- Email: [email protected]
Generation of monoclonal antibodies against human carcinosarcoma cell line for identification of myoepithelium and basement membrane
CLINICAL
IMMUNOLOGY
AND
IMMUNOPATHOLOGY
48, 78-84
(1988)
Generation of Monoclonal Antibodies against Human Carcinosarcoma Cell Line for Identification of Myoepithelium and Basement Membrane A. MEENAKSHI,*
M. UDAYACHANDER,* D. J. SCHOL,~ Jo HILGERS?
PH. HAGEMAN,~
AND
*Biochemistry Department, Cancer Institute, Madras-600 020, India; and fDivision of Turnour Biology, The Netherlands Cancer Institute, Antoni Van Leeuwenhoekhuis, Plesmanlaan 121, 1066 CX Amsterdam, The Netherlands An immunohistochemical method is described for identification of myoepithelial cells and basement membrane for cryostat tissue sections of normal, benign, and in situ carcinomas of the breast using two monoclonal antibodies 155Cl and 155DlO generated against human breast carcinosarcoma cell line HS578T. In the majority of infiltrating ductal carcinomas of the breast, there was a discontinuity in the myoepithelial cell layer, as a result an intact basement membrane could not be visualized. The reactivity of these two monoclonal antibodies might prove useful in the study of myoepithelial differentiation antigens and in the delineation of basement membrane. Among the other types of tissues studied, prominent staining was present with soft tissue tumors like leiomyosarcoma and synovial sarcoma. 0 1988 Academic press. Inc.
INTRODUCTION
Monoclonal antibodies which identify the differentiation antigens of human mammary gland have been reported by several workers (l-3). Antibodies to human milk fat globule membrane have proved useful for identification of breast epithelial cells and serve as potential markers for breast carcinomas (4, 5). Mammary tumor cell lines and metastatic mammary tumors have also been used to generate antibodies (6, 7). Most of these antibodies are directed against luminal epithelial cells. The relative contribution of myoepithelial cells for both benign and malignant breast disease is very meager. Characteristic properties of myoepithelial cells are the presence of actin, myosin, and keratin containing intermediate filaments and production of basement membrane proteins like libronectin, laminin, and type IV collagen. But none of these proteins is specific to myoepithelial cells. The present study describes two monoclonal antibodies against human breast carcinosarcoma cell line which might prove useful in the study of basal lamina and myoepithelial cells. MATERIALS
AND METHODS
HS578T cell line derived from a rare type of highly malignant breast tumor of epithelial origin with histopathology as infiltrating ductal carcinoma was provided by Professor Hackett, University of California, and was cultured in Dulbecco’s modified Eagle’s medium supplemented with NaHCO, (36 mA4), streptomycin (100 pg/ml), penicillin (100 U/ml), and 10% heat-inactivated fetal calf serum. 78
0090-1229/88 $1.50 Copyright 0 1988 by Academic Press, Inc. Ail rights of reproduction in any form reserved.
MONOCLONAL
ANTIBODIES
TO HUMAN
CARCINOSARCOMA
79
SP2/0 myeloma cells and hybrid cells were cultured in the same medium supplemented with 10% donor horse serum. Immunization
of BALBIc
Mice
The HS578T cells were harvested by centrifuging at 3OOgfor 5 min, and cells were washed with phosphate-buffered saline (PBS) and resuspended in PBS to a concentration of lo7 cells/ml. Cell aggregate (0.5 ml) was administered ip to inbred BALB/c mice every week for a period of 3 weeks. The mice were reinoculated 4 weeks after the last immunization and sacrificed 4 days later. Hybridization
Procedure
The immune spleen was dissected out aseptically, and spleen cell suspension was prepared by final mincing and was washed twice in serum-free DMEM. SP2/0 myeloma cells in logarithmic growth were washed twice in the same medium. The immune splenocytes and myeloma cells were fused in the ratio of 5:l in the presence of 50% polyethylene glycol (MW 1500) by the method of Milstein (8). The fused cells were resuspended in HAT medium supplemented with 20% fetal calf serum and seeded over 96-well Falcon culture plates containing a layer of feeder cells (10’ cells/well splenocytes). Screening
Procedure
An enzyme-linked immunosorbent assay (ELISA) was perfected to detect production of antibody by the hybrid cells using 96well polyvinyl microtiter assay plates coated with 1 &well of the membrane preparation consisting of the antigenie determinant of the cells in culture. Fifty microliters of hybridoma supernatant of wells with clones was added to each well. After incubation for 45 min at 37°C the wells were washed three times with PBS-Tween 20 and then three times with PBS, incubated for 30 min at 37°C with 50 ~1 of rabbit anti-mouse immunoglobulin conjugated with horse radish peroxidase (DAK0 P260), and diluted thousandfold in 10% NRS (normal rabbit serum). After a wash with PBS-Tween 20, bound peroxidase activity was visualized by adding 100 yl of a solution containing 0.06 mg TMB/ml and 0.03% (v/v) H,O, in 0.1 M sodium phosphate buffer, pH 6.0. After a IO-min incubation in the dark, the reaction was arrested by adding 50 ~1 2 M H,SO,. The extinction was measured at 450 nm using a BIO-TEK ELISA reader. Clones which showed reproducible positive signal against minimal background were considered positive. Zmmunohistochemical
Analysis
An indirect immunoperoxidase test was carried out on acetone-fixed frozen sections or on formalin-fixed paraffin embedded sections of normal, benign, or malignant breast tissues using culture supemates of positive clones and peroxidase-labeled rabbit anti-mouse IgG as second antibody. Normal rabbit serum (10%) was used as negative control. For frozen sections 3-amino-9-ethylcarbozole in 0.05 M sodium acetate buffer pH 4.9 containing 0.003% H,02 was used as
80
MEENAKSHI
ET AL.
substrate and for paraffin sections, diaminobenzidine hydrochloride in PBS containing 0.003% H202 was used. Sections were counterstained with hematoxylin. Cloning
of Hybrid
Cells
Positive clones with specific immunohistochemical reactivity were subcloned by limiting dilution to obtain stable clones elaborating monospecific antibodies. These hybrid cells were grown in large amounts, cryopreserved, and used for inducing hybridomas in BALBlc mice by ip injection of 5 x IO6 cells. Determination
of Antibody
Subclass
The immunoglobulin typing of each monoclonal was carried out by an agar gel diffusion test using rabbit anti-mouse immunoglobulins and then against antisera to IgG subclasses. Indirect
Zmmunojluorescence
Assay
HS578T cells were collected from confluent monolayers using 0.2% EDTA in PBS, washed twice, incubated for 30 min with monoclonal antibodies (ascites). NRS (MOO) was used as negative control. Cells were washed twice with PBS, incubated with FITC-labeled rabbit anti-mouse antibody for 30 min, and washed twice in PBS, and fluorescent cells were counted. RESULTS Of 37 positive clones identified by ELISA 15 clones revealed strong antibody activity. Immunostaining property of these clones studied by an indirect immunoperoxidase test using paraBin embedded and frozen sections of normal breast tissues as well as tissues of carcinoma of the breast with different histology indicated that two clones designated 155Cl and 155DlO exhibited staining reactivity with cryostatic sections of normal, benign, and malignant breast tissues. No positive staining was seen with paraffin sections. By an indirect immunofluorescence assay, more than 80% of the carcinosarcoma cells were found to be positive, further ascertaining the specificity of these two monoclonals. Isotyping of these monoclonals revealed them to be of IgG, class. The staining patterns of these two clones were identical and of great interest in that intense reactivity was present in the myoepithelial cells and basal lamina with very weak reaction in the epithelial cells. The tumor cells were positive. These two monoclonals thus facilitated visualization of intact basement membrane and myoepithelium and were stabilized. The other positive clones exhibited nonspecific reactivity. These two antibodies clearly delineated the myoepithelial cells in tissues of normal, benign, and in situ carcinoma of the breast. The luminal epithelial cells were weakly reactive. The basal membrane was clearly visible. In benign breast diseases like tibroadenoma, apocrine metaplasia, etc, the staining pattern was similar to those found in the normal breast tissues. For carcinoma in situ, the basement membrane was present intact all around the duct. These results are presented in Figs. la and b. In the majority of infiltrating ductal carcinomas, there
MONOCLONAL
ANTIBODIES
TO HUMAN
CARCINOSARCOMA
81
FIG. 1. (a) The duct of a normal mammary gland stained with monoclonal antibody 155Cl. The peripheral myoepithelial cells are clearly delineated; x 10. (b) Ducts of in siru carcinoma of the breast stained with 155Cl-represents intact basement membrane all around the duct; x25.
was a discontinuity in the myoepithelial cell layer and loss of definition of the basement membrane proteins, as a result the myoepithelium and basement membrane cannot be visualized compared to a normal duct and the duct of a carcinoma in situ. Among the other tissues studied, for salivary gland, the basal side and myoepithelium were positive for normal tissue with weak reaction in peripheral nerves.
MEENAKSHI
82
ET AL.
For normal lung tissue, alveoli and bronchus epithelium were positive, striated muscle in general was nonreactive, and smooth muscle exhibited positive staining. For ovarian cancer, tumor cells were weakly reactive with strong staining in the epitheiium and for carcinoma of the colon, tumor cells were weakly reactive with prominent staining in the luminal epithelium. For lymphomas and melanomas, few cases studied revealed negative results. Of the soft tissue tumors, leiomyosarcoma and synovial sarcoma were reactive, whereas liposarcoma and Ewing’s sarcoma were nonreactive. These results are presented in Table 1 and Table 2. Further studies on the distribution of antigen recognized by these two monoclonal antibodies are being carried out using various types of other normal and malignant tissues. DISCUSSION
An immunocytochemical technique is described using two monoclonal antibodies, 15X1 and 155D10, which facilitates the identification of myoepithelial cells and delineation of an intact basement membrane. Myoepithelial celis constitute a large proportion of benign breast lesions while in the majority of infiltrating carcinomas, very few myoepithelial cells could be identified. These results are in agreement with earlier morphological studies (9, 10). The reactivity of these two monoclonal antibodies with breast tissues of in sifu and invasive carcinomas suggests that during transition from in situ to invasive carcinoma, tumor cells penetrate the epithelial basement membrane causing disintegration and disorganization of the basement membrane, as a result an intact myoepithelium could not be visualized. Also, loss of myoepithelial differentiation is a general finding in invasive carcinomas (11, 12). These findings are in good correlation with those reported by other workers (13, 14). TABLE IMMUNOPEROXIDASE
REACTION __-
Tissue type _.~--~
1
PATTERN OF ANTIBODIES 155Cl AND HUMAN TISSUES ___..___-_ .- __-.. -~~.
-Breast Ductal epithelium Acinar epithelium Myoepithelium Basal membrane Salivary gland Acini Ducts striated basal side Myoepithelium Lung Alveoli Bronchus epithelium basal side Striated muscle (leg] Smooth muscle (skin) Peripheral nerve (salivary gland/skin) Note. - , Negative; W + , weak positive; + , positive.
15SDlO -~~--
ON NORMAL ..~--~~~
.-
Staining pattern -lW+ -IW+ + + + + + + + w+
-
MONOCLONAL
ANTIBODIES
TO HUMAN
TABLE IMMUNOPEROXIDASE
REACTION
83
CARCINOSARCOMA
2
PATTERN OF ANTIBODIES HUMAN TISSUES
15X1 AND 155DlO ON MALIGNANT Staining patterns
Tissue type --.--. Benign Fibroadenoma Apocrine metaplasia Malignant Carcinoma in situ Infiltrating carcinoma Other types of Cancer Ovary Colon Melanoma Lymphoma Soft tissue tumors Liposarcoma Leiomyosarcoma Synovial sarcoma Ewing’s sarcoma -__ Note.
.NR,
nonrelevant;
Epithelium
NR NR
-
+ +
+ +
w+
+
+
+ + w+ + +
Myoepithelium
Basement membrane
Tumor cells - -.-
+ + NR
NR
-
NR NR
Epithelial cells + Spindle cells -
-
NR
- , negative; W + , weak positive; + , positive.
Alterations of basement membrane during tumor invasion has been extensively reviewed by Liotta (15). Basement membrane is a ubiquitous extracellular matrix that separates epithelium from stroma. Normal epithelium and myoepithelium rest on a continuous basement membrane. Benign proliferative disorders of the human mammary gland are characterized by disorganization of the normal epithelial pattern with proliferation of epithelial and stromal elements. However, a continuous basement membrane is always present. In situ carcinoma is characterized by proliferation of neoplastic epithelium within the ducts with a firm basement membrane . During transition from in situ to invasive carcinoma, tumor cells penetrate the epithelial basal lamina, enter the underlying interstitial stroma, and gain access to lymphatics and blood vessels. These characteristics are used to type mammary carcinomas. Unfortunately, 85% of all breast cancers lack such features and cannot be classified other than undifferentiated infiltrating carcinomas. The HS578T carcinosarcoma cell line with a histopathology of both epithelial and myoepithelial origin has facilitated generation of two monoclonal antibodies which might prove clinically useful in studying the differentiation antigens of the basal lamina and myoepithelium. The presence of an intact myoepithelium often proves to be a good prognostic index. ACKNOWLEDGMENTS The financial support from the Ministry of Science, Netherlands, toward the above study is gratefully acknowledged. The technical assistance rendered by Mrs. Greet is acknowledged.
84
MEENAKSHI
ET Al..
REFERENCES I. 2. 3. 4. 5. 6. 7. 8. 9. 10. II. 12. 13. 14. IS.
Edwards, P. A. W.. Brit. .I. Crmw 51, 149. 1985. Levy, R., and Miller. R. A., Arm. Rev. Med. 34, 107. 1983. Neville. A. M., Foster. C. S., Moshakis, V., and Gore, M.. HWI. Pdd. 13, 1067, 1982. Hilkens,J.. Buijs, F.. Hilgers, J.. Hageman. Ph.. Calafat. J., Sonnenberg. A.. and Vander Valk. M., Int. .I. Cancer 34, 197. 1984. Iman, A., Lawrence, D. J. R., and Neville, A. M., Biochem. J. 193, 47, 1981, Menard. S., Tagilabue, E., Canevari, S., Fossati, G., and Colnaghi, M. I., Cancer Re.s. 43, 1295. 1983. Colcher, D., Hand, P.. Nuti, M.. and Schlom, J.. Prdc. Nat/. Ac&. Sci. UsA 78, 3199, 1981. Milstein, C., Sci. Am. 243, 66, 1980. Ahmed, A., J. Pathol. 113, 129, 1973. Hamperi, H., Curt-. Top. Purhol. 53, 161, 1970. Afbrechtsen, R., Nielsen, M., Wewer, U.. Engvall. E., and Ruoslahti, E., Cancer Re.c. 41, 5076, 1981. Barsky, S. H., Siegal, G. P., Jannotta, F., and Liotta, L. A., Lub. Invest. 46, 7, 1982. Siegal, G. P., Barsky, S. H., Terranova, V. P., and Liotta, L. A., lnvusion Metastasis 1, 54, 1981. Gould, V. E., Jao, W., and Battifora, H., Pathol. Res. Prucf. 167, 45, 1980. Liotta, L. A., Rao, C. N., and Barsky, S. H., Lab. ImBesr. 49, 636, 1983.
Received September 21, 1987; accepted with revision March 3, 1988
IMMUNOLOGY
AND
IMMUNOPATHOLOGY
48, 78-84
(1988)
Generation of Monoclonal Antibodies against Human Carcinosarcoma Cell Line for Identification of Myoepithelium and Basement Membrane A. MEENAKSHI,*
M. UDAYACHANDER,* D. J. SCHOL,~ Jo HILGERS?
PH. HAGEMAN,~
AND
*Biochemistry Department, Cancer Institute, Madras-600 020, India; and fDivision of Turnour Biology, The Netherlands Cancer Institute, Antoni Van Leeuwenhoekhuis, Plesmanlaan 121, 1066 CX Amsterdam, The Netherlands An immunohistochemical method is described for identification of myoepithelial cells and basement membrane for cryostat tissue sections of normal, benign, and in situ carcinomas of the breast using two monoclonal antibodies 155Cl and 155DlO generated against human breast carcinosarcoma cell line HS578T. In the majority of infiltrating ductal carcinomas of the breast, there was a discontinuity in the myoepithelial cell layer, as a result an intact basement membrane could not be visualized. The reactivity of these two monoclonal antibodies might prove useful in the study of myoepithelial differentiation antigens and in the delineation of basement membrane. Among the other types of tissues studied, prominent staining was present with soft tissue tumors like leiomyosarcoma and synovial sarcoma. 0 1988 Academic press. Inc.
INTRODUCTION
Monoclonal antibodies which identify the differentiation antigens of human mammary gland have been reported by several workers (l-3). Antibodies to human milk fat globule membrane have proved useful for identification of breast epithelial cells and serve as potential markers for breast carcinomas (4, 5). Mammary tumor cell lines and metastatic mammary tumors have also been used to generate antibodies (6, 7). Most of these antibodies are directed against luminal epithelial cells. The relative contribution of myoepithelial cells for both benign and malignant breast disease is very meager. Characteristic properties of myoepithelial cells are the presence of actin, myosin, and keratin containing intermediate filaments and production of basement membrane proteins like libronectin, laminin, and type IV collagen. But none of these proteins is specific to myoepithelial cells. The present study describes two monoclonal antibodies against human breast carcinosarcoma cell line which might prove useful in the study of basal lamina and myoepithelial cells. MATERIALS
AND METHODS
HS578T cell line derived from a rare type of highly malignant breast tumor of epithelial origin with histopathology as infiltrating ductal carcinoma was provided by Professor Hackett, University of California, and was cultured in Dulbecco’s modified Eagle’s medium supplemented with NaHCO, (36 mA4), streptomycin (100 pg/ml), penicillin (100 U/ml), and 10% heat-inactivated fetal calf serum. 78
0090-1229/88 $1.50 Copyright 0 1988 by Academic Press, Inc. Ail rights of reproduction in any form reserved.
MONOCLONAL
ANTIBODIES
TO HUMAN
CARCINOSARCOMA
79
SP2/0 myeloma cells and hybrid cells were cultured in the same medium supplemented with 10% donor horse serum. Immunization
of BALBIc
Mice
The HS578T cells were harvested by centrifuging at 3OOgfor 5 min, and cells were washed with phosphate-buffered saline (PBS) and resuspended in PBS to a concentration of lo7 cells/ml. Cell aggregate (0.5 ml) was administered ip to inbred BALB/c mice every week for a period of 3 weeks. The mice were reinoculated 4 weeks after the last immunization and sacrificed 4 days later. Hybridization
Procedure
The immune spleen was dissected out aseptically, and spleen cell suspension was prepared by final mincing and was washed twice in serum-free DMEM. SP2/0 myeloma cells in logarithmic growth were washed twice in the same medium. The immune splenocytes and myeloma cells were fused in the ratio of 5:l in the presence of 50% polyethylene glycol (MW 1500) by the method of Milstein (8). The fused cells were resuspended in HAT medium supplemented with 20% fetal calf serum and seeded over 96-well Falcon culture plates containing a layer of feeder cells (10’ cells/well splenocytes). Screening
Procedure
An enzyme-linked immunosorbent assay (ELISA) was perfected to detect production of antibody by the hybrid cells using 96well polyvinyl microtiter assay plates coated with 1 &well of the membrane preparation consisting of the antigenie determinant of the cells in culture. Fifty microliters of hybridoma supernatant of wells with clones was added to each well. After incubation for 45 min at 37°C the wells were washed three times with PBS-Tween 20 and then three times with PBS, incubated for 30 min at 37°C with 50 ~1 of rabbit anti-mouse immunoglobulin conjugated with horse radish peroxidase (DAK0 P260), and diluted thousandfold in 10% NRS (normal rabbit serum). After a wash with PBS-Tween 20, bound peroxidase activity was visualized by adding 100 yl of a solution containing 0.06 mg TMB/ml and 0.03% (v/v) H,O, in 0.1 M sodium phosphate buffer, pH 6.0. After a IO-min incubation in the dark, the reaction was arrested by adding 50 ~1 2 M H,SO,. The extinction was measured at 450 nm using a BIO-TEK ELISA reader. Clones which showed reproducible positive signal against minimal background were considered positive. Zmmunohistochemical
Analysis
An indirect immunoperoxidase test was carried out on acetone-fixed frozen sections or on formalin-fixed paraffin embedded sections of normal, benign, or malignant breast tissues using culture supemates of positive clones and peroxidase-labeled rabbit anti-mouse IgG as second antibody. Normal rabbit serum (10%) was used as negative control. For frozen sections 3-amino-9-ethylcarbozole in 0.05 M sodium acetate buffer pH 4.9 containing 0.003% H,02 was used as
80
MEENAKSHI
ET AL.
substrate and for paraffin sections, diaminobenzidine hydrochloride in PBS containing 0.003% H202 was used. Sections were counterstained with hematoxylin. Cloning
of Hybrid
Cells
Positive clones with specific immunohistochemical reactivity were subcloned by limiting dilution to obtain stable clones elaborating monospecific antibodies. These hybrid cells were grown in large amounts, cryopreserved, and used for inducing hybridomas in BALBlc mice by ip injection of 5 x IO6 cells. Determination
of Antibody
Subclass
The immunoglobulin typing of each monoclonal was carried out by an agar gel diffusion test using rabbit anti-mouse immunoglobulins and then against antisera to IgG subclasses. Indirect
Zmmunojluorescence
Assay
HS578T cells were collected from confluent monolayers using 0.2% EDTA in PBS, washed twice, incubated for 30 min with monoclonal antibodies (ascites). NRS (MOO) was used as negative control. Cells were washed twice with PBS, incubated with FITC-labeled rabbit anti-mouse antibody for 30 min, and washed twice in PBS, and fluorescent cells were counted. RESULTS Of 37 positive clones identified by ELISA 15 clones revealed strong antibody activity. Immunostaining property of these clones studied by an indirect immunoperoxidase test using paraBin embedded and frozen sections of normal breast tissues as well as tissues of carcinoma of the breast with different histology indicated that two clones designated 155Cl and 155DlO exhibited staining reactivity with cryostatic sections of normal, benign, and malignant breast tissues. No positive staining was seen with paraffin sections. By an indirect immunofluorescence assay, more than 80% of the carcinosarcoma cells were found to be positive, further ascertaining the specificity of these two monoclonals. Isotyping of these monoclonals revealed them to be of IgG, class. The staining patterns of these two clones were identical and of great interest in that intense reactivity was present in the myoepithelial cells and basal lamina with very weak reaction in the epithelial cells. The tumor cells were positive. These two monoclonals thus facilitated visualization of intact basement membrane and myoepithelium and were stabilized. The other positive clones exhibited nonspecific reactivity. These two antibodies clearly delineated the myoepithelial cells in tissues of normal, benign, and in situ carcinoma of the breast. The luminal epithelial cells were weakly reactive. The basal membrane was clearly visible. In benign breast diseases like tibroadenoma, apocrine metaplasia, etc, the staining pattern was similar to those found in the normal breast tissues. For carcinoma in situ, the basement membrane was present intact all around the duct. These results are presented in Figs. la and b. In the majority of infiltrating ductal carcinomas, there
MONOCLONAL
ANTIBODIES
TO HUMAN
CARCINOSARCOMA
81
FIG. 1. (a) The duct of a normal mammary gland stained with monoclonal antibody 155Cl. The peripheral myoepithelial cells are clearly delineated; x 10. (b) Ducts of in siru carcinoma of the breast stained with 155Cl-represents intact basement membrane all around the duct; x25.
was a discontinuity in the myoepithelial cell layer and loss of definition of the basement membrane proteins, as a result the myoepithelium and basement membrane cannot be visualized compared to a normal duct and the duct of a carcinoma in situ. Among the other tissues studied, for salivary gland, the basal side and myoepithelium were positive for normal tissue with weak reaction in peripheral nerves.
MEENAKSHI
82
ET AL.
For normal lung tissue, alveoli and bronchus epithelium were positive, striated muscle in general was nonreactive, and smooth muscle exhibited positive staining. For ovarian cancer, tumor cells were weakly reactive with strong staining in the epitheiium and for carcinoma of the colon, tumor cells were weakly reactive with prominent staining in the luminal epithelium. For lymphomas and melanomas, few cases studied revealed negative results. Of the soft tissue tumors, leiomyosarcoma and synovial sarcoma were reactive, whereas liposarcoma and Ewing’s sarcoma were nonreactive. These results are presented in Table 1 and Table 2. Further studies on the distribution of antigen recognized by these two monoclonal antibodies are being carried out using various types of other normal and malignant tissues. DISCUSSION
An immunocytochemical technique is described using two monoclonal antibodies, 15X1 and 155D10, which facilitates the identification of myoepithelial cells and delineation of an intact basement membrane. Myoepithelial celis constitute a large proportion of benign breast lesions while in the majority of infiltrating carcinomas, very few myoepithelial cells could be identified. These results are in agreement with earlier morphological studies (9, 10). The reactivity of these two monoclonal antibodies with breast tissues of in sifu and invasive carcinomas suggests that during transition from in situ to invasive carcinoma, tumor cells penetrate the epithelial basement membrane causing disintegration and disorganization of the basement membrane, as a result an intact myoepithelium could not be visualized. Also, loss of myoepithelial differentiation is a general finding in invasive carcinomas (11, 12). These findings are in good correlation with those reported by other workers (13, 14). TABLE IMMUNOPEROXIDASE
REACTION __-
Tissue type _.~--~
1
PATTERN OF ANTIBODIES 155Cl AND HUMAN TISSUES ___..___-_ .- __-.. -~~.
-Breast Ductal epithelium Acinar epithelium Myoepithelium Basal membrane Salivary gland Acini Ducts striated basal side Myoepithelium Lung Alveoli Bronchus epithelium basal side Striated muscle (leg] Smooth muscle (skin) Peripheral nerve (salivary gland/skin) Note. - , Negative; W + , weak positive; + , positive.
15SDlO -~~--
ON NORMAL ..~--~~~
.-
Staining pattern -lW+ -IW+ + + + + + + + w+
-
MONOCLONAL
ANTIBODIES
TO HUMAN
TABLE IMMUNOPEROXIDASE
REACTION
83
CARCINOSARCOMA
2
PATTERN OF ANTIBODIES HUMAN TISSUES
15X1 AND 155DlO ON MALIGNANT Staining patterns
Tissue type --.--. Benign Fibroadenoma Apocrine metaplasia Malignant Carcinoma in situ Infiltrating carcinoma Other types of Cancer Ovary Colon Melanoma Lymphoma Soft tissue tumors Liposarcoma Leiomyosarcoma Synovial sarcoma Ewing’s sarcoma -__ Note.
.NR,
nonrelevant;
Epithelium
NR NR
-
+ +
+ +
w+
+
+
+ + w+ + +
Myoepithelium
Basement membrane
Tumor cells - -.-
+ + NR
NR
-
NR NR
Epithelial cells + Spindle cells -
-
NR
- , negative; W + , weak positive; + , positive.
Alterations of basement membrane during tumor invasion has been extensively reviewed by Liotta (15). Basement membrane is a ubiquitous extracellular matrix that separates epithelium from stroma. Normal epithelium and myoepithelium rest on a continuous basement membrane. Benign proliferative disorders of the human mammary gland are characterized by disorganization of the normal epithelial pattern with proliferation of epithelial and stromal elements. However, a continuous basement membrane is always present. In situ carcinoma is characterized by proliferation of neoplastic epithelium within the ducts with a firm basement membrane . During transition from in situ to invasive carcinoma, tumor cells penetrate the epithelial basal lamina, enter the underlying interstitial stroma, and gain access to lymphatics and blood vessels. These characteristics are used to type mammary carcinomas. Unfortunately, 85% of all breast cancers lack such features and cannot be classified other than undifferentiated infiltrating carcinomas. The HS578T carcinosarcoma cell line with a histopathology of both epithelial and myoepithelial origin has facilitated generation of two monoclonal antibodies which might prove clinically useful in studying the differentiation antigens of the basal lamina and myoepithelium. The presence of an intact myoepithelium often proves to be a good prognostic index. ACKNOWLEDGMENTS The financial support from the Ministry of Science, Netherlands, toward the above study is gratefully acknowledged. The technical assistance rendered by Mrs. Greet is acknowledged.
84
MEENAKSHI
ET Al..
REFERENCES I. 2. 3. 4. 5. 6. 7. 8. 9. 10. II. 12. 13. 14. IS.
Edwards, P. A. W.. Brit. .I. Crmw 51, 149. 1985. Levy, R., and Miller. R. A., Arm. Rev. Med. 34, 107. 1983. Neville. A. M., Foster. C. S., Moshakis, V., and Gore, M.. HWI. Pdd. 13, 1067, 1982. Hilkens,J.. Buijs, F.. Hilgers, J.. Hageman. Ph.. Calafat. J., Sonnenberg. A.. and Vander Valk. M., Int. .I. Cancer 34, 197. 1984. Iman, A., Lawrence, D. J. R., and Neville, A. M., Biochem. J. 193, 47, 1981, Menard. S., Tagilabue, E., Canevari, S., Fossati, G., and Colnaghi, M. I., Cancer Re.s. 43, 1295. 1983. Colcher, D., Hand, P.. Nuti, M.. and Schlom, J.. Prdc. Nat/. Ac&. Sci. UsA 78, 3199, 1981. Milstein, C., Sci. Am. 243, 66, 1980. Ahmed, A., J. Pathol. 113, 129, 1973. Hamperi, H., Curt-. Top. Purhol. 53, 161, 1970. Afbrechtsen, R., Nielsen, M., Wewer, U.. Engvall. E., and Ruoslahti, E., Cancer Re.c. 41, 5076, 1981. Barsky, S. H., Siegal, G. P., Jannotta, F., and Liotta, L. A., Lub. Invest. 46, 7, 1982. Siegal, G. P., Barsky, S. H., Terranova, V. P., and Liotta, L. A., lnvusion Metastasis 1, 54, 1981. Gould, V. E., Jao, W., and Battifora, H., Pathol. Res. Prucf. 167, 45, 1980. Liotta, L. A., Rao, C. N., and Barsky, S. H., Lab. ImBesr. 49, 636, 1983.
Received September 21, 1987; accepted with revision March 3, 1988