Generation of multiple monoclonal antibodies for diagnostic use from a single hybridoma fusion.

BALB/c mice were exposed simultaneously to three nonrelated immunogens, myelin basic protein, uridyl-galactosyl transferase, and tissue obtained from a formalin-fixed, paraffin-embedded block containing a pilomatrixoma. Standard hybridoma techniques were used and antibody generation assayed using an unlabelled antibody biotin-avidin method with sections of human cerebellum, liver, and pilomatrixoma as the substrates. Using the above assay, clones were selected that secreted antibodies with selective specificities for each of the immunogens. By ELISA assay, the monoclonal antibodies reacting with cerebellum also reacted to the myelin basic protein preparation used as immunogen and the monoclonal antibody reacting to hepatocytes bound to the preparation of uridyl-galactosyl transferase used as immunogen. Our data suggests that the generation of monoclonal antibodies with a variety of diagnostic applications can be obtained from a single fusion following immunization with multiple nonrelated antigens, requiring considerably less laboratory cost and effort than would be required to obtain similar monoclonal antibodies in separate fusions.

Monoclonal antibody ERY-1 identifies an antigen in erythroid cells, hepatocellular and renal cell carcinomas.

We have identified a monoclonal antibody (ERY-1), which reacts with erythrocytes, erythroid precursor cells, and with embryonal yolk sac, and normal liver and kidney. The antibody also decorates the neoplastic cells of hepatocellular, renal, and yolk sac carcinomas. No reactivity was seen in a variety of other epithelial or mesenchymal neoplasms. It is possible that ERY-1 recognizes an erythropoiesis-associated antigen present in yolk sac, kidney, liver, and bone marrow, all of which are involved in erythropoiesis in various stages of human development. Furthermore, ERY-1 has proved to be extremely useful in the histopathologic diagnosis of hepatocellular and renal cell carcinomas.

Intestinal metaplasia of the gallbladder: a morphologic and immunocytochemical study.

The morphologic spectrum of intestinal metaplasia was studied in 49 gallbladders that had been excised because of cholelithiasis. Based on the absence or presence of endocrine cells, the cases of intestinal metaplasia were arbitrarily divided into two groups. The gallbladders from the first group (26 cases) contained isolated or small clusters of mature goblet cells, while those from the second group (23 cases), in addition to the goblet cells, contained argyrophil and argentaffin cells and, less frequently, Paneth cells and gland-like structures similar to colonic crypts. Pseudopyloric glands and superficial gastric-type epithelium were present in both groups. Argyrophil cells outnumbered argentaffin cells by a ratio of 4 to 1. By immunocytochemical methods serotonin-containing cells were found to be the most common endocrine cells. Other endocrine cells showed immunoreactivity for somatostatin, cholecystokinin, gastrin, and pancreatic polypeptide. The presence of gut endocrine cells and Paneth cells in the pseudopyloric glands suggests that these glands are also an integral component of intestinal metaplasia of the gallbladder. The findings support the hypothesis that cholelithiasis induces the appearance of a stem endodermal cell that, in turn, may differentiate into cells with mature intestinal or gastric phenotypes.

Immunohistologic identification of fungi in systemic and cutaneous mycoses.

Using specific antibodies and the peroxidase-antiperoxidase technique, we were able to demonstrate a variety of fungal organisms in smears and sections of formaldehyde-fixed, paraffin-embedded tissue. The procedure is simple, fast, and accurate and may be used as an alternative to, or in conjunction with, cultural methods to identify fungi specifically.

Delineation of nevus cell nests in inflammatory infiltrates by immunohistochemical staining for the presence of S100 protein.

Fifteen halo nevi were stained for the presence of S100 protein by an unlabelled antibody peroxidase-antiperoxidase method. S100 protein was clearly identifiable within nevus cell nests in the inflammatory infiltrate. The presence of this substance helped to identify nevus cells in dense inflammatory infiltrates and confirm the histologic diagnosis of halo nevus.

The value of carcinoembryonic antigen in differentiating sclerosing epithelial hamartoma from syringoma.

We examined 4 cases of sclerosing epithelial hamartoma for the presence of carcinoembryonic antigen (CEA) by the unlabelled antibody peroxidase-antiperoxidase method but could not detect CEA in these lesions. Because syringomas contain CEA, the detection of this antigen is a useful marker for differentiating sclerosing epithelial hamartomas from syringomas.

Prekeratin immunohistochemistry in the diagnosis of undifferentiated carcinoma of the nasopharyngeal type.

We investigated the value of prekeratin immunostaining in establishing the diagnosis of undifferentiated carcinoma of the nasopharyngeal type (lymphoepithelioma). As in squamous cell carcinoma of the nasopharynx, all seven lymphoepithelial carcinomas stained for prekeratin whereas other look-alike but histogenetically different neoplasms (malignant lymphoma, esthesioneuroblastoma, and small round cell sarcomas) did not stain. In addition to confirming the squamous epithelial nature of lymphoepithelioma, our findings indicated that immunoperoxidase staining for prekeratin is a useful diagnostic tool in the differential diagnosis of nasopharyngeal carcinomas in general and lymphoepithelioma in particular.

Prekeratin in spindle cell tumors of the skin.

A variety of benign and malignant skin lesions was stained for the presence of prekeratin using an unlabeled antibody peroxidase-antiperoxidase method and an antibody raised against human prekeratin protein. No prekeratin could be detected in benign or malignant lesions derived from melanocytes, and prekeratin could not be found in atypical fibroxanthoma, Kaposi's sarcoma, angiosarcoma, dermatofibrosarcoma protruberans, and leiomyosarcoma. Paget's cells did not contain prekeratin. Varying staining intensities were observed in keratinocytic atypical hyperplasias. Of particular interest is the observation that many spindle cell lesions that were difficult to identify with 100% assurance from routine stains could be positively identified as squamous cell carcinoma after prekeratin was found in the atypical cells.

Granular cell tumors of the skin contain myelin basic protein.

Myelin basic protein, a substance found in neural structures, has been demonstrated in cutaneous granular cell tumors using a monoclonal antibody generated against myelin basic protein and an immunoperoxidase method. The substance was not found in fibrohistiocytic skin lesions. The presence of myelin basic protein in granular cell lesions of the skin supports the concept that this lesion is related closely to nerve structures.

Kaposi's sarcoma: immunoperoxidase staining for cytomegalovirus.

An immunoperoxidase technique was used to test for cytomegalovirus (CMV) proteins in lesions of Kaposi's sarcoma from 19 patients with acquired immune deficiency syndrome (AIDS) and 3 patients without AIDS. CMV proteins could not be detected in any of the cases of Kaposi's sarcoma. Thus, while CMV genetic material may be present in these cells, the virus is not actively directing the synthesis of proteins. This eliminates reactivation of latent CMV in tumor tissue, or late infection of Kaposi's sarcoma cells as explanations for the elevated antibody titers to CMV and other evidence of CMV infection in patients with Kaposi's sarcoma. However, the possibility remains that exposure to CMV occurs as a primary event in the development of Kaposi's sarcoma. These findings indicate that, as in Kaposi's sarcoma of the elderly, the virus is inactive in Kaposi's sarcoma cells of patients with AIDS, and there is no difference between the two lesions in immunoperoxidase staining for CMV proteins.

Paget's disease of the skin. A unifying concept of histogenesis.

Utilizing the peroxidase antiperoxidase technique we explored the presence and the distribution of carcinoembryonic antigen (CEA) in seven cases of mammary and sixteen of extramammary Paget's disease of skin. In every case positive immunostaining was observed in all the Paget's cells and the underlying tumor (where one was present), whereas intervening keratinocytes and melanocytes did not stain. CEA was also present in the cells and secretions of normal eccrine and apocrine glands. Our observations confirm that Paget's cells are of glandular origin.

Carcinoembryonic antigen in sweat-gland carcinomas.

Using standard immunoperoxidase techniques, four cases of malignant sweat-gland carcinoma were examined for the presence of carcinoembryonic antigen (CEA). This substance was universally present in these lesions in spite of their wide spectrum of histologic features. CEA was mainly localized within the cytoplasm of tumor cells and in the lumina of neoplastic glands. The findings indicate that the presence of CEA in malignant skin adnexal tumors could help to identify and categorize these lesions as derived from eccrine or apocrine adnexal epithelium.

Immunochemical demonstration of immunoreactive insulin in human breast cancer.

The immunoreactive insulin contents of the cytosol extracts of 57 human mammary carcinomas were measured by specific radioimmunoassay. All of these extracts contained measurable insulin (detection limit 1.3 ng/g tumor). In some cases (23) the levels were quite greater than the normal human serum levels. Immunoperoxidase technique revealed the presence of immunoreactive insulin within the cytoplasm of tumor cells in 6 cases.

Tissue immunoreactive prolactin hormone in breast cancer.

The prolactin contents of the cytosol extracts of 159 human mammary carcinomas was measured by specific radioimmunoassay. Of these extracts 157 contained measurable prolactin (detection limit, 7.8 ng/g tumor). The levels measured were in most cases greater than would be expected to be due to only serum contamination. The levels of prolactin did not correlate with either estrogen or progesterone receptor content of the tumors, or menopausal status of the patient.