A tumor-associated fibronectin isoform generated by alternative splicing of messenger RNA precursors.

Fibronectin (FN) represents the mixture of a number of structurally different molecules (isoforms) whose make-up varies depending on the FN sources. FN from cultured transformed human cells has a very different isoform composition with respect to its normal counterpart. In fact, SV-40-transformed WI-38VAI3 human fibroblasts produce high levels of a FN isoform (B-FN) which is very poorly expressed in their normal, WI-38, counterpart. We have recently demonstrated that the B-FN isoform derives from a differential splicing pattern of the FN primary transcript which leads, in transformed cells, to a high level expression of the exon ED-B (Zardi, L., B. Carnemolla, A. Siri, T. E. Petersen, G. Paolella, G. Sebastio, and F. E. Baralle. 1987. EMBO (Eur. Mol. Biol. Organ.) J. 6:2337-2342). Here we report on the production and characterization of a monoclonal antibody (BC-1) which recognizes an epitope within the protein sequence coded for by the ED-B exon. This monoclonal antibody makes it possible to carry out immunohistochemical analysis of the distribution of the ED-B-containing FN isoform (B-FN) in human tissues. The results show that while in normal, adult, human tissues total FN has a widespread distribution, the B-FN isoform is restricted only to synovial cells, to some vessels and areas of the interstitium of the ovary, and to the myometrium. On the contrary, the B-FN isoform has a much greater expression in fetal and tumor tissues. These results demonstrate that, in vivo, different FN isoforms have a differential distribution and indicate that the B-FN isoform may play a role in ontogenesis and oncogenetic processes.

Phenotyping of lesions of melanocyte origin with monoclonal antibodies to melanoma-associated antigens and to HLA antigens.

The antigenic profiles of a large number of surgically removed human benign and malignant lesions of melanocyte origin have been analyzed with the use of monoclonal antibodies (MoAb) against la antigens, against the HLA-A,B,C-beta 2-microglobulin molecular complex, against a cytoplasmic melanoma-associated antigen (MAA), and against membrane-bound MAA. Membrane-bound MAA include a high-molecular-weight MAA (HMW-MAA), a 115K MAA, and a 100K MAA. Appearance of the HMW-MAA and of the cytoplasmic MAA, as well as cytoplasmic distribution or loss of HLA-A,B,C antigens, occurs in benign lesions. Additional appearance of Ia antigens is associated with malignant transformation of melanocytes. The antigenic profile defined by the battery of MoAb used displays differences among benign lesions of different histogenesis, between benign and malignant lesions, and among malignant lesions with different histopathologic properties. These results suggest that phenotyping of surgically removed lesions with anti-MAA and anti-HLA MoAb may contribute to the understanding of the steps involved in tumor progression of melanocytes and may aid in the diagnosis of lesions with unusual histopathologic features.

A 94,000-dalton glycoprotein expressed by human melanoma and carcinoma cells.

The IgG2a monoclonal antibody (MoAb) 376.96S, secreted by a hybridoma derived from a mouse immunized with cultured human melanoma cells COLO 38, reacts with a single 94,000-dalton glycoprotein that is peripherally associated with the plasma cell membrane of cultured melanoma cells. Indirect immunofluorescence analysis with cryostat thin sections of human tissues showed that this antigen is absent from a large variety of normal tissues but is readily detectable on melanomas, nevi, and several different carcinomas. The MoAb 376.96S binds with cultured melanoma and carcinoma cell lines to a similar extent and can mediate both complement-dependent and cell-dependent lysis of these cells. The 94,000-dalton glycoprotein detected by MoAb 376.96S is distinct in its tissue distribution, antigenicity, and molecular profile from several structures previously identified with monoclonal antibodies that have similar molecular weights.

Antigenic heterogeneity of skin tumors of nonmelanocyte origin: analysis with monoclonal antibodies to tumor-associated antigens and to histocompatibility antigens.

Surgically removed benign and malignant human skin lesions of nonmelanocyte origin have been tested with monoclonal antibodies to la antigens, to the HLA-A,B antigenic molecular complex, and to melanoma-associated antigen(s) (MAA). MAA include a high-molecular-weight (HMW) MAA, a 115,000-molecular-weight MAA, a 94,000-molecular-weight MAA, and a cytoplasmic MAA. Indirect immunofluorescence was used as the assay system because of the limited amount of tissue available. When the amount of tissue available was sufficient, double determinant immunoassays (DDIA) were used to quantitate the level of the HMW MAA and of the cytoplasmic MAA. The results of the DDIA were in agreement with those of indirect immunofluorescence in more than 75% of the cases. Malignant skin tumors of various histiotypes displayed three types of changes: 1) appearance of la antigens and cytoplasmic MAA, 2) increase in the level of the HMW MAA, of a 115,000- and a 100,000-molecular-weight MAA, and 3) reduction in the level of HLA-A,B antigens and beta 2-microglobulin. A significant heterogeneity was found in the antigenic profile among various lesions of a given histiotype as well as among tumor cells within a given lesion.

Structural properties and tissue distribution of the antigen recognized by the monoclonal antibody 653.40S to human melanoma cells.

The hybridoma 653.40S, constructed with splenocytes from an inbred BALB/c mouse immunized with cultured human melanoma cells, secreted an antibody that had been shown to recognize an antigenic determinant restricted to human melanoma cell lines. The monoclonal antibody (MoAb) 653.40S showed immunoprecipitation of two glycopolypeptides synthesized by melanoma cells, one with the apparent molecular weight of 280,000 and the other one with a molecular weight larger than 500.000. These two glycopolypeptides were not bridged by disulfide bonds and were peripheral rather than integral to the plasma cell membrane. Comparison of the reactivity of cells of the melanocyte lineage with the MoAb 653.40S and with the MoAb Q5/13 to human Ia-like antigens showed that the former reacted with proliferating melanocytes and melanoma cells, whereas the latter reacted only with melanoma cells. The MoAb 653.40S did not react with a large variety of surgically removed normal and tumor tissues except for some instances of basal cell and squamous cell carcinomas. These results suggested that double staining of pigmented skin lesions with the MoAb 653.40S and with an MoAb to Ia-like antigens may help to solve controversial diagnosis of melanoma. Furthermore, the MoAb 653.40S may be useful for radioimaging and immunotherapy of melanoma.

Tissue distribution, molecular profile, and shedding of a cytoplasmic antigen identified by the monoclonal antibody 465.12S to human melanoma cells.

The mouse immunoglobulin G2 monoclonal antibody (MoAb) 465.12S reacts with a cytoplasmic antigen present in human melanoma cells but not detectable in melanocytes. Indirect immunofluorescent staining of a large number of surgically removed normal adult and fetal tissues with the MoAb 465.12S detected the cytoplasmic antigen in epithelial cells from several organs. The intensity of staining was greater in adult tissues than in the corresponding fetal tissues. Furthermore, the MoAb 465.12S stained nearly all of the surgically removed tumors tested but did not stain many of the normal tissues from which they originated. In almost all cases, the intensity and frequency of staining wa greater for tumor cells than for corresponding normal tissues. From cultured carcinoma and melanoma cells, the MoAb 465.12S immunoprecipitated four glycopolypeptides with molecular weights of 94,000, 75,000, 70,000, and 25,000. Incorporation of 3H-labeled sugars into the various components of the cytoplasmic antigen revealed that the M.W. 75,000 component was unusual in that it contained only glucosamine and mannose. The antigenic determinant defined by the MoAb 465.12S appears to be protein rather than carbohydrate in nature since it is heat sensitive and is expressed on the antigens synthesized by cells in presence of tunicamycin. Analysis of the spent culture medium of carcinoma and melanoma cell lines revealed that the cytoplasmic antigen is readily shed by these cells and consists of a major M.W. 94,000 and a minor M.W. 72,000 component. Treatment of cultured melanoma cells with the antibiotic tunicamycin showed that glycosylation of the cytoplasmic antigen is required for its shedding and/or stability in the spent culture medium.

Expression of CAR-3 and TAG-72 macromolecules in normal and transformed endometrium: potential diagnostic application in postmenopausal patients.

Mass cytoscreening for the early diagnosis of endometrial carcinoma has thus far been hampered by the low diagnostic accuracy of current cytopathology. In this study we have analyzed the reactivity of the two monoclonal antibodies, AR-3 and B72.3, recognizing two distinct glycosylated high molecular weight carcinoma associated antigens on histological specimens from normal, hyperplastic, and transformed endometrium with the aim of establishing their diagnostic potential. Because women with a high risk of endometrial cancer are frequently postmenopausal, where normal endometrium is characterized by atrophy and cystic glandular hyperplasia, the following findings were of interest. Both antibodies reacted with variable and apical staining patterns with a minority of specimens of normal cycling endometrium from premenopausal women. However, they were constantly negative when tested on normal atrophic postmenopausal endometrium, and only monoclonal antibody AR-3 occasionally stained glandular cystic hyperplasia. By contrast, lesions with atypical hyperplasia, which represent a preneoplastic condition, were stained by both antibodies in 89 or 67% of the cases depending on the monoclonal antibody used (100% if used in combination). Furthermore, 98% of the endometrial carcinomas tested were found to react with the combination of two monoclonal antibodies. If these findings are confirmed in a multicentric study, the use of the two reagents could be a valuable adjunct in the cytodiagnosis of endometrial cancer, especially in providing a guideline to selecting patients for endometrial curettage and additional diagnostic procedures.

Transformation of thyroid epithelium is associated with loss of c-kit receptor.

The receptor for the stem cell factor encoded by the c-kit proto-oncogene is expressed by a number of epithelial cells including thyrocytes. Since malignant transformation may be associated with loss of this receptor (melanoma and breast cancer), we have analyzed its expression in benign (38 cases) and malignant (31 cases) thyroid lesions. While low levels of c-kit are expressed in normal thyroids and in 60% of benign lesions, the receptor is undetectable in 60 and 90% of the follicular and papillary carcinomas, respectively. Northern blot analysis from surgical specimens of carcinomas and from carcinoma cell lines has demonstrated a lack of specific c-kit transcripts. These findings indicate that the c-kit receptor may be involved in the growth control of thyroid epithelium and that this function may be lost following malignant transformation.

Expression of c-kit receptor in normal and transformed human nonlymphoid tissues.

The protooncogene c-kit encodes a tyrosine kinase with a molecular weight of 145,000, highly related to the platelet derived growth factor/colony stimulating factor receptors. Mutations of the murine gene result in impairment of hematopoiesis, gametogenesis, and of the melanocyte cell lineage. In order to elucidate c-kit functions in development and oncogenesis we have analyzed immunohistochemically its expression in human normal and transformed nonlymphoid tissues. The receptor has been detected in spermatogonia, melanocytes, and unexpectedly, in astrocytes, renal tubules, parotid cells, thyrocytes, and breast epithelium. While the gene product is expressed in seminoma, lung tumors, and melanoma of low invasiveness, no detectable levels have been detected in thyroid and breast carcinomas, astrocytomas, and invasive melanomas. In breast tumors these findings were confirmed by paired, Northern blot analysis of RNA preparations from normal and transformed tissue. The present results demonstrate that the c-kit receptor plays a role in the development of a larger spectrum of cell lineages. Furthermore, on the basis of the transformation associated changes, we speculate that, while in some cell types, c-kit expression positively regulates mitogenesis and is selected for in neoplastic transformation, in other tissues the c-kit pathway is involved in morphogenesis and differentiation and is, therefore, negatively selected in the course of tumor progression.

Heterogeneity in the expression of HLA and tumor-associated antigens by surgically removed and cultured breast carcinoma cells.

Surgically removed normal and malignant mammary tissues and human breast carcinoma cell lines were tested in binding assays with monoclonal antibodies to HLA-A,B,C antigens, beta 2-microglobulin, HLA-DR antigens, and tumor-associated antigens; the latter included a Mr 280,000, a Mr 94,000, and a Mr 85,000 membrane-bound glycoprotein and a cytoplasmic antigen. HLA-A,B antigens, beta 2-microglobulin, HLA-DR antigens, and the cytoplasmic antigen are expressed by normal mammary cells. Their malignant transformation may be associated with quantitative changes in the expression of these antigens and with the appearance of Mr 94,000 and Mr 85,000 glycoproteins. The Mr 280,000 glycoprotein was detected on only one of the breast carcinoma cell lines tested. Analysis of primary tumors and autologous axillary lymph node metastasis from 13 patients has shown differences in the expression of all the antigens tested between primary and metastatic lesions.

Distribution of human Class I (HLA-A,B,C) histocompatibility antigens in normal and malignant tissues of nonlymphoid origin.

Indirect immunofluorescence and immunoperoxidase staining of surgically removed tissues of nonlymphoid origin with monoclonal antibodies to the heavy and light chain of HLA-A,B,C antigens have shown that they have a more restricted tissue distribution than previously assumed. HLA-A,B,C antigens were not detected in brain cortex, cerebellum, sympathetic ganglia, hypophysis, parathyroid gland, thyroid, exocrine pancreas, hepatocytes, sperm, seminiferous tubules, or skeletal or smooth muscle. Malignant transformation of cells may be associated with appearance, changes in cellular distribution of HLA-A,B,C antigens, and/or dissociation in the expression of the two subunits. Analysis of primary tumors and of autologous metastases showed heterogeneity in the expression of HLA-A,B,C antigens among lesions removed from different sites. The degree of heterogeneity did not correlate with the site of origin of metastases.

Analysis of the antigenic profile of uveal melanoma lesions with anti-cutaneous melanoma-associated antigen and anti-HLA monoclonal antibodies.

The reactivity of 12 surgically removed uveal melanoma lesions with monoclonal antibodies (MoAb) to 14 membrane-bound and 2 cytoplasmic cutaneous melanoma-associated antigens (MAA), to the 2 subunits of HLA Class I antigens and to the gene products of the HLA-D region was compared with that of cutaneous melanoma lesions and correlated with their histiotype. The membrane-bound determinants defined by the anti-Mr 92,000 and 45,000 MAA MoAb TP39.1, anti-Mr 110,000 MAA MoAb M111, anti-Mr 118,000 MAA MoAb TP36.1, anti-Mr 115,000 MAA MoAb 345.134, anti-ICAM-1 MoAb CL203.4 and anti-Mr 31,000 MAA MoAb M2590, and the cytoplasmic determinants defined by the anti-MAA MoAb 465.12 and 2G-10 display a distribution in uveal melanoma lesions similar to that in cutaneous melanoma lesions. On the other hand, membrane-bound determinants defined by the anti-Mr 100,000 MAA MoAb 376.96, anti-9-O-acetyl-GD3 ganglioside MoAb ME311 and anti-GD2-GD3 ganglioside MoAb ME361 were not detected in the uveal melanoma lesions tested. Furthermore, the membrane-bound determinants defined by the anti-GD3 MoAb R24, anti-nerve growth factor receptor MoAb ME20.4, anti-Mr 97,000 MAA MoAb 140.240, anti-carcinoembryonic antigen MoAb B1.1 and anti-HMW-MAA 149.53, 225.28, and 763.74 have a markedly lower expression in uveal than in cutaneous melanoma lesions. Incubation of uveal melanoma lesions with the pool of the MoAb 149.53, 225.28, and 763.74 recognizing distinct and spatially distant determinants of the HMW-MAA increased the intensity of staining of six lesions and stained four lesions which were not stained by the individual monoclonal antibodies. The distribution of HLA Class I antigens in uveal melanoma lesions resembles that in cutaneous melanoma lesions, since they are expressed in all the lesions of the mixed and epithelioid type but were not detected in those of the spindle type, i.e., the counterparts of nevocellular nevi. HLA Class II antigens are expressed with a lower frequency in uveal than in cutaneous melanoma lesions, since they were detected only in 2 of the 12 lesions. One of them is of the mixed type and the other one of the epithelioid type. Besides HLA antigens the determinants defined by the anti-carcinoembryonic MoAb B1.1, anti-ICAM-1 MoAb CL203.4, and anti-GD3 MoAb R24 displayed a differential distribution in the different histiotypes of uveal melanoma, since they are preferentially expressed in lesions of the mixed and epithelioid type.

Differential expression of intercellular adhesion molecule 1 in primary and metastatic melanoma lesions.

Immunochemical studies have shown that the monoclonal antibody (MoAb) CL203.4, elicited with immune interferon treated cultured human melanoma cells Colo 38, recognizes intercellular adhesion molecule 1 (ICAM-1). The determinant defined by MoAb CL203.4 is distinct and spatially distant from that defined by anti-ICAM-1 MoAb RR1/1, which had been elicited with Epstein-Barr virus-transformed B-lymphocytes from a lymphocyte function associated antigen 1 deficient patient. Immunohistochemical testing with MoAb CL203.4 of surgically removed lesions of melanocyte origin has shown a markedly lower reactivity with benign than with malignant lesions. Among the latter, a higher percentage of metastatic than of primary lesions was stained by MoAb CL203.4. The higher expression of ICAM-1 in metastases than in primary lesions is unique to melanoma, since no difference was found in its distribution in primary and metastatic lesions of a variety of malignancies of different embryological origin. Reactivity with MoAb CL203.4 of primary lesions removed from patients with stage I melanoma showed a highly significant correlation with the lesion thickness and with the clinical course of the disease. The disease free interval in patients without detectable reactivity of their primary lesion with MoAb CL203.4 was significantly (P = 0.004) longer than that of patients whose primary lesion was stained with MoAb CL203.4. These results suggest that ICAM-1 may be a useful marker in the analysis of the molecular mechanism underlying the association between lesion thickness and clinical course of the disease.

Heterogeneous expression of melanoma-associated antigens and HLA antigens by primary and multiple metastatic lesions removed from patients with melanoma.

Indirect immunofluorescence staining with a large battery of monoclonal antibodies of primary and autologous metastatic lesions removed from seven patients with melanoma has detected heterogeneity in the expression of various types of melanoma-associated antigens (MAAs), of distinct determinants of the high molecular weight melanoma-associated antigen (HMW-MAA), of the two subunits of Class I HLA antigens, and of the gene products of the HLA-D region. Among the 10 MAAs tested, the HMW-MAA had the highest frequency and the Mr 87,000 MAA the lowest. Furthermore, the HMW-MAA displayed the lowest heterogeneity. These findings, in conjunction with the restricted tissue distribution of the HMW-MAA, its lack of susceptibility to antibody-mediated modulation, and the high affinity of the available anti-HMW-MAA monoclonal antibodies, indicate that this antigen may be a useful marker for radioimaging and immunotherapy in patients with melanoma. The common acute lymphoblastic leukemia antigen was detected only in five lesions. Class I HLA antigens were detected in a larger number of lesions than HLA-DR antigens, which had a significantly higher frequency than HLA-DQ antigens. The degree of antigenic heterogeneity did not appear to correlate with the histopathological features of the lesions and/or with the clinical course of the disease. The results of the present study indicate that immunodiagnostic and immunotherapeutic approaches to melanoma should rely on the use of combinations of monoclonal antibodies to distinct MAAs.

Distribution of a cross-species melanoma-associated antigen in normal and neoplastic human tissues.

In previous studies the monoclonal antibody (MoAb) M2590 elicited in C57/BL6 mice with the syngeneic melanoma cell line B16 has been shown to recognize a 31K glycoprotein expressed by human melanoma cell lines. The present study has shown that the MoAb M2590 cross-reacts with surgically removed benign and malignant lesions of melanocyte origin. The reactivity pattern of the MoAb M2590 with these lesions is different from that of the anti-high-molecular weight-melanoma associated antigen (HMW-MAA) MoAb 225.28S, of the anti-115K MAA MoAb 345.134S, and of the anti-100K MAA MoAb 376.96S, which were elicited with human melanoma cell lines. In particular, the MoAb M2590 reacts with blue nevi. The MoAb M2590-defined MAA, like other types of MAA, is heterogeneous in lesions removed from different patients, in autologous lesions removed from different anatomic sites, and in cells within a lesion. The distribution of the MoAb M2590-defined MAA in normal tissues and in tumors of nonmelanocyte origin is broader than that of the HMW-MAA, but is similar to that of the 115K MAA and of the 100K MAA. The results of this investigation suggest that immunization with xenogeneic melanoma cells may broaden the range of specificity of antihuman MAA MoAbs and provide information about the phylogenetic evolution of MAAs.

Gene products of the HLA-D region in normal and malignant tissues of nonlymphoid origin.

Indirect immunofluorescence staining with monoclonal antibodies has shown a differential distribution of HLA-DR, DQ, and DP antigens in normal tissues of nonlymphoid origin. The distribution of HLA-DP antigens is similar to that of HLA-DR antigens, while that of HLA-DQ antigens is more restricted. Malignant transformation of cells of nonlymphoid origin may be associated with the appearance of the gene products of the HLA-D region. HLA-DR antigens appear more frequently than the other two types of HLA class II antigens and HLA-DP antigens more frequently than HLA-DQ antigens. Differential expression of the gene products of the HLA-D region was also found in autologous metastases removed from different anatomic sites from patients with melanoma. The HLA class II phenotype of surgically removed malignant lesions did not correlate with the degree of differentiation of tumor cells and/or with the expression and/or cellular distribution of HLA class I antigens. Furthermore, in melanoma lesions, no relationship was found between the HLA class II phenotype and the expression of 3 membrane bound and 1 cytoplasmic melanoma associated antigen recognized by monoclonal antibodies. The functional significance and the practical implications of the differential expression of the gene products of the HLA-D region by tumor cells are discussed.

Immunocytochemical diagnosis of amelanotic metastatic melanoma using monoclonal antibodies HMB-45 and Ep1-3.

The analysis of fine-needle aspirates and effusions has proved to be a valuable tool for the cytological identification of metastatic melanoma. Nevertheless, while malignant pigmented cells can be easily detected on cytological specimens, their identification may be very difficult in patients bearing amelanotic lesions. In the present study we have evaluated whether this limitation can be overcome using two monoclonal antibodies (mAbs) HMB45 and Ep1-3, which are highly specific for the melanocyte lineage. These reagents were assayed in immunocytochemical tests performed on cytological material obtained from 50 patients with a past history of melanoma and 463 bearing metastases from a cryptic primary tumour. These mAbs, employed in combination with a panel of monoclonal reagents with well-defined tumour specificity, may improve the accuracy of the conventional cytopathological diagnosis of melanoma metastases in the two groups of patients.

Expression of glutathione transferase pi in benign and malignant lesions of the melanocyte lineage.

Intrinsic and acquired resistance to chemotherapeutic agents represents the major clinical obstacle in the control of most tumours. In vitro studies have established that multiple mechanisms, including changes in drug uptake and efflux and in detoxifying enzymes, are responsible for drug resistance. Among the latter, glutathione S transferases (GST) have been recognized to play a relevant role. In the present study we have evaluated GST pi immunohistochemically as well as enzymatically in benign and malignant primary and metastatic lesions of the melanocyte lineage. A parallel analysis of the multiple drug resistance (MDRI) gene product was performed in a representative number of specimens. Results of this study demonstrate that while GST pi is constitutively expressed by the melanocyte lineage, independently from the transformed stage, MDRI p-glycoprotein is detected with a significantly lower frequency. These findings clearly indicate that GST pi represents the major detoxifying metabolic pathway of the melanocyte lineage and may be responsible for the high degree of inherent resistance of malignant melanoma to available cytostatic treatments.

Vla-3 distribution in normal and neoplastic non-lymphoid human tissues.

Using monoclonal antibody (mAb) M-Kid 2 to the alpha 3 beta 1 heterodimer, we have evaluated immunohistochemically the in vivo expression of the Vla-3 integrin in normal and transformed non-lymphoid human tissues. In normal tissues the alpha 3 beta 1 complex displays a polarized distribution at the baso-lateral aspect of most keratinizing and glandular epithelia. In addition the integrin is detected in perineurium, basal lamina of smooth muscular fibers, vascular media, podocytes and Bowman's capsule, myoepithelial cells of the parotid and breast, and in pulmonary alveoli. Neoplastic transformation is associated with qualitative and quantitative changes in expression of this integrin. The loss of polarized distribution often occurs in various malignancies. Furthermore, a significant decrease in expression occurs in 13% of the colon-rectum carcinomas, 75% of the ductal invasive, and 40% of the lobular invasive breast carcinomas. Among the lung malignancies tested, the small cell lung carcinomas (SCLC) were found to be consistently unreactive with mAb M-Kid 2. Analysis of Vla-3 expression in established tumor cell lines demonstrated that the integrin is almost invariably expressed by the plastic adherent cell subpopulations.