concentration of each tumor specimen determined by DCC. A scoring system was chosen which took into account only the proportion of stained cells, as reported extensively elsewhere. RESULTS

Specific staining showed a different subcellular localization according to the antibody used (Fig. I). While ER-ICA stained only nuclei, DS gave a selective cytoplasmic positivity. With JS34/32 antibody the most prevalent pattern of staining was only nuclear, both in normal and in neoplastic cells. However, neoplastic ceils often showed also a cytoplasmic staining, either isolated or associated with the nuclear one. The degree of correlation between the bio-

Estrogen receptors in breast pathology

Fig. I, Staining of estrogen receptors in proljferaeive lesions of the breast with different monclonai antibodies. (af Infiltrating ductal carcinoma. The positivity is mainly nuclear with an evident hetecogeneity. JS34/32 antibody, ABC complex. (b) Normal lobule with mild hyperplasia. Evident heterogeneous nuclear positivity. ER-ICA antibody, PAP complex. (c. d) Intraductal in situ carcinoma(c) and severe atypical hyperplasia clinging carcinoma-like fd). Enhanced and homogeneous nuclear staining of entire cell population. ER-ICA antibody, PAP complex. (e) Lobular in situ carcinoma. Homogeneous and deep cytoplasmic staining of the entire cell population. IX antibody, PAP complex. (f) Lobular in siru carcinoma in the left; on the right a lobule with a mild hyperplasia. Note the difference between homogeneous and deep cytoplasmic staining of lobular carcinoma and the heterogeneous and fainter cytoplasmic staining of lobular hyperplasia. D5 antibody, PAP complex.

173

G. FABRIS

174

chemical and immunohistochemical assay methods of receptors on cancer specimens was highly significant for JS34/32[26] and for ER-ICA (unpublished results). The small amount of cancer specimens tested with D5 antibody does not allow a comparison between the amount of stained cells and the estrogen receptor assay by DCC. However, a good correlation is reported by others [25]. Despite these variations of the subcellular localization of the staining, the overall staining of all the specimens tested showed good agreement with the different antibodies. In fact, a highly heterogeneous staining pattern was a constant feature of all the cancer specimens studied, which could be appreciated with all the antibodies. This heterogeneity concerned not only the presence or absence of specific staining but also the degree of staining intensity. Moreover, this variation of specific staining was present not only within a single tumor but also among different tumors. This feature was especially prominent at the growing edge of each tumor. Some degree of staining heterogeneity was always present also among non-neoplastic cells of normal lobules and ducts, with a variable number of negative cells also seen among normal cells (Fig. I b, f). Normal breast structures displayed in general a high proportion of negative cells, with positive cells showing a degree of staining intensity lower than neoplastic cells. In normal lobules specific staining involved only epithelial cells, no staining was ever observed in myoepithelial cells and non-epithelial cells, except for some fibrocytes of the specialized lobular stroma. This highly heterogeneous staining pattern was also a feature of the majority of proliferative lesions of fibrocystic disease, with a variable proportion of stained and unstained cells. This was true for the various kinds of adenosis, mild ductal and lobular hyperplasias and hyperplastic epithelia within radial scars: all these lesions were constituted of positive and negative cells, positive cells always showing a degree of staining intensity lower than neoplastic cells. Also. the hyperplastic lesions usually responsible for clinically appreciable lumps, like fibroadenomas. papillomas and florid hyperplasias showed a heterogeneous pattern of receptor expression. On the contrary, staining of receptor appeared to differ with respect to various features in other hyperplastic lesions. Staining intensity looked more like that of neoplastic cells than that of normal cells. Above all, staining was not hererogeneous, but homogeneous, involving the whole cell population (Fig. lc, d). This pattern was appreciable more often in proliferative lesions from specimens harbouring an invasive tumor; however, it could be found also in biopsies

from

benign

breast

disease.

Such a homogeneous receptor expression was a constant finding of atypical hyperplasia and of non-

etal. invasive, common

in situ carcinoma, whereas it was an unof less severe hyperplastic lesions.

feature

DISCUSSION

All the antibodies used, though displaying a different staining at the subcellular level, gave almost overlapping results among the different breast specimens. Several reasons can account for the different subcellular localization of staining obtained with the various antibodies. Cytoplasmic staining of DS is in accord with the reported pattern of staining of this antibody, which is known to recognize a receptor-related antigen with a cytoplasmic localization, with physico-chemical properties different from the estrogen receptor. JS34/32 antibody and ER-ICA specifically react with estrogen receptor [20,23]; yet, although in normal tissue estrogen receptors are traced at the nuclear level by both the antibodies, in neoplastic cells a cytoplasmic localization is often detected by JS34/32 antibody. With respect to ER-ICA, the different reactivity of JS34/32 antibody could be ascribed to a variety of reasons[26], such as different sources of the antibody and of the immunogen, and minor variations in purification procedures of the immunogen. A selective recognition of different antigenic domains is also suggested by the selective reactivity with frozen or fixed tissues of each antibody, respectively. Evidence for specific receptor recognition comes not only from the control experiments of the immunocytochemical procedure, but also from the good correlation between the number of receptorpositive cells traced by JS34/32 or ER-ICA and the amount of estrogen receptor determined by the DCC method, in each tumor [26,27]. A heterogeneous receptor expression can be appreciated both in normal mammary lobules and in tumors with all the antibodies; on the other hand, receptor expression is quite homogeneous and involves the entire cell population in hyperplastic atypical lesions and in in situ carcinomas [ 17,281. The significance and the role played by tumor heterogeneity in the natural history of tumor has been investigated for a long time [29-311. The heterogeneity of architectural and cytological features, of biological, biochemical, endocrinological and immunological properties and of caryotype, are a clue to the capacity of each tumor to give origin to new cellular clones which are responsible of the unique behaviour of the tumor: its continuous adaptability ot environmental pressure and the capacity to progression. Heterogeneity is the phenotypic expression of the clonal selection which takes place in a tumor to acquire and maintain a proliferative advantage [32,33]. However, immunocytochemical staining of receptors shows that the heterogeneity of

Estrogen

receptors

receptor expression is a constant feature of the invasive tumor, while, as a rule, receptor expression is homogeneous in in situ carcinoma and in atypical hyperplasias. Not only estrogen receptor but also other features are expressed with a homogeneous pattern in proliferative and pre-invasive lesions. The monotonous cellular appearance of the in situ carcinoma both of the breast and of other organs have long been recognized and represent a diagnostic clue for the pathologist [34-361. Other features related to the functional state of the cell are expressed with a homogeneous pattern in proliferative lesions of the breast, as for instance in the M, 52,000 protein [37]. These observations show strong similarities between mammary carcinogenesis and some steps of experimental liver carcinogenesis [38,39]. In this model, in fact, one of the earliest identifiable lesions is the formation of hepatocyte nodules which are characterized by homogeneous architectural, cytological and biochemical features with the appearance of a new population of hepatocytes with a distinctive, possibly constitutive, phenotype. Present results show that receptor expression is enhanced and becomes homogeneous, maybe constitutive, in atypical hyperplasia and in in situ carcinoma and that this phenomenon could subserve important changes of proliferative capacity which are necessary and possibly sufficient for autonomous growth. In this respect it is interesting to note that the constitutive activation of a gene is not an uncommon phenomenon of the natural history of breast cancer: the constitutive activation of genes for casein and for lactalbumin has in fact been demonstrated in some experimental models [40,4 11. This homogeneous expression of the receptor differs not only from the heterogeneity of invasive carcinoma, but also from receptor heterogeneity which, as a rule, is present also in normal mammary tissue. This points again to a similarity with liver carcinogenesis: in this model, many properties are expressed in a homogeneous way in the hepatocyte nodules, at variance with a heterogeneous expression in normal liver tissue [42]. As mentioned above, epidemiological observations and morphological evidence point to atypical proliferative lesions and in situ carcinoma as the true precursors to invasive growth, meanwhile not disregarding the possibility of regression. This means that if the constitutive activation of the gene for estrogen receptor results from the initiation steps of mammary carcinogenesis, then some mechanisms must act in the cell to control this phenomenon and therefore to prevent neoplastic progression. Again, some analogies can be drawn from liver carcinogenesis: hepatocyte nodules, which are constituted by a homogeneous population of initiated hepatocytes, show a high rate of regression through

in breast

pathology

175

remodeling which is due to a redifferentiation phenomenon [39]. In this respect, it is interesting to remember that within proliferative lesions of the mammary gland a distinction has been made by the pathologist, who recognizes two groups-organoid and non organoid lesions--only the latter being endowed with neoplastic potential [34]. It seems therefore an attractive hypothesis that the enhanced receptor expression brings about focal hyperplasia. The restored regulation of receptor expression should induce a process of remodeling or redifferentiation with the appearance of organoid lesions; on the contrary, if receptor expression becomes constitutive, then the focal hyperplasia could evolve into a non-organoid lesion, thus acquiring all the features of atypical epithelial hyperplasia and of in situ carcinoma. Present results underscore, in accord with clinical and experimental evidence, the striking difference between in situ and invasive carcinomas, as highlighted by homogeneity and heterogeneity, respectively. Progression from in situ to invasive growth should result from a substantial charge of the genome, resulting in heterogeneity and in infiltrative capacity. The carcinogenesis process is a multistep process, which is today understood as a sequential activation of cellular oncogenes [38,43]; in this context, it is interesting to recall the homology between the gene for estrogen receptor and v-Erb A oncogene [44] and to suggest that among all the activated, or mutated, genes of mammary carcinogenesis a pivotal role be played by estrogen receptor gene. Acknowledgemenls-This work was supported in part by erant no. 85.02147.44 from Italian National Research Council. A.B. and P.Q. are recipients of a fellowship from Associatione Italiana per la Ricerca sul Cancro. REFERENCES I. Ewing J.: Neopkzstic Discuses. W. B. Saunders. Philadelphia, 1st edn (1919). 2. Lagios M.D.: Human breast precancer: current status. Cancer Suru. 2 (1983) 382-402. 3. DuPont W. D. and Page D. L.: Risk factors for breast cancer in women with proliferative breast disease. New Engl. J. Med. 312 (1985) 146-151. 4. Bartow S. A.. Black W. C., Waeckerlin R. W. and Mettler F. A.: Fibrocystic disease: a continuing enigma. Path. Ann. 18 (1983) 93-l 11. 5. Schierch C. III, Rosen P. P., Horota T., Itabashi M., Yamamoto H., Kinne D. W. and Beattie E. J.. Jr: A pathologic study of benign breast diseases in Tokyo and New York. Cancer 50 (1982) 1899-1903. 6. Kodlin D., Winger E. E., Morgenstern N. L. and Chen U.: Chronic m&topathy and breast cancer. A followuo studv. Cancer 39 (1977) 2603-2607. 7. Pgge D.-L., Van der Zwaag k., Rogers L. W., Williams L. T., Walker W. E. and Hartmann W. H.: Relation between component parts of fibrocystic disease complex and breast cancer. J. natn. Cancer Inst. 61(1978) 1055-1063.

G. FABRIS et al.

176

8. Alpers C. E. and Wiellings 9.

10.

11.

12.

13. 14.

15. 16. 17.

18.

19.

20.

21

22.

23.

24.

25.

26.

S. R.: The prevalence of carcinoma in situ in normal and cancer-associated breasts. Hum. Path. 16 (1985) 796-807. Petrakis N. L., Ernster V. L., King E. B. and Sacks S. T.: Epithelial dysplasia in nipple aspirates of breast fluid: association with family history and other breast cancer risk factors. J. natn. Cancer Inst. 68 (1982) 9-13. Dao T. L., Sinha D., Christakos S. and Varela R.: Biochemical characterization of carcinogen-induced mammary hyperplastic alveolar nodule and tumor in the rat. Cancer Res. 35 (1975) 1128-1134. Sinha D. and Dao T. L.: A direct mechanism of mammary carcinogenesis induced by 7,12-dimethylbenz(a)anthracene. J. natn. Cancer Inst. 53 (1974) 841-846. Thomas D. B.: Hormones and hormone receptors in the etiology of breast cancer. Breast Cancer Res. Treat. 7 Suppl. (1986) 1 l-22. Brennan M.-j.: Endocrinology in cancer of the breast. Am. J. clin. Path. 64 (1975) 797-809. Moore D. H., Moore D.~H. B’and Moore C. T.: Breast carcinoma ethiological factors. Ado. Cancer Res. 40 (1983) 189-253. Cikes M.: Expression of hormone receptors in cancer cells: a hypothesis. Eur. J. Cancer 14 (1978) 211-215. Israel L. and Band P.: Hypothesis. Hormones as cancer growth factors. Lancet II (1984) 843-844. Nenci I., Marchetti E., Querzoli P. and Bagni A.: Toward understanding human mammary preneoplasia. The estrogen receptor-promotion hypothesis. (in press). Sporn M. B. and Roberts A. B.: Autocrine growth factors and Cancer. Nature, Land. 313 (1985) 745747. Goustin A. S., Leof E. B., Shipley G. D. and Moses H. L.: Perspectives in cancer research. Growth factors and cancer. Cancer Res. 46 (1986) 1015-1029. Moncharmont B. and Parikh I.: Monoclonal antibodies against estrogen receptors: interaction with different molecular forms and functions of the receptor. Biochemistry 21 (1982) 6916-6921. Moncharmont B., Anderson W. L.. Rosenberg B. and Parikh I.: Interaction of estrogen receptor of calf uterus with a monoclonal antibody: probing of various molecular forms. Biochemistry 23 (1984) 3907-3912. Jensen E. V., Greene G. L., Closs L. E., DeSombre E. R. and Nadji M.: Receptors reconsidered: a 20-year perspective. Rec. Prog. Horm. Res. 38 (1982) l-40. Greene G. L., Sobel N. B., King W. J. and Jensen E. V.: Immunochemical studies of estrogen receptors. J. steroid Biochem. 20 (1984) 51-56. Coffer A. I., Spiller G. H., Lewis K. M. and King R. J. B.: Immunoradiometric studies with monoclonal antibody against a component related to human estrogen receptor. C’ancer Res. 45 (1985) 3694-36YX. King R. J. B., Coffer A. I., Gilbert J., Lewis K., Nash R., Millis R., Raju S. and Taylor W. T.: Histochemical studies with a monoclonal antibody raised against a partially purified soluble estradiol receptor preparation from human myometrium. Cancer Res. 45 (1985) 5728-5733. Marchetti E., Querzoli P., Moncharmomt B., Parikh I., Bagni A., Marzola A., Fabris G. and Nenci I.: Immunocytochemical demonstration of estrogen receptors by monoclonal antibodies in human breast cancer:

27.

28.

29. 30. 31.

32.

33.

34. 35.

36.

37.

38. 39.

40.

41.

42.

43. 44.

correlation with estrogen receptor assay by dextrancoated charcoal method. Cancer Res. 47 (1987) 25082513. King W. J., DeSombre E. R., Jensen E. V. and Greene G. L.: Comparison of immunocytochemical and steroid-binding assays for estrogen receptor in human breast tumors. Cancer Res. 45 (1985) 2Y3-304. Nenci I.: I. Biological and clinical rationale for a combined approach to breast cancer. Expression and modulation-of estrogen receptors in human breast cancer. .I. steroid Biochem. 23 (1985) 1093-1096. Heppner G.: The challenge of tumor heterogeneity. Commun. Res. Breast Dis. 1 (1979) 177-192. Heppner G. H.: Tumor heterogeneity. Cancer Res. 44 (1984) 2259-2265. Woodruff M. F. A. and McGuire W. L.: Steroid hormone receptors in human breast cancer. Br. J. Cancer 47 (1983) 589-594. Osborne C. K.: Heterogeneity in hormone receptor status in primary and metastatic breast cancer. Sem. Oncol. 12 (1985) 317-326. Epstein F. H.: Clinical implication of tumor cell heterogeneity. New Engl. J. Med. 314 (1986) 14231431. Azzopardi G. J.: Problems in Breast Pathology. W. B. Saunders, London (1979). Jensen H. M.: Breast pathology, emphasizing precancerous and cancer associated lesions. Commun. Res. Breast Dis. 2 (1981) 41-86. Squartini F. and Sarnelli R.: Structure, functional changes and proliferative pathology of the human mammary lobule in cancerous breasts. J. natn. Cancer Inst. 67 (1981) 33-46. Garcia M., Salazar-Retana G., Pages A., Richer G., Domergue J., Pages A. M., Cavalie G., Martin J. M., Lamarque J.-L., Pau B., Pujol H. and Rochefort H.: Distribution of the M, 52,000 estrogen-regulated protein in benign breast diseases and other tissues by immunohistochemistry. Cancer Res. 46 (1986) 37343738. Farber E.: The multistep nature of cancer development. Cancer Res. 44 (1984) 4217-4223. Farber E.: Cellular biochemistry of the stepwise development of cancer with chemicals: G. H. A. Clowes Memorial Lecture. Cancer Res. 44 (1984) , 5463-5474. Smith G. H., Vonderhaar B. K., Graham D. E. and Medina D.: Expression of pregnancy-specific genes in preneoplastic mouse mammary tissues from virgin mice. Cancer Res. 44 (1984) 3426-3437. Banerjee M. R., Ganguly R., Mehta N. M. and Ganguly N.: Hormonal regulation of casein gene expression in normal and neoplastic cells in murine mammary glands. In Hormonal Regulation of Mammary Tumors (Edited by B. S. Leung). Eden Press, (1982) pp. 229-283. Rappaport A. M.: Physioanatomical basis of toxic liver injury. In Toxic Injury of the Liver (Edited by E. Farber and M. M. Fisher), Marcel Dekker. New York. Part A (1979) pp. l-57. Duesberg P. H.: Activated proto-one genes: sufficient or necessary for cancer? Science 228 (1985) 669-677. Green S., Walter P., Kumar V., Krust A., Bornert J.-M.. Argos P. and Chambon P.: Human oestrogen receptor cDNA: sequence, expression and homology to v- erb -A. Nature, Land. 320 (1986) 134-139.