pS2 expression as a possible diagnostic marker of colorectal carcinoma in ulcerative colitis

This study was performed to evaluate the significance of pS2 and MUC1 expressions in ulcerative colitis (UC)-associated colorectal neoplasias. Tissues were collected from 6 patients with UC-associated colorectal neoplasias treated surgically. Specimens were 13 adenocarcinomas, 40 dysplasias (20 high-grade dysplasias, 20 low-grade dysplasias), and 60 normal mucosae. Tissues were also collected from patients without UC treated surgically or endoscopically. pS2, p53, and MUC1 expressions were examined immunohistochemically and compared. The K-ras codon 12 mutation was investigated by single-strand conformation polymorphism analysis. In patients with UC, the incidence of pS2 expression was significantly higher (p<0.01) in adenocarcinomas than it was in dysplasias, and no pS2 expression was seen in normal mucosae. p53 overexpression was detected in 50% (10/20) even in low-grade dysplasias. MUC1 expression was seen only in invasive carcinomas, but it was seen in 100% of cases (3/3). K-ras gene mutations were detected in 2 (20%) of 10 carcinomas. In low and high-grade dysplasias, the incidences of pS2 expression were significantly (p<0.01) lower than the incidences of p53 overexpression, however, in UC-associated carcinomas there was no significant difference; pS2 expression and p53 overexpression were detected in 13 of 13 (100%) cases and in 12 of 13 (92%) cases, respectively. These results suggest that p53 overexpression may be a diagnostic marker of neoplasia, and that pS2 expression may be a diagnostic marker of colorectal carcinoma in case of UC.     

Identification of HLA-A2-restricted epitopes of the tumor-associated antigen MUC2 recognized by human cytotoxic T cells

Previous studies have shown that self-antigens overexpressed in malignant tissue can provide a basis for a tumor-specific immune response. The mucin MUC2 is strongly overexpressed in all mucinous tumors of colon, breast, ovary and pancreas. In the corresponding normal tissue it is either not expressed (breast, ovary, pancreas) or it is expressed at considerably lower levels than in the mucinous tumors (colon). We therefore investigated whether the MUC2 molecule comprises HLA-A2-binding epitopes recognized by human cytotoxic T cells. Four MUC2 peptides with high affinity and stable binding to HLA-A2 were identified. Those peptides and additionally 3 peptides with moderate binding to HLA-A2 were loaded onto dendritic cells, which were used for stimulation of autologous T cells from healthy donors. Two MUC2 peptides, which belonged to the group of stable binders, induced specific cytotoxic T-cell lines. Target cells loaded with these peptides were strongly lysed in a concentration-dependent and HLA-A2-restricted manner. Our data show that the tumor-associated mucin MUC2 has potential as a target antigen for cytotoxic T cells in patients with mucinous carcinomas. Int. J. Cancer 75:688–693, 1998.© 1998 Wiley-Liss, Inc.     

MUC6 expression in breast tissues and cultured cells: Abnormal expression in tumors and regulation by steroid hormones

Neoplastic transformation of epithelial cells is commonly associated with alterations in the expression of mucin genes. The mechanisms involved in this process are largely unknown. MUC6, isolated from a stomach cDNA library, is mainly expressed in stomach antral glands, as detected by using in situ hybridization and immunohistochemistry. We examined MUC6 expression in normal and pathological breast tissues using immunohistochemistry with MUC6-specific antibodies and in cultured breast cancer cells using immunocytochemistry and Northern blotting. MUC6 was generally not detected in normal breast (1/11) but was detected in fibrocystic disease without atypia (7/17, 41%), in atypical fibrocystic disease (11/11, 100%) and in carcinoma (57/60, 95%). To study the mechanisms involved in mucin gene up-regulation in breast cancer, we examined baseline, growth-related and steroid-induced levels of MUC1, MUC3 and MUC6 in 4 breast cancer cell lines, 2 of which express estrogen receptors. MUC6 levels were up-regulated at post-confluence in 2/4 cell lines, whereas no changes were detected for the other mucin genes examined. MUC6 and MUC3 were constitutively expressed, and steroid-induced, in BT-474 and MCF-7 cells, respectively. As a control, pS2 was induced in both cell lines. Our results indicate that (1) MUC6 is overexpressed in breast cancer and in benign breast disease, (2) in vitro, MUC6 and MUC3 are up-regulated by steroids and (3) abnormal expression of MUC6 in breast cancers may, in part, be explained by hormonal changes associated with tumor development. Int. J. Cancer 77:193–199, 1998.© 1998 Wiley-Liss, Inc.     

Mucin gene expression and cell differentiation in human normal, premalignant and malignant esophagus

Esophageal carcinoma includes squamous cell carcinoma and Barrett's adenocarcinoma. The latter usually develops from a premalignant lesion named Barrett's esophagus. MUC genes are known to be specifically expressed in the normal, premalignant and malignant epithelia of various tissues. The aim of this study was to establish the pattern of MUC gene expression in the esophageal mucosa under normal conditions, and under pathological conditions such as squamous cell carcinoma, Barrett's esophagus and adenocarcinoma. Samples of esophageal control mucosa, metaplastic and malignant tissues were obtained from 40 patients undergoing esophagectomy for squamous cell carcinoma (n = 17), or Barrett's esophagus with adenocarcinoma (n = 23). In situ hybridization and northern blot were used with probes specific for the MUC1, MUC2, MUC3, MUC4, MUC5AC, MUC5B, MUC6 and MUC7 genes to assess their expression in these samples. Submucosal glands of control esophageal mucosa expressed MUC5B, whereas MUC1 and MUC4 were found in both control epithelium and squamous cell carcinoma. MUC4 expression correlated with squamous cell differentiation. Barrett's adenocarcinoma exhibited various patterns of MUC gene expression, the strongest being in the well-differentiated mucinous adenocarcinomas. Barrett's metaplasia was also associated with a specific MUC gene expression pattern, since the gastric apomucin mRNAs, MUC5AC and MUC6, were expressed in gastric metaplasia, and the intestinal apomucin mRNAs, MUC3, MUC4 and mostly MUC2, in intestinal metaplasia. Residual expression of gastric apomucin mRNAs was found in intestinal metaplasia. From these results, we conclude that MUC genes can be considered reliable phenotypic markers of the esophageal cell differentiation, thus providing new insight into the development of Barrett's esophagus.     

Immunohistochemical study of MUC5AC expression in human gastric carcinomas using a novel monoclonal antibody

In order to investigate the expression of MUC5AC mucin in normal gastric mucosa and gastric carcinomas, we produced 3 monoclonal antibodies (MAbs) using a MUC5AC synthetic peptide. The immunohistochemical study was performed using one of these MAbs (CLH2) which reacted with the different designs of peptides based on the MUC5AC tandem repeat and with native and deglycosylated mucin extracted from gastric tissues. CLH2 immunoreactivity was restricted to foveolar and mucopeptic neck cells in normal gastric mucosa. No reactivity was observed in type-1 intestinal metaplasia. Out of 66 gastric carcinomas, 42 (63.6%) expressed MUC5AC. Most diffuse carcinomas were positive (83.3%), whereas only 59.3% of intestinal and 40.0% of atypical carcinomas expressed MUC5AC (p < 0.05). Gastric carcinomas with mixed pattern showed immunoreactivity in diffuse areas and decreased immunoreactivity in intestinal areas. Every early gastric carcinoma expressed MUC5AC, in contrast to 58.6% of advanced carcinomas (p < 0.05). A trend toward decreased immunoreactivity was observed in deep areas of advanced carcinomas in comparison with the respective superficial areas. Taking together the specific staining of foveolar and mucopeptic neck cells and the absence of immunoreactivity in intestinal metaplasia, we conclude that MUC5AC expression may be used as a marker of gastric differentiation. This assumption is further supported by the finding of MUC5AC immunoreactivity in most diffuse carcinomas, which usually display morphologic and histochemical signs of gastric differentiation. The expression of MUC5AC in early gastric carcinomas, regardless of their histologic type, suggests that all gastric carcinomas retain at least some cells with a gastric phenotype during the first steps of neoplastic development. Int. J. Cancer 74:112–121. © 1997 Wiley-Liss, Inc.     

Novel breast-tumor-associated MUC1-derived peptides: characterization in Db-/- x beta2 microglobulin (beta2m) null mice transgenic for a chimeric HLA-A2.1/Db-beta2 microglobulin single chain

The MUC1 protein was found to be up-regulated in a spectrum of malignant tumors. T-cell responses to the MUC1 extracellular tandem repeat array (TRA) were observed in murine models as well as in breast-carcinoma patients. In the present study, we evaluated the anti-tumor potential of HLA-A2.1-motif-selected peptides from non-TRA domains of the molecule. Peptide immunogenicity was examined in the Db-/- x beta2 microglobulin (beta2m) null mice transgenic for a modified HLA-A2.1/Db-beta2 microglobulin single chain (HHD mice). Our results show the existence of 3 novel HLA-A2.1-restricted MUC1-derived cytotoxic T-lymphocyte (CTL) epitopes. These peptides are processed and presented by the HHD-transfected breast-tumor cell line MDA-MB-157. Moreover, CTL induced by these 3 peptides show higher lysis of target cells pulsed with breast-carcinoma-derived peptides than of targets pulsed with normal breast-tissue-derived peptides. These data suggest an important role for non-TRA MUC1-derived peptides as inducers of a MHC-restricted CTL reaction to a breast-carcinoma cell line and patient-derived tumor extracts.     

Monoclonal antibodies to a MUC 4 peptide react with lung cancer

The cDNA of mucin 4 (MUC 4) has been cloned from human tracheo-bronchial cDNA library and MUG 4 was demonstrated to be expressed in normal tracheobronchial and colonic mucosa. However, as there are no antibodies available, there is little information on the distribution of MUC 4 on normal and malignant tissues. A 22 amino-acid peptide (M4.22), derived from the MUC 4 cDNA sequence was synthesized and used to immunise mice and 2 monoclonal antibodies were produced and characterized by ELISA, immunoperoxidase staining and flow cytometry. By ELISA the two anti-MUG 4 mAbs reacted with the immunizing peptide (M4.22), but not with peptides derived from other mucins. By immunoperoxidase staining the mAbs reacted with native mucin expressed in normal colon and to a lesser extent with stomach, but not with normal lung. By contrast, the mAbs reacted most strongly (similar to 70%) with lung cancers. Using six-mer overlapping peptides the minimum number of amino acids in the epitopes detected by the mAbs were TPL (M4.171) and PLPV (M4.275). MUC 4 detected by the mAbs was of a molecular mass (180 kDa) by Western blotting. Using anti-MUC 4 mAbs non-large cell lung cancers were found to be MUC 4 positive compared with normal lung tissues which were negative. By contrast, normal colon was strongly positive but the expression was decreased in colon cancer. These findings could have diagnostic and therapeutic implication for lung cancer.     

Expression of MUC1, MUC2 and oligosaccharide epitopes in breast cancer: prognostic significance of a sialylated MUC1 epitope

MUC1 represents a promising marker in breast cancer. However, due to the structural complexity of the MUC1 glycoprotein, multiple epitopes can be detected by monoclonal antibodies. This fact may be responsible for the contradictory results of previous investigations regarding the clinical and prognostic relevance of MUC1 expression in breast cancer. Therefore, we tried to evaluate the role of different glycosylated and non-glycoslyated MUC1 epitopes as well as other mucin-associated peptides (MUC2) and carbohydrates (Thomsen-Friedenreich antigen, sialyl-Lewisa, sialyl-Lewisx) as predictors of the clinical course and prognosis in mammary carcinomas. An immunohistochemical study applying numerous monoclonal antibodies (mabs) was performed to characterize the expression of a selected panel of MUC1 epitopes, and of MUC2, Thomsen-Friedenreich antigen, sialyl-Lewisa, and sialyl-Lewisx in a series of 140 patients with breast cancer. The results were correlated with clinicopathological variables as well as overall survival. Generally, more than 90% of the mammary cancers, were strongly stained with the MUC1-specific mabs. Especially ductal and lobular carcinomas were strongly MUC1- and sialyl Lewisa-positive, whereas MUC2 binding was significantly elevated in mucinous neoplasms. Associations between the immunoreactivity of any mab under study and tumor progression as reflected by pTNM staging could not be observed. However, expression of the sialylated MUC1 epitope detected by mab MY1.E12 was revealed as a favourable independent prognostic factor. These results confirm that MUC1 is generally strongly expressed in mammary carcinomas. As an exception, mucinous carcinomas are significantly less MUC1 reactive, but strongly express MUC2. Our data suggest that only the presence of a sialylated short-chain MUC1 glycoform is associated with a better prognosis, whereas the other molecules under study are not correlated with the course of disease and survival probability.     

Intestinal metaplasia of human stomach displays distinct patterns of mucin (MUC1, MUC2, MUC5AC, and MUC6) expression

Intestinal metaplasia is a well-established premalignant condition of the stomach that is characterized by mucin carbohydrate modifications defined by histochemical methods. The purpose of the present study was to see whether the expression of mucin core proteins was modified in the different types of intestinal metaplasia and to evaluate the putative usefulness of mucins as "molecular markers" in this setting. We used a panel of monoclonal antibodies with well-defined specificities to MUC1, MUC2, MUC5AC, and MUC6 to characterize the expression pattern of mucins. In contrast to normal gastric mucosa, the complete form or type I intestinal metaplasia (n = 20) displayed little or no expression of MUC1, MUC5AC, or MUC6 in the metaplastic cells and strong expression of the intestinal mucin MUC2 in the goblet cells of all cases. The incomplete forms of intestinal metaplasia, type II (n = 25) and type III (n = 16), expressed MUC1 and MUC5AC in every case, both in goblet and in columnar cells. MUC6 was also expressed in 16 cases of type II intestinal metaplasia and in 11 cases of type III intestinal metaplasia. The intestinal mucin MUC2 was expressed in every case of incomplete intestinal metaplasia, mostly in goblet cells. The mucin expression profile in the different types of intestinal metaplasia allows the identification of two patterns: one defined by decreased levels of expression of "gastric" mucins (MUC1, MUC5AC, and MUC6) and expression of MUC2 intestinal mucin, which corresponds to type I intestinal metaplasia, and the other defined by coexpression of "gastric mucins" (MUC1, MUC5AC, and MUC6) together with the MUC2 mucin, encompassing types II and III intestinal metaplasia. Our results challenge the classical sequential pathway of intestinal metaplasia (from type I to type III via a type II intermediate step).     

Monoclonal antibodies reacting with the MUC2 mucin core protein.

This study sought to produce monoclonal antibodies (MAbs) which reacted with the MUC2 core protein. Two MAbs [3A2 (IgG1) and 4F1 (IgM)] were produced by immunising female BALB/c mice with gel-formed mucin from the LS174T colon cancer cell line followed by a KLH conjugate of a 29 amino acid synthetic peptide whose sequence was derived from the variable number of tandem repeats (VNTR) region of a MUC2 cDNA clone. The MAbs reacted with synthetic MUC2 VNTR peptides but not synthetic MUC1 or MUC3 VNTR peptides, and showed specific reactivity in Western blotting with a high molecular weight protein produced by the LS174T colon carcinoma cell line. The use of shorter peptides indicated that the minimum peptide epitopes for these MAbs were different. Mab 3A2 reacted with amino acids 5-19 of the MUC2 VNTR by inhibition ELISA but not by direct ELISA, while 4F1 reacted with this peptide in both assays. Furthermore, 4F1 reacted in direct ELISA when a larger (29 amino acid) MUC2-derived peptide was coated onto the assay plate by incubating in carbonate buffer or by drying the peptide onto the assay plate, while 3A2 only reacted when this peptide was coated in carbonate buffer. The different specificity of the MAbs was also illustrated by the reactivity of 4F1 but not 3A2 with partially deglycosylated cystic fibrosis mucin. Immunohistochemical analysis with these MAbs revealed a strong reactivity with lung, gastric and colon tumours relative to normal tissue, with some breast and ovarian tumours also reacting. Both MAbs stained some normal goblet cells in the perinuclear region but not the mucin droplet or secreted mucin, indicating a reaction with immature (poorly glycosylated) mucin in the endoplasmic reticulum and/or golgi, but not with mature (fully glycosylated) mucin. In contrast, tumours showed strong diffuse cytoplasmic staining. 4F1 also showed weak apical cytoplasmic staining in some goblet cells and stained some tumours which showed no reactivity with 3A2. These antibodies should prove useful in the study of MUC2 structure and function, and in the diagnosis of some tumours.     

Mucin expression profile in Barrett's, dysplasia, adenocarcinoma sequence in the esophagus

BACKGROUND: The molecular events that accompany the progression to adenocarcinoma (ADC) of the esophagus are poorly understood. Aberrant mucin receptor expression can contribute to increased cell growth and metastatic ability. AIM: The aim of this study was to establish a pattern for mucin (MUC) gene expression in the esophageal mucosa under normal and pathological conditions. SETTING: University Hospital Cancer Center Laboratory. Archived tissue samples studied in a retrospective fashion. MATERIALS AND METHODS: Tissue samples were obtained from the archives of patients with histological evidence of Barrett's esophagus (BE) progressing to ADC. Immunohistochemical analysis was performed using mouse monoclonal antibodies for MUC1, MUC2, MUC5AC, MUC6. Semiquantitative scoring of histological staining was performed using a linear scoring system: 0-staining absent; 1-staining in 0-25%; 2-staining in 25-50%; and 3-staining in 50-75% of the epithelium. The Binomial test was used to explore trends and differences in frequency of mucin expression along the pathological sequence. RESULTS: Only mild superficial staining of MUC1 was seen in normal squamous epithelium. MUC1 and MUC2 were uniformly expressed in all samples (7/7) of BE and dysplasia (P=0.008). MUC1 expression was upregulated (7/7) in progression to adenocarcinoma (P=0.008). The secretory mucins, MUC5AC and MUC6 showed a decrease in expression with progression from BE to dysplasia to ADC (P< 0.05). CONCLUSIONS: Downregulation of MUC5AC and MUC6 decreases mucosal protection against gastric acid. Increasing MUC1 expression is associated with progression from dysplasia to ADC. Upregulation of MUC2 reflects intestinal metaplasia in BE.     

Lewis x is highly expressed in normal tissues: A comparative immunohistochemical study and literature revision

An immunohistochemical analysis was employed to determine the expression of carbohydrate antigens associated to mucins in normal epithelia. Tissue samples were obtained as biopsies from normal breast (18), colon (35) and oral cavity mucosa (8). The following carbohydrate epitopes were studied: sialyl-Lewis x, Lewis x, Lewis y, Tn hapten, sialyl-Tn and Thomsen-Friedenreich antigen. Mucins were also studied employing antibodies against MUC1, MUC2, MUC4, MUC5AC, MUC6 and also normal colonic glycolipid. Statistical analysis was performed and Kendall correlations were obtained. Lewis x showed an apical pattern mainly at plasma membrane, although cytoplasmic staining was also found in most samples. TF, Tn and sTn haptens were detected in few specimens, while sLewis x was found in oral mucosa and breast tissue. Also, normal breast expressed MUC1 at a high percentage, whereas MUC4 was observed in a small number of samples. Colon specimens mainly expressed MUC2 and MUC1, while most oral mucosa samples expressed MUC4 and MUC1. A positive correlation between MUC1VNTR and TF epitope (τ=0.396) was found in breast samples, while in colon specimens MUC2 and colonic glycolipid versus Lewis x were statistically significantly correlated (τ=0.28 and τ=0.29, respectively). As a conclusion, a defined carbohydrate epitope expression is not exclusive of normal tissue or a determined localization, and it is possible to assume that different glycoproteins and glycolipids may be carriers of carbohydrate antigens depending on the tissue localization considered.     

Mucin gene expression in colonic tissues and cell lines.

Complementary DNA clones encoding four different mucin core peptides have been isolated. However, the expression of these mucin genes in the colon has not been systematically studied. The present investigation used Northern blot analysis to study the expression of MUC1, MUC2, MUC3, and MUC4 mRNA in paired normal and cancerous colonic tissues, and nine colon cancer cell lines. Results were correlated with the clinicopathological features of the tumors and with the immunohistochemical expression of several carbohydrate tumor-associated antigens that may reside on mucins. MUC1 mRNA was expressed in all colonic tissues, and levels in paired normal and cancer tissues were similar in most cases. MUC2 and MUC3 mRNAs were expressed in both normal and cancer tissues, but levels were often decreased in the cancers. MUC4 mRNA was present in normal mucosa with comparable or sometimes greater expression in cancers. There was no apparent correlation between the expression of any particular mucin gene or pattern of mucin genes and the site, stage, or histological type of tumor. In addition, the expression of mucin-associated carbohydrate antigens did not correlate with any individual mucin gene or group of mucin genes. In colon cancer cell lines all four MUC genes were expressed rather weakly or not at all. These results indicate that the human colon expresses a broad repertoire of mucin genes which are differentially regulated in malignancy. Whether this differential regulation of mucin genes affect the behavior of the tumor and results in the altered glycosylation commonly seen in these requires further investigation.     

MUC1, MUC2, MUC5AC, and MUC6 expressions in gastric carcinomas: their roles as prognostic indicators.

BACKGROUND: Although mucin expressions appear to be correlated with prognoses in patients with various cancers, several studies have reported conflicting and inconclusive results on the prognostic significance of mucin expression in gastric carcinomas. METHODS: To clarify the correlations between clinicopathologic profiles and the patients' survival, the expression of MUC1, MUC2, MUC5AC, MUC6 mucins and the p53 protein were evaluated immunohistochemically in 300 consecutive gastric carcinomas using the tissue-array method. In addition, 59 gastric adenomas and 57 adenoma-associated carcinomas were investigated. RESULTS: MUC1 was expressed in 2 (3.4%) cases of gastric adenoma, and MUC2 in 19 (32.2%) cases of gastric adenoma, out of a total of 59 lesions. In consecutive gastric carcinomas, 24.3% of gastric carcinomas expressed MUC1, 27.3% expressed MUC2, 38.0% expressed MUC5AC and 12.7% expressed MUC6. The rate of MUC1 expression in gastric carcinomas was significantly higher than in associated gastric adenomas (P < 0.01). The patients with MUC1-positive carcinomas showed significantly poorer survival than those with MUC1-negative carcinomas. On the other hand, MUC2, MUC5AC and MUC6 expressions were not significantly associated with patient survival. Interestingly, combined evaluation revealed that the group with the MUC1-negative plus p53-negative expression pattern showed a better prognosis than the remaining cases. In contrast, the group with the MUC2-negative plus p53-positive pattern showed a worse prognosis. CONCLUSIONS: Mucin expression is altered in gastric adenoma and carcinoma, and MUC1 mucin expression is significantly associated with poorer outcome in gastric carcinomas. A MUC1-negative plus p53-negative pattern or a MUC2-negative plus p53-positive pattern may predict outcome in patients with gastric carcinomas.     

Identification of three non-VNTR MUC1-derived HLA-A*0201-restricted T-cell epitopes that induce protective anti-tumor immunity in HLA-A2/K(b)-transgenic mice

The human epithelial mucin MUC1 is over-expressed in more than 90% of carcinomas of the breast, ovary, and pancreas as well as in some other tumours, making it a potential target for tumour immunotherapy. We have identified several MUC1-derived peptides mapping outside the variable number tandem repeat region that comply with the peptide-binding motif for HLA-A*0201 and that become processed into stable major histocompatibility complex-peptide complexes as assessed by in vitro assays. In A2/K(b) transgenic mice, 3 peptides, namely MUC(79-87) (TLAPATEPA), MUC(167-175) (ALGSTAPPV) and MUC(264-272) (FLSFHISNL) elicit peptide-specific cytotoxic T lymphocyte (CTL) immunity, which protects these mice against a challenge with MUC1, A2/K(b)-expressing tumour cells. These peptides therefore represent naturally processed MUC1-derived CTL epitopes that could be used as components in peptide-based vaccines and for the analysis of anti-MUC1 CTL responses in A*0201-positive patients with MUC1-expressing tumours.