The gut microbiota, bile acids and their correlation in primary sclerosing cholangitis associated with inflammatory bowel disease

Background

Patients with primary sclerosing cholangitis associated with inflammatory bowel disease (PSC-IBD) have a very high risk of developing colorectal neoplasia. Alterations in the gut microbiota and/or gut bile acids could account for the increase in this risk. However, no studies have yet investigated the net result of cholestasis and a potentially altered bile acid pool interacting with a dysbiotic gut flora in the inflamed colon of PSC-IBD.

Aim

The aim of this study was to compare the gut microbiota and stool bile acid profiles, as well as and their correlation in patients with PSC-IBD and inflammatory bowel disease alone.

Methods

Thirty patients with extensive colitis (15 with concomitant primary sclerosing cholangitis) were prospectively recruited and fresh stool samples were collected. The microbiota composition in stool was profiled using bacterial 16S rRNA sequencing. Stool bile acids were assessed by high-performance liquid chromatography tandem mass spectrometry.

Results

The total stool bile acid pool was significantly reduced in PSC-IBD. Although no major differences were observed in the individual bile acid species in stool, their overall combination allowed a good separation between PSC-IBD and inflammatory bowel disease. Compared with inflammatory bowel disease alone, PSC-IBD patients demonstrated a different gut microbiota composition with enrichment in Ruminococcus and Fusobacterium genus compared with inflammatory bowel disease. At the operational taxonomic unit level major shifts were observed within the Firmicutes (73%) and Bacteroidetes phyla (17%). Specific microbiota-bile acid correlations were observed in PSC-IBD, where 12% of the operational taxonomic units strongly correlated with stool bile acids, compared with only 0.4% in non-PSC-IBD.

Conclusions

Patients with PSC-IBD had distinct microbiota and microbiota-stool bile acid correlations as compared with inflammatory bowel disease. Whether these changes are associated with, or may predispose to, an increased risk of colorectal neoplasia needs to be further clarified.

A rat model to study the role of STn antigen in colon cancer

Expression of the mucin-associated sialyl-Tn (STn) antigen has been associated with a decreased survival in patients with colorectal, gastric, and ovarian cancer. To better understand the role of STn antigen in tumor biology, we developed STn(+) (called LP) and STn(-) (called LN) clonal cell lines from a parental metastatic rat colon carcinoma cell line (LMCR). Both derivative cell lines exhibited identical proliferation rates in vitro. LP cells strongly expressed STn antigen both in vitro and in vivo, and were poorly tumorigenic when given to syngeneic rats. LN cells did not express STn antigen in vitro, but as in vivo tumors these cells rapidly acquired STn expression, readily formed tumors, and were highly lethal. When rats were given an otherwise lethal inoculum of i.p. LN cells, pre-immunization with synthetic STn antigen conjugated to keyhole limpet hemocyanin (STn-KLH) resulted in a 60% survival rate. When LN cells were injected subcutaneously in the presence of STn-KLH-sensitized lymphocytes, tumor growth was decreased. Distribution of STn antigen in normal organs of host rats is quite similar to that of humans. This model mimics human disease and should facilitate studies of mucin-associated antigens in tumor biology and the development of immunotherapeutic agents based on mucin-related antigens.

The Mount Sinai division of gastroenterology at the beginning of the 21st century

The Mount Sinai Division of Gastroenterology has an international reputation for outstanding contributions to the study of digestive diseases, especially inflammatory bowel disease. A discussion of the current structure of the gastroenterology (GI) fellowship training program is provided, along with an overview of the GI Division at the turn of the 21st century.

Redefined substrate specificity of ST6GalNAc II: a second candidate sialyl-Tn synthase

The acceptor substrate specificities of ST6GalNAc I and II, which act on the synthesis of O-linked oligosaccharides, were reexamined using ovine submaxillary mucin, [Ala-Thr(GalNAc)-Ala]n polymer (n = 7-11). It has been suggested that only ST6GalNAc I can synthesize carbohydrate structures of sialyl-Tn-antigen; i.e., NeuAc alpha2-6GalNAc-O-Thr/Ser [Kurosawa et al., J. Biol. Chem. 269, 19048-19053 (1994)] based on the result that ST6GalNAc I, not ST6GalNAc II, exhibited activity toward asialoagalacto-fetuin. In this study, we present evidence that both ST6GalNAc I and II exhibit activity toward asialo-OSM (ovine submaxillary mucin) and [Ala-Thr(GalNAc)-Ala]n polymer (n = 7-11) which have only the GalNAc-O-Thr/Ser-structures. These results strongly indicate that not only ST6GalNAc I but also II are candidates for sialyl-Tn synthases.

GLUT1 glucose transporter expression in colorectal carcinoma: a marker for poor prognosis

BACKGROUND

Malignant cells exhibit increased glycolytic metabolism, and in many cases increased glucose transporter gene expression. The authors hypothesized that GLUT1 glucose transporter expression is increased in colorectal carcinoma, and that the degree of expression might have prognostic significance.

METHODS

GLUT1 glucose transporter immunostaining was studied in normal colon and benign colon adenomas and in 112 colorectal carcinomas from patients for whom long term clinical outcome was known.

RESULTS

GLUT1 immunostaining was absent in normal colorectal epithelium and tubular adenomas, and absent or only weakly apparent in tubulovillous adenomas. The majority of carcinomas (101 of 112; 90%) had GLUT1 immunostaining. Tumors from 92 patients had low GLUT1 expression (< 50% of cells were GLUT1 positive) and 19 of these patients (21%) died of disease during follow-up. In contrast, tumors from 20 patients had high GLUT1 expression (> 50% of cells were GLUT1 positive) and 9 of these patients (45%) died of disease during follow-up. Disease specific mortality was greater in patients with high GLUT1 tumors (relative risk of 2.4; P=0.02). In a multivariate analysis to assess whether high GLUT1 staining correlated with increased mortality independently of Dukes stage, the risk of death from colon carcinoma in the group with high GLUT1 staining was 2.3 times that in the group with low GLUT1 staining, a difference that approached statistical significance (P=0.07).

CONCLUSIONS

GLUT1 glucose transporter expression is associated strongly with neoplastic progression in the colon, and assessment of the extent of GLUT1 immunostaining in colorectal carcinoma identifies patients with a poorer prognosis.

GLUT1 glucose transporter expression in colorectal carcinoma: a marker for poor prognosis

BACKGROUND

Malignant cells exhibit increased glycolytic metabolism, and in many cases increased glucose transporter gene expression. The authors hypothesized that GLUT1 glucose transporter expression is increased in colorectal carcinoma, and that the degree of expression might have prognostic significance.

METHODS

GLUT1 glucose transporter immunostaining was studied in normal colon and benign colon adenomas and in 112 colorectal carcinomas from patients for whom long term clinical outcome was known.

RESULTS

GLUT1 immunostaining was absent in normal colorectal epithelium and tubular adenomas, and absent or only weakly apparent in tubulovillous adenomas. The majority of carcinomas (101 of 112; 90%) had GLUT1 immunostaining. Tumors from 92 patients had low GLUT1 expression (< 50% of cells were GLUT1 positive) and 19 of these patients (21%) died of disease during follow-up. In contrast, tumors from 20 patients had high GLUT1 expression (> 50% of cells were GLUT1 positive) and 9 of these patients (45%) died of disease during follow-up. Disease specific mortality was greater in patients with high GLUT1 tumors (relative risk of 2.4; P=0.02). In a multivariate analysis to assess whether high GLUT1 staining correlated with increased mortality independently of Dukes stage, the risk of death from colon carcinoma in the group with high GLUT1 staining was 2.3 times that in the group with low GLUT1 staining, a difference that approached statistical significance (P=0.07).

CONCLUSIONS

GLUT1 glucose transporter expression is associated strongly with neoplastic progression in the colon, and assessment of the extent of GLUT1 immunostaining in colorectal carcinoma identifies patients with a poorer prognosis.

Liver metastasis and adhesion to the sinusoidal endothelium by human colon cancer cells is related to mucin carbohydrate chain length

Mucin production by human colon cancer cells correlates with liver metastasis in animal models, but it is not known which steps in metastasis depend on specific alterations in mucin synthesis. Clonal variants of cell line LS174T selected for differences in mucin core carbohydrate expression have been further characterized biochemically, and tested for their ability to participate in metastasis-related events. LS-C mucin contains truncated carbohydrates enriched for sialyl Tn and these cells bind to basement membrane matrix to a greater extent than LS-B cells. This binding is partially inhibitable by antibody to sialyl Tn. LS-B produces more fully glycosylated mucin and preferentially binds to hepatic sinusoidal endothelial cells and E-selectin through sialylated peripheral mucin-associated carbohydrate structures. Adhesion of LS-B to endothelial cells is inhibited by neutralizing antibody to E-selectin, and inhibition of glycosylation or desialylation of LS-B mucin abrogates binding to E-selectin in vitro. LS-B cells spontaneously metastasized from cecum to liver and colonized the liver of athymic mice after splenic-portal injection to a significantly greater extent than LS-C, suggesting that expression of peripheral mucin carbohydrate structures is most important for metastasis of human colon cancer cells.

Enzymatic basis for sialyl-Tn expression in human colon cancer cells

Sialyl-Tn antigen (SAalpha2-6 GalNAc alpha-Ser/Thr) is expressed as a cancer-associated antigen on the surface of cancer cells and its expression correlates with a poor prognosis in patients with colorectal and other adenocarcinomas. To understand the enzymatic basis of sialyl-Tn (STn) antigen expression, we used two clonal cell lines, LSB and LSC, derived from LS174T human colonic cancer cells. LSC cells express only the truncated carbohydrate antigen Tn (GalNAc alpha-Ser/Thr) and sialyl-Tn on their mucin molecules, whereas LSB cells express elongated oligosaccharide chains. Both cell lines demonstrated similar activities of glycosyltransferases involved in the biosynthesis of elongated and terminal structures of complex O-glycans. However, LSC cells were unable to synthesize core 1 (Gal beta1-3GalNAc-) because the ubiquitous enzyme activity of UDP-Gal:GalNAc-R beta3-Gal-transferase (core 1 beta3-Gal-transferase) was lacking. Core 1 beta3-Gal-transferase could not be reactivated in LSC cells by treatment with sodium butyrate or by in vivo growth of LSC cells in nude mice. In contrast, LSB cells were able to synthesize and process core 1 and core 2 (GlcNAc beta1-6 (Gal beta1-3) GalNAc-). LSC cells represent the first example of a non-hematopoietic cell line which lacks core 1 beta3-Gal-transferase activity. The lack of core 1 beta3-Gal-transferase in LSC cells explains why they are incapable of forming the common mucin O-glycan core structures and are committed to synthesizing the short Tn and STn oligosaccharides. These findings suggest that the activity of core 1 beta3-Gal-transferase is an important determinant of the STn phenotype of colon cancer cells.

Liver metastasis and adhesion to the sinusoidal endothelium by human colon cancer cells is related to mucin carbohydrate chain length

Mucin production by human colon cancer cells correlates with liver metastasis in animal models, but it is not known which steps in metastasis depend on specific alterations in mucin synthesis. Clonal variants of cell line LS174T selected for differences in mucin core carbohydrate expression have been further characterized biochemically, and tested for their ability to participate in metastasis-related events. LS-C mucin contains truncated carbohydrates enriched for sialyl Tn and these cells bind to basement membrane matrix to a greater extent than LS-B cells. This binding is partially inhibitable by antibody to sialyl Tn. LS-B produces more fully glycosylated mucin and preferentially binds to hepatic sinusoidal endothelial cells and E-selectin through sialylated peripheral mucin-associated carbohydrate structures. Adhesion of LS-B to endothelial cells is inhibited by neutralizing antibody to E-selectin, and inhibition of glycosylation or desialylation of LS-B mucin abrogates binding to E-selectin in vitro. LS-B cells spontaneously metastasized from cecum to liver and colonized the liver of athymic mice after splenic-portal injection to a significantly greater extent than LS-C, suggesting that expression of peripheral mucin carbohydrate structures is most important for metastasis of human colon cancer cells.

Amaranthin lectin binding in the rat colon: response to dietary manipulation

BACKGROUND

In human colon, binding of the lectin Amaranthus caudatus has been considered to be a marker of cellular proliferation and malignant progression. We studied regional amaranthin binding in rat colon and correlated this with physiologic manipulations of proliferation.

METHODS

Binding of amaranthin in segments of proximal and distal colon was studied in starved, refed, and control Wistar rats and was compared to tritiated thymidine labeling and proliferating cell nuclear-antigen expression.

RESULTS

Amaranthin bound mainly to cells in the lower crypt of distal colon and midcrypt of proximal colon, paralleling the distribution of proliferative markers. Binding occurred in the supranuclear region in distal colon and the pericellular membrane in proximal colon. Starvation/refeeding was associated with a change in amaranthin binding intensity in distal colon, but not in proximal colon.

CONCLUSION

The pattern of amaranthin binding during starvation/refeeding seems to reflect physiologic changes in several areas of the colon.

Different modes of sialyl-Tn expression during malignant transformation of human colonic mucosa

Monoclonal antibodies TKH2 and B72.3, which react with the mucin-associated sialyl-Tn(STn) antigen, preferentially bind to cancerous but not normal colonic tissues. If O-acetyl groups are removed by saponification of tissues, MAb TKH2 will react with normal colonocytes, whereas MAb B72.3 remains non-reactive. To explain this difference in binding specificity, we tested both MAbs against synthetic constructs of single (monomeric) or clustered (trimeric) STn epitopes by enzyme immunoassay. Both MAb TKH2 and MAb B72.3 reacted with trimeric STn, but MAb TKH2 demonstrated greater binding than MAb B72.3 to monomeric STn. This suggests that normal colonic mucosa expresses monomeric STn epitopes, but that with transformation to malignancy, clustered STn epitopes appear. The appearance of clustered STn epitopes during colonic carcinogenesis represents a novel pattern of carbohydrate antigen expression and implicates alterations at the level of apomucins and/or glycosyltransferases responsible for cluster epitope formation.

Intestinalization of gastric signet ring cell carcinomas with progression

Recent developments in mucin histochemistry and immunohistochemistry have made reliable determination of the gastric and intestinal phenotypes of gastric carcinoma cells possible. Phenotypic expression changes from gastric epithelial cell type to intestinal epithelial cell type with the growth of gastric tumours in experimental animals. We studied cell differentiation in gastric signet ring cell carcinomas with progression in 203 surgically obtained specimens. The results showed that the proportion of gastric phenotype carcinomas, in which over 90% of the tissue consists of gastric epithelial cell type cells, decreases with the depth of invasion. The proportion of mixed phenotype carcinomas (between 10% and 90% of the tissue made up of gastric and/or intestinal epithelial cell type cells) increases. The intestinal phenotype (over 90% intestinal epithelial cell type carcinoma cells) was found in four carcinomas (about 2%) involving the serosa. No clear relationship was evident between phenotypic expression of carcinoma cells and the degree of intestinal metaplasia of the surrounding mucosa. Progression of gastric signet ring cell carcinomas is associated with a phenotypic shift from gastric to intestinal type expression.

Inflammatory bowel disease and cancer

Colorectal cancer is a significant clinical problem for patients with long-standing ulcerative colitis and Crohn's disease. Traditional risk factors include long disease duration and greater extent of colonic disease, but newer factors such as associated primary sclerosing cholangitis, folate deficiency, and family history of colon cancer may help to refine risk stratification. Molecular pathogenesis of colitis-associated colon cancer shares some of the same genetic abnormalities as sporadic colon cancer, but the timing of certain alterations suggests different carcinogenic pathways. Some molecular alterations show promise as complementary markers to dysplasia. Clinical studies of colonoscopic surveillance indicate that colon cancers can be detected early and that mortality may therefore be improved. A suggested surveillance strategy is proposed.

Expression of MUC2 and MUC3 mRNA in human normal, malignant, and inflammatory intestinal tissues

MUC2 and MUC3 are prominent mucin genes expressed in the human intestine. Using in situ hybridization with RNA probes, we examined the cellular distribution of MUC2 and MUC3 mRNA in normal, malignant, and inflammatory human intestinal tissues. In normal small intestine and colon, MUC2 mRNA was expressed exclusively in goblet cells and occurred throughout the entire height of the mucosa. MUC3 mRNA was expressed by goblet and columnar cells but was restricted to the villous compartment of the small intestine and the surface epithelium of the colon. Expression of MUC2 and MUC3 mRNA were both markedly decreased in poorly, moderately, and well-differentiated colon cancers but were preserved in mucinous colon cancers. In ulcerative colitis and Crohn's colitis tissues, MUC2 and MUC3 mRNA expression displayed a normal pattern regardless of whether the mucosa manifested active or quiescent inflammation. These findings indicate that MUC2 is goblet cell-specific, whereas MUC3 is related to maturation of intestinal epithelial cells. In colon cancers, the genetic regulation of MUC2 and MUC3 is different depending on the histological type of tumor. The constitutive expression of MUC2 and MUC3 mRNA in inflammatory bowel diseases suggests that these genes may be necessary for maintenance of normal epithelial cell function during inflammation.

Gastrointestinal adenomatous polyps

Adenomatous polyps of the gastrointestinal tract are dysplastic precursor lesions of adenocarcinoma. The features of an adenoma that are associated with a greater tendency to progress to carcinoma include larger polyp size, high grade dysplasia, and increased villous glandular architecture. Alterations in particular oncogenes and tumor-suppressor genes have been correlated with various stages of colonic adenoma formation, thereby facilitating our knowledge of adenoma pathogenesis. Heredity and environment contribute to the risk of developing adenomatous polyps of the colon. The incidence of colorectal cancer can be decreased by the endoscopic removal of adenomas, thereby providing good rationale for screening and surveillance programs.

Sialosyl-Tn antigen is prevalent and precedes dysplasia in ulcerative colitis: a retrospective case-control study

BACKGROUND & AIMS

Expression of the carbohydrate-associated sialosyl-Tn (STn) antigen has been correlated with carcinogenesis in sporadic colon cancer. This retrospective study analyzed surveillance colonoscopy biopsy specimens to determine whether STn antigen could serve as a marker for neoplasia risk in patients with long-standing ulcerative colitis.

METHODS

Eleven patients who developed dysplasia or cancer and who had undergone at least three surveillance colonoscopies were matched with 11 controls who had not developed neoplasia. Sections from 969 available surveillance biopsy specimens were stained immunohistochemically with monoclonal antibody TKH2 and were interpreted blindly.

RESULTS

Compared with controls, patients expressed STn antigen more frequently in their biopsy specimens (44% patient biopsy specimens vs. 11% control biopsy specimens). Patients were also more likely to undergo colonoscopies with two or more colonic segments expressing STn antigen (73% vs. 22%), repeated STn antigen expression in two or more consecutive colonoscopies (91% vs. 27%), and STn antigen expression proximal to the hepatic flexure (64% vs. 0%). STn antigen appearance preceded the detection of neoplasia by 7 or fewer years, and its preferential expression in patients was not confounded by the degree of active inflammation.

CONCLUSIONS

Repeated STn antigen expression in nondysplastic colonic mucosa may presage the development of neoplasia in long-standing ulcerative colitis and may be a useful adjunct to dysplasia during colonoscopic surveillance.

Immunogenicity of synthetic TF-KLH (keyhole limpet hemocyanin) and sTn-KLH conjugates in colorectal carcinoma patients

Mucins of colorectal carcinomas overexpress the cancer-associated disaccharides Thomsen-Friedenreich antigen (TF) and sialyl-Tn antigen (sTn), making these antigens suitable for active specific immunotherapy. Patients at high risk for recurrent colon cancer, but free from disease after surgical resection, were immunized with synthetic TF and sTn covalently attached by a two-carbon crotyl linker to keyhole limpet hemocyanin (KLH). Four groups of patients were treated with TF-KLH without adjuvant, TF-KLH plus the immunological adjuvant Detox, sTn-KLH plus Detox, or sTn-KLH plus the immunological adjuvant QS-21, and the serological response was monitored. Enzyme-linked immunosorbent assay (ELISA), dot-blot immunostains, and inhibition assays were used to identify antibody responses against synthetic TF and sTn epitopes and against natural antigens, including asialoglycophorin expressing TF antigen, and ovine submaxillary mucin and the human colon cancer line LS-C expressing sTn antigen. Our results demonstrate that vaccines containing TF or sTn-KLH conjugates plus immunological adjuvants Detox and especially QS-21 induced high IgM and IgG antibody titers against the respective synthetic disaccharide epitopes. However, when tested against natural antigens expressing these disaccharide epitopes, IgM antibodies showed weak to moderate reactivity, while IgG antibodies were almost totally unreactive. On the basis of these results we are continuing to test modifications of synthetic TF and sTn epitopes to identify those that induce IgM and IgG antibodies that are more reactive with these antigens as they are expressed on tumor mucins.

Sialosyl-Tn antigen: initial report of a new marker of malignant progression in long-standing ulcerative colitis

BACKGROUND & AIMS

Expression of the mucin-associated carbohydrate antigen sialosyl-Tn (STn) correlates with malignant transformation in sporadic colonic neoplasms. The aim of this study was to analyze STn antigen expression in patients with long-standing ulcerative colitis (UC).

METHODS

STn antigen was assessed by immunohistochemistry in archival tissues. Study A was a retrospective chronological case-control study. Serial surveillance colonoscopic biopsy specimens without inflammation or dysplasia were analyzed in 7 patients who developed colon cancer and in 8 controls who did not develop colon cancer. Study B analyzed the anatomic distribution of STn expression in 17 cancer-bearing (case) and 6 cancer-free (control) colectomy specimens from patients with UC. In some colectomy specimens, STn was compared with aneuploidy, which was determined by flow cytometry.

RESULTS

In study A, among the 7 patients with UC who developed cancer, 6 patients (86%) expressed STn in at least one prior nondysplastic surveillance biopsy specimen from the same site. Only 3 of 8 control patients (38%) expressed STn. In study B, STn was expressed in 40 of 82 specimens (49%) from cancer-bearing colons but only 8 of 62 specimens (13%) from cancer-free colons. STn was expressed in most aneuploid areas but was also found in diploid, nondysplastic mucosa.

CONCLUSIONS

STn antigen seems to be a promising marker of cancer risk in patients with UC.

Immune sera and monoclonal antibodies define two configurations for the sialyl Tn tumor antigen

Sialyl Tn (sTn) is a mucin-associated carbohydrate antigen expressed in most types of human adenocarcinoma. Defining the configuration of tumor cell surface sTn recognized by antibodies is important for understanding the basis for the cancer cell specificity of sTn-reactive mAbs, for the development of more effective mAbs, and for designing cancer vaccines against sTn. In this study, we compared the immunogenicity of synthetic single sTn disaccharide epitopes and clusters [sTn(C)] of 3 sTn epitopes covalently linked via serine to keyhole limpet hemocyanin [KLH; sTn-KLH and sTn(C)-KLH, respectively]. The cell surface sTn configurations were analyzed with the use of sera from mice immunized with these neoglycoproteins and a panel of sTn-reactive mAb. Sera from mice immunized with sTn-KLH reacted in direct and inhibition assays with sTn-human serum albumin (HSA) but only weakly with sTn(C)-HSA, whereas sera from mice immunized with sTn(C)-KLH reacted with sTn(C)-HSA but not with sTn-HSA. Both anti-sTn and anti-sTn(C) sera reacted with ovine submaxillary mucin (a natural source of sTn) and with sTn-positive human tumor cell line LS-C but not with sTn-negative LS-B cells. With regard to the sTn-reactive mAbs, B72.3 reacted exclusively with clustered sTn, whereas mAb B195.3R11 reacted preferentially with unclustered sTn. Results with mAbs TKH2, B239.1, and CC49 were less clear, although all reacted more strongly with clustered sTn than with unclustered sTn. These results suggest that sTn is recognized at the tumor cell surface in at least two quite distinct configurations, as clustered and nonclustered arrays.

Tumor-associated sialylated antigens are constitutively expressed in normal human colonic mucosa

Immunohistochemical studies have indicated that sialylated carbohydrate antigens such as sialyl-Tn, sialyl-Le(a), and sialyl-Le(x) are expressed in a tumor-associated fashion in human colon. Since sialic acid residues are O-acetylated more extensively in normal colonic epithelium than in colon cancer cells, we examined whether deacetylation of colonic tissues might enable monoclonal antibodies to recognize these tumor-associated sialylated antigens. In normal colon, deacetylation turned most cases (82%) positive with anti-sialyl-Tn mAb TKH2; and in colon cancers, it increased the number of TKH2-positive cells. Sialyl-Le(a) and sialyl-Le(x) detection was also increased after deacetylation of normal and malignant colonic tissues so that the frequency of positive cases in normal tissues was similar to that in the cancers. However, in the stomach and pancreas, the same treatment rarely increased the detection of the sialylated epitopes in normal or cancerous tissues. Thus, the same sialylated epitopes can be expressed in a tumor-associated fashion by different mechanisms in different gastrointestinal organs; in the colon, these antigens are constitutively expressed and O-acetylated, whereas in the upper gastrointestinal tract, they are rarely O-acetylated, suggesting that other mechanisms such as differences in glycosylation account for the cancer-associated expression.

Monoclonal antibody COL-1 reacts with restricted epitopes on carcinoembryonic antigen: an immunohistochemical study

We used a monoclonal antibody, MAb COL-1, which recognized a restricted epitope on the carcinoembryonic antigen (CEA) molecule, to stain a wide variety of human normal and cancerous tissues. None of the 35 different types of normal tissue stained with COL-1. Of 59 types of benign and malignant tissues, COL-1 reacted with neoplasms of epithelial origin, especially the gastrointestinal tract, breast, lung, and bladder. In benign adenomatous colon polyps, villous adenomas were more frequently stained than tubular adenomas. Normal colon tissue from individuals without colon disease was unreactive, but very weak reactivity was noted in normal-appearing mucosa several centimeters remote from colon cancers. In contrast, another anti-CEA antibody with a less restricted epitope reacted frequently with both normal and remote colon mucosa. These results indicate that MAb COL-1 recognizes a restricted CEA epitope expressed only on pre-malignant or malignant cells and therefore may be a useful reagent for immunopathology.

Use of model cell lines to study the biosynthesis and biological role of cancer-associated sialosyl-Tn antigen

Sialosyl-Tn (STn) is a mucin-associated carbohydrate antigen that is not expressed by most normal epithelial cells but becomes expressed in several types of adenocarcinomas, where it is often associated with a poor prognosis. Little is known about the regulation of the STn phenotype in tumor cells and the immune response to STn antigen. In the present study, we established clonal cell lines in which virtually all of the cells were STn positive (designated LS-C) or STn negative (designated LS-B). These two cell lines, derived from a single parental cell line, LS174T, have the same total protein electrophoretic profiles but carry markedly different oligosaccharide structures on their mucin; the mucin from LS-C cells has only the Tn and STn structures, whereas LS-B cell mucin lacks these simple structures but carries more complex oligosaccharides. These results indicate that lack of STn expression by cells can be due to the lack of STn synthesis rather than inaccessibility of antibodies to bind to STn by steric hindrance. Both clones were similar in their growth rates, response to gamma-interferon, and sensitivity to lysis by lymphokine-activated killer cells. These cells may be important models for understanding the regulation of glycosylation at the cellular level and for further studies of tumor biology and immune response to STn antigen.

Mucin-associated sialosyl-Tn antigen expression in gastric cancer correlates with an adverse outcome

The expression of sialosyl-Tn (STn) antigen was evaluated by immunohistochemistry in primary gastric cancers. Twenty-one of 31 (68%) gastric cancers expressed STn, regardless of tumour location, stage or histological type. Eighty-one per cent of patients with STn-positive tumours died of their disease or had recurrent cancer, compared with 20% of patients with STn-negative tumours (P < 0.002). STn may be a useful prognostic marker in patients with gastric cancer.

Expression of sialosyl-Tn in intestinal type cancer cells of human gastric cancers

A sample of 219 primary stomach cancers, 143 advanced cancers and 76 early cancers were examined for mucin histochemical staining (the paradoxical concanavalin A method, the galactose oxidase-Schiff [GOS] reaction, and the sialidase-GOS reaction) and immunohistochemical reactivity (pepsinogen [Pg] I, Pg II, SH-9 and TKH-2). Gastric cancer cells were clearly classified according to mucin histochemistry into a gastric type, including mucus neck cell, pyloric gland cell and surface mucus cell types, and an intestinal type, including goblet-cell, and intestinal absorptive cell types. TKH-2 monoclonal antibody, which recognizes the mucin-associated sialosyl-Tn antigen, reacted with the mucin of goblet cells in both the normal small intestine and in the intestinal metaplasia of the stomach. Sixty-five of 106 (61%) differentiated adenocarcinomas and 76 of 113 (67%) undifferentiated adenocarcinomas had over 10% of their cancer cells positive for TKH-2. The TKH-2-positive cancers were primarily classified as a goblet-cell type by mucin-histochemical staining and the other immunohistochemical staining methods. Therefore, it is concluded that sialosyl-Tn is an excellent marker of small intestinal mucins and is indicative of a small intestinal type of differentiation in two-thirds of gastric cancers.

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.

Mucin oligosaccharide biosynthesis in human colonic cancerous tissues and cell lines

BACKGROUND

The Tn, sialosyl-Tn, and T antigens are carbohydrate-associated antigens that represent initial steps in mucin O-linked glycosylation. Previous immunohistochemical studies have shown that these three antigens are rarely, if ever, expressed in normal colonic mucosa; however, most colonic cancerous tissues express these structures. Little is known about the factors that control the expression of these antigens in colonic tissues or cell lines. One hypothesis is that cancers have increased levels of the glycosyltransferase activities responsible for synthesizing these antigens.

METHODS

The current study analyzed antigen expression by immunohistochemistry and glycosyltransferase enzyme activities for Tn, sialosyl-Tn, and T antigens in colonic tissues and cell lines to (1) compare values between normal and cancerous tissues and (2) correlate these results with tumor stage, histologic findings, and location.

RESULTS

All nine colonic cancer cell lines expressed Tn antigen; sialosyl-Tn and T antigens were expressed by the more mucin-producing cell lines. Sialosyl-Tn transferase activity was higher in the more mucinous cell lines; T transferase activity was higher in those with less mucin. In paired specimens of normal and cancerous tissues, levels of each of the three glycosyltransferases were similar. In cancerous tissues, enzyme activity did not correlate with tumor location, stage, or histologic type. There also was no correlation between glycosyltransferase activity and expression of the relevant antigen.

CONCLUSIONS

Thus, because normal and malignant colonic tissues have comparable levels of Tn, sialosyl-Tn, and T glycosyltransferases, the absence of these antigens in normal mucosa apparently is related to other factors such as antigen masking.

Mucins bearing the cancer-associated sialosyl-Tn antigen mediate inhibition of natural killer cell cytotoxicity

The sialosyl-Tn (STn) antigen is a mucin-associated carbohydrate antigen expressed by a variety of adenocarcinomas. In the colon, expression of this antigen has been associated with a poor prognosis, independent of tumor stage or histology. The present study was performed to determine whether this adverse clinical outcome might be due to an interaction between STn-positive mucin and natural killer (NK) cell cytotoxicity. Ovine submaxillary mucin (OSM), a mucin highly rich in STn antigen, partially inhibited NK cell cytotoxicity against K562 target cells, but only at high concentrations. Low concentrations of OSM were not inhibitory but became markedly inhibitory in the presence of ammonium ions. Two other STn-positive submaxillary mucins also markedly inhibited NK cytotoxicity when combined with ammonium ions. Removal of sialic acid from OSM reversed the OSM/ammonium-mediated inhibition of NK cell activity. Unlike the submaxillary mucins, two mucins derived from human breast and lung cancer cells which lack the STn antigen, did not inhibit NK cell activity in this system. Likewise, four other non-mucin glycoproteins which lack STn expression did not inhibit NK cells despite having levels of sialic acid that were, in some cases, comparable to submaxillary mucin. These results indicate that mucins bearing the cancer-associated STn antigen can effectively inhibit NK cell cytotoxicity in the presence of ammonium ions. While this NK cell inhibition is likely to be caused by ammonium, mucin markedly enhances this effect, thereby implicating a novel immunomodulatory property of mucin.

Mucin associated Tn and sialosyl-Tn antigen expression in colorectal polyps

Sialosyl-Tn antigen and its immediate precursor, Tn antigen, are carbohydrate structures associated with the earliest steps of mucin O-linked glycosylation. Both antigens have been shown previously to be highly sensitive and specific markers of colorectal cancer. One hundred and three colorectal polyps (79 adenomatous; 24 hyperplastic) were examined for expression of Tn antigen using vicia villosa isolectin B4, and for sialosyl-Tn antigen by monoclonal antibody TKH2. Tn antigen was expressed by all of the polyps studied. Sialosyl-Tn, on the other hand was expressed weakly by a few cells in 7 of 24 (29%) hyperplastic polyps. Among the adenomatous polyps, 56% expressed sialosyl-Tn and expression correlated with larger adenoma size, greater villous component, and more severe grades of dysplasia. In individuals with two or more synchronous adenomas, the level of sialosyl-Tn expression within an adenoma was associated with the severity of cytological atypia. All the adenomas that contained a focus of invasive carcinoma expressed sialosyl-Tn. These results indicate that colorectal polyps manifest incomplete glycosylation, exposing antigens in the innermost region of mucin oligosaccharides. In addition, the correlation of sialosyl-Tn antigen expression with the adenoma-carcinoma sequence may make this a useful marker for studying malignant progression in the colon.

Blood group-related carbohydrate antigen expression in malignant and premalignant colonic neoplasms

Cell surface glycoconjugates of colonic epithelial cells carry certain carbohydrate antigens related to blood group substances. During the progression to malignancy, these oligosaccharide immunodeterminants undergo specific types of alterations. In colon cancers, the blood group antigens A, B, H, and Le(b), which are normally expressed only in the proximal colon, can be re-expressed in distal colon cancers or deleted in proximal colon cancers. Also, an antigen which is incompatible with the individual's blood type can be expressed. Similar alterations occur in adenomatous polyps, but with reduced frequency. The simple form of blood group-related Le(x) and Le(y) antigens found in normal mucosa can undergo modification by oligosaccharide elongation, internal fucosylation, and sialylation into novel structures found in carcinomas as well as in adenomas with greatest malignant potential. Finally, antigens representing the first steps of glycosylation, Tn, T, sialosyl-Tn (STn), which are normally cryptic in the colon, can be unmasked due to incomplete glycosylation in adenomatous polyps and cancers. Several of these antigens, such as extended Le(x), extended Le(y), T, and sialosyl-Tn, are quite cancer-specific in that they are rarely expressed in normal mucosa or hyperplastic polyps, but preferentially occur in adenomas of greatest malignant potential. As such, these antigens might be useful as candidate intermediate endpoint biomarkers.

Expression of native and deglycosylated colon cancer mucin antigens in normal and malignant epithelial tissues

Immunohistochemical techniques were used to determine the distribution and cellular location of the mature and precursor forms of a colonic-type mucin in normal and malignant epithelial tissues. The antisera used in this study were prepared against native human colon cancer mucin (LS), partially deglycosylated mucin (HFA or GalNAc-apomucin), and fully deglycosylated mucin (HFB or apomucin). These antisera reacted with most mucin-producing cells of the normal gastrointestinal tract, salivary ductular cells, bronchial epithelial cells, some bronchial mucous glands, and squamous epithelial cells of the esophagus. Breast, endometrium, ovary, prostate, liver, and thyroid were nonreactive. In most normal organs, HFB reactivity was present in the supranuclear and perinuclear cytoplasm and LS and HFA were located primarily in goblet cell vacuoles, apical cytoplasm, and luminal secretions. These findings are consistent with the expected subcellular locations of apomucin and more "mature" mucins. LS, HFA, and HFB were frequently expressed in adenocarcinomas of the colon, stomach, pancreas, and lung. Lymphoma, sarcoma, and melanoma specimens were nonreactive. Alterations in the expression of these mucin antigens in malignant tissues included loss of subcellular compartmentalization, increased intensity of staining, and disappearance of staining. In addition, de novo expression of HFB was observed in one of five breast carcinomas and three of five ovarian mucinous cystadenocarcinomas. These data demonstrate that LS, HFA, and HFB are useful for studying the organ specificities and biosynthetic pathways of one type of mucin in normal and malignant tissues.

Sialosyl-Tn. A novel mucin antigen associated with prognosis in colorectal cancer patients

Colon cancers typically produce mucin. However, it is not known whether tumor mucin plays a biological role in cancer cell behavior. To address this issue, the expression of a mucin-associated antigen, sialosyl-Tn, was examined by immunohistochemical study in 128 primary colorectal carcinoma specimens from 137 patients who underwent curative surgical resection. Antigen expression was correlated with disease-free and overall 5-year survival. Sialosyl-Tn antigen expression occurred in 112 (87.5%) tumors, and was independent of age, gender, tumor location, Dukes' stage, depth of invasion, degree of differentiation, and ploidy status. Survival at 5 years for patients with sialosyl-Tn-negative versus sialosyl-Tn-positive tumors was 100% versus 73% (P less than 0.05) and disease-free survival was 94% versus 73%, respectively (P = 0.12). Although more advanced Dukes' stage, deeper invasion, and aneuploidy were all associated with poorer overall 5-year survival, antigen-negative tumors within each of these groups had much better prognoses than antigen-positive tumors. Multivariate regression analysis revealed that tumor ploidy (P less than 0.001) and sialosyl-Tn expression (P less than 0.05) were the two variables of most importance for predicting both disease-free and overall survival. The authors conclude that sialosyl-Tn expression is an independent predictor of poor prognosis in colon cancer, and therefore suggest that qualitative mucin alterations may reflect important differences in the biological behavior of these neoplasms.

Blood group antigen synthesis and degradation in normal and cancerous colonic tissues

ABH antigens are expressed by colonic epithelial cells throughout the colon during fetal life but only in proximal segments during adulthood. Malignant and premalignant colonic tumors frequently exhibit ABH reappearance (distal lesions) or ABH deletion (proximal lesions) and occasionally express incompatible A or B substances. Mechanisms governing these developmental and cancer-associated alterations are unknown. Therefore, experiments were performed to assess the activities of biosynthetic (glycosyltransferase) and degradative (glycosidase) enzymes in normal and cancerous tissues of the proximal and distal colon. In normal colonic mucosa, A, B, and H transferase activities were similar in proximal and distal segments. Analysis of enzyme substrate affinities and product characterization confirmed that the ABH transferases in colonic tissues were similar to the gene-specified transferases in human serum. Glycosidase enzyme activities were also comparable in proximal and distal normal colon. Cancers had lower A and B transferase but similar H transferase activities compared with paired normal mucosa. Thus, the absence of ABH antigen expression in normal distal colon is not caused by insufficient glycosyltransferase activity or excessive glycosidase activity.

ABH blood group antigen expression, synthesis, and degradation in human colonic adenocarcinoma cell lines

The synthesis of blood group ABH antigens is under genetic control, where the primary gene products are glycosyltransferases. Several studies have demonstrated cancer-associated alterations in ABH antigen expression in human colon cancer tissues. However, the mechanism(s) responsible for these alterations has not been elucidated. Therefore, experiments were conducted using nine established colon cancer cell lines (four type O, three type A, and two type B) to examine ABH antigen expression by immunocytochemistry and correlate this with activities of ABH biosynthetic (glycosyltransferase) and degradative (glycosidase) enzymes. The products of the glycosyltransferase enzymes were characterized by high performance liquid chromatography and paper chromatography, and substrate affinities (apparent Km values) of the cancer cell-derived glycosyltransferases were analyzed. The present data demonstrate: (a) all cell lines except H-498 (blood type A) expressed the appropriate ABH glycosyltransferase as well as all three glycosidases; (b) product characterization and substrate dependence experiments suggested that the cancer cell-derived ABH glycosyltransferase enzymes had properties that were similar to those of the ABH enzymes in human serum; (c) H-498 cells exhibited A antigen deletion with accumulation of H precursor substance, most likely due to insufficient A transferase activity; (d) SW1417 cells (blood type B) demonstrated B antigen deletion without precursor accumulation, despite adequate levels of B transferase and low alpha-galactosidase activity; and (e) weak incompatible A antigen expression occurred in LoVo (type B) and SW1116 (type O) cells, and weak incompatible B antigen expression occurred in H-498 (type A) and SW1116 cells. However, since these cells lacked incompatible A or B transferase activity, these incompatible antigens are probably not the true A or B antigens. Thus, the colon cancer cell lines used in this study exhibit all of the ABH alterations previously described in colon cancer tissues and appear to be useful experimental models for studying the molecular events involved in cancer-associated ABH expression.

Cell lineage markers in premalignant and malignant colonic mucosa

Normal colonic epithelial cells consist of several cell types or lineages that are thought to arise from a common stem cell precursor. Neoplastic transformation may occur at different stages in the differentiation of a colonic stem cell to produce tumors that may retain characteristic cell lineage phenotypes. In this study, immunohistochemical techniques were used to identify cell lineage-related markers in fetal, normal, hyperplastic, adenomatous, and cancerous colonic tissue. These markers consisted of secretory component (columnar cells), a purified mucin antigen (mucous or goblet cells), chromogranin A (enteroendocrine cells), lysozyme (Paneth cells), and carcinoembryonic antigen (panepithelial cell marker). Colonic neoplasms, like normal mucosa, predominantly expressed the markers of columnar and goblet cell lineages. Chromogranin A was expressed in a small population of cells in most normal and fetal colonic crypts. Chromogranin A reactive cells were found in 55% of hyperplastic polyps, 31% of adenomatous polyps, and 33% of carcinomas. Lysozyme reactivity was rare in fetal, normal, and hyperplastic specimens, but was present in 86% of adenomas and 40% of carcinomas. Of 42 primary carcinomas, 9% were "pluripotent" and expressed markers of all four cell lineages. In addition to columnar and goblet cell markers, 7% expressed both enteroendocrine and Paneth cell markers, 17% expressed enteroendocrine cell markers, and 24% expressed Paneth cell markers. Two cases (5%) lacked expression of any of the cell lineage markers. The remainder expressed only columnar and goblet cell markers. The markers used in this study appear to identify the major cell lineages of fetal and normal colonic epithelium and can be used to delineate the altered cell lineage phenotypes in premalignant and malignant colonic mucosa.

Deglycosylation of mucin from LS174T colon cancer cells by hydrogen fluoride treatment

Mucin from xenografts of LS174T human colon cancer cells was treated with anhydrous HF for 1 h at 0 degree C to give a product (HFA) with over 80% of the glucosamine and hexose removed, but retaining some galactosamine, and for 3 h at room temperature to give a product (HFB) devoid of carbohydrate. Rabbit antibodies against HFA bound to HFA much more than to HFB, and bound to native mucin to an intermediate extent. Antibodies to HFB bound to HFB more than to HFA, and did not bind to native mucin. Both HFA and native mucin bound a number of lectins, but HFB did not. By SDS/polyacrylamide-gel electrophoresis and size-exclusion h.p.l.c., native mucin and HFA are of apparent molecular mass greater than 400 kDa, whereas HFB is heterogeneous and of low molecular mass. On Western blots, antibody to HFA detected both high-molecular-mass mucin and a 90 kDa protein in homogenates of LS174T cells. Antibody to HFB detected a major 70 kDa band as well as higher-molecular-mass species. In tissue sections of normal colon and colon cancers, antibody to HFA showed both cytoplasmic and extracellular staining, whereas antibody to HFB generally stained only cytoplasmic antigens. These results indicate that anti-HFB antibody is specific for apo-mucin, whereas anti-HFA antibody is specific for GalNAc-apo-mucin.

The expression of colorectal carcinoma-associated antigens in the normal colonic mucosa. An immunohistochemical analysis of regional distribution

Immunohistologic studies were performed to evaluate the expression and regional distribution of 20 colorectal carcinoma-associated antigens in the colonic mucosa of 12 normal adults. A distinct regional variation was seen in the expression of blood group A, B, Lewis(b), and extended Lewis(y) antigens, which were expressed predominantly in the right colon, whereas Lewis(a), Lewis(y), and H-type 2 were prevalent throughout. Lewis(x) and X-like antigens were only occasionally expressed. Two antibodies against sialylated Lewis(a) showed different staining patterns, with CA19.9 positive in only two biopsies and CO29.11 intensely positive in most. Two of the three antibodies detecting carcinoembryonic antigen epitopes (3d6 and COL-4) stained many biopsies, whereas COL-12 was rarely detected. No regional gradient was found in the expression of the other antigens studied (B72.3, ME491; GA73.3, ND1, and ND4). This mapping data will provide an important baseline for future studies of epitope distribution in the colon in premalignant and neoplastic conditions.

Tumor-associated glycoprotein (TAG-72) expression in ulcerative colitis

Monoclonal antibody (MAb) B72.3 reactive with the high-molecular-weight (Mr greater than 10(6) tumor-associated glycoprotein (TAG)-72 is being increasingly utilized in vivo and in vitro for a variety of purposes in colon cancer patients. Recent evidence has suggested that the TAG-72 antigen expression may be enhanced in inflammatory bowel disease, particularly ulcerative colitis (Thor et al., 1986a: Cancer Res., 46, 3118-3124). We have utilized 117 paraffin-embedded formalin-fixed colonic specimens from 56 ulcerative colitis patients which demonstrate a spectrum of epithelial abnormalities (reactive atypia, dysplasia, and carcinoma) as well as 11 inflammatory controls to evaluate TAG-72 expression. Our selected patient population all had pan-colitis and demonstrated a generally increasing incidence of dysplasia or carcinoma with duration of disease (20% at 0 to 10 years, 50% at 11 to 20 years, 59% at 21 to 30 years, and 100% at more than 31 years). TAG-72 expression was similar in the control and non-dysplastic colonic epithelia, and increased with low- or high-grade dysplasia as well as carcinomatous lesions (mean cellular reactivities 23.7%, 26.5%, 36.7%, 70% and 84.3%, respectively). Epithelium with low-grade dysplasia exhibited a focal perinuclear, superficial crypt staining (when present). High-grade dysplastic epithelium showed pancytoplasmic, pan-cryptic reactivity. Invasive disease showed cytoplasmic as well as extracellular mucin staining. Biopsies from patients with active disease showed significantly more immunoreactive cells for TAG-72 than patients with quiescent disease. For any given biopsy specimen the percentage of cells reactive did not always correlate with the degree of dysplasia. TAG-72 expression in quiescent disease generally increased with duration of disease, in contrast to active disease which showed no correlation between MAb B72.3 staining and duration of disease. The frequent expression of TAG-72 in actively inflamed colonic mucosa (ulcerative colitis and other colitides) may limit the clinical utility of this antigen for detecting colon cancer in ulcerative colitis patients by serological assay or in vivo radiolocalization techniques. The tendency for TAG-72 expression to correlate with disease duration in patients with quiescent disease and to increase with more severe degrees of dysplasia suggests that the expression of this gene product correlates with the dysplasia-to-carcinoma sequence.

Expression of Tn, sialosyl-Tn, and T antigens in human colon cancer

Mucin glycoproteins are major secretory products of the colon and contain O-linked oligosaccharides synthesized on a polypeptide backbone. The initial step in the synthesis of O-linked oligosaccharides is the addition of N-acetylgalactosamine to serine or threonine residues forming the Tn antigen. This substance can then receive additional carbohydrate residues such as sialic acid to form sialosyl-Tn antigen, or galactose to form T antigen. In the colon, the T antigen is an oncodevelopmental cancer-associated antigen but little is known about Tn and sialosyl-Tn expression. The present comparative immunohistochemical study was performed to analyze the expression of these antigens in fetal, normal adult, and malignant colorectal tissues with an aim toward elucidating whether Tn and sialosyl-Tn are also oncodevelopmental colon cancer-associated antigens and to gain insight into the earliest steps of mucin glycosylation in colonocytes. We used three reagents to detect Tn antigen (two monoclonal antibodies ETn1.01 and CU-1, and one lectin Vicia villosa), two reagents to detect sialosyl-Tn (monoclonal antibodies TKH2 and B72.3) and one to detect T antigen (monoclonal antibody AH9-16). Except for occasional reactivity with VVA and CU-1, cells of normal colonic mucosa did not express Tn, sialosyl-Tn, or T antigens. However, in the transitional mucosa immediately adjacent to cancer, all three antigens were expressed (ranging from 35 to 67% of cases depending upon the reagent). In colon cancers, the percentage of cases expressing each antigen were as follows: Tn 72-81%, sialosyl-Tn 93-96%, and T 71%. Unlike T antigen, which was preferentially expressed by moderately well- and well-differentiated adenocarcinomas, both Tn and sialosyl-Tn antigens were expressed by most histological subsets of colon cancers, including poorly differentiated adenocarcinomas and mucinous (colloid and signet ring cell type) carcinomas. The majority of cancers expressed both Tn and sialosyl-Tn, usually in association with T antigen. Only one cancer lacked all three antigens. Fetal colonic mucosal cells expressed all three antigens, particularly in goblet cell mucin. These results indicate that like T antigen, Tn and sialosyl-Tn are oncodevelopmental cancer-associated antigens in the colon. Moreover, Tn and sialosyl-Tn antigens appear to be useful markers of poorly differentiated adenocarcinomas and mucinous carcinomas: two histological subsets that often fail to express other cancer-associated antigens and that are often associated with a poor clinical outcome.(ABSTRACT TRUNCATED AT 400 WORDS)

Immunohistochemical comparison of Lea, monosialosyl Lea (CA 19-9), and disialosyl Lea antigens in human colorectal and pancreatic tissues

The CA 19-9 antigen is a monosialosyl Lea blood group antigen which has been shown to be a useful tumor-associated antigen for the diagnosis of gastrointestinal cancers. Recently, a sialylated derivative of this antigen, disialosyl Lea, was isolated from a colon cancer liver metastasis and a monoclonal antibody (FH7) recognizing this novel determinant was developed. The present study simultaneously compared the expression of Lea, monosialosyl Lea, and disialosyl Lea antigens in a variety of nonmalignant, premalignant, and malignant tissues of the colorectum and pancreas with an aim toward elucidating whether disialosyl Lea is expressed as a tumor-associated antigen. In normal colonic mucosa, disialosyl Lea expression closely resembled Lea expression in overall frequency, segmental distribution, and cellular localization whereas monosialosyl Lea (CA 19-9) was essentially absent. Along the crypt axis, Lea was more often expressed in goblet cells of the upper crypt whereas disialosyl Lea was found in goblet cells along the entire crypt. Fetal colonic mucosa expressed all three antigens, as did most colorectal cancers regardless of location within the colon or degree of differentiation. The majority of hyperplastic polyps and practically all adenomatous polyps also expressed these three antigens, and in adenomas, antigen expression was independent of polyp size, villous morphology, or degree of dysplasia. In the normal pancreas, the three antigens were expressed on ductal, ductular and centroacinar cells of all specimens. The majority of pancreatic cancers expressed all three antigens. Thus, in the normal colon, the absence of monosialosyl Lea (CA 19-9) in the presence of disialosyl Lea suggests that an alpha 2,6 sialyltransferase is active, which results in the masking of CA 19-9 antigen expression. These results further support the concept that specific sialyltransferases play a role in regulating the expression of tumor-associated antigens.

A comparison of three immunoperoxidase techniques for antigen detection in colorectal carcinoma tissues

We compared the streptavidin-peroxidase conjugate (SP) method of immunoperoxidase histochemistry to the unlabeled antibody (PAP) and avidin-biotin-peroxidase complex (ABC) techniques in human colorectal carcinoma tissues stained with a monoclonal antibody for expression of carcinoembryonic antigen. Compared to the ABC and PAP method, the SP method produced stronger staining intensity and very low background staining. This was true when other antibody isotypes, other antibody species, other organs, and another tumor-associated antigen were used. Moreover, the SP procedure time could be reduced to one third that of the ABC or PAP methods without compromising accuracy, and the SP reagent is stable for several months. The chemical nature of the streptavidin molecule accounts, in large part, for the advantages of the SP method.

Lex and Ley antigen expression in human pancreatic cancer

Carbohydrate antigens are useful markers for the serological detection of pancreatic cancer. However, data concerning the expression of structurally well-defined carbohydrate antigens in normal and malignant pancreatic tissue is quite limited. The Lex and Leg antigens are closely related carbohydrate antigens synthesized on type 2 blood group oligosaccharide side chains of glycolipids and glycoproteins. Monoclonal antibodies anti-SSEA-1 and AH6 recognize "simple" Lex and Ley epitopes, respectively, regardless of the length of the carrier carbohydrate. Other monoclonal antibodies recognize Lex (FH4), sialyl Lex (FH6, IB9) or Ley (KH1, CC-1, CC-2) carried only by elongated type 2 side chains with or without internal alpha 1,3 fucosyl substitution. The present comparative immunohistochemical study used tissues of normal pancreas, chronic pancreatitis, and pancreatic cancer to determine the normal expression of Lex and Ley antigens in the pancreas and to elucidate any cancer-associated alterations. Lex-related antigens were not expressed in normal pancreas, expressed in only 10-20% of chronic pancreatitis tissues, but expressed in 50-70% of pancreatic cancer tissues. The frequency of Lex-related antigen expression in pancreatic cancer tissues was lowest in poorly differentiated cancers. Within a given specimen, at least three or all four of the Lex recognizing monoclonal antibodies were simultaneously expressed. Unlike Lex antigens, Ley-related antigens were expressed in 32-77% of specimens of normal pancreas, with similar frequencies in specimens of chronic pancreatitis and pancreatic cancer. In normal pancreas, simple Ley was expressed by both ductal and acinar cells, but extended Ley antigens were expressed only by acinar cells. In pancreatic cancer, extended Ley antigen expression was found in less than 10% of poorly differentiated tumors. Coexpression among the Ley-related antigens was less common than with the Lex-related antigens. Also in cancer specimens, simple Lex and simple Lex antigens were often concordantly expressed, whereas extended Lex and extended Ley antigen expression was often discordant. Hyperplastic ducts and ductules associated with pancreatic cancer expressed Lex-related antigens more frequently than morphologically similar lesions associated with chronic pancreatitis. These results demonstrate that Lex-related antigens are cancer-associated determinants in the human pancreas. The discrepant expression between Lex and Ley antigens in these tissues implies altered regulation of fucosyltransferase activity associated with the malignant state.

Cancer-associated alterations of blood group antigen expression in the human pancreas

Certain alterations of blood group substance (BGS) expression have been observed in gastrointestinal cancer tissues. However, in the pancreas little is known about BGS expression by normal or malignant tissue. The present immunohistochemical study analyzed simultaneously the expression of A, B, H, Lewisa (Lea), and Lewisb (Leb) antigens in specimens of normal pancreas, chronic pancreatitis, and pancreatic carcinoma (primary and metastatic). In normal pancreas all five antigens were expressed in ducts, ductules, and acini, but not in islets. Acinar cells expressed A, B, H, and Leb in supranuclear cytoplasm, whereas Lea was found mainly on centroacinar cells. Only BGSs that were appropriate for the host's blood type were expressed, except for one case of Lea deletion. BGS expression by chronic pancreatitis tissue closely resembled that by normal tissue. In primary pancreatic cancer two cancer-associated alterations were noted that were not found in either normal pancreas or chronic pancreatitis. Deletion of an expected A, B, H, or Leb antigen occurred in approximately 25% of cases, particularly in more poorly differentiated cancers. Incompatible expression of an unexpected A or B antigen occurred in 33% of cases, regardless of degree of differentiation. Metastatic pancreatic cancers also exhibited BGS deletion and incompatibility. In both primary and metastatic cancers the incidence of incompatible A or B expression was higher in cancers from the United States than in cancers from Japan, but the incidence of BGS deletion was similar between the two countries. It was concluded that deletion of A, B, H, or Leb antigens and incompatible expression of A or B antigens are cancer-associated events in the pancreas.

Expression of LewisX and sialylated LewisX antigens in human colorectal polyps

The LewisX (LeX) antigen [characterized by trisaccharide Gal beta 1----4 (Fuc alpha 1----3)N-acetylglucosamine] is an oncodevelopmental antigen in the human colon. Monoclonal antibodies (MoAbs), anti-SSEA-1 and AH8-183, which recognize LeX antigen either on short oligosaccharide side chains or as a terminal immunodeterminant on longer carbohydrate side chains of glycoconjugates, bind to most colon cancer tissues but also to some normal colon mucosae. However, the monoclonal antibodies FH1, FH4, FH6, and IB9, which recognize extended difucosylated and trifucosylated LeX structures or their sialylated derivatives, are more cancer-associated because they rarely bind to normal colon mucosa. In the present study, these MoAbs were used to compare the expression of various LeX-related antigens in premalignant (adenomatous) and nonpremalignant (hyperplastic) colorectal polyps. Antigen expression in polyps was also compared to antigen expressions of normal colon mucosa and colon cancer tissues. The four MoAbs recognizing extended LeX antigens bound to adenomatous polyps (APs) significantly more than to hyperplastic polyps (HPs). In contrast, anti-SSEA-1 and AH8-183 recognizing monofucosyl LeX were less able to distinguish between APs and HPs. In APs, staining with the four MoAbs recognizing extended LeX antigens correlated with the premalignant parameters of larger polyp size, more severe dysplasia, and increased villose component. However, staining with AH8-183 correlated only with polyp size, and anti-SSEA-1 correlated only with polyp size and degree of dysplasia. In general, the staining frequency of HPs was similar to that of normal colon mucosa, although FH6, which did not stain any specimens of normal mucosa, stained a few HPs. The staining frequency of APs was less than that of colon cancer tissues, but these differences were generally not statistically significant. In conclusion, extended LeX antigens and their sialylated derivatives are cancer-associated antigens that are expressed preferentially in premalignant colon polyps, that tend to correlate with malignant potential in these polyps, and that may eventually help to define mechanisms involved in the polyp-to-cancer sequence.

An animal model for colon cancer metastasis: establishment and characterization of murine cell lines with enhanced liver-metastasizing ability

A detailed understanding of the pathogenesis of colon cancer metastasis has been hindered by the lack of appropriate animal models which accurately reflect events in this complex process. An animal model for colon cancer metastasis is described in which spontaneously metastasizing colonic tumors are formed after injection of murine colon cancer cells into the cecal wall of BALB/c mice. Using this model, tumor cells with different liver-metastasizing potential were selected and shown to possess several properties known to be associated with other metastatic cell lines. The ability of tumor cells to invade a reconstituted basement membrane and to secrete type IV collagenase was directly proportional to their metastatic ability. In addition, liver-metastasizing cells preferentially migrated toward liver extracts in a Boyden chamber assay, as compared to extracts of brain or lung, and adhered rapidly to highly purified hepatic sinusoidal endothelial cells versus hepatic parenchymal cells in vitro. This model may thus be useful for studying many aspects of the pathogenesis of colon cancer metastasis.