Advances of MUC1 as a target for breast cancer immunotherapy

MUC1 is a potential target in breast cancer immunotherapy as MUC1 is overexpressed in breast cancer, and is absent or expressed in low level in normal mammary gland. In addition, MUC1 is mostly aberrantly underglycosylated in cancer and the antigens on the cancer surface are different from normal cell. Therefore targeting MUC1 for cancer immunotherapy can exploit the difference between cancer and normal cells, and eliminating the cancerous cells while leaving the normal mammary cells unharmed. This review will focus on the recent advance of MUC1 breast cancer immunotherapy currently being investigated.

Anti-Cripto Mab inhibit tumour growth and overcome MDR in a human leukaemia MDR cell line by inhibition of Akt and activation of JNK/SAPK and bad death pathways

Doxorubicin (DOX) selection of CCRF-CEM leukaemia cell line resulted in multidrug resistance (MDR) CEM/A7R cell line, which overexpresses MDR, 1 coded P-glycoprotein (Pgp). Here, we report for the first time that oncoprotein Cripto, a founding member of epidermal growth factor-Cripto-FRL, 1-Criptic family is overexpressed in the CEM/A7R cells, and anti-Cripto monoclonal antibodies (Mab) inhibited CEM/A7R cell growth both in vitro and in an established xenograft tumour in severe combined immunodeficiency mice. Cripto Mab synergistically enhanced sensitivity of the MDR cells to Pgp substrates epirubicin (EPI), daunorubicin (DAU) and non-Pgp substrates nucleoside analogue cytosine arabinoside (AraC). In particular, the combination of anti-Cripto Mab at less than 50% of inhibition concentrations with noncytotoxic concentrations of EPI or DAU inhibited more than 90% of CEM/A7R cell growth. Cripto Mab slightly inhibited Pgp expression, and had little effect on Pgp function, indicating that a mechanism independent of Pgp was involved in overcoming MDR. We demonstrated that anti-Cripto Mab-induced CEM/A7R cell apoptosis, which was associated with an enhanced activity of the c-Jun N-terminal kinase/stress-activated protein kinase and inhibition of Akt phosphorylation, resulting in an activation of mitochondrial apoptosis pathway as evidenced by dephosphorylation of Bad at Ser136, Bcl-2 at Ser70 and a cleaved caspase-9.

Overexpression of Cripto and its prognostic significance in breast cancer: a study with long-term survival

INTRODUCTION

Cripto is a founding member of the EGF-CFC family, and plays an important role in tumourigenesis, tumour cell proliferation and migration. We aimed to determine the significance of Cripto expression on the survival of patients with breast cancer.

METHODS

Immunohistochemical detection of Cripto was performed by using mAb C13 on 120 formalin-fixed paraffin-embedded breast tumour specimens in tissue microarrays. This cohort comprises a series of 120 patients with primary operable breast cancer diagnosed between 1989 and 1995, retrieved from the Concord Repatriation General Hospital breast carcinoma database.

RESULTS

Using a cutoff value of 80%, Cripto overexpressed in 57 of the 120 (47.5%) patients. We found significant associations between overexpression of Cripto and the Nottingham Prognostic Index (NPI, p<0.01), histological grade (p<0.01), pathological tumour type (p=0.04), PR (p=0.02) as well as Ki-67 (p=0.02). Univariate analysis reveals that there is a significant correlation between overexpression of Cripto and survival (p=0.0003). Cox regression analysis indicates that the overexpression of Cripto is an independent prognostic factor in breast cancer (HR 2.79, 95%CI 1.20-6.50).

CONCLUSION

The unique epitope recognized by mAb C13 is overexpressed on breast tumour tissues. In this series of invasive breast cancers, overexpression of Cripto was more often found in high grade and poor prognosis tumours compared to low grade and good prognosis breast cancers. Moreover, overexpression of Cripto was significantly associated with decreased patient survival.

MUC1 epithelial mucin (CD227) is expressed by activated dendritic cells

The MUC1 mucin (CD227) is a cell surface mucin originally thought to be restricted to epithelial tissues. We report that CD227 is expressed on human blood dendritic cells (DC) and monocyte-derived DC following in vitro activation. Freshly isolated murine splenic DC had very low levels of CD227; however, all DC expressed CD227 following in vitro culture. In the mouse spleen, CD227 was seen on clusters within the red pulp and surrounding the marginal zone in the white pulp. Additionally, we confirm CD227 expression by activated human T cells and show for the first time that the CD227 cytoplasmic domain is tyrosine-phosphorylated in activated T cells and DC and is associated with other phosphoproteins, indicating a role in signaling. The function of CD227 on DC and T cells requires further elucidation.

Development of a fecal occult blood test using a monoclonal antibody to haptoglobin

Monoclonal antibodies (mAbs) were produced to human haptoglobin by immunising with fecal extracts from patients with colon cancer. An enzyme-linked immunosorbent assay was developed with one of the mAbs (FE14.1), and its ability to diagnose colorectal carcinoma evaluated. Patients with colorectal cancer were positive (43/46 = 93.5%) compared to normal individuals (4/211 = 1.9%). The assay has a specificity 93.5% and sensitivity 98.1% and has several advantages over current fecal occult blood tests. The test is potentially useful for bowel cancer diagnosis and to quantitate the level of haptoglobin in other body fluids such as urine and in effusions.

Anti-mucin monoclonal antibodies

Mucins are of major interest in cell biology, not only are they highly over-expressed in many adenocarcinomas (up to 40-fold increase), but also have important physiological function, and probably more to be determined (1-3). There is much information available on mucins - doubtless because of their unusual structure being heavily glycosylated, but also containing a repeat region rich in the amino acids serine, threonine and proline. This repeat region confers high immunogenicity of the mucins, and as a result, many antibodies (Abs) have been made to mucins of different species (4). Furthermore, the production of Abs led to the cloning of the cDNAs and armed with these reagents (antibodies, cDNA and genomic structures), advances in the knowledge of the structure and function of mucins has been rapid, together with the development of transgenic and gene knockout animals for biological studies (1-9). Here we describe monoclonal antibodies (Mabs) made to the different mucins, including Mucins 1-4, concentrating on human Mucin 1 (MUC1), to variants of MUC1, to regions outside the VNTR of MUC1, mouse Mucin1 (muc1), unusual features and cross reactions of anti-MUC1 Mabs and Abs made by patients in clinical trials. We will especially describe the Mabs produced in our laboratory.

Characteristics of immunoglobulin gene usage of the xenoantibody binding to gal-alpha(1,3)gal target antigens in the gal knockout mouse

BACKGROUND

Natural antibodies that react with galactose-alpha(1,3)galactose [galalpha(1,3)gal] carbohydrate epitopes exist in humans and Old World primates because of the inactivation of the alpha1,3-galactosyltransferase (alpha1,3GT) gene in these species and the subsequent production of antibodies to environmental microbes that express the galalpha(1,3)gal antigen. The Gal knockout (Gal o/o) mouse, produced by homologous disruption of the alpha1,3GT gene, spontaneously makes anti-galalpha(1,3)gal antibodies and can be used to study the genetic control of humoral immune responses to this carbohydrate epitope.

METHODS

Six hybridomas that produce monoclonal antibodies (mAbs) to galalpha(1,3)gal were generated in Gal o/o mice. The mAbs were tested to characterize the binding activity with flow cytometry using pig aortic endothelial cells and ELISA with galalpha(1,3)gal carbohydrates. The VH and VK genes of these hybridomas were cloned, sequenced, and analyzed.

RESULTS

The mAbs showed distinct patterns of antibody binding to galalpha(1,3)gal antigens. The VH genes that encode the mAb binding activity were restricted to a small number of genes expressed in their germline configuration. Four of six clones used closely related progeny of the same VH germline gene (VH441). Comparison of the mouse gene VH441 to the human gene IGHV3-11, a gene that encodes antibody activity to galalpha(1,3)gal in humans, demonstrates that these two genes share a nonrandom distribution of amino acids used at canonical binding sites within the variable regions (complimentary determining regions 1 and 2) of their immunoglobulin VH genes.

CONCLUSIONS

These results demonstrate the similarity of the Gal o/o mice and humans in their immune response to galalpha(1,3)gal epitopes. Gal o/o mouse can serve as a useful model for examining the genetic control of antibody/antigen interactions associated with the humoral response to pig xenografts in humans.

Glycoconjugate metabolism in a cystic fibrosis knockout mouse model

Cystic fibrosis knockout mice (cftr(-/-)) die prematurely of obstruction of the intestine which may result from accumulation of dehydrated glycoconjugate-containing mucus. We noted an increase in the specific activity of [(14)C]glucosamine-labeled high-molecular weight glycoconjugates, probably mucin, in the lumen of the intestine of cftr(-/-) (homozygous) mice compared to cftr(+/+) (wild-type) and cftr(+/-) (heterozygous) mice and a decrease in the turnover of glycoconjugates of several organs of the cftr(-/-) mice. No difference in the anionic composition of secreted intestinal glycoconjugates was detected and no difference in the amount of mucin 1 (Muc1) was found in the small intestine, colon, pancreas, and lungs of the different genotypes. In addition, the spleen of the cftr(-/-) mice was significantly smaller than that of control mice and the small intestine and colon were, respectively, longer and shorter compared to control mice. These results indicate modified glycoconjugate metabolism in cystic fibrosis knockout mice and morphologic changes to the spleen and intestine where the latter modifications are possibly related to the intestinal malabsorption associated with cystic fibrosis.

Breast cancer in mice: effect of murine MUC-1 immunization on tumor incidence in C3H/HeOuj mice

Mucin-1 (MUC-1), which is overexpressed in more than 90% of human breast cancers, is a potential target for immunotherapy. To develop a mouse model appropriate for the immunotherapy of human cancer, mouse mucin-1 (muc-1) fusion protein, containing ten tandem repeats, was made and used to immunize C3H/HeOuj mice, which supposedly have a high incidence of breast cancer. C3H/HeOuj mice were injected eight times with 5 microg oxidized mannan muc-1-glutathione-S-transferase (MMFP) with or without cyclophosphamide, which is used to increase cellular immunity. At 80 age weeks, only 12.1% (4 of 33) mice of the untreated C3H/HeOuj mice had mammary tumors. The reason for the low incidence of breast cancer in these mice is not known, but all the mammary tumors were MUC-1+ breast adenocarcinomas and were transplantable to C3H/HeOuj mice. The incidence was 11.4% (4 of 35) in mice injected with MMFP: 38.2% (13 of 34) in mice given cyclophosphamide; and 14.3% (2 of 14) in mice treated with glutathione-S-transferase. That is, cyclophosphamide increased the incidence of mammary tumors, and metastases were found in only these mice. Fewer tumors (6 of 34 or 17.6% compared with 13 of 34 or 38.2% with cyclophosphamide only) occurred in the group immunized with MMFP and cyclophosphamide. Mice immunized with MMFP had high levels of muc-1 antibodies and cellular immune responses (the frequency of the precursor of the cytotoxic Tlymphocyte cell was 1 of 40,000 to 1 of 100,000), which were not found in control groups. The occurrence of muc-1 immunity, particularly the presence of large amounts of anti-mucin-1 antibodies, had no effect on tumor incidence. Thus, the immunization with murine muc-1 reduced the tumor incidence in only cyclophosphamide-treated mice and led to strong muc-1 antibody production and to cellular responses. These findings have implications for human tumor immunotherapy in which strong antibody and weak cellular responses are to be expected and, indeed, have been found.

Construction and characterization of a chimeric receptor containing the cytoplasmic domain of MUC1 mucin

MUC1 mucin is a transmembrane glycoprotein that is highly expressed in various cancer cell lines and is also present in most of the glandular epithelial cells including the airway. Although the presence of numerous phosphorylation sites in its cytoplasmic domain suggests its potential role as a receptor, the unavailability of a ligand for MUC1 mucin has limited our understanding of its function. In this paper, we tried to circumvent this problem by constructing a chimeric receptor containing the cytoplasmic domain of MUC1 mucin, which can be phosphorylated on activation. To this end, we constructed a chimeric plasmid vector (pCD8/MUC1) by replacing the extracellular and transmembrane domains of human MUC1 mucin with those of human CD8. Transient transfection of the vector into COS-7 cells resulted in expression of the chimeric receptor on the surface of the COS-7 cells as judged by immunologic assays with various antibodies as well as by fluorescence-activated cell-sorting analysis. Treatment of the transfected COS-7 cells with an anti-CD8 antibody resulted in a significant increase in phosphorylation of tyrosine moieties of the chimeric receptor. This chimeric receptor will serve as a powerful tool in elucidating the signaling mechanism as well as the functional role of MUC1 mucin in the airway.

Expression of MUC2 gene in gastric regenerative, metaplastic, and neoplastic epithelia

It has been reported that MUC2 mucin is expressed in goblet cells of gastric intestinal metaplasia, but not in its normal epithelium. To confirm this finding, we have examined the expression of the MUC2 gene by reverse transcriptase-polymerase chain reaction and immunohistochemical methods in gastric tissues obtained by routine upper gastrointestinal tract endoscopy and compared the results with pathological findings based on hematoxylin and eosin (H&E) staining. In 16.7% of the tissue specimens tested, MUC2 mRNA was detected in spite of the absence of intestinal metaplasia in HE specimens. A possible explanation for this was the identification by immunohistochemistry of MUC2 protein in regenerative gastric mucosal cells in biopsies that did not contain intestinal metaplasia. Sialyl-Le(x) epitope, which is suggested to be located on MUC2 mucin core protein (MUC2 protein), was also immunohistochemically detected in both goblet cells of intestinal metaplasia and regenerative epithelium. With regard to carcinoma, MUC2 protein was predominantly expressed in intestinal-type adenocarcinoma. These data indicate that MUC2 mucin is expressed in gastric regenerative epithelium in addition to intestinal metaplasia and intestinal type adenocarcinoma.

Expression of MUC2 gene in gastric regenerative, metaplastic, and neoplastic epithelia

It has been reported that MUC2 mucin is expressed in goblet cells of gastric intestinal metaplasia, but not in its normal epithelium. To confirm this finding, we have examined the expression of the MUC2 gene by reverse transcriptase-polymerase chain reaction and immunohistochemical methods in gastric tissues obtained by routine upper gastrointestinal tract endoscopy and compared the results with pathological findings based on hematoxylin and eosin (H&E) staining. In 16.7% of the tissue specimens tested, MUC2 mRNA was detected in spite of the absence of intestinal metaplasia in HE specimens. A possible explanation for this was the identification by immunohistochemistry of MUC2 protein in regenerative gastric mucosal cells in biopsies that did not contain intestinal metaplasia. Sialyl-Le(x) epitope, which is suggested to be located on MUC2 mucin core protein (MUC2 protein), was also immunohistochemically detected in both goblet cells of intestinal metaplasia and regenerative epithelium. With regard to carcinoma, MUC2 protein was predominantly expressed in intestinal-type adenocarcinoma. These data indicate that MUC2 mucin is expressed in gastric regenerative epithelium in addition to intestinal metaplasia and intestinal type adenocarcinoma.

Expression of mucin 1 (MUC1) in esophageal squamous-cell carcinoma: its relationship with prognosis

Using 2 anti-mucin 1 (MUC1) monoclonal antibodies (MAbs), DF3 and BCP8, we examined MUC1 expression immunohistochemically in 192 esophageal squamous-cell carcinomas (SCCs). In normal squamous epithelium of the esophagus, DF3 was not expressed, but BCP8 was expressed on the cell membrane, mainly in the surface layer. In esophageal SCCs, DF3 and BCP8 were expressed mainly on the cell membrane of SCC cells, but also in the cytoplasm in several cases. To analyze the correlation of MUC1 expression and the prognosis of the patients, the 192 cases were divided into 2 groups: high-expression group (HEG, > 50% of the neoplastic cells stained) and low-expression group (LEG, < 50% of neoplastic cells stained). DF3-HEG (24 patients) showed a significantly poorer survival rate than DF3-LEG (168 patients), whereas there was no significant difference in survival between BCP8-HEG (43 patients) and BCP8-LEG (149 patients). Also, in the analysis of 162 patients with advanced stage (submucosal or deeper invasion) to exclude the influence of low expression of DF3 and BCP8 in 30 patients with early stage (up to the level of muscularis mucosae), DF3-HEG (24 patients) showed significantly poorer survival than DF3-LEG (138 patients), whereas there was no significant difference in survival between BCP8-HEG (42 patients) and BCP8-LEG (120 patients). The results of our study on esophageal SCC suggest that the expression of sialyl oligosaccharides detected by DF3 is related to poor prognosis.

Expression of mucin 1 (MUC1) in esophageal squamous-cell carcinoma: its relationship with prognosis

Using 2 anti-mucin 1 (MUC1) monoclonal antibodies (MAbs), DF3 and BCP8, we examined MUC1 expression immunohistochemically in 192 esophageal squamous-cell carcinomas (SCCs). In normal squamous epithelium of the esophagus, DF3 was not expressed, but BCP8 was expressed on the cell membrane, mainly in the surface layer. In esophageal SCCs, DF3 and BCP8 were expressed mainly on the cell membrane of SCC cells, but also in the cytoplasm in several cases. To analyze the correlation of MUC1 expression and the prognosis of the patients, the 192 cases were divided into 2 groups: high-expression group (HEG, > 50% of the neoplastic cells stained) and low-expression group (LEG, < 50% of neoplastic cells stained). DF3-HEG (24 patients) showed a significantly poorer survival rate than DF3-LEG (168 patients), whereas there was no significant difference in survival between BCP8-HEG (43 patients) and BCP8-LEG (149 patients). Also, in the analysis of 162 patients with advanced stage (submucosal or deeper invasion) to exclude the influence of low expression of DF3 and BCP8 in 30 patients with early stage (up to the level of muscularis mucosae), DF3-HEG (24 patients) showed significantly poorer survival than DF3-LEG (138 patients), whereas there was no significant difference in survival between BCP8-HEG (42 patients) and BCP8-LEG (120 patients). The results of our study on esophageal SCC suggest that the expression of sialyl oligosaccharides detected by DF3 is related to poor prognosis.

Immunohistochemical detection of MUC2 mucin core protein in ulcerative colitis

MUC2 mucin is predominantly expressed in the colon and is considered to play an important role in the protection of that organ. Recent findings suggested that MUC2 protein levels are significantly decreased in active ulcerative colitis (UC). We therefore performed an immunohistochemical study to reveal if the expression of MUC2 protein is altered in UC. Seventy-nine biopsy tissue specimens from 31 UC patients, along with normal colon tissues, were immunostained with anti-MUC2 mucin core protein monoclonal antibody (MoAb) CCP58 (IgG1). UC tissue specimens were divided into two groups based on the histological severity of inflammation, i.e., 64 with active inflammation (grades 2-5) and 15 without (grade 1). In the former group, 52 out of 64 (81.3%) tissue specimens contained sections of glands with a few cells positive for MoAb CCP58. These glands were small and consisted of MUC2 negative-short cells and a few positive cells without apparent mucus formation, considered to be immature regenerative glands. In contrast, the staining pattern was almost the same as that of the normal colon and no immature glands were seen in the group without active inflammation. The sections of immature regenerative glands with a few MUC2-positive cells were exclusively found in the UC tissues with active inflammation, but not in those without it, suggesting that the expression of MUC2 protein may be decreased in active UC.

Anti-MUC1 antibodies react directly with MUC1 peptides presented by class I H2 and HLA molecules

Peptides bound in the groove of MHC class I molecules and detected by CTLs are not normally accessible to Ab. We now report that MUC1 peptides that are bound within the groove of MHC class I molecules (H2 and HLA) and that can be detected by CTLs can also be detected by anti-MUC1 Abs. mAbs to the middle and C-terminal regions of the class I-associated peptides but not to the N terminus were able to react with MUC1 peptides bound to H2Kb and HLA-A*0201, and only to the mid-region for H2Db, by flow cytometry and also to block CTL activity. Molecular modeling showed that the N terminus is buried (and not accessible), whereas the midpeptide residues form a loop and the C terminus is free, making these two regions accessible to Ab. The findings demonstrate for the first time that peptides associated with class I molecules can be detected by anti-peptide Abs.

Mouse mucin 1 (MUC1) defined by monoclonal antibodies

Mucins are highly expressed in many different human cancers and numerous murine monoclonal antibodies (MAbs) to human mucins, particularly Mucin 1 (MUC1), have been produced. However, no such antibodies to murine mucin 1 (muc1) have been described and we now describe 6 different antibodies produced to murine muc1 and to human MUC1 cytoplasmic tail, either by immunising rats, or muc1 o/o mice with synthetic peptides or a fusion protein composed of glutathione-s-transferase (GST) linked to the tandem repeat region of muc1. The antibodies to both the extracellular tandem repeat region and to the cytoplasmic tail were found to react with mucin-containing murine tissues such as breast, stomach, colon, ovary, kidney and pancreas, and the staining patterns were similar to those found in humans. The reagents reacted specifically with muc1 peptides and tissues; however, some cross reactivity with other mucin-derived peptides was noted, particularly those containing the amino acid sequence TSS. Three different epitopes (TSS, TAVLSGTS and LSGTSSP) of the M30, M70 and MFP25 MAbs were detected. Of interest was the finding that some of the antibodies reacted with murine lymphocytes; it was not clear whether these reactions were due to mucin 1 on mouse lymphocytes (MUC1 was considered to be absent from human lymphocyte), or due to cross reaction with a sialic adhesion molecule on lymphocytes. The antibodies should prove valuable reagents when studying differentiation and expression in murine glandular tissues and the ontogeny of mucin-secreting tumours.

Oxidised mannan antigen conjugates preferentially stimulate T1 type immune responses

It is desirable to be able to produce either T1 or T2 responses and we have found that, in mice, mannose--coupled antigens stimulated T2 type responses antibodies and CTLs, whereas if oxidized, mannose--coupled antigens stimulated T1 responses little antibody and a potent CTL response. In addition, the cytokine profiles support the T1rT2 differentiation with these immunizations, in that oxidized mannan antigen gives IFNg, IL-2 and IL-12 production, whereas in the absence of oxidization, IL-4 and not the other cytokines is produced. A number of antigens have been examined--particularly Mucin 1 and the delivery method using mannose may be applicable to the other antigens.

Antibody and T cell responses of patients with adenocarcinoma immunized with mannan-MUC1 fusion protein

Mucin 1 (MUC1) is a large complex glycoprotein that is highly expressed in breast cancer, and as such could be a target for immunotherapy. In mice, human MUC1 is highly immunogenic, particularly when conjugated to mannan, where a high frequency of CD8(+) MHC-restricted cytotoxic T lymphocytes is induced, accompanied by tumor protection. On this basis, a clinical trial was performed in which 25 patients with advanced metastatic carcinoma of breast, colon, stomach, or rectum received mannan-MUC1 in increasing doses. After 4 to 8 injections, large amounts of IgG1 anti-MUC1 antibodies were produced in 13 out of 25 patients (with antibody titers by ELISA of 1/320-1/20,480). Most of the antibodies reacted to the epitopes STAPPAHG and PAPGSTAP. In addition, T cell proliferation was found in 4 out of 15 patients, and CTL responses were seen in 2 out of 10 patients. Mannan-MUC1 can immunize patients, particularly for antibody formation, and to a lesser extent, cellular responses. It remains to be seen whether such responses have antitumor activity.

Mucins (MUC1 and MUC3) of gastrointestinal and breast epithelia reveal different and heterogeneous tumor-associated aberrations in glycosylation

In a comprehensive study, we examined the expression of the membrane and secretory mucins MUC1 and MUC3, respectively, in normal and neoplastic gastrointestinal and breast epithelia before and after specific alterations of their glycan structures by neuraminidase, alpha-fucosidase, or carbohydrate-specific periodate oxidation. MUC1 mRNA was also identified in normal colorectal tissues by in situ hybridization. The data revealed that normal colorectal epithelia express both MUC1 mRNA and protein, which were detectable after periodate oxidation with all tested MUC1-specific antibodies. During tumorigenesis in the colon, MUC1 became recognizable without periodate treatment concomitantly with highly dysplastic lesions and the malignant state. In the breast, in which MUC1 is detectable with most antibodies in normal epithelium as well as in carcinomas, staining could be enhanced by pretreatment with periodate and casually by enzyme treatments. MUC3 was detectable in normal and neoplastic colorectal tissues and was more intensely stained after periodate oxidation. It was absent in normal breast even after pretreatment but was expressed in seven of 20 breast carcinomas. Therefore, incomplete glycosylation, abnormal distribution, and ectopic expression of mucins are characteristics of malignancy. Periodate oxidation may be widely applicable to immunohistochemistry for examining changes in glycosylation and for detecting antigens masked by glycans.

Correlative histochemical study providing evidence for the dual nature of human colorectal cancer mucin

Luminal secretions within colorectal cancers have been assumed to be the counterpart of normal goblet cell mucin. The aim of this study was to establish whether secretory material within colorectal cancers may in fact be traced to different lineages: goblet cells and columnar cells. The distribution of the apomucins MUC1 and MUC2, non-O-acetylated sialic acid and the carbohydrate structures sialosyl Tn, Tn, Lewis(x), sialosyl Lewis(x) and Lewis(y) was studied in normal colorectal mucosa and colorectal cancer specimens using standard histochemical techniques. Unmasking of MUC1 and MUC2 was achieved using periodic acid and saponification-neuraminidase-periodic acid pretreatment respectively. Within normal and malignant epithelium, correlations and/or co-localization could be demonstrated for goblet cells with MUC2, non-O-acetylated sialic acid, sialosyl Tn, Tn (Golgi region) and sialosyl Lewis(x), and for columnar cells with MUC1, Lewis(x), sialosyl Le(x), Tn (cytoplasm) and Lewis(y) (UEA-1). The goblet cell spectrum was associated with mucin-like (type I) luminal secretions within cancers, whereas the columnar cell spectrum characterized non-mucin-like (type II) secretions and intracytoplasmic lumina. These data indicate that colorectal cancer mucin can be broadly separated into two types: secretory mucin linked to cells of goblet lineage and up-regulated membrane-associated mucin of presumed columnar cell origin.

Natural human anti-Gal alpha(1,3)Gal antibodies react with human mucin peptides

We have recently demonstrated that both antibodies to Gal alpha(1,3)Gal, and the Gal alpha(1,3)Gal binding lectin (IB4), bind a synthetic peptide (DAHWESWL), there being a similar recognition of carbohydrate and peptide structures. We now report that the anti-Gal alpha(1,3)Gal antibodies and IB4 lectin also react with peptides encoded by mucin genes (MUC 1, 3, 4)-sequences known to be rich in serine, threonine and proline. This activity was demonstrated (1) by the ability of mucin derived peptides to block the reaction of anti-Gal alpha(1,3)Gal antibodies and IB4 lectin with a Gal alpha(1,3)Gal+ pig endothelial cell line; the reactions were specific and did not occur with a random peptide containing the same sequences or with other mucin peptides; (2) by the fact that anti-mucin1 antibodies could react with the Gal alpha(1,3)Gal expressed after transfection of COS cells (Gal alpha(1,3)Gal-,Muc1-) with cDNA encoding the pig alpha, 3galactosyltransferase; and (3) that the IB4 lectin and anti-Gal alpha(1,3)Gal antibodies could react with mucin 1 found on the surface of human breast cancer cells. Thus natural occurring anti-Gal alpha(1,3)Gal antibodies found in all human serum can react with self (Muc1) peptides expressed in large amounts on the surface of tumour cells but not on normal cells. The findings are of interest and serve to explain the previously reported findings that human cells can, at times, express Gal alpha(1,3)Gal; such expression is an artefact, the reaction is due to the phenomenon described herein, i.e. that anti-Gal alpha(1,3)Gal antibodies react with mucin peptides.

A new approach to fecal occult blood testing based on the detection of haptoglobin

BACKGROUND

Colorectal carcinoma is a common disease, occurring in 1 in 20 adults in Western society, and there is a compelling need for an effective early diagnostic test. Several serum tests, including carcinoembryonic antigen have been used, but none are sufficiently sensitive for the early diagnosis of the disease.

METHODS

In a novel approach using fecal extracts from patients with colorectal cancer as the antigen for immunization, several MoAbs were produced. One (FE14.1) was found to react with the feces from patients with colon cancer, but not with those from normal subjects. A sandwich enzyme-linked immunoadsorbent assay was developed, and its ability to diagnose colorectal carcinoma evaluated.

RESULTS

Of the patients with colorectal carcinoma, 91.5% (43/46) were positive compared with 1.9% of normal individuals (4/211). Analysis of the N-terminal amino acid sequence of a subunit of the molecule detected by FE14.1 shows it to be the beta chain of haptoglobin.

CONCLUSIONS

The assay developed in this study has several advantages compared with current fecal occult blood tests, including no requirement for dietary restriction and the ability to distinguish between upper and lower gastrointestinal bleeding, while retaining the sensitivity and specificity of the current tests. Furthermore, the sensitivity of the tests increases to 100% if the FE14.1 and HemeSelect are combined. In addition, the study shows the potential to produce anticancer agents by immunizing with fecal material.

Anti-peptide monoclonal antibodies to intestinal mucin 3

Second generation anti-peptide monoclonal antibodies were produced using synthetic peptide SIB35 (C-HSTPSFTSSITTTETTSHSTPSFTSSITTTETTS) as an antigen which contains two of the MUC3 tandem repeats. Three monoclonal antibodies were produced (M3.1, M3.2 and M3.3) which reacted with the immunizing peptide, but also reacted strongly with colon tissues. Using the immunoperoxidase staining technique, two of the monoclonal antibodies, M3.1 and M3.2, reacted with both colon carcinoma and normal colon tissue whereas the antibody M3.3 reacted with normal colon tissues but very weakly, if not at all, with colon carcinoma. MUC3 was found to be distributed in colon and rectum, and was also present to a lesser extent in breast, lung and salivary gland tissues. Analysis of mucin molecules by Western blotting, revealed the antigen detected by the antibodies (M3.1 and M3.2) to be of a high relative molecular mass. M3.1 and M3.3 reacted with eptiopes SITTTE and PSFTSS, respectively. M3.2 did not react with any of the 6-mer peptides, even though it reacted with the full length VNTR (2 repeat peptide--SIB35). Anti-MUC3 peptide antibodies appear to react with colon, rectum, breast, salivary gland and lung tissues, and represent a new method of producing anti-tumour antibodies.

The immunogenicity of MUC1 peptides and fusion protein

Mucin 1 (MUC1) is highly expressed in breast cancer, has an ubiquitous distribution and, due to altered glycosylation, peptides within the VNTR are exposed. These peptides are the target for anti-MUC1 antibodies, which give a differential reaction on cancer compared with normal tissue. The amino acids, APDTR or adjacent amino acids, are highly immunogenic in mice for antibody production (after immunisation with either breast cancer cells, human milk fat globule (HMFG) or the VNTR peptide). In addition, human studies show that this region of the MUC1 VNTR functions as target epitopes for cytotoxic T cells. We have performed preclinical and clinical studies to examine the immune responses to MUC1 in mice and humans: (a) MUC1+ 3T3 or P815+ 3T3 cells in syngeneic mice are rejected, with the generation of both cytotoxic T lymphocyte (CTL) and DTH responses and a weak antibody response and a weak antibody responses; this type of immunity gives rise to total resistance to re-challenge with high doses of these tumors; (b) immunisation with peptides (VNTR x 2), a fusion protein (VNTR x 5), or HMFG leads to no CTLs, DTH, good antibody production and weak tumour protection (to 10(6) cells, but not 5 x 10(6) cells) (possibly a TH2 type response); (c) immunisation with mannan-fusion protein (MFP) gives rise to good protection (resistance to 50 x 10(6) cells), CTL and DTH responses and weak antibody responses (possibly a TH1 type response, similar in magnitude to that obtained after tumor rejection); (d) established tumors can be rapidly rejected by delayed treatment of MFP; (e) the CTL responses are MHC restricted (in contrast to the human studies); (f) APDTR appears not to be the T cell reactive epitope in mice. On the basis of these findings, two clinical trials are in progress: (a) VNTR x 2 (diphtheria toxoid) which gives rise to some T cell proliferation, DTH and antibody responses in some patients and (b) an MFP trial. The ability to alter the immune response towards cellular immunity with mannan or to humoral immunity with peptides, allows the immune response to be selectively manipulated.

Expression of MUC2-mucin in colorectal adenomas and carcinomas of different histological types

The expression of mucin MUC2 was investigated in normal colonic tissue, in colonic adenomas and in carcinomas of the mucinous and non-mucinous type. The latter were subdivided into carcinomas originating from the adenoma-carcinoma sequence (ACS) and de novo (DN) carcinomas. The expression was assayed by immunohistochemistry with the monoclonal anti-MUC2 antibody CCP58 and by mRNA semiquantitation. MUC2 protein epitope CCP58 was strongly expressed in 21% of normal colonic tissues, in 40% of villous and in 48% of tubular adenomas. Mucinous carcinomas exhibited strong expression in 72%, ACS carcinomas in 21% and DN adenocarcinomas in none of the tumors investigated. Compared with the adjacent non-malignant tissue (transitional mucosa), CCP58 epitope expression in the tumor was higher in 74% of mucinous carcinomas, but equal or lower in 69% of ACS carcinomas and in 100% of de novo carcinomas. The alterations of MUC2 expression detected by immunohistochemistry in adenocarcinomas were confirmed on mRNA level. These data indicate that the MUC2 expression pattern is different in the 3 carcinoma types investigated. MUC2 over-expression occurs in the adenomatous tissue. It is always maintained in mucinous carcinomas, but frequently decreased in non-mucinous ACS carcinomas. DN carcinomas are most frequently associated with decreased expression of MUC2.

Murine immune response to cells transfected with human MUC1: immunization with cellular and synthetic antigens

Humans with breast cancer have T-cells in their lymph nodes which recognize a peptide sequence within the variable number of tandem repeats of the mammary mucin, MUC1, which is overexpressed in breast cancer. To find means of making this recognition event into a potent immune response to breast cancer, we used a murine tumor model and have examined the parameters of the immune response to human mucin (MUC1) expressed in murine BALB/c 3T3 cells. We then sought to boost this response with MUC1-containing synthetic peptides, fusion proteins, and natural mucin (HMFG). MUC1+3T3 cells were found to be rejected by BALB/c mice by day 15 due to a cellular [CD3+, Ly2+ (CD8+)] response. The cellular rejection response was accompanied by the generation of CD8+ cytotoxic T-cells, CD4+ delayed-type hypersensitivity, and little anti-MUC1 antibody. This immune response is presumably of the TH1 type (which occurs in CD8 as well as CD4 cells) of CD8+ cytotoxic cells. By contrast, mice immunized with the MUC1 synthetic peptide, a fusion protein, or HMFG have good antibody responses, a delayed-type hypersensitivity reaction, but no cytotoxic T-cells and less tumor protection, possibly a TH2 type response. We conclude that CD8+ cytotoxic anti-mucin cells can produce significant antitumor responses in vivo to a human "tumor" antigen expressed in murine cells; immunization with soluble synthetic or native materials leads to the "humoral" (TH2) type of immunity, and efforts need to be made to convert this to a TH1-type response.

Discrimination between alternatively spliced STP-A and -B isoforms of CD46

CD46 (membrane cofactor protein; MCP) is ubiquitously expressed on nucleated human cells; it has a protective function, binding C3b and C4b, which are then cleaved by serum factor I. CD46 molecules (55,000-65,000 MW) have four short consensus repeats (SCR): the function of SCR-1 and -2 is unknown; SCR-3 and -4 bind C3b and C4b. These are succeeded by the STP region, which can contain three separate regions (STP-A, -B, -C) rich in serine, threonine and proline and which are heavily glycosylated, succeeded by transmembrane and cytoplasmic tail regions (of which there are several). Multiple isoforms exist due to the different splicing of exons: STP-A and -B can thus be present or absent. So far these products can only be detected separately by polymerase chain reaction (PCR) and RNA studies; we now describe their detection by anti-peptide antibodies. Peptides whose sequences corresponded with those of STP-A and STP-B were synthesized and used for the immunization of mice; although they differ in only seven of 21 amino acids, monoclonal antibodies (mAb) that reacted specifically with STP-A but not with STP-B, and mAb that reacted specifically with STP-B but not with STP-A, were produced; these reacted specifically with native CD46 on human tissues and cell lines. STP-A mAb reacted with tissues in which STP-A RNA had been found, some leukaemias and cell lines; in normal tissue expression was mainly found in the intestine (large and small) and salivary gland. Anti-STP-B reacted with most tissues and cell lines. The antibodies should be of use in defining the expression and function of CD46 in different tissues.

Distribution of the major xenoantigen (gal (alpha 1-3)gal) for pig to human xenografts

We have previously demonstrated that the major epitope in pig tissues detected by naturally occurring human IgM antibodies is galactose (alpha 1-3)galactose. Subsequent biochemical studies demonstrated this epitope to be present on molecules (Mr40-220kDa) on both endothelial cells and lymphocytes. The objective of the present study was to define the distribution of gal(alpha 1-3)gal in different pig tissues, concentrating on those of relevance for the potential transplantation of pig organs or tissues to humans. Adult pig tissues were obtained fresh, fixed, and stained by the immunoperoxidase technique using biotinylated Griffonia simplicifolia lectin (IB4) which binds only to gal(alpha 1-3)gal, and examined histologically. Endothelial cells in all small vessels (capillaries, arterioles and venules) had a unifrom and dense expression of gal(alpha 1-3)gal; in larger vessels, like the aorta, they were less reactive. The highest concentrations were found in the liver parenchyma which stained uniformly, and in the kidney, where the highest amounts were found in the brush border of the proximal convoluted tubules. There was no staining of collecting ducts or glomeruli (except for endothelium) and moderate staining of the distal convoluted tubules. Heart muscle was nonreactive, although the high density of capillaries indicated a reasonable content of gal(alpha 1-3)gal. In contrast to these tissues was the distribution in the pancreas, which, apart from vessels and the lining of ducts, was nonreactive, i.e. islet cells were essentially lacking in gal(alpha 1-3)gal. Other tissues such as the lung contained moderate amounts of material lining the alveoli and bronchioles.(ABSTRACT TRUNCATED AT 250 WORDS)

Peptide binding sites recognized by anti-mucin (MUC2) monoclonal antibodies

Multiple genes coding for human mucins have been identified (MUC 1-5) and here monoclonal antibodies (MoAb) to a gastrointestinal mucin--MUC2 are examined. The antibodies were made to a synthetic peptide representing a single repeat in the core protein of the variable number of tandem repeat region. Using the six-mer overlapping peptides synthesized on polyethylene pins, different binding sites were detected by five anti-MUC 2 MoAbs. These contained amino acids: STTT, PTT, GTQTP, TPTP and PTTT (one antibody), and TPTPT. The repeat region of MUC2 essentially is hydrophobic, but contain useful immunogenic sites. This information will be useful for studying the structure and function of MUC2.

Mucin 1 antigens in the serum and bronchial lavage fluid of patients with lung cancer

BACKGROUND

Lung cancer (LC) is the most common fatal malignancy, but there are no useful tumor markers for diagnosis or monitoring. Mucin 1 has an established role as a marker in other malignancies, but has undergone limited assessment in LC.

METHODS

Serum from 86 patients with LC and 24 with benign pulmonary disease (BPD), and bronchial lavage fluid from 55 LC patients and 21 BPD patients were tested using the Mucin 1 assays mammary serum antigen (MSA) and cancer-associated serum antigen (CASA).

RESULTS

For LC, serum CASA achieved sensitivity of 57%, specificity of 93% relative to normals, and 63% specificity relative to BPD. For MSA the same parameters were 19%, 95%, and 92%. Serum CASA levels were significantly higher in LC patients compared with BPD (P = 0.024) but there was no difference for MSA (P = 0.635). CASA showed excellent correlation with tumor stage and in patients with changing status of disease, while MSA did not. By contrast there was no difference in bronchial lavage fluid tumor marker levels from LC and BPD patients (CASA, P = 0.87; MSA, P = 0.89).

CONCLUSIONS

In a small series serum CASA appears to be a useful agent in detecting LC because it is elevated in all types and stages of LC, and its level correlates with stage and progress of disease. Some patients with BPD have elevated levels suggesting a greater value for monitoring rather than diagnosis. Both serum MSA testing and measurements of either marker in bronchial lavage fluid are of no value.

Altered glycosylation of the MUC-1 protein core contributes to the colon carcinoma-associated increase of mucin-bound sialyl-Lewis(x) expression

The mucin carbohydrate epitope sialyl-Le(x), detected with the monoclonal antibody AM-3, is strongly overexpressed in > 90% of human colon carcinomas. We show here that in colon carcinoma one of the mucin cores bearing the sialyl-Le(x) group is MUC-1, whereas sialyl-Le(x) present in normal colon is not detectable on MUC-1. The amounts of MUC-1 core detectable with the monoclonal antibody BC3 in extracts of tumor tissue are 60-180% of those in normal tissue. Two other carbohydrate epitopes located on MUC-1 in mucins from normal and tumor tissue have also been characterized. In contrast to sialyl-Le(x), their expression on MUC-1 is variable and does not correlate with the malignant transformation of colonic mucosa. The transfer of the sialyl-Le(x) group onto the MUC-1 core contributes to the colon carcinoma-associated overexpression of the sialyl-Le(x) epitope.

Immunohistochemical study of mucin carbohydrates and core proteins in hepatolithiasis and cholangiocarcinoma

The expression of mucin carbohydrates [Tn, sialosyl-Tn(STn), and T antigens] and core proteins [MUCI-apomucin-related antigen (ARA) and MUC2-ARA] was examined immunohistochemically in tissues from 40 patients with hepatolithiasis and 26 patients with intrahepatic bile-duct carcinoma. Tn and STn antigens were expressed in most of the carcinomas, and were also often expressed in the atypical bile-duct epithelium of the patients with hepatolithiasis or carcinoma, whereas they were rarely or never expressed in the normal bile duct, suggesting that they are effective tumor markers. T antigen was less useful as a marker for intrahepatic bile-duct carcinoma or the atypical epithelium, because it was expressed in normal bile-duct of some cases. Regarding the expression of ARAs in the carcinomas, non-invasive bile-duct cyst adenocarcinomas with favorable prognosis either expressed no MUCI-ARA with [DF3(-), MUSEII(-) and 139H2(-)] staining pattern or expressed MUCI-ARA with [DF3(-), MUSEII(+) and 139H2(+)] staining pattern. However these tumors often expressed MUC2-ARA with [anti-MRP(+) and CCP58(+)] staining pattern. In contrast, most invasive non-papillary cholangiocarcinomas with poor prognosis expressed MUCI-ARA with [DF3(+), MUSEII(+) and 139H2(+)] staining pattern, but expressed no MUC2-ARA with [anti-MRP(-) and CCP58(-)] staining pattern. These results suggests that different apomucins are produced by bile-duct cystadenocarcinomas and cholangiocarcinomas with differing prognosis. Furthermore, expression of Tn and STn antigens is a useful indicator of malignancy in the intrahepatic duct.

Production of anti-breast cancer monoclonal antibodies using a glutathione-S-transferase-MUC1 bacterial fusion protein

Two murine Mabs VA1(IgG1) and VA2(IgG1) were produced against a bacterial fusion protein comprising glutathione S-transferase and five tandem repeats of the MUC1 protein. Using the immunoperoxidase staining technique, VA1 detected 46/53 and VA2 detected 48/53 breast cancers and both also reacted with a range of other human epithelial carcinomas. In addition VA1 gave weak reactions with normal breast tissues whereas VA2 was non-reactive and could be a relatively tumour specific antibody for breast cancer. The antibodies were also tested by ELISA-VA1 reacted weakly with glycosylated HMFG but strongly with deglycosylated HMFG, whereas VA2 reacted strongly with both forms of HMFG. The reactivities of the two Mabs with synthetic peptides of the MUC1 tandem repeat were used to map the epitopes recognised by VA1 (amino acids RPAPGS) and VA2 (amino acids DTRPA). The use of fusion proteins provides another means of immunisation to produce anti-tumour antibodies.

Second-generation monoclonal antibodies to intestinal MUC2 peptide reactive with colon cancer

BACKGROUND

Antitumor antibodies have traditionally been made to whole tumors or tumor extract. The use of defined synthetic antigens would be desirable for producing monoclonal antibodies.

PURPOSE

Our purpose was to determine if antipeptide antibody to MUC2 had antitumor activity and specificity.

METHODS

A 29-amino-acid peptide to MUC2 was synthesized and monoclonal antibodies were produced after immunizing BALB/c mice with peptide-keyhole-limpet hemocyanin in complete Freund's adjuvant, and the monoclonal antibodies were tested on peptides and human tissues.

RESULTS

CCP31, CCP37, and CCP58 monoclonal antibodies were produced using MUC2 MI-29 (KYPTTTPISTTTMVTPTPTPTGTQTPTTT) containing one repeat unit of 23 amino acids and part of the next repeat of four amino acids. These antibodies reacted with the MUC2-derived peptide but not with MUC1- or MUC3-derived peptides. One of the monoclonal antibodies, CCP58, reacted strongly with human colon cancer and normal intestine in both fresh and formalin-fixed tissues; two other antibodies, CCP37 and CCP31, reacted only with fresh human tissues of normal colon and malignant colon tumors by immunoperoxidase staining. In addition, CCP37 and CCP58 reacted strongly with human gastric cancer; all antibodies reacted weakly with human salivary gland, and none reacted with tissues from normal human lung, kidney, stomach, pancreas, or endometrium. By analysis of mucin molecules by Western blotting, the antigen detected by monoclonal antibodies CCP37 and CCP58 was found to be of a high relative molecular mass (520 kd).

CONCLUSIONS

Anti-MUC2 peptide antibodies appear to be relatively tissue specific and represent a new method of producing antitumor antibodies.

Epitope mapping of anti-breast and anti-ovarian mucin monoclonal antibodies

Anti-breast cancer antibodies (BC2, HMPV and 4B6) and an anti-ovarian cancer antibody (OM1) were found to react with mucins--indeed with the protein core encoded by the MUC1 gene. This gene contains a VNTR (variable number of tandem repeats) encoding a 60 bp (= 20 amino acids) repeat sequence and within this amino acid sequence SAPDTRPAP was predicted, by hydrophilicity analysis, to be the immunogenic peptide sequence. The four antibodies were shown to react with MUC1 VNTR encoded peptides in direct binding and inhibition studies. The precise reactivity of the 4 mAbs was mapped using ELISA in both solid and liquid phase, and demonstrated the epitopes to be: APDTR (BC2 and HMPV), PDTR (4B6) and DTRPA (OM1). By using the pepscan method, the epitopes were shorter (PDTR, DTR and DTRP). However when these short peptides (except DTR) were synthesized they did not react; flanking amino acids are needed for the epitopes. Clearly several different methods should be used to define the reactive epitope. Within (S)APDTR, major amino acid substitutions could be made--even of three to four amino acids without altering antibody binding, provided that P and R were not substituted. It was of interest that an anti-ovarian cancer antibody gave similar anti-peptide reactions to the anti-breast cancer antibodies; apparently MUC1 peptides in ovarian cancer are the same as in breast cancer.

Second generation anti-MUC1 peptide monoclonal antibodies

Second generation antibodies to mammary mucins were produced by immunizing mice with a peptide with a sequence deduced from that of the MUC1 complementary DNA sequence (PAHGVTSAPDTRPAPGSTAP). Four monoclonal antibodies (BCP7-10) were produced which gave different reactions. BCP8 was similar in tissue reactivity (by immunoperoxidase staining) to anti-breast cancer or anti-human milk fat globule membranes (HMFG) antibodies and reacted strongly with most breast cancers and a proportion of other adenocarcinomas, whether formalin fixed or fresh, and reacted less strongly with some normal tissues. The three other antibodies (BCP7, BCP9, BCP10) reacted only with fresh tissues or a single cell line (LS174T of colon cancer origin) and gave variable weak reactions. Like many anti-mucin antibodies BCP8 reacted with HMFG, but more strongly with deglycosylated HMFG; analysis with peptides by enzyme-linked immunosorbent assay indicated reactivity with an epitope contained in the amino acid motif PDTR and using the pepscan method, the minimum epitope was DTR. MAbs BCP7, BCP9, and BCP10 did not react with HMFG; substantial reactions were obtained with deglycosylated HMFG for BCP7 and weaker reactions with BCP9 and BCP10. The finding that BCP7 reacted with breast cancer tissues and deglycosylated HMFG suggested that the epitope recognized by BCP7 was masked in native form and exposed in cancer, indicating that BCP7 could be a useful agent for analyzing differences between normal and cancer mucins. The amino acid epitopes for these antibodies were VTSA (BCP7), GSTAP (BCP9), and RPAP (BCP10). For BCP8, amino acid substitution analysis of SAPDTR indicated that substitutions were poorly tolerated (except Q for T and L/Y for R), contrasting with the substitution analysis of anti-mucin antibody reactions where virtually any amino acid can be substituted for T, indicating that in the native state T (threonine) may be O-glycosylated. The use of synthetic peptides to produce antibodies similar to those produced using crude mucins or tumor extracts represents a major advance in the production of antitumor reagents.

Expression of MUC1 and MUC2 mucins by human tumor cell lines

The secretion and nature of mucins produced from a panel of recently available new gastric and colon carcinoma cell lines (LIM1839, LIM1215, LIM1863, LIM1899, LIM2099, LIM2405, LIM2408, LIM2412, LIM2463), as well as other colon (LS174T, HT29, HT29-SB, COLO533, COLO206), breast (T47D, MCF-7, BT20, ZR75-1) and ovarian (COLO316) tumor cell lines, was investigated. ELISA and Western blotting of the culture supernatants with novel anti-MUC1 and anti-MUC2 monoclonal antibodies (MAbs) specific for mucin core proteins showed their secretion by most of these cell lines. In addition, mucins produced by these cell lines expressed the tumor-associated carbohydrate detected by MAb 3E1.2 (glycolylsialyl-Tn, mammary serum antigen or MSA) and the Tn or T antigens reactive with lectin SSA-M. SSA-M detected MUC1 or MUC2 captured by MAbs BC2 or CCP58, while 3E1.2 only detected MUC1-associated carbohydrate, indicating that the MAb may react with a conformationally dependent epitope, or that the sialyl/glycolyl-transferases involved in MSA production may be sequence specific. In addition, the BC2/SSA-M and CCP58/SSA-M assays detected mucins in some samples which were not detected by BC2/BC2 or CCP58/CCP58 dual determinant assays, indicating that this format may be more appropriate for the detection of tumor-associated mucins in body fluids. These new cell lines and assays should be of use in the investigation of mucin core proteins, particularly LIM2463 and LIM1839 which express significant quantities of both MUC1 and MUC2.

Production of MUC1 and MUC2 mucins by human tumor cell lines

A mucus secreting, clonal derivative (HT29-SB) of the human colonic adenocarcinoma cell line HT29, and the LS174T colon cancer cell line, secrete mucin into the culture medium as a viscoelastic gel. Mab BC2, which defines a peptide epitope present in the variable number of tandem repeats (VNTR) of the MUC1 core protein, reacted with this material after deglycosylation. Two high molecular weight bands were detected in TFMSA treated gel-formed mucin from HT29-SB and LS174T by western blotting (Mr 580 kDa and 420 kDa). A similar pattern of reactivity was seen with the culture supernatants from HT29-SB, the ovarian tumor cell line COLO-316, and the breast cancer cell line MCF-7. Mab CCP58 (anti-MUC2 VNTR) reacted with a 580 kDa band in gel-formed mucin produced by LS174T, but was not reactive with mucin produced by the other cell lines. The findings indicate that human colonic cell lines, in addition to breast and ovarian cell lines, may both express and secrete the MUC1 protein core, and that the LS174T cell line expresses and secretes both the MUC1 and MUC2 core proteins.

Effect of variations in peptide sequence on anti-human milk fat globule membrane antibody reactions

Monoclonal anti-mucine antibodies BC1, BC2 and BC3 produced using human milk fat globule membrane react with a synthetic peptide p1-24 (PDTRPAPGSTAPPAHGVTSAPDTR) representing the repeating amino acid sequence of the mucin core protein. The minimum epitope recognized by these three monoclonal antibodies (mAb) in p1-24 was contained in the five amino acids APDTR. To analyse the variation of position of the epitope, various modifications of the APDTR sequence were made by synthesizing peptides and testing by direct binding and inhibition enzyme-linked immunosorbent assays. Firstly, peptides p13-32 and C-p13-32, in which the epitope APDTR was placed in the middle instead of the C-terminal as in p1-24, were examined. These peptides had a greater reaction with mAb BC1, BC2 and BC3 compared with the reaction with p1-24. Secondly, A-p1-24 and TSA-p1-24 were made wherein two APDTR epitopes were present--these peptides were shown to bind two IgG antibody molecules. Finally, the contribution of each amino acid in the APDTR epitope was studied using the pepscan polyethylene rods, making all 20 of the amino acid substitutions in each position for SAPDTR (the minimum epitope APDTR with an adjacent amino acid S). In the 120 peptides examined there were some 'permissible' substitutions in A, D and T but not in P or R for BC1 and BC2; there were more 'permissible' substitutions for BC3; different substitution patterns were found with each antibody and some substitutions gave an increased reaction compared with the native peptide SAPDTR. The studies are of value in analysing the reaction of antibodies with epitopes expressed in breast cancer and in determining the antigenicity of synthetic peptides.

Characteristics of a breast cancer-associated antigen defined by RCC-1 antibody

This study reports the characterization of a breast cancer-associated antigen identified by murine monoclonal antibody (MoAb) RCC-1 (formerly called 24-17.1). Immunoperoxidase staining indicated that RCC-1 recognized an antigen highly expressed in malignant tumours of breast origin, and no reactivity was noted with connective tissue, muscle or lymph nodes, which is an important consideration in its successful use in immunolymphoscintigraphy. The RCC-1 was shown to consist of 94,000 dalton disulfide-bonded dimers which were shown to be different from the transferrin receptor. In addition, the antibody RCC-1 did not react with components of human milk or with an antigenic peptide derived from the core protein of a mammary mucin. Chemical treatment and enzymatic digestion suggested that the epitope recognized by antibody RCC-1 was protein as it was resistant to neuraminidase and periodate treatment but was sensitive to trypsin. The RCC-1-defined antigen detects a novel breast cancer associated antigen.

Anti-colorectal carcinoma monoclonal antibodies reactive with human milk fat globular membranes

Two monoclonal antibodies (MoAbs) were produced against human colorectal cancer and were found to react with many colorectal carcinomas (MoAb DH-1 detected 25 of 32 and MoAb SMA-1 detected 28 of 32) by the immunoperoxidase technique. These two antibodies were found to detect epitopes present on human milk fat globules (HMFG) by enzyme-linked immunosorbent assays (ELISAs). There have been few reports of anti-mucin antibodies being produced to colon cancer and so we examined the reactions of these and 13 other anti-HMFG MoAbs with colonic cancers. Thirty-two colonic tumours were examined and different staining patterns were noted. Staining was particularly marked on the cell membranes and glandular deposits of the tumours rather than in the cytoplasm of the cells. All colonic tumours tested were positive with at least one MoAb, but no single MoAb stained all tumours; this suggested that at least one epitope of HMFG could be found on all colonic tumours but no epitope detected by any one MoAb could be found on all colonic tumours. The 15 MoAbs were also tested on a panel of adult tissues, and with the exception of two, all had unique reaction patterns. Thus, at least 13 different epitopes associated with HMFG could be detected by the panel of MoAbs based on their different tissue distributions. However, no normal tissue expressed all the epitopes detected by all the MoAbs (i.e. none was positive with all the MoAbs) and no 'monomorphic' HMFG epitope could be found. The antibodies were shown to react with carbohydrate or peptide-based epitopes, but this reactivity had no relationship with the pattern of tissue reaction and the significance of the expression of different epitopes is not clear. Preliminary results of serum tests with three different anti-HMFG MoAbs showed that approximately 10% of colorectal cancer patients had elevated HMFG levels but with the cut-off level selected, 10% of normal subjects also had raised levels. From this study, antibodies to HMFG, commonly associated with breast cancer, can also react with carcinoma of the colon and can indeed be useful in detecting this disease histologically (e.g. MoAbs CC3, 4 and 5 which detect greater than 80% of colonic tumours). In addition it is clear that some of the mucins produced in carcinoma of the colon are similar to those produced by breast tumours.

Mucins in breast cancer: recent immunological advances

Mucins are heavily glycosylated proteins that are produced in excessive amounts in breast cancers and other adenocarcinomas. These proteins are potent immunogens; indeed, most monoclonal antibodies raised against extracts of breast cancer cells or tumors are directed toward a single family of mucins. The anti-mucin antibodies currently available are valuable diagnostic aids for detecting advanced stages of breast cancer. To improve the sensitivity of these antibodies so that they might be suitable for use in screening and early diagnosis, tumor imaging, and therapy of breast cancer, recent studies have focused on identifying the precise epitopes they recognize.

Synthetic peptides reactive with anti-human milk fat globule membrane monoclonal antibodies

Mammary mucins are increased in amounts in breast cancer patient sera, and most anti-breast cancer antibodies react with such mucins. One such mucin is found in human milk fat globule membrane and consists predominantly of O-linked sugars and a protein core. Partial complementary DNA clones for the protein core have recently been obtained. The nucleotide sequence is of interest as it contains a 60-base pair repeat, giving rise to a repeated 20-amino acid sequence (PDTRPAPGSTAPPAHGVTSA). Peptides with various lengths were synthesized using this sequence and the adjacent 4 amino acids (PDTR). Three anti-human milk fat globule membrane antibodies produced in our laboratory (BC1, BC2, and BC3) were tested to determine their reactivity with these synthetic peptides. Using three different assays (direct enzyme-linked immunosorbent assay test on peptides, direct enzyme-linked immunosorbent assay test on bovine serum albumin-conjugated peptides, and an inhibition test with the peptides in liquid, rather than solid phase), it was shown that APDTR was the minimum amino acid sequence required to form a reactive epitope with all 3 antibodies, although individual differences in the reactivities of the antibodies were noted. The addition of alanine (A) converted a nonreactive PDTR peptide to a reactive one, and the deletion of arginine (R) did the reverse; thus APDTR is the smallest peptide which reacts with these anti-human milk fat globule membrane antibodies.