Antigen Expression in Normal and Neoplastic Canine Tissues Defined by a Monoclonal Antibody...

Vet Pathol 31:663-673 (1994)

EXPERIMENTAL DISEASE

Antigen Expression in Normal and Neoplastic Canine Tissues Defined by a Monoclonal Antibody Generated

against Canine Mesothelioma Cells

K. X. LIU, A. E. CHURCH BIRD, S. D. LENZ, S. P. MCDONOUGH, AND L. G. WOLFE

Department of Pathobiology, College of Veterinary Medicine, Auburn University, Auburn, AL (KXL, AECB, SDL, LGW); and

Department of Veterinary Pathology, School of Veterinary Medicine, University of California, Davis, CA (SPM)

Abstract. Monoclonal antibody (MAb) 3B5 generated against canine mesothelioma cells was applied to canine tumors and normal tissues via immunohistochemical and immunoblotting techniques to evaluate antigen binding. By use of an avidin-biotin immunoperoxidase complex (ABC) method, immunoreactivity was noted in reactive mesothelial cells and in normal tissues was observed primarily in mesothelial cell linings, endothelial cells, and smooth muscle of blood vessels and soft tissues; the reactivity was nearly equivalent in frozen or formalin-fixed, paraffin-embedded tissue sections. Use of the ABC method on formalin-fixed, paraffin-embedded tumors yielded moderate to strong cytoplasmic immunostaining of neoplastic cells in 10/11 (9 1%) mesotheliomas, 18/23 (78%) hemangiosarcomas, 4/ 10 (40%) intestinal and lung carcinomas, and 5 20% of hemangiomas, leiomyosarcomas, leiomyomas, mammary carcinomas, and squamous cell carcinomas. No immunostaining of tumor cells was observed in fibrosarcomas, hemangiopericytomas, perianal gland carcinomas, and melanomas. Immunoblotting was performed on samples that demonstrated strong immunoreactivity with MAb 3B5 by the ABC method: mesothelioma, hemangiosarcoma, urinary bladder (smooth muscle), and lung (alveolar capillaries). These analyses showed that MAb 3B5 bound a major antigen of 78 kilodaltons (kd) and minor antigens at 56 and 54 kd in normal and neoplastic tissues. The preliminary immunohistochemical results suggest that MAb 3B5 may possess utility in diagnosis of mesotheliomas and hemangiosarcomas, discrimination of cell types in proliferative serosal lesions, and demonstration of vascularity or angiogenesis in neoplastic and inflammatory lesions.

Key words: Dogs; hemangiosarcoma; immunoblotting; immunohistochemistry; mesothelioma; monoclonal antibodies; tumor-associated antigen.

Hybridoma technology25 has been applied by many research groups in recent years to generate numerous monoclonal antibodies (MAbs) that are reactive against human tumor-associated antigens (TAAs).~' ,~ ' Among the clinical applications of MAbs, i.e., immunodiagnosis, monitoring of disease progression, and im- m ~ n o t h e r a p y , ~ ~ the predominant application of MAbs in medical oncology continues to be in the area of immunodiagnosis. Use of MAbs in immunodiagnosis include detection of TAA in body fluids, nuclear scan- ning with radiolabeled MAbs, cytologic evaluations, and immunostaining of tissue sections to aid in dif- ferential diagnosis of tumor types and in determinations of biological behavior and prognosis. The types of antigens that frequently serve as targets in immu-

nohistochemical tests applied to tissue sections are oncofetal antigens (e.g., carcinoembryonic antigen [CEA],5 tumor-associated glycoprotein-72 [TAG-72]),44 differentiation antigens (e.g., endocrine pep tide^),^^ onco- gene products (e.g., ~ - e r b B - 2 ) , ~ and cytoskeletal markers (e.g., intermediate filaments).6

The use of MAbs in veterinary oncology thus far has been limited primarily to immunohistochemical ap-

and therein has been limited primarily to the use of commercially available MAbs for which applications with human cancer have been well documented. The MAbs apparently most frequently employed have been those that react with the protein antigens of intermediate filaments. Because the expression of in-

plications in tumor pathology1,3,8,9.~3,~6,18-20,30,32,34,35,37,40,43

663

by guest on September 24, 2016vet.sagepub.comDownloaded from

http://vet.sagepub.com/

664 Liu, Church Bird, Lenz, McDonough, and Wolfe Vet Pathol 31~6, 1994

termediate filament proteins characteristically oc- curs in both normal and neoplastic cells, MAbs against intermediate filaments have been particular- ly used to identify the histogenesis of anaplastic neoplasms, e.g., cytokeratin, vimentin, and desmin intermediate filaments for identification of epithelial, mesenchymal, and muscle origins, respectively. Un- til recently, coexpression of intermediate filament proteins was thought to be restricted to only a few tumor types, e.g., cytokeratin and vimentin expression in me~othe l iomas , ’ .~~ .~~ and thereby the demonstration of a single type of intermediate filament was considered a reliable indicator of histogenesis. However, recent findings with human tumors have suggested that coexpression of intermediate filament proteins may be a much more common occurrence than once thought. Reports of coexpression include the demonstration of both cytokeratins and vimentin in some melanoma^^^,^' and carcinomas33 and the suggestion that coexpression of intermediate filament proteins may have some prognostic value.21

Most MAbs that possess novel utility in immunodiagnosis of human tumors were generated by utilizing a human tumor cell preparation as the immunogen. Exclusive of virus-encoded cellular antigens, this strat- egy for generation of MAbs reactive with TAA of animal tumors, i.e., use of animal tumor cell preparations as immunogens, apparently has been employed very infrequently. 1,16,35,43 Relevant to cancer in dogs, MAbs have been shown to possess diagnostic potential with regard to lymph0mas,4~ several epithelial cancers, I 6 and rnelanorna~.~~ Two MAbs were reported to show prom- ise in distinguishing between benign and malignant tumors, namely perianal gland tumors16 or melano- m a ~ . ~ ~

In this report, we describe the tissue reactivity of an MAb, designated 3B5, which was generated against canine mesothelioma cells. The expectancy was that the antibody would show preferential reactivity for malignant versus reactive and normal mesothelial cells and/or for mesothelial cells versus carcinoma cells and thereby be useful in exfoliative cytology and surgical pathologic analysis. MAbs generated against human mesothelioma cells demonstrate such preferential reac- t i v i t i e ~ . ~ ~ , ~ ~ Human mesotheliomas lack the frequently measured epithelial tumor markers such as CEA, TAG- 72, and epithelial membrane antigen.2.11.26

MAb 3B5 failed to distinguish mesothelioma cells from reactive or normal mesothelial cells, but it did show preferential reactivity for mesothelial/mesothelioma cells versus carcinoma cells. The antibody also demonstrated high binding specificity for antigen expressed by normal and malignant endothelial cells and for smooth muscle cells of blood vessels and soft tissues.

Materials and Methods

Monoclonal antibodies

The murine MAb examined in this study, designated MAb 3B5, was generated by established The immunogen was cryopreserved whole cells harvested from fluid obtained by thoracentesis from a 7-year-old German Short- haired Pointer dog with histopathologically confirmed pleural mesothelioma. The pleural effusion (PE) cells were pre- dominantly cells of mesothelial origin as determined by cytologic examination, ultrastructural features (long, slender microvilli covering the cell surface^),^*.^^ and immunostaining for vimentin and cytokeratin intermediate filament proteins. Immunization of a mouse was accomplished by intra- peritoneal injections of 2 x lo6 PE cells 3 times at 2-week intervals. Hybridoma colonies were formed by fusion of mouse splenic cells with the nonsecreting mouse myeloma cell line P3x63Ag8.653 (American Type Culture Collection, Rockville, MD). Screening of hybridoma supernates was con- ducted against PE cells, canine peripheral blood lymphocytes, and a canine kidney cell line (Department of Pathobiology, Auburn University) by the avidin-biotin peroxidase complex (ABC) methodz4 in an enzyme-linked immunosorbent assay (ELISA) per the manufacturer’s instructions (Vector Labo- ratories, Burlingame, CA). The hybridoma culture 3B5 was selected for expansion and further study because of high sig- nal strength against PE cells, lack of binding with normal canine peripheral blood lymphocytes and kidney cells, and excellent growth properties. The 3B5 culture was subcloned four times by using limiting dilutions, and the antibody pro- duced by the culture was designated MAb 3B5. The isotype of MAb 3B5 was determined by use of ELISA kits (GIBCO, Grand Island, NY) to be IgG1 with kappa light chains.

An IgG 1 MAb (MAb-TG 1 DE4) developed against Toxo- plasma gondiiz8 (supplied by Dr. D. Lindsay, Department of Pathobiology, Auburn University) was used as a control for nonspecific binding. This control antibody was run as a primary antibody in parallel with MAb 3B5 on all tissue sections examined by immunohistochemical methods.

Immunohistochemical staining

Normal and neoplastic tissues were studied. Normal tissues (lung, liver, large intestine, kidney, urinary bladder, brain, heart, skeletal muscle, spleen. lymph node, skin) were obtained from six healthy dogs at necropsy. From each tissue, some samples were fixed in 10°/o buffered formalin and embedded in paraffin, and other samples were frozen immediately in liquid nitrogen and stored at -70 C. Tumor specimens, except for several mesotheliomas, were obtained from the file of tissue blocks in the Department of Pathobiology, Auburn University. These tissues had been fixed in 10% buffered formalin and embedded in paraffin in a routine man- ner; sections of each tissue block had been examined and a histopathologic diagnosis rendered. Additional sections from each block were cut for use in this study. From one hemangiosarcoma in this group, tumor samples also were available that had been frozen in liquid nitrogen immediately after surgical resection and stored at -70 C. With both normal and neoplastic tissues, frozen samples and tissue blocks were cut at 5 pm thickness and mounted onto poly-L-lysine-treated



Vet Pathol 31:6. 1994 Anti-canine Mesothelioma Monoclonal Antibody Immunoreactivity 665

or gelatin-coated glass slides, respectively. Important sup- plements to case materials at Auburn were tissue sections of several mesotheliomas supplied as follows: nine canine, De- partment of Veterinary Pathology, University of California, Davis, CA; one human and two mouse, Dr. D. L. Dillehay, Emory University School of Medicine, Atlanta, GA; and one hamster, National Institutes of Health, Bethesda, MD.

Immunohistochemical staining in this study was performed using a modification of the ABC immunoperoxidase (IP) methodz4 per instructions accompanying the ABC kit. Formalin-fixed, paraffin-embedded tissue sections were de- paraffinized in xylene and rehydrated in graded ethanols. Frozen sections were fixed in cold methanol for 4 minutes and then submersed in phosphate-buffered saline (PBS) in Tween-20 (PBST) for 10 minutes. Rehydrated formalin-fixed and frozen tissue sections were treated with methanol containing 0.6% H20, for 20 minutes to block endogenous peroxidase. After two washes in PBST solution (1 0 minutes per wash), sections were treated with 1% normal horse serum in PBS for 30 minutes at room temperature and incubated with MAb 3B5 (undiluted supernate of hybridoma culture) or MAb-TG 1 DE4 (isotype control performed on sections of each tissue in the same assay) at 4 C overnight. The sections were then washed with PBST and incubated with biotinylated horse anti-mouse IgG immunoglobulin for 30 minutes at room temperature, followed by three washes and incubation with avidin and biotinylated horseradish peroxidase for 30 minutes at room temperature. Sections were washed twice, and the peroxidase reaction was initiated using 0.06% diami- nobenzidine (DAB; Sigma Chemical Co., St. Louis, MO) in 0.1 M Tris buffer (pH 7.2) with 0.03% H202 for 5 minutes. After a rinse with tap water for 10 minutes, the sections were counterstained with hematoxylin for 3 minutes, rinsed with tap water, dehydrated in ethanol, cleared in xylene, and mounted under a coverslip using Permount. With each staining series, tissue sections of a canine mesothelioma or hemangiosarcoma that had been previously shown to bind MAb 3B5 were reacted with MAb 3B5 (positive control) and with MAb-TG 1 DE4 (isotype control).

Each section was evaluated for cell-associated brown staining and for staining of extracellular material. The staining intensity was subjectively graded as 3+ (intense), 2+ (moderate), and 1 + (weak). The percentage of stained cells and the pattern of staining (focal or diffuse) were recorded. Neoplastic samples were considered positive for immunostaining when L 5% of tumor cells stained.

Immunoblotting analysis

Cell lysates were prepared from samples of PE (mesothelioma) cells, hemangiosarcoma, urinary bladder, and lung that were positive for MAb 3B5 immunoreactivity in IP assays and from a sample of brain that was negative in IP assay, except for a few blood vessels. Cryopreserved PE cells were washed five times and incubated with a lysis buffer (PBS with 0.02% sodium azide, 1 mM phenylmethylsulfonylfluo- ride [Sigma], 0.0005% Tween-20, 0.00 1% NP-40 [Sigma], pH 7.4) at 4 C for 30 minutes. One sample each of hemangiosarcoma, urinary bladder, lung, and brain, which had been snap-frozen in liquid nitrogen and stored at -70 C, was pulverized using mortar and pestle (liquid nitrogen cooled)

and then homogenized with a Dounce homogenizer in the lysis buffer. The lysates of PE cells and tissues were centri- fuged at 13,000 x g at 4 C for 10 minutes. Supernates were collected, 10 ~1 of 12 mM deoxycholate (Sigma) was added per milliliter of supernate, and the solution was filtered through a 0.22-wm syringe filter, dispensed into aliquots, and stored at -20 C. The concentration of total protein was determined by the modified Lowry method.31

Sodium dodecyl sulfate (SDS)-polyacrylamide gel electro- phoresis was performed on the lysates according to the method of Laemmli.27 Samples were mixed with an equal volume of 2x Laemmli’s sample buffer, heated in a boiling water bath for 2 minutes, then run at 10 C and a constant current of 25 mA for the stacking gel (2.0% polyacrylamide) and 35 mA for the running gel (5-1 5% gradient polyacrylamide). Molecular weight standards (Bio-Rad Laboratories, Melville, NY) and prestained molecular weight standards (Bio-Rad Laboratories) were loaded onto each gel. Proteins were transferred from the gels to Immobilon P PVDF membranes (Mil- lipore, Bedford, MA) under conditions of 14 volts and 4 C for 15 The strip of membrane containing molecular weight standards was stained with amido black. Strips containing test samples were incubated with 0.5 9’0 casein (Sigma) in PBS on a rocking apparatus for 1 hour at room temperature. Then 3B5 hybridoma supernatant, 1 : 7 dilution in PBS or 0.5% casein in PBS, or an irrelevant MAb, was added to the membrane and incubated for 1.5 hours on a rocking apparatus at room temperature. The membrane was washed three times for 10 minutes in PBS and incubated in alkaline phosphatase-conjugated goat anti-mouse IgG (Cappel Re- search Products, Durham, NC) for 1 hour. The membrane was washed again and immersed in substrate reagent (Cappel) until the color development was complete. The reaction was stopped by rinsing in distilled water.

Results

Immunoperoxidase staining

Canine mesothelioma cells comprised the immunogen in generation of MAb 3B5 so initial efforts were directed toward determining the reactivity of MAb 3B5 against mesothelial and mesothelioma cells. Immu- noperoxidase assays were performed on formalin-fixed, paraffin-embedded tissue sections of normal mesothelium covering lung, liver, or spleen from 10 dogs, hyperplastic mesothelial cells on the surfaces of liver or spleen from seven dogs, 1 1 canine mesotheliomas (five pleural, three pericardial, three peritoneal), and 11 peritoneal mesotheliomas derived from five species other than dogs (Table 1). Intense brown staining in- dicative of strong reactivity was consistently observed in normal mesothelial cell linings and in the cytoplasm of hyperplastic mesothelial cells. Cytoplasmic binding of MAb 3B5 also was evident in 10 of 1 1 (9 1%) canine mesotheliomas, as demonstrated by moderate to intense staining in 50-100% of the neoplastic cells (Figs. 1, 2). In contrast, immunostaining was absent in the mesotheliomas obtained from six cattle, one cat, two



666 Liu, Church Bird, Lenz, McDonough, and Wolfe Vet Pathol 31:6, 1994

Table 1. Immunohistochemical reactivity of MAb 3B5 with canine mesothelial cells and mesotheliomas.

Histologic Diagnosis Staining

No. Positive/ Yo Positive

Cells No. Tested Pattern* Inteiisityt

Normal mesothelium 10/10 Patchy or diffuse 3' NDS Mesothelial cell hyperplasia 7/7 Diffuse 3' 90-100

Mesothelioma

Epithelial Sarcomatoid Mixed

Canine 718 Focal or diffuse 2-3 +

2/2 Diffuse 2-3 +

1/1 Diffuse 3+

50-100 50-75

75

Bovine 016 Feline o/ 1 Mouse 012 Hamster o/ 1 Human o/ 1

* Patchy = intermittent staining; diffuse = continuity of mesothelial cell staining; stained tumor cells distributed throughout the tissue

t 2+ = moderate; 3' = strong. + ND = not determined.

section; Focal = staining localized to clusters of tumor cells.

Fig. 2. Epithelial mesothelioma reacted with MAb 3B5; dog. There is heterogeneity of antigen expression within the tissue section. Note sharp demarcation between the areas of reactive (top) and nonreactive (below) tumor cells. Immu- noperoxidase, DAB. Bar = 20 gm.

Fig. 1. Epithelial mesothelioma reacted with MAb 3B5; dog. Dark cytoplasmic staining in the mesothelioma cells denotes antigen-antibody binding. Immunoperoxidase, DAB. Bar = 20 pm.



Vet Pathol 31:6, 1994 Anti-canine Mesothelioma Monoclonal Antibody Immunoreactivity 667

Table 2. Immunohistochemical reactivity of MAb 3B5 with canine tumors other than mesotheliomas.

Tumor* Staining

No. Positive/ Yo Positive

Intensity4 Tumor Cells No. Tested Pattern?

Hemangiosarcoma Leiom yosarcoma Fibrosarcoma Hemangioma Leiom yoma Hemangiopericytoma

Intestinal carcinoma Lung adenocarcinoma Mammary carcinoma Squamous cell carcinoma Perianal gland carcinoma

19§/23 Diffuse 1 / 6 Focal 0/12 2/15 Diffuse 115 Diffuse 0/15

2-3 + 5 0-8 0 2-3 + 50

2+ 30,80 3+ 90

4/10 Focal 2-3 + 5-20 4/10 Focal or diffuse 2-3 + 10-50 2/10 Diffuse 2+ 80 2/10 Focal, diffuse 2+ 20,80 0110

Melanoma 0/10

* Formalin-fixed, paraffin-embedded tissue sections of all tumors and frozen sections of one hemangiosarcoma were evaluated. t Diffuse = stained tumor cells distributed throughout the tissue section; Focal = staining localized to clusters of tumor cells. 4 2+ = moderate; 3+ = strong. 8 One hemangiosarcoma was strongly positive in frozen sections but negative in formalin-fixed samples.

mice, one hamster, and one human. The study of canine tissue sections containing mesothelial cells and mesotheliomas also provided the opportunity to evaluate underlying tissues for immunostaining with MAb 3B5, i.e., lung, liver, spleen, and intestine. Staining graded moderate to intense was present in smooth muscle cells of most veins and arteries, endothelial cells in about 10% of veins and arteries in each tissue type, pulmonary alveolar linings (interpreted to be capillary linings and not alveolar epithelial cells), peribronchial smooth muscle, smooth muscle of the splenic capsule and trabeculae, smooth muscle of the intestinal wall and muscularis mucosae, and intestinal mucosal epithelia (focal, < 5 % total cells). Staining was absent in bronchial and bronchiolar epithelia, hepatic parenchyma, splenic parenchyma, stromal or connective tissue, erythrocytes, and blood leukocytes. Little or no background staining was observed. In all assays performed in this study, a section of each tissue was treated with an irrelevant primary antibody in parallel with MAb 3B5; nonspecific staining was not observed.

A variety of formalin-fixed, paraffin-embedded canine tumors were analyzed for antigen expression recognized by MAb 3B5 (Table 2). Of the mesenchymal tumors evaluated, hemangiosarcomas demonstrated the greatest reactivity; 18 of 23 tumors (78%) were positive. Fifty to 80% of the neoplastic endothelial cells in these tumors showed moderate to intense staining (Fig. 3). Frozen tissue was available from one of the hemangiosarcomas, which was negative when evaluated in formalin-fixed, paraffin-embedded tissue sections; frozen sections of this tumor were distinctly positive

m (. .

Fig. 3. Hepatic hemangiosarcoma reacted with MAb 3B5; dog. Malignant endothelial cells (arrows) lining vascular spaces and in tissue fragments show a moderate to intense degree of immunoreactivity. Also note the intense staining of hyperplastic mesothelial cells on the surface of the liver (top). Immunoperoxidase, DAB. Bar = 20 pm.




Table 3. Immunoreactivity of MAb 3B5 in normal formalin-fixed and frozen canine tissue sections.

Tissue* Positive Tissue Componentst Negative Tissue Components$

Lung Alveolar capillaries, penbronchial smooth Bronchial and bronchiolar epithelium, alveo-

Liver

Spleen Capsule, trabecular smooth muscle Lymphocytes, monocyte/phagocyte system Lymph node Subcapsular lymphatics Lymphocytes Colon

muscle lar epithelial cells, stroma

tive cells), bile Bile duct epithelia (apical granules, < 1% reac- Hepatocytes, Kupffer cells, stroma

Muscularis mucosae, circular and longitudinal smooth muscle, mucosal epithelia (multifo- cal, one of six formalin-fixed samples)

Mucosal epithelia (five of six), stroma

Kidney Glomeruli, tubular epithelia, stroma Urinary bladder Smooth muscle Urothelium, stroma Heart Cardiac muscle Skeletal muscle Skeletal muscle Skin Epidermis and adnexal epithelia (two of six Epidermis and adnexal epithelia (four of six),

formalin-fixed samples) stroma * Formalin-fixed and snap-frozen tissue sections from the same six dogs were examined by an avidin-biotin immunoperoxidase complex

method. t Vascular smooth muscle was stained in all tissues. Endothelial cells were stained in about 10% of the blood vessels in formalin-fixed

tissues; endothelial cell staining was not determined in frozen sections. Mesothelial cell linings, whenever present in the tissue sections, were consistently positive.

$ Blood leukocytes and erythrocytes were negative in all tissues.

(3+, >50% of cells). In contrast to hemangiosarcomas, only 2 of 15 (13%) hemangiomas showed immunostaining of the endothelial cells that lined dilated vascular spaces comprising the tumors, Of the remaining mesenchymal tumors, immunostaining was observed infrequently in smooth muscle tumors (one of six leiomyosarcomas, one of five leiomyomas) and was absent in fibrosarcomas (zero of 12) and hemangiopericytomas (zero of 15). In the tumors that showed no reactivity, a positive internal control for staining was nearly always provided by the intense staining of smooth muscle in blood vessels or soft tissues. In hemangiopericytomas, capillaries in the centers of tissue whorls were sometimes stained. Tissue sections of 10 hemangiosarcomas previously shown to possess strong immunoreactivity with MAb 3B5 were predigested with 0.19'0 trypsin for 20 minutes at room temperature. Treatment with trypsin resulted in the loss of all immunoreactivity in all 10 tumors.

Reactivity of MAb 3B5 also was evaluated in IP assays (formalin-fixed, paraffin-embedded sections) against melanomas and primary carcinomas originat- ing from five different tissue types (10 tumors evaluated per type): intestinal adenocarcinomas (large intestine); lung carcinomas (bronchiolar-alveolar); mammary carcinomas; squamous cell carcinomas of the skin; and perianal gland carcinomas. Immuno- staining was observed in four intestinal and four lung carcinomas (4Oo/o), and in two mammary and two squa- mom cell carcinomas (20%); no staining was detected

in perianal gland carcinomas or melanomas. With the intestinal carcinomas, focal immunostaining was present in 10-20% of neoplastic cells in two tumors and in about 5% of tumor cells in two tumors. The staining was diffusely distributed in the cytoplasm and was usually intense (3+ ). Intense cytoplasmic staining was present in about 50% of the tumor cells in one lung carcinoma and in 10-20% of the cells in three other lung carcinomas. Immunostaining in the two mammary carcinomas was diffusely distributed throughout the tissue sections and was moderate to intense in 75- 80% of the neoplastic cells. Secreted material in glan- dular lumina also stained in these two tumors and in some tumors in which carcinoma cells failed to stain. Reactivity in squamous cell carcinomas was focal in one tumor, with about 20% ofthe cells intensely stained, and was diffuse in a second tumor, with 75-80% of the cells showing intense cytoplasmic staining. In all tissue sections of carcinomas and melanomas evaluated, staining of blood vessels and smooth muscle cells of soft tissue was consistently present as previously de- scribed. The lining of lymphatics in the intestine also was reactive. The control MAb assayed simultaneously with MAb 3B5 demonstrated no reactivity in the tissues.

Normal adult canine tissues obtained from six dogs, prepared both as formalin-fixed, paraffin-embedded sections and frozen sections, were analyzed in IP assays for immunoreactivity with MAb 3B5 (Table 3). The reactivity was determined to be nearly equivalent in




A B C D E F G H I A B C D E

116- ww

97- ”

66-

45- -5 6 -54

31- - 21- _. 14- LL,.

7-

4 Fig. 4 Immunoblot. Mesothelioma (PE) cell lysate re-

acted with MAb 3B5; dog. One-microgram samples of total protein in lanes (B-I) were electrophoretically separated on a 5-1 5% SDS-polyacrylamide gel, transferred overnight to nitrocellulose, and incubated with MAb 3B5. Time for color development: lane B, 1 minute; C, 2 minutes; D, 3 minutes; E, 5 minutes; F, 8 minutes; G, 10 minutes; H, 15 minutes; I, 20 minutes. Prestained molecular mass standards were loaded onto lane A: 200 kd, myosin; 1 16 kd, (3-galactosidase; 97 kd, phosphorylase B; 66 kd, bovine serum albumin; 45 kd, ovalbumin, 3 1 kd, carbonic anhydrase; 2 1 kd, trypsin inhibitor; 14 kd, lysozyme; 7 kd, aprotinin. A major protein band at 78 kd, already apparent at 1 minute of color development, increased in intensity over time. Minor bands at 56 and 54 kd, barely discernible in the photograph at 20 minutes (lane I), were first observed at 8 minutes of color development.

fixed and frozen tissue samples. Moderate to intense (2-3+ ) staining was consistently observed in pleural and peritoneal mesothelial cell linings, smooth muscle of soft tissues and blood vessels, and pulmonary alveolar capillaries. Endothelial cells were stained in about 10% of veins and arteries, irrespective of the tissue, and heterogeneity of staining was usually observed within individual vessels. Epidermis and adnexal epithelia of two dogs and colonic mucosal epithelia of one dog showed diffuse, moderate (2+ ) staining in formalin-fixed samples only.

Immunoblot analysis

Immunoblot analyses were used to determine the molecular weights of antigen@) recognized by MAb 3B5 in canine cell or tissue lysates of PE (mesothelioma) cells, hemangiosarcoma, urinary bladder, lung,

-- - 78- - 56-

5 Fig. 5. Immunoblot. Normal or neoplastic cell or tissue

lysates reacted with MAb 3B5; dog. Lane A: 1 pg total protein of PE cells. Lane B: 30 pg of brain. Lane C: 30 pg of urinary bladder. Lane D: 30 wg of lung. Lane E: 30 pg of hemangiosarcoma. Electrophoretic separation and immunoblotting were performed as in Fig. 4. Color was developed for 10 minutes. The major 78-kd band was demonstrable in lysates of PE cells, hemangiosarcoma, urinary bladder, and lung but not of brain. The lanes positive for the 78-kd band at 10 minutes also contained faint 56- and 54-kd bands at 20 minutes of color development.

and brain. Total protein in samples subjected to elec- trophoresis was 1-50 pg. With 1-pug samples, only the PE cell lysate showed immunoreactivity with MAb 3B5. A major band was observed at 78 kd within 1 minute of color development, and minor bands at 56 and 54 kd were apparent at 8-15 minutes of color development (Fig. 4). Molecular masses assigned to the three bands are mean values calculated from seven immunoblot analyses. In 30-pg samples, the 78-kd band was demonstrated in hemangiosarcoma, urinary bladder, and lung at 10 minutes of color development (Fig. 5) , and the 56- and 54-kd bands were observed in each sample by 20 minutes (data not shown). In contrast, 30-pg samples of brain showed no immunoreactivity and 50-pug samples yielded only the 56-kd band.

The immunoreactivity of MAb 3B5 with lysates of PE cells as presented previously was compared with




that of lysates in which the PE cells had been treated with 0.1% trypsin for 60 minutes at 37 C. No bands were observed in blots derived from lysates of trypsin- treated cells.

Discussion

Monoclonal antibody 3B5, generated with canine mesothelioma cells as immunogen, was shown to bind with high sensitivity to cytoplasmic antigen expressed by canine mesothelial cells. Immunohistochemical staining of formalin-fixed, paraffin-embedded tissue sections revealed intense staining of normal mesothelial cell linings and of reactive mesothelial cells and mesotheliomas. Although different biologic types of mesothelial cells cannot be distinguished by this antibody, it possesses some utility in histopathologic in- terpretation of serosal lesions. It can be useful in iden- tifying reactive mesothelial cells in proliferative lesions and in distinguishing epithelial mesotheliomas from carcinomas. Moreover, by virtue of its failure to stain fibroblasts, fibrosarcomas, and stroma, the antibody can be useful in distinguishing sarcomatoid mesotheliomas from sarcomas or reactive fibrosis. Although not evaluated in this study, MAb 3B5 also may have applications in exfoliative cytology for identification of mesothelial cells. Depending on the diagnostic prob- lem, it might be advantageous to utilize MAb 3B5 in parallel with staining for vimentin or cytokeratins, intermediate filaments known to be expressed in canine rneso the l i~mas .~~

In initial immunohistochemical studies of canine tissue sections to evaluate MAb 3B5 immunoreactivity against mesothelial cells, specificity of antigen-antibody binding also was observed in endothelial cells, vascular smooth muscle, and smooth muscle of soft tissues. The staining was intense, with little or no background staining. These observations prompted the examination of formalin-fixed, paraffin-embedded sections of benign and malignant tumors of endothelial cell or smooth muscle cell origin for immunoreactivity with MAb 3B5. Somewhat surprisingly, there was a distinct difference in sensitivity of MAb 3B5 reactivity among these tumor types; 78% of hemangiosarcomas (1 8 of 23) were positive compared with i 20% of hemangiomas, leiomyosarcomas, or leiomyomas. One hemangiosarcoma that was negative in formalin-fixed sections was strongly positive in frozen sections, sug- gesting that the sensitivity with hemangiosarcomas may even have been substantially higher in frozen sections. With a level of sensitivity at around 80% or higher with hemangiosarcomas, combined with little or no background staining, MAb 3B5 might be superior to anti-factor VIII antibodies currently available commercially for enhancing diagnostic acuity relevant to

canine hemangiosarcoma. Hemangiopericytomas, the histogenesis of which remains c o n t r ~ v e r s i a l , ~ ~ , ~ ~ were consistently negative for im munostaining when reacted with MAb 3B5, except for the staining of capillaries at the center of the tissue whorls characteristic of this tumor.

Immunoblot analyses of lysates prepared from PE (mesothelioma) cells, hemangiosarcoma, urinary bladder, and lung demonstrated that MAb 3B5 recognized antigens of the same molecular masses (78, 56, and 54 kd) in the tumor samples and normal tissues. The predominant reactivity was with the 78-kd antigen in each sample. Correlating the immunohistochemical staining in these samples with the immunoblot analyses, the antigens detected in the immunoblots were apparently derived from mesothelioma cells in the PE cell sample, malignant endothelial cells in the hemangiosarcoma, smooth muscle in urinary bladder, and alveolar capillaries in lung. Normal endothelial cells and vascular smooth muscle in the hemangiosarcoma and urinary bladder and peribronchial smooth muscle in lung also were stained in immunohistochemical assays, but they represented very minor components. The mesothelioma cells appeared to contain much more of the major 78-kd antigen than did the other tissues, based on comparative reactivity in microgram quan- tities of total protein analyzed in immunoblotting, but the sample of mesothelioma cells harvested from pleural effusion was much more homogeneous than the tissue samples, which made conclusive comparisons impossible. In immunoblots of brain, only the 56-kd antigen was demonstrated at the highest total protein concentration evaluated (50 wg). Because blood vessels were the only component that stained in immunohistochemical assays of brain, they apparently were the source of the 56-kd antigen detected in the immunoblots.

Except for molecular weight determinations, defin- itive studies have not been performed thus far regard- ing characterization of the antigen recognized by MAb 3B5. However, the observations that treatment of hemangiosarcoma tissue sections and mesothelial cell lysates with trypsin completely abrogated the immunoreactivity of MAb 3B5 in immunohistochemical and immunoblot analyses, respectively, suggest that the epitopes are protein. Consistent with these observations, preliminary studies in vitro have shown the antigen to be highly sensitive to treatment with trypsin and proteases. The three antigens reactive with MAb 3B5 may be subunits of the same proteins or different proteins that share a common epitope. In very limited immunohistochemical studies of mesotheliomas and other tissues from a variety of species, it appears that the antigen recognized by MAb 3B5 may be canine species specific.



Vet Pathol 31:6, 1994 Anti-canine Mesothelioma Monoclonal Antibody Immunoreactivity 67 1

Human thrombomodulin, a glycoprotein receptor for thrombin, has a molecular mass (75 kd) and distribution pattern10,50 similar to those of the antigen reactive with MAb 3B5. By immunohistochemical staining, thrombomodulin has been demonstrated in human vascular endothelial cells, angiosarcoma cells, mesothelial cells, 100% of mesotheliomas, and a low percentage of certain carcinomas (e.g., squamous cell carcinomas). However, thrombomodulin is a membrane antigen, in contrast to the cytoplasmic antigen recognized by MAb 3B5.

Intratumoral heterogeneity with regard to antigen expression in tumor cells has been well documented in other ~ y s t e m s ’ ” ~ ~ , ~ ~ ~ ~ ~ and was demonstrated in this study with immunostaining by MAb 3B5 (Fig. 2). Het- erogeneous antigenic expression by tumor cells that appear identical morphologically may be related to different functional states or levels of differentiation. l 2

Tissue fixation and processing of specimens may be contributing factors. However, the antigenic determi- nants reactive with MAb 3B5 are apparently very sta- ble in formalin-fixed specimens; some of the strongest reactions were observed with tissue blocks stored for more than 10 years. Antigenic heterogeneity of vascular endothelium from different sites in the cardio- vascular system and between capillaries, veins, or arteries also has been In this study, immunoreactivity with MAb 3B5 was observed in all types of blood vessels and in all tissues examined. However, the percentage of veins and arteries that contained endothelial cell staining was rarely greater than 10% in a given tissue, and the staining of endothelial cells within individual vessels was heterogeneous.

The use ofMAb 3B5 does not appear to be warranted for immunodiagnosis of canine carcinomas. Much higher levels of sensitivity and greater specificity than achievable with MAb 3B5 have been reported with other ant ibodie~.~J~ Results obtained thus far with MAb 3B5 suggest several immunohistochemical applications in which this antibody could assist the pathologist in diagnostic interpretations. These applications include diagnosis of mesotheliomas and hemangiosarcomas, cellular identifications in proliferative serosal lesions, and demonstration of vascularity or angiogenesis.

Acknowledgements

We express our sincere appreciation for the assistance provided by Ms. Misako Hwang and Ms. Beth Landreth in the preparation of tissue specimens and by Dr. Jianyi Wang in the generation of hybridoma 3B5. This report is published as Publication No. 2458, College of Veterinary Medicine, Auburn University.

References

1 Aida Y , Okada K, Amanuma H: Phenotype and ontog- eny of cells carrying a tumor-associated antigen that is expressed on bovine leukemia virus-induced lympho- sarcoma. Cancer Res 53:429-437, 1993

2 Al-Nafussi A, Carder PJ: Monoclonal antibodies in the cytodiagnosis of serous effusions. Cytopathology 1: 1 19- 128, 1990

3 Andreasen CB, Mahaffey EA, Duncan JR: Intermediate filament staining in the cytologic and histologic diagnosis of canine skin and soft tissue tumors. Vet Pathol25:343- 349, 1988

4 Baak JPA, Chin D, Van Diest PJ, Ortiz R, Matze-Cok P, Bacus SS: Comparative long-term prognostic value of quantitative HER-2heu protein expression, DNA ploidy, and morphometric and clinical features in paraffin-embedded invasive breast cancer. Lab Invest 64:

5 Benchimol S, Fuks A, Jothy S, Beauchemin N, Shirota K, Stanners CP: Carcinoembryonic antigen, a human tumor marker, functions as an intercellular adhesion molecule. Cell 57:327-334, 1989

6 Benz EW: Intermediate filament proteins: a molecular basis for tumor diagnostics. BioTechniques 3:4 12-420, 1985

7 Blobel GA, Moll R, Franke WW, Kayser KW, Gould VE: The intermediate filament cytoskeleton of malignant mesotheliomas and its diagnostic significance. Am J Pathol 121:235-247, 1985

8 Brown PJ: Immunohistochemical localization of myo- globin in connective tissue tumors in dogs. Vet Pathol

9 Clemo FAS, DeNicola DB, Delaney LJ: Immunohis- tochemical evaluation of canine carcinomas with monoclonal antibody B72.3. Vet Pathol 30: 140-145, 1993

10 Collins CL, Ordonez NG, Schaefer R, Cook CD, Xie SS, Granger J, Hsu PL, Fink L, Hsu SM: Thrombomodulin expression in malignant pleural mesothelioma and pulmonary adenocarcinoma. Am J Pathol 141:827-833, 1992

11 Daste G, Serre G, Mauduyt MA, Vincent C, Caveriviere P, Soleihavoup J: Immunophenotyping of mesothelial cells and carcinoma cells with monoclonal antibodies to cytokeratins, vimentin, CEA and EMA improves the cytodiagnosis of serous effusions. Cytopathology 2: 19-28, 1991

12 Dempsey PJ, Feleppa FP, Brown RW, de Kretser TA, Whitehead RH, Jose DG: Development of monoclonal antibodies to human breast carcinoma cell line PMC 42. J Natl Cancer Inst 77: 1-9, 1986

13 Destexhe E, Lespagnard L, Degeyter M, Heymann R, Coignoul F Immunohistochemical identification of myoepithelial, epithelial, and connective tissue cells in canine mammary tumors. Vet Pathol30: 146-154, 1993

14 Esteban JM, Battifora H: Tumor immunophenotype: comparison between primary neoplasm and its metas- tases. Mod Pathol 2: 192-1 97, 1990

15 Fidler IJ: Tumor heterogeneity and the biology ofcancer invasion and metastasis. Cancer Res 38:265 1-2660, 1978

2 15-223, 199 1

24:573-574, 1987




16

17

18

19

20

21

22

23

24

25

26

27

28

29

Gangopadhyay A, Wolfe LG: Conserved antigen expression in epithelial tumors recognized by monoclonal antibody 4A9 generated against canine mammary carcinoma cells. Hybridoma 8:175-186, 1989 Greiner JW, Horan Hand P, Colcher D, Weeks M, Thor A, Noguchi P, Pestka S, Schlom J: Modulation ofhuman tumor antigen expression. J Lab Clin Med 109:244-26 1, 1987 Griffey SM, Madewell BR, Dairkee SH, Hunt JE, Naydan DK, Higgins RJ: Immunohistochemical reactivity of basal and luminal epithelium-specific cytokeratin antibodies within normal and neoplastic canine mammary glands. Vet Pathol 30:155-161, 1993 Haines DM, Matte G, Wilkinson AA, Noujaim AA, Tur- ner C, Longenecker BM: Immunohistochemical staining and radionuclide imaging of canine tumors, using a monoclonal antibody recognizing a synthetic carbohy- drate antigen. Am J Vet Res 50:875-881, 1989 Hellmen E, Lindgren A: The expression of intermediate filaments in canine mammary glands and their tumors. Vet Pathol 26:420-428, 1989 Hendrix MJC, Seftor EA, Chu YW, Seftor REB, Nagle RB, McDaniel KM, Leong SPL, Yohem KH, Leibovitz AM, Meyskens F'L, Conaway DH, Welch DR, Liotta LA, Stetler-Stevenson W: Coexpression of vimentin and keratins by human melanoma tumor cells: correlation with invasive and metastatic potential. J Natl Cancer Inst 84:

Horan Hand P, Nuti M, Colcher D, Schlom J: Definition ofantigenic heterogeneity and modulation among human mammary adenocarcinoma cell populations using monoclonal antibodies to tumor associated antigens. Cancer Res 43:728-735, 1983 Hsu SM, Hsu PL, Zhao X, Chien Song KS, Jacqueline WP: Establishment of human mesothelioma cell lines (MS-1, -2) and production of a monoclonal antibody (anti-MS) with diagnostic and therapeutic potential. Can- cer Res 48:5228-5236, 1988 Hsu SM, Raine L, Fanges H: Use of avidin-biotin-peroxidase complex (ABC) in immunoperoxidase techniques: a comparison between ABC and unlabeled antibody (PAP) procedures. J Histochem Cytochem 29:

Kohler G, Milstein C: Continuous cultures of fused cells secreting antibody of predefined specificity. Nature 256: 495497, 1975 Kuhlmann L, Berghauser KH, Schaffer R: Distinction of mesothelioma from carcinoma in pleural effusions: an immunocytochemical study on routinely processed cy- toblock preparations. Pathol Res Pract 187:46747 1, 199 1 Laemmli UK: Cleavage of structural proteins during the assembly of the head of bacteriophage T,. Nature 227:

Lindsay DS, Dubey JP, Blagburn BL: Characterization of monoclonal antibodies generated against Toxoplasma gondzz tachyzoites. J Ala Acad Sci 61:144, 1990 Lloyd RV, Sisson JC, Shapiro B, Verhofstad AM: Im- munohistochemical localization of epinephrine, norepi- nephrine, catecholamine-synthesizing enzymes, and

165-174, 1992

577-580, 198 1

680-685, 1970

chromogranin in neuroendocrine cells and tumors. Am J Pathol 125:45-54, 1986

30 Lowseth LA, Gillett NA, Chang IY, Muggenburg BA, Boecker BB: Detection of serum a-fetoprotein in dogs with hepatic tumors. J Am Vet Med Assoc 199:735-74 1, 1991

31 Markwell MAK, Hass SM, Bieber LL, Tolbert NE: A modification of the Lowry procedure to simplify protein determination in membrane and lipoprotein samples. Anal Biochem 87:206-2 10, 1978

32 McDonough SP, MacLachlan NJ, Tobias AH: Canine pericardial mesothelioma. Vet Pathol29:256-260, 1992

33 McNutt MA, Bolen JW, Gown AM, Hammar SP, Vogel AM: Co-expression of intermediate filaments in human epithelial neoplasms. Ultrastruct Pathol 9:3 1-44, 1985

34 Moore AS, Madewell BR, Lund JK: Immunohistochem- ical evaluation of intermediate filament expression in canine and feline neoplasms. Am J Vet Res 50:88-92, 1989

35 Oliver JL 111, Wolfe LG: Antigen expression in canine tissues, recognized by a monoclonal antibody generated against canine melanoma cells. Am J Vet Res 53: 123- 128, 1992

36 Page C, Rose M, Yacoub M, Pigott R: Antigenic heterogeneity of vascular endothelium. Am J Pathol 141:

37 Patnaik AK, Lieberman PH: Gross, histologic, cytochemical, and immunocytochemical study of medullary thyroid carcinoma in sixteen dogs. Vet Pathol 28:223- 233, 1991

38 Porter PL, Bigler SA, McNutt M, Gown AM: The immunophenotype of hemangiopericytomas and glomus tumors, with special reference to muscle protein expression: an immunohistochemical study and review of the literature. Mod Pathol4:46-52, 199 1

39 Pulley LT, Stannard AA: Tumors of the skin and soft tissues. In: Tumors in Domestic Animals, ed. Moulton JE, 3rd ed., pp. 23-87. University of California Press, Berkeley, CA, 1990

40 Sandusky GE, Wightman KA, Carlton WW: Immuno- cytochemical study of tissues from clinically normal dogs and of neoplasms, using keratin monoclonal antibodies. Am J Vet Res 52:613-618, 1991

4 1 Schlom J: Antibodies in cancer therapy: basic principles of monoclonal antibodies. In: Biologic Therapy of Can- cer, ed. De Vita VT Jr, Hellman s, and Rosenberg SA, pp. 464-48 1. J.B. Lippincott, Philadelphia, PA, 199 1

42 Stahel RA, O'Hara CJ, Waibel R, Martin A: Monoclonal antibodies against mesothelial membrane antigen dis- criminate between malignant mesothelioma and lung adenocarcinoma. Int J Cancer 41:218-223, 1988

43 Steplewski Z, Jeglum KA, Rosales C, Weintraub N: Ca- nine lymphoma-associated antigens defined by murine monoclonal antibodies. Cancer Immunol Immunother

44 Thor A, Ohuchi N, Szpak CA, Johnston WW, Schlom J: Distribution of oncofetal antigen tumor-associated glycoprotein-72 defined by monoclonal antibody B 72.3. Cancer Res 46:3 1 18-3 124, 1986

45 Towbin H, Staehelin T, Gordon J: Electrophoretic trans-

673-683, 1992

24: 197-20 1, 1987




fer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc Natl Acad Sci USA 76:4350-4354, 1979

46 Trigo FJ, Morrison WB, Breeze RG: An ultrastructural study of canine mesothelioma. J Comp Pathol 91:53 1- 537, 1981

47 Urban JL, Schreiber H: Tumor antigens. Annu Rev Im- munol 10:6 17-644, 1992

48 Wu YJ, Parker LM, Binder NE, Beckett MA, Sinard JH, Griffiths CT, Rheinwald J G The mesothelial keratins: a new family of cytoskeletal proteins identified in cul- tured mesothelial cells and nonkeratinizing epithelia. Cell

49 Yokoyama WM: Production of antibodies. In: Current 31~693-703, 1982

Protocols in Immunology, ed. Colligan JE, Kruisbeek AM, Margulies DH, Shevach EM, and Strober W, vol. 1, pp. 2.4.1-2.5.17. Greene/Wiley-Interscience, New York, NY, 1992

50 Yonezawa S, Marurama I, Sakae K, Igata A, Majerus PW, Sat0 E: Thrombomodulin as a marker for vascular tumors: comparative study with factor VIII and Ulex europaeus I lectin. Am J Clin Pathol 88:405-411, 1987

5 1 Zarbo RJ, Gown AM, Nagle RB, Visscher DW, Cnssman JD: Anomalous cytokeratin expression in malignant melanoma: one- and two-dimensional western blot analysis and immunohistochemical survey of 100 melanomas. Mod Pathol 4:494-501, 1990

Request reprints from Dr. L. G. Wolfe, Department of Pathobiology, College of Veterinary Medicine, Auburn University, Auburn, AL 36849-5519 (USA).



Antigen Expression in Normal and Neoplastic Canine Tissues Defined by a Monoclonal Antibody...

Documents

Transcript of Antigen Expression in Normal and Neoplastic Canine Tissues Defined by a Monoclonal Antibody...