Primary structure of the human melanoma-associated antigen p97 (melanotransferrin) deduced from the mRNA sequence.

p97 is a cell-surface glycoprotein that is present in most human melanomas but only in trace amounts in normal adult tissues. To determine the structure of this tumor-associated antigen and to identify its functional domains, we have purified and cloned p97 mRNA and determined its nucleotide sequence. The mRNA encodes a 738-residue precursor, which contains the previously determined N-terminal amino acid sequence of p97. After removal of a 19-residue signal peptide, the mature p97 molecule comprises extracellular domains of 342 and 352 residues and a C-terminal 25-residue stretch of predominantly uncharged and hydrophobic amino acids, which we believe acts as a membrane anchor. Each extracellular domain contains 14 cysteine residues, which form seven intradomain disulfide bridges, and one or two potential N-glycosylation sites. Protease digestion studies show that the three major antigenic determinants of p97 are present on the N-terminal domain. The domains are strikingly homologous to each other (46% amino acid sequence homology) and to the corresponding domains of human serum transferrin (39% homology). Conservation of disulfide bridges and of amino acids thought to compose the iron binding pockets suggests that p97 is also related to transferrin in tertiary structure and function. We propose that p97 be renamed melanotransferrin to denote its original identification in melanoma cells and its evolutionary relationship to serotransferrin and lactotransferrin, the other members of the transferrin superfamily.     

Monoclonal antibodies that demonstrate specificity for several types of human lung cancer.

Monoclonal antibodies with selectivity for human lung cancer were produced by immunizing BALB/c mice with an established line of human small cell lung cancer (NCI-H69) and fusing the mouse spleen cells to mouse myeloma line X63-Ag8.653. The resulting hybrid cells were initially screened by immunoautoradiography for production of antibodies that would react with NCI-H69 and another small cell lung cancer line (NCI-H128) but not its autologous B-lymphoblastoid line (NCI-H128BL). Stable monoclonal antibody-producing lines were isolated by repeated cloning. Three independently derived monoclonal antibodies, designated 525A5, 534F8, and 538F12, were found to react with three of the major types of human lung cancer (small cell, adenocarcinoma, and squamous carcinoma). They did not react with bronchioloalveolar and large cell lung cancers, myeloma, lymphomas, leukemias, osteogeneic sarcoma, mesothelioma, hypernephroma, malignant melanoma, simian virus 40-transformed human fetal lung cells, skin fibroblast lines, human B-lymphoblastoid lines, human erythrocytes, and rodent cells. Interestingly, these antibodies also bound to three out of three human neuroblastomas and two out of three breast cancers but failed to react with mouse neuroblastoma and rat pheochromocytoma. The monoclonal antibodies reacted with human small cell lung cancer tumors obtained at autopsy, but had insignificant reactions with normal human lung, liver, spleen, and skeletal muscle. We conclude that monoclonal antibodies have been generated that react with common antigenic determinants expressed on several human lung cancer types, neuroblastoma, and some breast cancers, but are not detectable by our current assays on a variety of other human tumors or normal adult human tissues. Such antibodies are of potential clinical and biological importance.     

Identification of a cell surface protein, p97, in human melanomas and certain other neoplasms.

BALB/c mice were immunized with a human melanoma cell line, SK-MEL 28, and their spleen cells were fused with mouse NS-1 myeloma cells. Hybrid cells were tested in an indirect 125I-labeled protein A assay for production of antibodies that bound to surface antigens of SK-MEL 28 melanoma cells but not to autologous skin fibroblasts. One hybridoma, designated 4.1, had the required specificity. It was cloned and grown in mice as an ascites tumor. The monoclonal IgG1 antibody produced by the hybridoma was purified from the ascites fluid and labeled with 125I. The labeled antibody bound, at significant levels, to approximately 90% of the melanomas tested and to approximately 55% of other tumor cells, but not to three B-lymphoblastoid cell lines or to cultivated fibroblasts from 15 donors. Immunoprecipitation and sodium dodecyl sulfate gel electrophoresis were used to detect the target antigen in 125I-labeled cell membranes of both cultivated cells and tumor biopsy samples. A protein with a molecular weight of 97,000 was identified. This protein, designated p97, was present in both cultured cells and biopsy material from melanomas and certain other tumors, but it was not detected in eight different samples of normal adult epithelial or mesenchymal tissues obtained from five donors.     

Serological analysis of Melan-A(MART-1), a melanocyte-specific protein homogeneously expressed in human melanomas.

Recent progress in the structural identification of human melanoma antigens recognized by autologous cytotoxic T cells has led to the recognition of a new melanocyte differentiation antigen, Melan-A(MART-1). To determine the properties of the Melan-A gene product, Melan-A recombinant protein was produced in Escherichia coli and used to generate mouse monoclonal antibodies (mAbs). Two prototype mAbs, A103 and A355, were selected for detailed study. Immunoblotting results with A103 showed a 20-22-kDa doublet In Melan-A mRNA positive melanoma cell lines and no reactivity with Melan-A mRNA-negative cell lines. A355, in addition to the 20-22-kDa doublet, recognized several other protein species in Melan-A mRNA-positive cell lines. Immunocytochemical assays on cultured melanoma cells showed specific and uniform cytoplasmic staining in Melan-A mRNA-positive cell lines. Immunohistochemical analysis of normal human tissues with both mAbs showed staining of adult melanocytes and no reactivity with the other normal tissues tested. Analysis of 21 melanoma specimens showed homogenous staining of tumor cell cytoplasm in 16 of 17 Melan-A mRNA-positive cases and no reactivity with the three Melan-A mRNA-negative cases.     

Changes in cell surface antigens during stem cell ontogeny

A panel of monoclonal antibodies that bind to the murine pluripotential stem cell CFU-s was used to examine the antigenic profile of the stem cell during ontogeny. The results show that the stem cell surface changes dramatically during development. One group of three independently derived monoclonal antibodies binds to subpopulations (50%-70%) of stem cells at plateau values, and these populations increase marginally during development. A second group of four monoclonal antibodies, including anti-H-2Kk (11-4.1), define stem cell antigens that increase from low levels in the fetal liver to high levels in adult bone marrow. The presence of these two classes of antigens on adult splenic stem cells was in general similar to that observed on adult bone marrow. Antigens defined by the first group of monoclonal antibodies were present in similar amounts on CBA, C57B1/6, and Balb/c bone marrow stem cells, whereas antigens of the second group showed mouse strain variations. Quantitative absorption analysis was used to distinguish H-2Kk (11-4.1) from 9F6, which showed a similar developmental profile. Monoclonal antibodies recognizing subpopulations of stem cells were shown to be distinct by complementation studies and recognized antigens not present on brain tissue.     

Murine monoclonal anti-idiotope antibody breaks unresponsiveness and induces a specific antibody response to human melanoma-associated proteoglycan antigen in cynomolgus monkeys.

The mouse monoclonal antibody MEM136 (mAb1) is directed against an epitope on human melanoma-associated proteoglycan antigen (MPG). This epitope is also present on various normal human and subhuman tissues. A monoclonal murine anti-idiotope (anti-Id) antibody (mAb2), designated I-Mel-2, was generated against MEM136 and used as a surrogate antigen for the MPG molecule. I-Mel-2 was tested in cynomolgus monkeys (Macaca fascicularis) for its ability to induce anti-MPG humoral responses. All monkeys immunized with Ab2 developed specific anti-anti-idiotype (Ab3) responses that were capable of inhibiting binding of Ab2 to Ab1. Furthermore, I-Mel-2 immune monkey serum contained anti-MPG antibodies (Ab1') that bound to MPG-positive but not to MPG-negative melanoma cell lines. Monkeys immunized with Colo38 melanoma cells (membrane-bound MPG antigen) did not contain anti-MPG antibodies that inhibited the binding of two distinct anti-MPG mAb 125I-labeled MEM136 or 125I-labeled 225.28 to Colo38 cells. The induction of anti-MPG responses in monkeys did not cause any apparent side effects in animals, despite the fact that the MPG antigen is expressed by many normal tissues. The affinity-purified, I-Mel-2 idiotype-specific, Ab3 immunoprecipitated MPG antigen from melanoma cells. Furthermore, the I-Mel-2-induced Ab3 inhibited melanoma cell invasion in an in vitro assay, implying that these antibodies have biological significance.     

Monoclonal antibodies inhibit the adhesion of mouse B 16 melanoma cells in vitro and block lung metastasis in vivo.

Seven monoclonal antibodies against mouse B 16 melanoma cells (produced in syngeneic C57BL/6 mice) were selected that blocked the adhesion of melanoma cells to tissue culture dishes. These antibodies were found to be directed against antigens on the surface of mouse B 16 melanoma cells but not on normal mouse cells such as 3T3 fibroblasts. Similarly, the antigens could not be detected in normal mouse tissues (e.g., lung, kidney, liver) but were found in lungs colonized by B 16 melanoma cells. Significantly, three of these antibodies virtually abolished lung colonization of highly invasive B 16 sublines injected into the animals' bloodstream. They exerted their effect both when preabsorbed by the melanoma cell in vitro and when delivered to the animals prior to the tumor cells. It is suggested that monoclonal antibodies might be a promising tool for preventing metastasis.     

Serological survey of normal humans for natural antibody to cell surface antigens of melanoma.

Sera of 106 normal adult men were tested for antibodies reacting with cell surface antigens of three established lines of cultured malignant melanoma. Positive reactions with a protein A assay for IgG antibodies were extremely rare (1-2%). The frequency of positive reactions with assays for IgM antibodies was higher: 5-15% in immune adherence assays and 55-82% in anti-C3 mixed hemadsorption assays. After low-titered sera and sera reacting with fetal calf serum components, conventional alloantigens, and widely distributed class 3 antigens were excluded, sera from seven individuals (one with IgG antibody and six with IgM antibodies) were selected for detailed analysis. The serum containing the IgG antibody came from a healthy 65-year-old Caucasian man; titers of antibody in his serum ranged from < 1/10 to 1/40,000 in tests with different melanoma cell lines. This IgG antibody identifies a differentiation antigen of melanocytes, provisionally designated Mel 1, that distinguishes two classes of melanomas: 22 melanoma cell lines typed Mel 1+ and 17 types Mel 1-. Mel 1 is expressed by fetal fibroblasts but not adult fibroblasts and can be found on a proportion of cultured epithelial cancer cell lines (5 out of 23) but not on glioma or B-cell lines. The melanoma antigens detected by the naturally occurring IgM antibodies are serologically unrelated to Mel 1 but, like Mel 1, appear to be differentiation antigens that distinguish subsets of melanoma. These IgM antibodies detect antigens that are identical or closely related to the AH antigen, a melanoma surface antigen that was initially defined by autologous antibody in a patient with melanoma. In view of the immunogenicity of both Mel 1 and the AH antigens in humans and their occurrence on more than 50% of melanomas, it remains to be seen whether antibody to these antigens can be elicited by specific vaccination of seronegative melanoma patients and whether this will have an influence on the clinical course of the disease.     

Recombinant vaccinia virus vaccine against the human melanoma antigen p97 for use in immunotherapy.

We have constructed a recombinant vaccinia virus, v-p97NY, which expresses the human melanoma-associated glycoprotein p97. Immunization with v-p97NY could induce humoral and cell-mediated immunity to p97, including delayed-type hypersensitivity, in mice and in two of two monkeys (Macaca fascicularis). The fact that an immune response was induced also in monkeys is important because normal cells from monkeys, but not from mice, express a low level of cross-reactive p97. Mice immunized with v-p97NY rejected transplants of syngeneic mouse melanoma expressing p97. A rejection response could be detected also when immunization was started 2 days after tumor transplantation, irrespective of whether the transplanted cells grew subcutaneously or as lung metastases. Evidence was obtained that melanoma cells lacking p97 may be killed as "bystanders" at the site of an immune response to melanoma cells expressing p97.     

Induction of immunity to a human tumor marker by in vivo administration of anti-idiotypic antibodies in mice.

Anti-idiotypic antibodies are described that were raised against murine monoclonal antibody 8.2, an antibody specific for a human melanoma-associated cell surface marker called p97. The 8.2 idiotopes recognized by this anti-idiotypic antiserum are binding site-associated and are shared by other monoclonal anti-p97 antibodies with the same specificity as antibody 8.2. Mice immunized with the anti-idiotype demonstrate delayed-type hypersensitivity reactions when challenged with melanoma (p97-positive) tumor cells.     

Immunoperoxidase detection of pancreatic oncofetal antigen in human tissues

Normal human adult tissues (from the pancreas, stomach, small intestine, colon and rectum), human fetal pancreas and various human tumor tissues were investigated for the expression of pancreatic oncofetal antigen (POA) by the peroxidase-antiperoxidase complex method. The antibody used in this study was raised by immunization of rabbit with POA purified from ascites of a patient with pancreatic cancer; it was the same as in our previous report on enzyme immunoassay for serum POA. POA was localized in the ductal cells of human fetal pancreas and ductal cell type pancreatic cancer tissue. It was hardly seen in acinar cells of human fetal or adult pancreas, or in acinar cell type pancreatic cancer tissue. Several cancer tissues other than the pancreas revealed positive POA-staining but less frequently. POA was also detected in small amounts in the small duct of normal adult pancreas. In other normal human tissues, POA was recognized in deep crypts of the small intestine and colon. These results indicate that POA is present abundantly in immature ductal cells of the pancreas such as in fetal pancreas and pancreatic cancer, and that high serum levels of POA seen in patients with pancreatic cancer and other malignant tumors may originate from these malignant tissues.     

Analysis of two new leukemia-associated antigens detected on human T- cell acute lymphoblastic leukemia using monoclonal antibodies

Two monoclonal antibodies (anti-3-3 and anti-3-40) were produced, which identify two new leukemia-associated antigens. Both antibodies reacted with most cell lines derived from patients with T lymphoblastic leukemia (T-ALL), but were not detected on suspensions of normal hematopoietic cells (including thymocytes) by cytotoxicity, absorption, or indirect immunofluorescence assays. Analysis of fresh leukemic cells indicated that anti-3-3 only reacted with T-ALL cells, while anti-3-40 also reacted with some non-T, non-B ALL cells and a few acute myelocytic leukemia (AML) cells. The 3-40 antigen was also found histopathologically in frozen sections of several normal tissues, including the epithelial cells and a few lymphoid cells of the thymus, and some malignant tissues. The 3-3 antigen was not found in any tissue studied. A "double absorption"assay provided additional serologic evidence that the two antibodies identify different antigenic determinants. Biochemical analysis indicated that the molecules immunoprecipitated by anti-3-3 and anti-3-40 have molecular weights of 35,000-40,000 daltons. This study demonstrated that the 3-3 and 3-40 antigens are markers for human T-ALL and can be used along with the normal T-lymphocyte antigen, 3A1, to discriminate T-ALL from cutaneous T-cell lymphoma (CTCL), adult T-cell leukemia (ATL), and T-cell chronic lymphocytic leukemia (T-CLL).     

Class 1 (unique) tumor antigens of human melanoma: identification of unique and common epitopes on a 90-kDa glycoprotein.

Antibodies present in the serum of melanoma patient FD detect an antigenic determinant (FD) restricted to the autologous melanoma cell line SK-MEL-131. This cell-surface determinant is carried on a glycoprotein of 90 kDa, designated gp90. Mice were immunized with a partially purified preparation of gp90 derived from SK-MEL-131 clone 1.5, and three murine hybridomas (KF23, KF26, and KF104) secreting monoclonal antibodies (mAbs) detecting this antigen have been generated. Sequential immunoprecipitation experiments demonstrate that the three mAbs and human FD serum react with the same gp90 species. The mAbs, in contrast to FD serum, react with a broad range of cultured cells in assays for cell-surface antigens and immunoprecipitate a gp90 component from radiolabeled extracts of these cells, including autologous FD B cells. We conclude that gp90 from SK-MEL-131 has two types of determinants: restricted (detected by FD serum) and common (detected by the mouse mAbs). gp90 molecules from cell lines other than SK-MEL-131 carry only the common determinant(s). Immunoperoxidase analysis of frozen tissue sections with mAb KF23 demonstrated a restricted gp90 expression in normal tissues (capillary endothelial cells, duct epithelium in sweat glands, breast, and pancreas). Melanomas and sarcomas showed strong gp90 expression, suggesting up-regulation of gp90 synthesis in certain human cancers.     

Localization of the growth hormone receptor, identified by immunocytochemistry, in second trimester human fetal tissues and in placenta throughout gestation.

Pituitary GH secretion appears largely unnecessary for the attainment of normal birth size in many species, including man. This is believed to be due to an immaturity and/or an absence of GH receptors in many fetal tissues. However, in vitro studies using late first trimester human fetal tissues have demonstrated mitogenic actions of GH on liver and stimulation of insulin biosynthesis in pancreas. To resolve this discrepancy, we have employed immunocytochemistry to identify the presence and distribution of GH receptors in various human fetal tissues. Fetuses of 14-16 weeks gestation were obtained after therapeutic abortion, tissues were fixed, and immunocytochemistry was performed using monoclonal antibodies against purified rat or rabbit GH receptor. The specificity of staining was confirmed by preabsorption of the antibodies with 1) adult rat liver membranes or 2) human fetal liver membranes, both of which possess specific GH-binding sites, or 3) human fetal skeletal muscle membranes, which do not specifically bind GH. Positive staining was seen in a subpopulation of liver parenchymal cells, in the ductal and endocrine tissue of pancreas, in the germinal layer of the epidermis and the deeper dermal layers of skin, and in the tubular epithelium of kidney. No immunopositive staining was seen in skeletal or cardiac muscle, epiphyseal growth plate, lung, intestine, or adrenal. Positive staining was present in the neuronal cell bodies of the cerebral cortex. GH receptor was also detectable as early as 8 weeks gestation in syncytial layers of the placenta and was maintained until term. Results demonstrate the presence of immunoreactive GH receptor/binding protein in some human fetal tissues early in development. In particular, these results would support a role for GH in the growth and function of liver and pancreas.     

Reactivity with erythroid and non-erythroid tissues of a murine monoclonal antibody to a synthetic peptide having amino acid sequence common to cytoplasmic domain of human glycophorins C and D

Three synthetic peptides encompassing the entire cytoplasmic polypeptide sequence (amino acid residues 82-128) of glycophorin C (GPC) and glycophorin D (GPD) were used to immunize mice for the production of monoclonal antibodies (MoAbs). Only the synthetic peptide (GPC-peptidel) corresponding to C-terminal residues 112-128 elicited a MoAb (named BGRL-100) which could react with native and denatured GPC and GPD. We characterized BGRL-100 by inhibition using GPC-peptide 1 and red cell sialoglycoproteins. The ability of BGRL-100 to interact with native GPC and GPD was assessed by immunoprecipitation with normal red cells (RBCs), and with denatured GPC and GPD by Western blotting of both normal RBCs and RBCs carrying GPC variants. Immunohistochemical staining of human tissue sections was performed using both BGRL-100 and a rat MoAb (named BRAC-1), which is specific for an extracellular domain of GPC and GPD. Both antibodies showed strong staining of erythroid lineage haemopoietic cells in fetal liver, sinusoids of adult liver and RBCs in the blood vessels of all tissues tested. Neither antibody reacted with epithelia from a range of human tissues. However, both MoAbs stained neural tissue in a distinctive fibrillar pattern. This suggests the presence of an analogue of erythroid GPC in neural tissues.     

Human uptake and incorporation of an immunogenic nonhuman dietary sialic acid

Humans are genetically unable to produce the sialic acid N-glycolylneuraminic acid (Neu5Gc), because of a mutation that occurred after our last common ancestor with great apes. Although Neu5Gc is presumed absent from normal humans, small amounts have been claimed to exist in human tumors and fetal meconium. We have generated an antibody with high specificity and avidity for Neu5Gc. Fetal tissues, normal adult tissues, and breast carcinomas from humans showed reactivity to this antibody, primarily within secretory epithelia and blood vessels. The presence of small amounts of Neu5Gc was confirmed by MS. Absent any known alternate pathway for its synthesis, we reasoned that these small amounts of Neu5Gc might originate from exogenous sources. Indeed, human cells fed with Neu5Gc incorporated it into endogenous glycoproteins. When normal human volunteers ingested Neu5Gc, a portion was absorbed and eliminated in urine, and small quantities were incorporated into newly synthesized glycoproteins. Neu5Gc has never been reported in plants or microbes to our knowledge. We found that Neu5Gc is rare in poultry and fish, common in milk products, and enriched in red meats. Furthermore, normal humans have variable amounts of circulating IgA, IgM, and IgG antibodies against Neu5Gc, with the highest levels comparable to those of the previously known anti-alpha-galactose xenoreactive antibodies. This finding represents an instance wherein humans absorb and metabolically incorporate a nonhuman dietary component enriched in foods of mammalian origin, even while generating xenoreactive, and potentially autoreactive, antibodies against the same molecule. Potential implications for human diseases are briefly discussed.     

Hamster pancreas acinar cell post-differentiation antigen detected by murine monoclonal antibody

For identification of hamster acinar cells, murine monoclonal antibodies to dispersed adult hamster acinar cells were developed. One of these, Ham-Acl, with strong affinity for a membrane-associated adult acinar cell antigen, was purified by HPLC, isotyped as IgG1 and used for the characterization of a species-specific, immunoprecipitable post-differentiation antigen on adult acinar cells. This antigen of 98 kDa was identified on paraffin sections of adult hamster acinar cells by an indirect immunofluorescence technique. It was undetectable in fetal hamster pancreas, but was present on 2-week-old and older hamster acinar cells.     

Regulation of plasminogen activation: a role for melanotransferrin (p97) in cell migration

We recently reported that human recombinant melanotransferrin (p97) presents a high transport rate across the blood-brain barrier that might involve the low-density lipoprotein receptor-related protein (LRP). We now report new interactions between p97 and another LRP ligand, the urokinase plasminogen activator (uPA) complex. By using biospecific interaction analysis, both pro-uPA and plasminogen are shown to interact with immobilized p97. Moreover, the activation of plasminogen by pro-uPA is increased by soluble p97. Because the uPA system plays a crucial role in cell migration, both in cancer and in angiogenesis, we also measured the impact of both endogenous membrane-bound and exogenous p97 on cell migration. The monoclonal antibody L235 (which recognizes a conformational epitope on p97) inhibited the migration of human microvascular endothelial cells (HMECs-1) and of human melanoma SK-MEL-28 cells, indicating that endogenous membrane-bound p97 could be associated with this process. In addition, low concentrations of exogenous p97 (10 and 100 nM) inhibited HMEC-1 and SK-MEL28 cell migration by more than 50%. These results indicate that membrane-bound and soluble p97 affect the migration capacity of endothelial and melanoma cells and suggest that p97 could be involved in the regulation of plasminogen activation by interacting with pro-uPA and plasminogen.     

Hamster pancreas acinar cell post-differentiation antigen detected by murine monoclonal antibody

Summary For identification of hamster acinar cells, murine monoclonal antibodies to dispersed adult hamster acinar cells were developed. One of these, Ham-Acl, with strong affinity for a membrane-associated adult acinar cell antigen, was purified by HPLC, isotyped as IgG₁ and used for the characterization of a species-specific, immunoprecipitable post-differentiation antigen on adult acinar cells. This antigen of 98 kDa was identified on paraffin sections of adult hamster acinar cells by an indirect immunofluorescence technique. It was undetectable in fetal hamster pancreas, but was present on 2-week-old and older hamster acinar cells.     

Cell surface antigens of human malignant melanoma: definition of six antigenic systems with mouse monoclonal antibodies.

Eighteen mouse monoclonal antibodies were selected for reactivity with cell surface antigens of the immunizing human melanoma cell line SK-MEL-28. Six distinct antigenic systems were defined by direct serological assays and absorption tests with a panel of 41 cell lines derived from normal and malignant human tissues. Biochemical analysis indicated that two of the antigens are glycoproteins with molecular sizes of 95,000 and 150,000 daltons (gp95 and gp150). Two other antigenic systems (O5 and the R24 group) are associated with heat-stable molecules having the characteristics of glycolipids. The remaining two antigens (M19 and R8) are heat labile, but molecular characterization has not been possible. Each of the antigenic systems has a distinctive pattern of distribution on various cell types, varying from a broad representation to a more restricted occurrence. O5 appears to be a species antigen, being present on virtually every human cell type tested. gp95, gp150, M19, and R8 are found on a characteristic proportion of melanomas, astrocytomas, and epithelial cancers and on normal kidney cells. The antigen defined by the R24 antibody has the most restricted distribution of all. Reactivity is found with melanomas and astrocytomas, whereas epithelial cell types, fibroblasts, and cells of hematopoietic origin lack R24. Although occurrence of gp95, gp150, M19, and R8 distinguishes a small subset of melanomas not expressing these antigens, R24 is found on all melanoma cells.