Search

To fulfil their function as APCs, dendritic cells (DC) and their precursors need to travel from blood to the peripheral tissues and, upon activation, migrate from tissues to draining lymph nodes. Because O-glycans play a role in T cell trafficking, we investigated the O-glycosylation profile of human monocyte-derived DC. Sialyl-Lewis(x) (sLe(x)), a glycan involved in extravasation via selectin binding, was found to be expressed exclusively on P-selectin glycoprotein ligand-1 in monocytes and immature DC. However, sLe(x) was lost from mature DC even though these cells retained expression of P-selectin glycoprotein ligand-1. Maturation of DC led to a rapid change in the expression of glycosyltransferases involved in O-linked glycosylation. A down-regulation of C2GnT1 mRNA and enzymatic activity was observed with a concurrent up-regulation of ST3Gal I and ST6GalNAc II mRNA resulting in a loss of the core 2 structures required for sLe(x) expression as a P-selectin ligand. Interestingly, the early regulation of these glycosyltransferases was mediated by PGE(2), which is known to be required for human DC migration. The pattern of O-glycosylation seen in mature cells was very similar to that expressed by naive T cells, which home to lymph nodes. Our data show that the regulation of O-glycosylation controls sLe(x) expression, and also suggest that O-glycans may have a function in DC migration.

In many carcinomas, infiltrating macrophages are commonly found closely associated with tumour cells but little is known concerning the nature or significance of adhesion molecules involved in these cellular interactions. Here we demonstrate in primary human breast cancers that sialoadhesin (Sn), a macrophage-restricted adhesion molecule, is frequently expressed on infiltrating cells that often make close contact with breast carcinoma cells. To determine whether Sn could act as a specific receptor for ligands on breast cancer cell lines, binding assays were performed with a recombinant form of the protein fused to the Fc portion of human immunoglobulin G1 (IgG1) (Sn-Fc). Sn-Fc was found to bind specifically and in a sialic acid-dependent manner to the breast cancer cell lines MCF-7, T47.D and BT-20 both in solid- and solution-phase binding assays. To investigate the nature of the sialoglycoproteins recognized by Sn on breast cancer cells, MCF-7 cells were labelled with [6-3H]glucosamine. Following precipitation with Sn-Fc, a major band of approximately 240000 MW was revealed, which was shown in reprecipitation and Western blotting experiments to be the epithelial mucin, MUC1.

We have investigated a mechanism of the regulation of mucin core polypeptide (MUC1) gene expression, which is induced by a soluble stimulatory factor, in KM12C human colon carcinoma cells. Conditioned media from normal human colon tissues elevated the level of expression of MUC1 mRNA. Transcriptional activation of the MUC1 gene was analyzed by transient expression of MUC1-CAT reporter plasmids containing the 5'-flanking sequence of the MUC1 gene fused to the bacterial chloramphenicol acetyltransferase (CAT) gene. A region between base pairs -531 and -520 of the 5'-flanking sequence of the MUC1 gene was sufficient for the induction of CAT activity by normal colon conditioned medium (NCCM). Mutagenesis of 3 base pairs within the region corresponding to sequence -531 to -517 from ACAGGGAGCGGTTAG to ACAGGGAGATTTTAG substantially decreased the induction of CAT activity by NCCM. Nuclear extracts from untreated or NCCM-treated KM12C cells were tested for their interaction with 32P-labeled oligonucleotides corresponding to this sequence. A specifically retarded band was identified after electrophoretic analysis. The quantity or mobility of this band was not changed by NCCM treatment. When an oligonucleotide with three point mutations was used as a competitor, the retarded band remained at the same position. This element (positions -531 to -520), which we call the responsive mucin element, does not contain any sequence that corresponds to previously described cis-acting elements. A protein component complexed with this sequence was identified with a molecular mass of approximately 70 kDa by SDS-polyacrylamide gel electrophoresis.

The product of the MUC1 gene, the polymorphic epithelial mucin (PEM), contains a large domain consisting of tandem repeats of 20 amino acids. Each repeat contains five potential sites for O-glycosylation, suggesting that this region forms a scaffold for the attachment of the O-linked carbohydrate which makes up more than 50% of the molecule. A number of monoclonal antibodies have been shown to recognize core-protein epitopes within the tandem repeat domain. One such antibody, SM3, reacts with the mucin expressed by breast carcinomas but shows little or no reaction with normal resting or lactating breast. Using primary mammary epithelial cells (HuME), an immortalized cell line derived from HuME (MTSV1-7) and a breast carcinoma cell line (BT20) the influence of the carbohydrate side chains on the binding of antibodies to core-protein epitopes has been investigated. We unequivocally show that the masking of the SM3 epitopes in normal breast epithelial cells is due to the carbohydrate side chains. In addition we demonstrate that the binding of two other antibodies (HMFG1, HMFG2) to core-protein epitopes is influenced by the carbohydrate side chains. The binding of HMFG-1 is particularly affected by sialic acid whereas the binding of HMFG2 is influenced by the length of the oligosaccharide side chains. Furthermore, inhibited elongation of O-linked carbohydrate side chains does not seem to interfere with the cell trafficking of the mucin to the cell surface.

The 5'-sequences flanking the human MUC1 gene have been analyzed for their ability to direct expression of a reporter gene (the chloramphenicol acetyltransferase gene (CAT)) in cell lines that normally express or do not express the MUC1 gene. A construct containing 2.9 kilobase pairs of MUC1 5'-flanking sequence sequence showed expression of CAT in breast and pancreatic cell lines but not in the non-epithelial cell lines HT-1080, SK23, and HTB96. Deletion analysis showed that maximum expression was obtained in ZR-75 (breast cancer line) and HPAF (pancreatic cancer line) with only 743 base pairs of 5'-flanking sequence. Sequences within 1.6 kilobase pairs of the transcriptional start site showed enhancing activity in a vector carrying an enhancerless SV40 promoter. Analysis of proximal 5'-sequences in a promoterless CAT vector carrying the SV40 enhancer showed that sequences between -60 and -150 were crucial for tissue-specific expression. An Sp1 site at -99/-90 and an E box (E-MUC1) at -84/-64 in this region were shown by mutational analysis to play a role in the regulation of transcription. Gel shift analysis with oligonucleotides and nuclear extracts of ZR-75 showed protein binding to both of these sites. Sp1 binding activity was similar in ZR-75 and HT1080 cells, whereas binding of factors to the E-MUC1 oligonucleotide revealed quantitative and qualitative differences between epithelial and non-epithelial cells.

The immune response of CAF1 mice to various synthetic peptides (SP) related to the amino acid sequence (PDTRPAPGSTAPPAHGVTSA) of the tandem repeat of the MUC1 human breast mucin core peptide was evaluated. The most immunogenic preparations of the synthetic peptides were those conjugated to keyhole limpet hemocyanin (KLH) or clustered in a dendritic multiple antigenic peptide (MAP-4) configuration. The mice were immunized subcutaneously with synthetic peptides emulsified in RIBI adjuvant, employing various immunization protocols. Equivalently high IgG responses were induced using SP-KLH conjugates (GVTSAPDTRPAPGSTA-KLH) or an SP--MAP-4 chimeric configuration (SP1-6), which also included a universal malarial CST-3 T-helper epitope (SP1-6 = SAPDTRPAEKKIAKMEKASSVFNVVNS--MAP-4). These IgG antibodies bound both the appropriate MUC1 synthetic peptides and the cell surface expressed MUC1 mucin on murine mammary cells that had been transfected with the human MUC1 gene and a human breast cancer cell line that expresses cell-surface MUC1. A MAP-4 molecule, which included the entire 20-amino-acid sequence of the MUC1 tandem repeat (SP1-5 = PDTRPAPGSTAPPAHGVTSA-MAP-4) induced a poor IgG response. In contrast, all three types of molecule: SP-KLH, SP1-6 and SP1-5, were found to be good immunogens for the induction of specific delayed-type hypersensitivity (DTH) reactions measured using either synthetic peptides or MUC1-transfected cells. In addition, immunization with irradiated MUC1-transfected cells induced strong DTH reactions measured using synthetic peptides that expressed the PDTRP sequence, which has been shown to be, or to overlap, a T cell epitope in humans and a B cell epitope in mice. Finally, it was demonstrated that synthetic MUC1 peptide "vaccines" could be used both prophylactically and therapeutically to inhibit the growth of MUC1-transfected tumor cells and prolong the survival of tumor-bearing mice.

The polymorphic epithelial mucin (PEM) is found as a cell associated transmembrane protein with an extracellular domain made up largely of tandem repeats and also as a soluble form in some body fluids and culture supernatants. To determine whether the soluble form can arise without the mechanism of alternative splicing mouse cells have been transfected with an expression construct containing the full-length cDNA, and the supernatants of the transfectants analyzed for the presence of the mucin. The presence of mucin in the supernatants could indeed be detected in a radioimmunoassay and by immunoprecipitation using monoclonal antibodies to the tandem repeat region of the core protein, indicating that release of the soluble form can occur without alternative splicing. The soluble form was not however precipitated with a polyclonal antiserum to the cytoplasmic tail, suggesting that it was released from the membrane by the action of a protease.

The human MUC1 gene codes for the core protein of a mucin which is expressed by glandular epithelia and the carcinomas which develop from these tissues. The core protein is aberrantly glycosylated in cancers, and some antibodies show specificity in their reactions with the cancer-associated mucin, which also contains epitopes recognized by T-cells from breast and pancreatic cancer patients. For evaluating the potential use of mucin-reactive antibodies and mucin-based immunogens in cancer patients, a mouse model, expressing the MUC1 gene product PEM (polymorphic epithelial mucin) as a self antigen, would be extremely useful. To this end, we have developed transgenic mouse strains expressing the human MUC1 gene product in a tissue-specific manner. The TG4 mouse strain was established using a 40-kilobase fragment containing 4.5 kilobases of 5' and 27 kilobases of 3' flanking sequence. The TG18 strain was developed using a 10.6-kilobase SacII fragment from the 40-kilobase fragment; this fragment contained 1.6 kilobases of 5' sequence and 1.9 kilobases of 3' flanking sequence. Both strains showed tissue specificity of expression of the MUC1 gene, which was very similar to the profile of expression seen in human tissues. The antibody SM-3 is directed to a core protein epitope, which is selectively exposed in breast cancers and which shows a more restricted distribution on normal human tissues. It was established that the distribution of the SM-3 epitope of PEM in the tissues of the transgenic mice is similar to that seen in humans. The transgenic mouse strains described here should form the basis for the development of a preclinical model for the evaluation of PEM-based antigens and of antibodies directed to PEM in cancer therapy.

Although mucins have been studied at the biochemical and biophysical level for some time, attempts to define their structures in detail were only partially successful because of their size and complexity. The advent of monoclonal antibodies reactive with these molecules introduced a new approach to structural studies by defining antigenic epitopes, by allowing purification of the mucin molecules by affinity chromatography, and by providing a means to clone genes coding for the core proteins. By their profile of reactivity with the normal and cancer-associated mucin in a particular tissue, the antibodies also defined a difference in the mucin derived from the two sources. It is now clear that this difference lies in the carbohydrate side chains, as the core proteins are identical. Because the mucins are tumor-associated antigens and the cancer mucins can express epitopes that are relatively tumor specific, this family of molecules is now being intensively studied. There is also considerable interest in elucidating the normal function of the mucin and in determining whether, through an altered structure, this function is subverted in malignancy. In the next few years we should expect that the structure of other mucins will be defined in the same detail as the product of the MUC1 gene. We should also expect to see the continued application of mucin-reactive antibodies in the clinic and the investigation of mucins as agents for immunotherapy of some cancers. As to the function(s) of these molecules, perhaps we will learn enough in the future to make a critical reappraisal of the name.

Human mammary cells present on the cell surface a polymorphic epithelial mucin (PEM) which is developmentally regulated and aberrantly expressed in tumors. PEM carries tumor-associated epitopes recognized by the monoclonal antibodies HMFG-1, HMFG-2, and SM-3. Previously isolated partial cDNA clones revealed that the core protein contained a large domain consisting of variable numbers of 20-amino acid repeat units. We now report the full sequence for PEM, as deduced from cDNA sequences. The encoded protein consists of three distinct regions: the amino terminus consisting of a putative signal peptide and degenerate repeats; the major portion of the protein which is the tandem repeat region; the carboxyl terminus consisting of degenerate tandem repeats and a unique sequence containing a transmembrane sequence and a cytoplasmic tail. Potential O-glycosylation sites (serines or threonines) make up more than one-fourth of the amino acids. Length variations in the tandem repeat result in PEM being an expressed variable number tandem repeat locus. Tandem repeats appear to be a general characteristic of mucin core proteins.

A new monoclonal antibody, SM3, against stripped mucin core protein has been evaluated for the radioimmunoscintigraphy of ovarian cancer. It was radiolabelled with In-111, I-123 and Tc-99m and results showed a sensitivity of 95%, 100% and 100% and an accuracy of 73%, 86% and 100% for malignancy; respectively.

Previously reported immunohistochemical studies using three monoclonal antibodies reactive with the human milk fat globule membrane, LICR.LON.M8 and HMFG1 and 2 from the ICRF, have shown that their antigenic determinants have similar tissue distributions. These distributions are also similar to those shown by polyclonal sera against epithelial membrane antigen. Using a radioimmunoassay, we have shown that all three monoclonal antibodies react with epithelial membrane antigen. Immunofluorescent staining of cultures of human breast cells with these antibodies demonstrated that they are heterogeneous in their distribution. Although the results emphasized their similarities, differences between the antibodies suggested that they react with different epitopes on the same molecule. The use of these reagents in diagnostic tumor pathology is discussed.

Carcinomatous meningitis is a devastating metastatic complication of systemic carcinoma, which may occur insidiously, accompanied by a confusing spectrum of clinical symptoms and signs. In the absence of reliable diagnostic tumour markers, the diagnosis is established by the demonstration of malignant cells within the cerebrospinal fluid (CSF). Cytological techniques requiring skillful interpretation are occasionally negative in the presence of established disease, and when positive may indicate leptomeningeal malignancy of such advanced nature that effective palliation is difficult. Biochemical tumour marker technology offers the potential of reliable diagnosis in early disease states, prior to the appearance of exfoliated malignant cells. In a series of 100 patients, we assayed for an epithelial associated glycoprotein (HMFGI antigen) in CSF obtained at lumbar puncture. In 18 of 20 patients with carcinomatous meningitis, this high-molecular-weight glycoprotein was detectable in the CSF. The antigen was also present in 2 patients with neoplastic meningitis complicating lymphoma and medulloblastoma, but was not detected in the CSF of the remaining 78 patients.