Exp Toxic Pathol 1992; 44: 235-238 Gustav Fischer Verlag Jena
Department of Pathologyl), The Kohno Clinical Medicine Research Institute Tokyo Department of 2nd Pathology2), Nara Medical University, Nara Department of Pathology 3), Kitasato University, School of Hygienic Sciences, Kanagawa
Immunohistochemical expression of monoclonal antibodies (epi-l and myo-l) derived from human breast cancer against rat mammary tumors
With 6 figures Received: December 20, 1990; Accepted: June 25, 1991 Address for correspondence: Dr. H. YAMAMOTO, Department of Pathology, The Kohno Clinical Medicine Research Institute, 34-4 Kitashinagawa, Shinagawa-Ku, Tokyo 140, Japan. Key words: Mammary tumors; Breast cancer; Antibodies, monoclonal; Monoclonal antibodies; Immunohistochemistry; DMBA.
Summary Immunhistochemical~· expression of monoclonal antibodies epi-l and myo-l derived from human breast cancer cell line (HBC-4W) was examined for DMBA-induced rat mammary tumors. Antibody epi-l reacted with luminal epithelial cells while antibody myo-l reacted with myoepithelial cells of the mammary glands in rats, respectively. The reactions with both antibodies were markedly visible, in particular, in the normal marnmarygland, tumor-like lesions and bel).ign epithelial mammary tumors in rats, which showed clear two-ceIl-type structures. Among malignant mammary tumors, adenocarcinoma was strongly positive with antibodies epi-l and myo-l. However, squamous cell carcinoma and adenoacanthoma mainly reacted with antibody epi-l. On the other hand, the intercellular matrices of pleomorphic cell sarcoma and stromal areas of the normal mammary gland or epithelial tumors were positive with antibody myo-l.
Introduction Studies of rat mammary tumors were reported by several workers, but the origin and biological characteristics of these tumors have not been elucidated in detail (LEUNG 1978; TSUKIDATE et al. 1988). Accordingly, specific monoclonal antibodies to rat mammary tumors are necessary for histogenetic clarification. Both monoclonal antibodies of epi-l and myo-l react specifically to tissues of normal mammary gland and diseased mammary glands including various breast cancers in humans (SEKIYA 1986). It is the specificity of these both antibodies that epi-l reacts with luminal epithelial cells, while myo-l reacts with myoepithelial cells (SEKIYA 1986; IMAI et al. 1986). 2*
This time, we applied antibodies epi-l and myo-l to the normal mammary gland, tumor-like lesions and mammary tumors in rats, and each distribution pattern of these two monoclonal antibodies was examined.
Materials and methods We used 7, l2-dimethylbenz[a]-anthracene (DMBA)-induced mammary tumors in female Sprague-Dawley (SD) rats. The following monoclonal antibodies were used for immunohistochemical staining: antibody epi-l, a mouse monoclonal antibody which specifically reacts with luminal epithelial cells in the human mammary gland, and antibody myo-l, a mouse monoclonal antibody specifically reactive with myoepithelial cells in the human mammary gland. These antibodies were kindly provided by Dr. IMAL Immunologicalization was performed using the VECTOR staining kit (Vector Laboratories Inc.; USA) of the avidin-biotin peroxidase complex (ABC) method producted by Hsu et al. (1981). Briefly, sections were treated with 0.3 % hydrogen peroxide to block endogenous peroxidase. After exposure to nonimmune horse serum, the sections were incubated with the first antibodies (anti-human epi-l and myo-l) for 30 minutes at 37 DC. Sections were subsequently washed with phosphatebuffered saline, incubated with anti-mouse IgO for 30 minutes at 37 DC, peroxidase conjugate and finally with the peroxidase substrate diamine benzidine tetrahydrochloride (DBA, 1.5 mg; H 2 0 2 , 10 I-tl; 0.1 M phosphate buffered saline, 15 ml). Sections were counterstained with MYER's hematoxylin, dehydrated, and amounted in HSR mount. Antibodies epi-l and myo-l were used as a 1: 1 000 dilution of phosphate buffered saline. Positive control stainings were performed using human mammary carcinoma. Exp Toxic Pathol 44 (1992) 5
235
Fig. 1. Immunostaining with antibody epi- l in fibroadenoma. The luminal epithelial cells are strongly positive. X 100. Fig. 2. Immunostaining with antibody myo-l in fibroadenoma . The myoid cells are moderately positive. x 250. Fig. 3. Immunostaining with antibody epi-l in adenocarcinoma. The luminal epithelial cells are strongly positive. x 400. Fig. 4. Immunostaining with antibody myo-l in adenocarcinoma. The myoepithelial cells layers are strongly positive. x 350.
236
Exp Toxic Pathol 44 (1992) 5
Fig. 5. Immunostaining with antibody epi-l in squamous cell carcinoma. The squamous epithelial cells are moderately positive. X250.
Fig. 6. Immunostaining with antibody myo-l in pleomorphic cell sarcoma. The intercellular matrices are moderately positive.
x 400.
Results In the nonnal mammary gland, antibody epi-I showed stronger reactivity than antibody myo-I. Antibody epi-I reacted specifically with the cytoplasm in luminal epithelial cells of mammary glands such as ducts, terminal ducts and lobules, while reaction of antibody myo-I was seen in spindle cells of basal layers of ducts and parts of the stroma. In the tumor-like lesions, ductular hyperplasia and adenosis were positive with antibodies epi-l and myo-l, while fibosis was stained with antibody myo-I, but not with antibody epi-l. In the cases of benign tumors, the reactivity of antibodies epi-l and myo-l was observed in adenoma and fibroadenoma. Among fibroadenoma, the tubular type was positive with luminal epithelial cells, myoid cells in basal layers and interstitial cells by the antibodies epi-l (fig. 1) and myo-l (fig. 2). In the sclerosing type, antibody epi-I was strongly positive in the luminal epithelial cells, but antibody myo-l showed weak reactivity in the stromal areas . On the other hand, fibroma was clearly positive with antibody myo-l, but not with antibody epi-l. In the malignant tumors, adenocarcinoma and adenoacanthoma were positive.· with antibodies epi-l and myo-l, though squamous cell carcinoma was positive with only antibody epi-l. Among adenocarcinoma; both of papillary and medullary types were strongly reactive for antibodies
epi-l and myo-l. With the antibody epi-I , a strong positive reaction was found diffusely in the luminal epithelial cells (fig. 3). On the other hand, antibody myo-l was unifonny reactive with myoepithelial cell layers (fig. 4). Epithelial tumor cells in the squamous cell carcinoma showed moderately positive reaction with antibody epi-l (fig. 5), but not with antibody myo-l. In the adenoacanthoma, the presence of antibody epi-l was moderately observed in the squamous metaplastic cells, while reactivity with antibody myo-l was seen in basal layers of squamous metaplastic epithelial and luminal epithelial areas. In contrast to the pleomorphic cell sarcoma, reactivity with antibody myo-l was observed in the intercellular matrices (fig. 6), but not with antibody epi-l.
Discussion This time, antibodies epi-l and myo-l of the human breast cancer origin were applied to DMBA-induced rat mammary tumors . In general, in the nonnal mammary gland, the 'epithelial cells of mammary ducts are showing two-cell-Iayer structure of luminal epithelial and myoepithelial cells (AZZOPARDI 1979). In our study, the number of cell layers confirmed the two-cell-type distribution of luminal epithelial and myoepithelial cells in the nonnal mammary gland or tumor-like lesions in rats. Especially, antibody epi-l reacted Exp Toxic Pathol 44 (1992) 5
237
with luminal epithelial cells of all mammary ducts, and antibody myo-l reacted with myoepithelial cells situated on the basal layer. On the other hand, reactivity of antibodies epi-l and myo1 to benign tumors showed two-cell-Iayer structure similar to the normal mammary gland. In other words, these fmdings are showing that there is not an essential difference in the constitution of luminal epithelial and myoepithelial cells among the normal mammary gland and benign mammary tumors. In the present study, the distribution of myoid cells was obvious by reaction with antibody myo-l. Most epithelial magliant mammary tumors were thought to be of luminal epithelial cell origin (AZZOPARDI 1979). On the other hand, several workers suggested that there is a possibility that certain mammary tumors may be of myoepithelial cell origin (HAGUENAU 1959; MURAD and SCRAPELLI 1967; HAYASHI and AOKI 1984). These observations suggest that there are great diversities of opinion in regard to the cell origin of the mammary tumors. In most DMBA-induced rat malignant mammary tumors examined in the present study, positive reactions with antibodies epi-l and myo-l were observed. DULBECCO (1982) described that most chemically-induced rat mammary tumors originated from stem cells of the mammary gland. They indicated that rat mammary tumors are differentiated into luminal epithelial or myoepithelial cells. Our results correspond to these findings. In the squamous cell carcinoma and adenoacanthoma, the presence of antibody epi-l was observed in the squamous metaplastic cells which are thought" to originate from the glandular epithelial cells. On the other hand, the present results showed that the intercellular matrices of pleomorphic cell sarcoma and stromal areas of the normal mammary gland or epithelial tumors were positive with myo-l, but negative with epi-l. It is possible that the sarcoma cells originated from transformed myogenic cells, but the reason for the reactivity with antibody myo-l is not clear. Additional studies will, therefore, be needed to establish the specificity of antibody myo-l. Myoepithelial cells have been reported to occur in the normal mammary gland (RANDOR 1972). LIOTTA et al. (1979) described that the basement membrane in the rat mammary gland is probably synthesized by myoepithelial cells. On the other hand, ARCHER et al. (1968) reported that the detailed ultrastructure of DMBA-induced rat mammary tumors indicates that they are composed of epithelial cells with a myoepithelial component. In the present study, reactivity with antibody myo-l was observed in normal mammary gland and tumor-like lesions including benign mammary tumors in rats. There findings might imply characteristics of myoepithelial cells. As mentioned above, our study showed that antibodies epi-l and myo-l established from the human breast cancer cell line reacted to normal mammary gland including various mammary tumors in rats. Therefore, we suppose that the monoclonal antibodies epi-l and myo-l will prove to be
238
Exp Toxic Patho144 (1992) 5
useful tools for the analysis of the interactions between luminal and myoepithelial cells in rat mammary tumors.
References ARCHER FL, ORLANDO RA: Morphology, natural history, and enzyme patterns in mammary tumors of the rat induced by 7,12-dimethylbenz(a)anthracene. Cancer Res 1986; 28, 217-224. AZZOPARDI JG: Problems in Breast Pathology. In: BENNINTON JL (Editor): WB Saunders Company Ltd. London, Philadelphia, Toronto, BIGGS R (1947). - The myoepithelium in certain tumors of the brest. J Pathol Bact 1979; 437. DULBECCO R : Immunological markers in the study of development and oncogenesis in the rat mammary gland. J Cell Physiol 1982; 2: 19-22. HAGUENAU F: Le cancer du sein chez la femme, electronique et au microscope optique. Bulletin de l' Association Fransaise pour l'Etude du Cancer. 1959; 46: 177-211. HA YASH! Y, AOKI Y: Findings of myoepithelial cells in human breast cancer. Acta Pathol Jpn 1984; 34: 537-552. Hsu SM, RAINE L, FANGER H: A comparative study of the peroxidase-antiperoxidase method and an avidin-biotin complex method for studying polypeptide hormone with radioimmunoassay antibodies. J Clin Pathol 1981; 75: 734-737. IMAI S, SEKIYA N, MORIMOTO J, KIYOZUKA Y, et al.: Establishment of monoclonal antibodies reactive with epithelial and myoepithelial cells in the breast. Acta Pathol Jpn 1986; 36: 1319-1326. LEUNG B: Hormonal dependency of experimental breast cancer. In Progress in Cancer Research and Therapy. Vol. 10. Hormones, receptors, and breast cancer. MCGURIE WL, ed., Raven Press, New York. LIOTTA LA, WICHA MS, FOIDART JM: Hormonal requirements for basement membrane collagen deposition by cultured rat mammary epithelium. Lab Invest 1979; 41: 511-518. MURAD TM, SCRAPELLI DG: The ultrastructure of medullary and scirrhous mammary duct carcinoma. Am J Pathol 1967; 50: 335-360. ORMEROD EJ, WARBURTON ML, GUSTERSON B, et al.: Abnormal deposition of basement membrane and connective tissue components in dimethylbenzanthracene-induced rat mammary tumors: An immunocytochemical and ultrastructural study. Histochem J 1985; 17: 1155-1166. RADNOR CJP: Myoepithelial cell differentiation in rat mammary glands. J Anat 1972; 111: 381-398. SEKIY A H: Production and analysis of monoclonal antibodies against human mammary carcinoma. J Nara Med Associ 1986; 37: 238-250. TUBURA A, INUI T, SENZAKI H, et al.: Immunohistochemical reactivity of a monoclonal antibody against cell membranes of human mammary carcinoma cells (l25B4) with rat normal tissues and DMBA-induced mammary tumors. Acta Histochem Cytochem 1989; 22: 561-565. TSUKIDATE K, TOIDA M, SOBUE M, et al.: Immunohistochemical studies of DMBA-induced rat mammary tumors. Acta Pathol Jpn 1988; 38: 129-139.
Department of Pathologyl), The Kohno Clinical Medicine Research Institute Tokyo Department of 2nd Pathology2), Nara Medical University, Nara Department of Pathology 3), Kitasato University, School of Hygienic Sciences, Kanagawa
Immunohistochemical expression of monoclonal antibodies (epi-l and myo-l) derived from human breast cancer against rat mammary tumors
With 6 figures Received: December 20, 1990; Accepted: June 25, 1991 Address for correspondence: Dr. H. YAMAMOTO, Department of Pathology, The Kohno Clinical Medicine Research Institute, 34-4 Kitashinagawa, Shinagawa-Ku, Tokyo 140, Japan. Key words: Mammary tumors; Breast cancer; Antibodies, monoclonal; Monoclonal antibodies; Immunohistochemistry; DMBA.
Summary Immunhistochemical~· expression of monoclonal antibodies epi-l and myo-l derived from human breast cancer cell line (HBC-4W) was examined for DMBA-induced rat mammary tumors. Antibody epi-l reacted with luminal epithelial cells while antibody myo-l reacted with myoepithelial cells of the mammary glands in rats, respectively. The reactions with both antibodies were markedly visible, in particular, in the normal marnmarygland, tumor-like lesions and bel).ign epithelial mammary tumors in rats, which showed clear two-ceIl-type structures. Among malignant mammary tumors, adenocarcinoma was strongly positive with antibodies epi-l and myo-l. However, squamous cell carcinoma and adenoacanthoma mainly reacted with antibody epi-l. On the other hand, the intercellular matrices of pleomorphic cell sarcoma and stromal areas of the normal mammary gland or epithelial tumors were positive with antibody myo-l.
Introduction Studies of rat mammary tumors were reported by several workers, but the origin and biological characteristics of these tumors have not been elucidated in detail (LEUNG 1978; TSUKIDATE et al. 1988). Accordingly, specific monoclonal antibodies to rat mammary tumors are necessary for histogenetic clarification. Both monoclonal antibodies of epi-l and myo-l react specifically to tissues of normal mammary gland and diseased mammary glands including various breast cancers in humans (SEKIYA 1986). It is the specificity of these both antibodies that epi-l reacts with luminal epithelial cells, while myo-l reacts with myoepithelial cells (SEKIYA 1986; IMAI et al. 1986). 2*
This time, we applied antibodies epi-l and myo-l to the normal mammary gland, tumor-like lesions and mammary tumors in rats, and each distribution pattern of these two monoclonal antibodies was examined.
Materials and methods We used 7, l2-dimethylbenz[a]-anthracene (DMBA)-induced mammary tumors in female Sprague-Dawley (SD) rats. The following monoclonal antibodies were used for immunohistochemical staining: antibody epi-l, a mouse monoclonal antibody which specifically reacts with luminal epithelial cells in the human mammary gland, and antibody myo-l, a mouse monoclonal antibody specifically reactive with myoepithelial cells in the human mammary gland. These antibodies were kindly provided by Dr. IMAL Immunologicalization was performed using the VECTOR staining kit (Vector Laboratories Inc.; USA) of the avidin-biotin peroxidase complex (ABC) method producted by Hsu et al. (1981). Briefly, sections were treated with 0.3 % hydrogen peroxide to block endogenous peroxidase. After exposure to nonimmune horse serum, the sections were incubated with the first antibodies (anti-human epi-l and myo-l) for 30 minutes at 37 DC. Sections were subsequently washed with phosphatebuffered saline, incubated with anti-mouse IgO for 30 minutes at 37 DC, peroxidase conjugate and finally with the peroxidase substrate diamine benzidine tetrahydrochloride (DBA, 1.5 mg; H 2 0 2 , 10 I-tl; 0.1 M phosphate buffered saline, 15 ml). Sections were counterstained with MYER's hematoxylin, dehydrated, and amounted in HSR mount. Antibodies epi-l and myo-l were used as a 1: 1 000 dilution of phosphate buffered saline. Positive control stainings were performed using human mammary carcinoma. Exp Toxic Pathol 44 (1992) 5
235
Fig. 1. Immunostaining with antibody epi- l in fibroadenoma. The luminal epithelial cells are strongly positive. X 100. Fig. 2. Immunostaining with antibody myo-l in fibroadenoma . The myoid cells are moderately positive. x 250. Fig. 3. Immunostaining with antibody epi-l in adenocarcinoma. The luminal epithelial cells are strongly positive. x 400. Fig. 4. Immunostaining with antibody myo-l in adenocarcinoma. The myoepithelial cells layers are strongly positive. x 350.
236
Exp Toxic Pathol 44 (1992) 5
Fig. 5. Immunostaining with antibody epi-l in squamous cell carcinoma. The squamous epithelial cells are moderately positive. X250.
Fig. 6. Immunostaining with antibody myo-l in pleomorphic cell sarcoma. The intercellular matrices are moderately positive.
x 400.
Results In the nonnal mammary gland, antibody epi-I showed stronger reactivity than antibody myo-I. Antibody epi-I reacted specifically with the cytoplasm in luminal epithelial cells of mammary glands such as ducts, terminal ducts and lobules, while reaction of antibody myo-I was seen in spindle cells of basal layers of ducts and parts of the stroma. In the tumor-like lesions, ductular hyperplasia and adenosis were positive with antibodies epi-l and myo-l, while fibosis was stained with antibody myo-I, but not with antibody epi-l. In the cases of benign tumors, the reactivity of antibodies epi-l and myo-l was observed in adenoma and fibroadenoma. Among fibroadenoma, the tubular type was positive with luminal epithelial cells, myoid cells in basal layers and interstitial cells by the antibodies epi-l (fig. 1) and myo-l (fig. 2). In the sclerosing type, antibody epi-I was strongly positive in the luminal epithelial cells, but antibody myo-l showed weak reactivity in the stromal areas . On the other hand, fibroma was clearly positive with antibody myo-l, but not with antibody epi-l. In the malignant tumors, adenocarcinoma and adenoacanthoma were positive.· with antibodies epi-l and myo-l, though squamous cell carcinoma was positive with only antibody epi-l. Among adenocarcinoma; both of papillary and medullary types were strongly reactive for antibodies
epi-l and myo-l. With the antibody epi-I , a strong positive reaction was found diffusely in the luminal epithelial cells (fig. 3). On the other hand, antibody myo-l was unifonny reactive with myoepithelial cell layers (fig. 4). Epithelial tumor cells in the squamous cell carcinoma showed moderately positive reaction with antibody epi-l (fig. 5), but not with antibody myo-l. In the adenoacanthoma, the presence of antibody epi-l was moderately observed in the squamous metaplastic cells, while reactivity with antibody myo-l was seen in basal layers of squamous metaplastic epithelial and luminal epithelial areas. In contrast to the pleomorphic cell sarcoma, reactivity with antibody myo-l was observed in the intercellular matrices (fig. 6), but not with antibody epi-l.
Discussion This time, antibodies epi-l and myo-l of the human breast cancer origin were applied to DMBA-induced rat mammary tumors . In general, in the nonnal mammary gland, the 'epithelial cells of mammary ducts are showing two-cell-Iayer structure of luminal epithelial and myoepithelial cells (AZZOPARDI 1979). In our study, the number of cell layers confirmed the two-cell-type distribution of luminal epithelial and myoepithelial cells in the nonnal mammary gland or tumor-like lesions in rats. Especially, antibody epi-l reacted Exp Toxic Pathol 44 (1992) 5
237
with luminal epithelial cells of all mammary ducts, and antibody myo-l reacted with myoepithelial cells situated on the basal layer. On the other hand, reactivity of antibodies epi-l and myo1 to benign tumors showed two-cell-Iayer structure similar to the normal mammary gland. In other words, these fmdings are showing that there is not an essential difference in the constitution of luminal epithelial and myoepithelial cells among the normal mammary gland and benign mammary tumors. In the present study, the distribution of myoid cells was obvious by reaction with antibody myo-l. Most epithelial magliant mammary tumors were thought to be of luminal epithelial cell origin (AZZOPARDI 1979). On the other hand, several workers suggested that there is a possibility that certain mammary tumors may be of myoepithelial cell origin (HAGUENAU 1959; MURAD and SCRAPELLI 1967; HAYASHI and AOKI 1984). These observations suggest that there are great diversities of opinion in regard to the cell origin of the mammary tumors. In most DMBA-induced rat malignant mammary tumors examined in the present study, positive reactions with antibodies epi-l and myo-l were observed. DULBECCO (1982) described that most chemically-induced rat mammary tumors originated from stem cells of the mammary gland. They indicated that rat mammary tumors are differentiated into luminal epithelial or myoepithelial cells. Our results correspond to these findings. In the squamous cell carcinoma and adenoacanthoma, the presence of antibody epi-l was observed in the squamous metaplastic cells which are thought" to originate from the glandular epithelial cells. On the other hand, the present results showed that the intercellular matrices of pleomorphic cell sarcoma and stromal areas of the normal mammary gland or epithelial tumors were positive with myo-l, but negative with epi-l. It is possible that the sarcoma cells originated from transformed myogenic cells, but the reason for the reactivity with antibody myo-l is not clear. Additional studies will, therefore, be needed to establish the specificity of antibody myo-l. Myoepithelial cells have been reported to occur in the normal mammary gland (RANDOR 1972). LIOTTA et al. (1979) described that the basement membrane in the rat mammary gland is probably synthesized by myoepithelial cells. On the other hand, ARCHER et al. (1968) reported that the detailed ultrastructure of DMBA-induced rat mammary tumors indicates that they are composed of epithelial cells with a myoepithelial component. In the present study, reactivity with antibody myo-l was observed in normal mammary gland and tumor-like lesions including benign mammary tumors in rats. There findings might imply characteristics of myoepithelial cells. As mentioned above, our study showed that antibodies epi-l and myo-l established from the human breast cancer cell line reacted to normal mammary gland including various mammary tumors in rats. Therefore, we suppose that the monoclonal antibodies epi-l and myo-l will prove to be
238
Exp Toxic Patho144 (1992) 5
useful tools for the analysis of the interactions between luminal and myoepithelial cells in rat mammary tumors.
References ARCHER FL, ORLANDO RA: Morphology, natural history, and enzyme patterns in mammary tumors of the rat induced by 7,12-dimethylbenz(a)anthracene. Cancer Res 1986; 28, 217-224. AZZOPARDI JG: Problems in Breast Pathology. In: BENNINTON JL (Editor): WB Saunders Company Ltd. London, Philadelphia, Toronto, BIGGS R (1947). - The myoepithelium in certain tumors of the brest. J Pathol Bact 1979; 437. DULBECCO R : Immunological markers in the study of development and oncogenesis in the rat mammary gland. J Cell Physiol 1982; 2: 19-22. HAGUENAU F: Le cancer du sein chez la femme, electronique et au microscope optique. Bulletin de l' Association Fransaise pour l'Etude du Cancer. 1959; 46: 177-211. HA YASH! Y, AOKI Y: Findings of myoepithelial cells in human breast cancer. Acta Pathol Jpn 1984; 34: 537-552. Hsu SM, RAINE L, FANGER H: A comparative study of the peroxidase-antiperoxidase method and an avidin-biotin complex method for studying polypeptide hormone with radioimmunoassay antibodies. J Clin Pathol 1981; 75: 734-737. IMAI S, SEKIYA N, MORIMOTO J, KIYOZUKA Y, et al.: Establishment of monoclonal antibodies reactive with epithelial and myoepithelial cells in the breast. Acta Pathol Jpn 1986; 36: 1319-1326. LEUNG B: Hormonal dependency of experimental breast cancer. In Progress in Cancer Research and Therapy. Vol. 10. Hormones, receptors, and breast cancer. MCGURIE WL, ed., Raven Press, New York. LIOTTA LA, WICHA MS, FOIDART JM: Hormonal requirements for basement membrane collagen deposition by cultured rat mammary epithelium. Lab Invest 1979; 41: 511-518. MURAD TM, SCRAPELLI DG: The ultrastructure of medullary and scirrhous mammary duct carcinoma. Am J Pathol 1967; 50: 335-360. ORMEROD EJ, WARBURTON ML, GUSTERSON B, et al.: Abnormal deposition of basement membrane and connective tissue components in dimethylbenzanthracene-induced rat mammary tumors: An immunocytochemical and ultrastructural study. Histochem J 1985; 17: 1155-1166. RADNOR CJP: Myoepithelial cell differentiation in rat mammary glands. J Anat 1972; 111: 381-398. SEKIY A H: Production and analysis of monoclonal antibodies against human mammary carcinoma. J Nara Med Associ 1986; 37: 238-250. TUBURA A, INUI T, SENZAKI H, et al.: Immunohistochemical reactivity of a monoclonal antibody against cell membranes of human mammary carcinoma cells (l25B4) with rat normal tissues and DMBA-induced mammary tumors. Acta Histochem Cytochem 1989; 22: 561-565. TSUKIDATE K, TOIDA M, SOBUE M, et al.: Immunohistochemical studies of DMBA-induced rat mammary tumors. Acta Pathol Jpn 1988; 38: 129-139.
