Cell surface-expressed Thomsen-Friedenreich antigen in colon cancer is predominantly carried on high molecular weight splice variants of CD44.

Increased mucosal expression of TF, the Thomsen-Friedenreich oncofetal blood group antigen (galactose beta1-3 N-acetylgalactosamine alpha-) occurs in colon cancer and colitis. This allows binding of TF-specific lectins, such as peanut agglutinin (PNA), which is mitogenic to the colorectal epithelium. To identify the cell surface TF-expressing glycoprotein(s), HT29 and Caco2 colon cancer cells were surface-labeled with Na[(125)I] and subjected to PNA-agarose affinity purification and electrophoresis. Proteins, approximately 110-180 kDa, present in HT29 but not Caco2 were identified by Western blotting as high molecular weight splice variants of CD44 (CD44v). Selective removal of TF antigen by Streptococcus pneumoniae endo-alpha-N-acetylgalactosaminidase substantially reduced PNA binding to CD44v. Immunoprecipitated CD44v from HT29 cell extracts also expressed sialyl-Tn (sialyl 2-6 N-acetylgalactosaminealpha-). Incubation of PNA 15 microg/ml with HT29 cells caused no additional proliferative effect in the presence of anti-CD44v6 mAb. In colon cancer tissue extracts (N = 3) PNA bound to CD44v but not to standard CD44. These data show that CD44v is a major PNA-binding glycoprotein in colon cancer cells. Because CD44 high molecular weight splice variants are present in colon cancer and inflammatory bowel disease tissue but are absent from normal mucosa, these results may also explain the increased PNA reactivity in colon cancer and inflammatory bowel disease. The coexpression of oncofetal carbohydrate antigens TF and sialyl-Tn on CD44 splice variants provides a link between cancer-associated changes in glycosylation and CD44 splicing, both of which correlate with increased metastatic potential.

Opposite effects on human colon cancer cell proliferation of two dietary Thomsen-Friedenreich antigen-binding lectins.

Increased cell surface expression of the Thomsen-Friedenreich antigen (TF antigen, Galbeta1-3GalNAcalpha-) is a common feature in malignant and pre-malignant epithelia. Our previous studies have shown that dietary TF-binding lectins from peanut (Arachis hypogea) and edible mushroom (Agaricus bisporus) produce marked but different effects on human intestinal epithelial cell proliferation. This study investigates the proliferative effects of the other two known dietary TF-binding lectins: jacalin (Artocarpus integrifolia, JAC) and amaranth lectin (Amaranthus caudatus, ACA). JAC produced dose-dependent and non-cytotoxic inhibition of proliferation in HT29 human colon cancer cells with maximal effects of 46 +/- 4% at 20 microg/ml, whereas ACA produced dose-dependent stimulation of proliferation with maximal effects of 22 +/- 3% at 20 microg/ml when assessed both by incorporation of [3H]thymidine into DNA and by cell counting. The lectin-mediated effects were inhibitable by the presence of appropriate Galbeta1-3GalNAc-expressing glycoproteins but differences existed between JAC and ACA in their patterns of inhibition by such substances. Ligand binding equilibrium studies using iodinated lectins revealed different Kd of the two lectins for HT29 cell surface glycoproteins. Lectin blots of cell membrane extracts showed different binding patterns in all the four TF-binding lectins. These results provide further evidence that dietary TF-binding lectins can have marked effects on the proliferation of human malignant gastro-intestinal epithelial cells and hence may play a role in intestinal cancer development, and also show that the biological effects of dietary lectins cannot be predicted solely from their carbohydrate binding properties.

Use of a biosensor to determine the binding kinetics of five lectins for Galactosyl-N-acetylgalactosamine.

The dietary lectins, edible mushroom (ABL) and Jacalin (JAC) inhibit the proliferation of colonic cancer cells, whereas Amaranth (ACL) and peanut (PNA) stimulate their proliferation. All these lectins share as their preferred ligand the Thomsen-Friedenreich (TF) antigen galactosyl beta1,3 N-Acetylgalactosamine (Galbeta1,3GalNAc), but differ in their finer specificities for modifications of this determinant and in their specificities for cancerous epithelia. We have investigated, using a resonant mirror biosensor, the kinetics of binding of these lectins, and Maclura pomifera lectin (MPL), which is similar to JAC, to two different Gal-GalNac bearing glycoproteins, antarctic fish antifreeze glycoprotein (AFG) and asialofetuin. JAC had the highest affinity for AFG [K(d) 0.027 microM] due to a fast association rate constant [k(ass) 610,000 (Ms)(-1)]. The other lectins had considerably lower affinities, with K(d) ranging from 0.16 microM (ABL) to 5.7 microM (PNA), largely due to slower k(ass) [ABL 74,000 (Ms)(-1) to PNA 2700 (Ms)(-1)]. Similarly, JAC had a much higher affinity for asialofetuin [K(d) 0.083 microM] than the other lectins [K(d) 1.0 microM-4.5 microM]. Affinities were also calculated from the extent of binding at equlibrium and were generally similar to those calculated from the kinetic parameters indicating the true nature of these values.

Edible mushroom (Agaricus bisporus) lectin, which reversibly inhibits epithelial cell proliferation, blocks nuclear localization sequence-dependent nuclear protein import.

The Galbeta1-3GalNAcalpha (TF antigen)-binding lectin (ABL) from the common edible mushroom (Agaricus bisporus) has a potent anti-proliferative effect without any apparent cytotoxicity. This unusual combination of properties prompted investigation of its mechanism of action. In contrast to soluble lectin, agarose-immobilized, and hence noninternalizable ABL had no effect on proliferation of HT29 colon cancer cells. Electron microscopy of HT29 cells incubated with fluorescein- and gold-conjugated ABL showed internalization of the lectin into endocytotic vesicles and multivesicular bodies. Confocal microscopy showed perinuclear accumulation of fluorescein isothiocyanate-conjugated lectin, which also inhibits HT29 cell proliferation, raising the possibility that the lectin might interfere with nuclear pore function. Transport of heat shock protein 70 into the nucleus in response to heat shock was blocked by preincubation of HT29 cells for 6 h with 40 micrograms/ml ABL. In digitonin-permeabilized cells, nuclear uptake of bovine albumin conjugated to a nuclear localization sequence (NLS)-containing peptide was also inhibited by a 15-min preincubation with 40-100 micrograms/ml ABL. In contrast, serum-stimulated nuclear translocation of mitogen-activated protein kinase, which is NLS-independent, was not affected by pretreatment of cells with the lectin. These results suggest that the anti-proliferative effect of ABL is likely to be a consequence of the lectin trafficking to the nuclear periphery, where it blocks NLS-dependent protein uptake into the nucleus.

Pancreatic tumour marker anti-mucin antibody CAM 17.1 reacts with a sialyl blood group antigen, probably I, which is expressed throughout the human gastrointestinal tract.

CAM 17.1 is an antimucin monoclonal antibody which has recently been proven valuable as a reagent for serological diagnosis of pancreatic cancer. A series of studies have been performed to characterise its epitope. First it was screened immunohistochemically against a wide range of formalin-fixed normal and neoplastic human tissues and showed widespread binding to mucin throughout the gastro-intestinal tract, in both normal and malignant tissues. In pancreas, strong intracellular staining of acinar and ductal cells was found in normal tissue and in carcinoma cells in tumours. Normal stomach showed only weak staining (n = 6), but gastritis with metaplasia showed strong staining (n = 4). Staining of colonic mucosa from patients of known Lewis phenotype showed Le(a+b-) (7/8) and Le(a-b+) (4/6) samples to be positive, but not Le(a-b-) (0/3) samples. CAM 17. 1 agglutinated all donor erythrocytes tested at 4 degreesC regardless of blood group, whereas cord blood red cells were not agglutinated. Since I antigen is the only antigen known to be present on all adult red blood cells but absent from cord blood, this suggests probable involvement of this antigen in the binding site. The agglutination was abolished by sialidase treatment of the red cells and immunoblotting with slot-blotted mucin showed that binding was both acid and sialidase sensitive indicating the involvement of sialic acid in the binding site. These studies show that CAM 17.1 binds to a sialic-acid-containing determinant of mucin, probably sialyl-I, which epitope shows wide distribution throughout the gastro-intestinal tract.

Quantification of intestinal mucins.

Intestinal mucins are glycoproteins containing a very high proportion of carbohydrate, up to 90%, and they can occur either in a membrane-bound or a soluble form (1). The general structure of these molecules includes discrete C-terminal and N-terminal regions, which are relatively poorly glycosylated, at either end of a long repeating structure, up to 50 repeats, which is heavily O-glycosylated on serine and threonine residues the first sugar in each oligosaccharide always being N-acetyl-galactosamine (2). These oligosaccharide chains can vary between 1 and about 15 saccharides. The C- and N-terminal regions are likely to be of importance in determining the membranous or secreted nature of the molecule (2).

Stimulation of proliferation in human colon cancer cells by human monoclonal antibodies against the TF antigen (galactose beta1-3 N-acetyl-galactosamine).

In many tissues, the TF (Thomsen-Friedenreich) blood group antigen (Galbeta1-3GalNAc alpha-) behaves as an onco-foetal carbohydrate antigen, showing increased expression in malignancy and hyperplasia. Dietary lectins which bind the TF antigen have marked effects on proliferation of epithelial cells without cytotoxicity. This led us to speculate that anti-TF antibodies, including those that naturally occur in humans, might have similar effects. Five anti-TF antibodies, TF2 (human), TF5 (human), 5A8 (mouse), 8D8 (mouse) and BM22 (mouse), but not TFI (human) or 49H.9 (mouse), showed marked dose-dependent stimulation (95-192%) of [3H]thymidine incorporation by HT29 human colon cancer cells. Similar stimulation of proliferation of HT29 cells by these monoclonal antibodies (MAbs) was found when cell count assessment was used. Antibody-stimulated proliferation was inhibited by co-incubation with glycoproteins expressing Galbeta1-3GalNAc alpha- (asialo glycophorin or [Galbeta1-3GalNAc alpha-O-p-aminophenyl]n-human serum albumin). A proliferative effect of these antibodies was also demonstrated on human colon cancer cell lines LS174T and HT29-MTX but not on Caco-2 cells. Although immunoblotting showed similar binding patterns of all the antibodies on HT29 cell membrane extracts, there was little correlation between cell surface binding assessed by immunofluorescence and proliferative response, and internalization of the biotinylated antibody TF5 was demonstrated by confocal microscopy. Our results provide further evidence that cell surface glycoproteins which express TF antigen may play an important role in the regulation of cell proliferation and also suggest that human anti-TF antibodies may have proliferative effects on cells which express TF antigen.

Altered expression of mucins throughout the colon in ulcerative colitis.

BACKGROUND/AIMS: Regional differences in the biology of the colonic epithelium may determine the extent of involvement by ulcerative colitis. Novel monoclonal antibodies (MAbs) were used in this study to investigate regional heterogeneity in the colonic mucosa. METHODS: MAbs generated using a method of tolerisation against common antigens in the proximal colon and distal colon were used for immunoperoxidase staining, comparative histochemistry, immunoblotting, and slot-blot analysis. RESULTS: The colon specific MAbs 5F1 (IgG3) and 6G4 (IgM) stained goblet cell contents throughout the normal distal colon but staining was markedly reduced in the proximal colon (p < 0.0001). In the distal colon of patients with ulcerative colitis, whether quiescent or actively inflamed, reactivity was reduced compared with controls (p < 0.05, p < 0.001 respectively). By contrast, an overall increase in staining was seen in the uninflamed proximal colon in ulcerative colitis compared with controls (p < 0.02). Comparative staining with high iron diamine and biochemical analyses indicated that MAb 6G4 was reactive with mucin bearing sulphate or O-acetylated sialic acid groups, or both. CONCLUSIONS: Regional differences in the staining characteristics of normal colonic mucin have been shown using novel monoclonal antibodies. The pattern of mucin expression throughout the colon in ulcerative colitis is altered even in the absence of histological changes.

Stimulation of colonic mucin synthesis by corticosteroids and nicotine.

1. We speculated that corticosteroids might cause beneficial stimulation of mucus synthesis, since this is a known action of carbenoxolone, itself a corticosteroid, and has also been proposed as a possible mechanism for the protective effect of smoking on ulcerative colitis. We have therefore compared the effects of corticosteroids including carbenoxolone, and nicotine on mucin synthesis, assessed by incorporation of N-[3H]acetylglucosamine into mucin by colonic epithelial biopsies in culture. 2. In histologically normal biopsies from the left colon, hydrocortisone and prednisolone caused a very marked concentration-dependent increase in mucin synthesis, with maximal effect (580 and 300% of control values respectively) at 6 mumol/l [P < 0.001, n = 35 biopsies (seven patients)] and 1.5 mumol/l [P < 0.001, n = 35 (seven patients)] respectively. The maximal effect of hydrocortisone was significantly greater than that of prednisolone (P < 0.05). Carbenoxolone, 0.17 mmol/l, also increased mucin synthesis in the left colon by 242% [P < 0.05, n = 15 (three patients)]. In contrast, these corticosteroids caused only a small, non-significant increase in mucin synthesis in the histologically normal right colon; fludrocortisone, 2 and 20 mumol/l, and aldosterone, 0.1-10 mumol/l, had no effect. Nicotine significantly increased mucin synthesis (180-220% of control values) between 62.5 nmol/l and 6.25 mumol/l (P < 0.05 at all concentrations) in both the right and left colon. In biopsies from the relatively uninvolved right colon of patients with ulcerative colitis, corticosteroids and nicotine caused relatively smaller increases in mucin synthesis. 3. The marked stimulation of mucin synthesis by corticosteroids suggests that this may account, at least in part, for their therapeutic effect in ulcerative colitis.

Interaction of socioeconomic status and provider practices as predictors of immunization coverage in Virginia children.

OBJECTIVE: To describe the pattern of immunization in the cohort of children who entered public schools in Virginia in 1992. DESIGN: This was a historic cohort study using stratified cluster sampling. Three strata were created based on the socioeconomic status (SES) of the children in the catchment area of each public school in Virginia. SETTING: The random sample included public elementary schools throughout Virginia. PARTICIPANTS: Immunization records were obtained for a randomly selected cohort of 2519 first-grade children in Virginia. OUTCOME MEASURES: Age at completion of recommended childhood vaccines was determined from birth to school entry by SES, race, and population density. Provider practices were assessed by ascertaining missed opportunities for simultaneous administration of vaccinations according to recommended schedules. RESULTS: Although immunization completion rates were high at school entry, low levels of immunization coverage were found in all areas of Virginia at 24 months of age regardless of SES (as measured by per capita income), population density, or race. However, under-immunization was more severe for poor children in urban areas (42.3% of children in low-SES urban areas were age-appropriately immunized at 24 months of age versus 64.0% in children in high-SES rural areas). By multivariate logistic regression, race and gender were not predictors of which children were appropriately immunized at 2 years of age after adjusting for the following: SES, population density, receiving the first DTP (diphtheria, tetanus, and pertussis) or OPV (oral polio) vaccination after 3 months of age, and failure to have the first DTP administered simultaneously with the first OPV or the second DTP administered simultaneously with the second OPV. Receiving the first DTP or OPV vaccination after 3 months of age and failure to have the first and second DTP and OPV administered simultaneously were the strongest predictors of not being age-appropriately immunized at 2 years of age. The effect of failure to vaccinate simultaneously on predicting vaccination coverage at 2 years of age was strongly modified by SES. Children who attended schools located in census tracts with per capita incomes less than $10,600 and who did not have the first and second doses of DTP and OPV administered simultaneously were 33.19 times more likely not to be age-appropriately immunized at 2 years of age compared with children who attended schools located in census tracts with per capita incomes greater than $18,800 and who received the first and second doses of DTP and OPV simultaneously (95% confidence interval: 18.29 to 60.22). CONCLUSIONS: Although beginning the immunization schedule at the recommended age was crucial to appropriate vaccination later in life, provider practices were important predictors of under-immunization. Failure to administer vaccinations simultaneously strongly influenced poorer children in Virginia. Serious delays in vaccine administration were observed not only for poor children in urban areas, but also in all areas of Virginia before school entry.

Immunohistochemical detection of O-acetylated sialomucins in intestinal metaplasia and carcinoma of the stomach.

Eighty-two selected gastric mucosal biopsy or resection specimens were stained both conventionally, to classify subtypes of intestinal metaplasia and carcinoma, and immunohistochemically with a mouse monoclonal antibody (MMM-17), raised against normal human colonic mucin, which has an affinity for di- and/or tri-O-acetylated sialomucin. The aims of the study were to reassess the prevalence of O-acetylated sialomucins in normal, metaplastic and carcinomatous gastric mucosa and to investigate whether the production of these mucins by intestinal metaplasia is related to its associated mucosal pathology. O-acetylated sialomucins were not seen in normal mucosa. They were, however, prevalent in all sub-types of metaplastic (64.8%) and carcinomatous (42.9%) mucosa. Type 1 intestinal metaplasia was significantly more likely to contain this type of mucin if Helicobacter pylori infection was identifiable in the adjacent gastric mucosa (81.0% v. 38.5%, P < 0.025). Type 3 showed a similar, albeit nonsignificant, relationship (100% v. 62.5%). O-acetylated sialomucins are, therefore, much more prevalent in gastric intestinal metaplasia and carcinoma than previously recognized by conventional staining techniques. The production of this type of mucin by intestinal metaplasia may reflect an adaptive response to alterations in the luminal environment such as an increase in bacterial content.

Increased faecal mucin sulphatase activity in ulcerative colitis: a potential target for treatment.

Colonic mucin is heavily sulphated and it has been shown that enzymatic desulphation by faecal bacterial sulphatases greatly increases its susceptibility to degradation by faecal glycosidases. A possible role for faecal mucin sulphatase in the pathogenesis of inflammatory bowel disease has therefore been explored. Faecal mucin sulphatase activity assayed using 35S mucin as substrate was increased in ulcerative colitis (median 80.2 units/g pellet weight (range 6.9-1063; 95% confidence intervals (CI): 45.2 to 293.8, n = 22) compared with 11.3 units/g (range 3.0-53.5; 95% CI: 8.7 to 29.8, n = 17) in healthy controls (p < 0.01), where one unit released 1000 dpm free sulphate/hour from 35S mucin (1680 dpm/microgram). Patients with active ulcerative colitis had higher sulphatase activity (median 146; 95% CI: 98 to 253 units/g, n = 10) than those with inactive ulcerative colitis (median 42.2; CI: 22.5 to 81.6 units/g, n = 12) (p < 0.05). Longitudinal studies in patients with ulcerative colitis show fluctuations of faecal mucin sulphatase activity corresponding to clinical disease activity in six of seven patients. Faecal mucin sulphatase activity was not significantly increased in Crohn's disease (median 36.6, range 5.7-106.6; 95% CI: 22.9 to 65.3 units/g, n = 14). The bismuth salts, bismuth subcitrate and bismuth subsalicylate were found to inhibit faecal mucin sulphatase activity at concentrations achievable therapeutically. The increased faecal mucin sulphatase activity in ulcerative colitis could be the result of greater intraluminal substrate (mucin) availability leading to bacterial enzyme induction, but would probably result in more rapid degradation of secreted mucin and represents a potential target for treatment.

Multicenter postmarketing surveillance of ondansetron therapy in pediatric patients.

OBJECTIVE: To identify prescribing patterns of ondansetron, to provide a general overview of the therapeutic responses and possible adverse effects to ondansetron in selected children's hospitals, and to evaluate this methodology of surveillance and assess its effectiveness as a means to collect postmarketing experience with a drug in pediatric patients. DESIGN: This survey examined the use of ondansetron in 210 children. Complete drug and medical histories, indications, doses, possible ondansetron-associated adverse reactions, and daily responses to ondansetron therapy were recorded by a study pharmacist for each patient. Patients were followed until discharged from the clinic or hospital and/or until ondansetron therapy was discontinued. SETTING: The survey was conducted in seven free-standing children's hospitals across the US. Hospitals ranged in size from 100 to 331 beds (average 234). One hospital was located on the West coast, one on the East coast, one in the Rocky Mountain region, one in the Southwest region, and three in the Midwest. PARTICIPANTS: The selection of study participants was limited to member free-standing children's hospitals of the Pediatric Pharmacy Administrative Group. Selection was based on geographic location and availability of a pharmacist to coordinate the study. One pharmacist at each study site served as surveillance coordinator. Each pharmacist monitored without intervention the use of ondansetron in 30 children. Patients were enrolled consecutively from physicians' orders for ondansetron. Enrollment was open to clinic and hospital patients. Patients were excluded if more than 48 hours of retrospective review was required. MAIN OUTCOME MEASURES: The survey queried patient demographics, type of antineoplastic therapy administered, indications and dosing regimen(s) for ondansetron, additional antiemetic agents administered, and clinical response. Adverse drug reactions and prescriptions for ondansetron on discharge were recorded. An evaluation of response rates in hospital patients based on exposure to antineoplastic regimens causing acute (within 24 h) or delayed emesis (after 24 h) was formulated after data collection. Off-label use was summarized. RESULTS: Surveys from 197 of the 210 patients enrolled were complete for evaluation. Ondansetron was used to treat chemotherapy-induced emesis in 88 percent of the patients and 12 percent received it for various other indications. Ondansetron dosing was off-label in 15 percent and 73 percent prior to and after an emetogenic exposure, respectively. Twenty-six percent of the patients were younger than four years. Dosages ranged from 0.15 to 0.45 mg/kg, given in various schedules. The injectable form was given both intravenously and orally. There was a significant difference in the mean number of doses in hospital (9 +/- 7.3) versus clinic (2 +/- 1.5) patients (p < 0.0001). Eighty-seven percent of all patients had a complete or major overall response. Possible ondansetron-associated adverse reactions were similar to those of previous reports for all patients, although some recorded reactions are not currently included in package labeling. CONCLUSIONS: This study documents off-label use of ondansetron in children. Further study of ondansetron use in children less than four years of age, and for indications other than chemotherapy-induced emesis, is needed. Additional evaluation into the most cost-effective dosing of ondansetron would also be valuable.

Reversible inhibition of proliferation of epithelial cell lines by Agaricus bisporus (edible mushroom) lectin.

Galactosyl beta-1,3-N-acetyl galactosamine (Gal beta-1,3-GalNAc) (Thomsen Friedenreich antigen), the Class I core sequence in O-linked oligosaccharide chains, behaves as an oncofetal antigen showing increased expression in many epithelial malignancies. Previous work has shown that peanut agglutinin (PNA), a lectin that binds Gal beta-1,3-GalNAc, stimulates proliferation in HT-29 (human colon cancer) cells and normal human colonic epithelium and this implies that cell surface glycoproteins which express Gal beta-1,3-GalNAc may play an important role in the regulation of epithelial cell proliferation. We have now studied the effect on epithelial cells of another dietary Gal beta-1,3-GalNAc-binding lectin, the edible mushroom Agaricus bisporus lectin (ABL). This differs from PNA in its ability to bind also to sialylated Gal beta-1,3-GalNAc. In contrast to PNA, ABL (25 micrograms/ml) inhibited incorporation of [3H]-thymidine into DNA of HT29 colon cancer cells by 87% (95% confidence limit, 85-89%), Caco-2 colon cancer cells by 16% (95% confidence limit, 12-20%), MCF-7 breast cancer cells by 50% (95% confidence limit, 47-52%), and Rama-27 rat mammary fibroblasts by 55% (95% confidence limit, 51-60%) when these cells were grown for 24 h in serum-free medium. When assessed by cell count, similar inhibition of proliferation of HT29 cells by ABL was found. In the presence of 2% fetal calf serum (which contains the ABL-binding glycoprotein fetuin), the inhibitory effect of ABL on cell proliferation was still demonstrable but at increased ABL concentration (60 micrograms/ml for 49% inhibition). Ten micrograms/ml ABL completely abolished the stimulatory effect on [3H]thymidine incorporation of epidermal growth factor (100 pg/ml) and PNA (25 micrograms/ml) and markedly inhibited the stimulatory effect of insulin (50 ng/ml). ABL (0.2 mg/ml) caused no cytotoxicity to HT29, MCF-7, and Rama-27 cells as measured by trypan blue exclusion, and inhibition of proliferation in HT29 cells caused by 50 micrograms/ml ABL was reversible after removal of the lectin. Binding studies with 125I-labeled ABL suggested a single class of binding site with an apparent Kd value of (4.12 +/- 0.29) x 10(-7) M with (3.6 +/- 0.3) x 10(7) binding sites/cell. A. bisporus lectin is a reversible noncytotoxic inhibitor of epithelial cell proliferation which deserves study as a potential agent for cancer therapy.

Distribution of O-acetylated sialomucin in the normal and diseased gastrointestinal tract shown by a new monoclonal antibody.

AIMS: To produce and characterise a monoclonal antibody specific for O-acetylated sialomucin and to assess its use in immunohistochemistry on a panel of normal and diseased intestinal tissue samples. METHODS: Mouse monoclonal antibodies were developed following immunisation with highly purified human colonic mucin. One of these (MMM-17) showed strong binding to mucin throughout the normal colon with relative lack of binding to colon cancer tissue. The binding epitope of MMM-17 was then characterised by screening for agglutination activity against a panel of human and animal erythrocytes and by assessment of its binding to a range of normal and chemically treated slot blotted mucins. Further immunohistochemical studies were then performed on formalin fixed, normal, and diseased human intestinal samples. RESULTS: Binding of MMM-17 to slot blotted human colonic mucin was reduced by 38 (SD 14%) (n = 4) by alkali treatment of the mucin, sequential alkali and sialidase treatment completely abolished binding. Sialidase treatment alone, however, caused only an 11 (11%) reduction in binding. MMM-17 failed to agglutinate any human, rabbit, rat or mouse erythrocytes. These findings were compatible with specificity of MMM-17 for sialomucins O-acetylated at the C-7 or C-8 positions on the sialic acid. Strong staining by MMM-17 was found in all goblet cells throughout all 40 normal colonic and rectal samples studied, but staining was absent in seven of 13 colorectal carcinomas. Normal duodenum (n = 16) and normal ileum (n = 3) all showed occasional positive goblet cells. The normal gastric antral mucosa was generally negative B MMM-17, but in all of 15 cases of gastritis with intestinal metaplasia the metaplastic glands were strongly positive for MMM-17. CONCLUSION: Monoclonal antibody MMM-17 has specificity for O-acetylated sialomucins and its binding depends both on the position of O-acetylation and on the adjacent oligosaccharide structure. Preliminary studies using the antibody on archival tissue samples support the previous reports of reduced O-acetylation in colon cancer demonstrated by indirect histochemistry and show the neo-formation of O-acetylated sialomucin in intestinal metaplasia in the stomach.

High performance gel filtration using monodisperse highly cross-linked agarose as a one-step system for mucin purification.

High performance gel filtration on monodisperse cross-linked agarose (Superose 6) has been assessed as a system for purification of mucus glycoproteins. Comparison with the conventional two-step purification of mucus glycoprotein by Sepharose CL4B gel filtration followed by caesium chloride density gradient centrifugation shows that purification by high performance gel filtration is at least as thorough, yielding mucin that is free from non-mucin glycoproteins as defined by buoyant density, mobility on sodium dodecyl sulphate-polyacrylamide gel electrophoresis and absence of Concanavalin A binding (mannose-containing) material. This technique allows mucus glycoprotein to be purified from lyophilized crude mucin in 120 min compared with approximately 72 h using the conventional techniques. This makes the comparative study of mucus glycoprotein changes in disease states much more feasible.

A new enzyme-linked lectin/mucin antibody sandwich assay (CAM 17.1/WGA) assessed in combination with CA 19-9 and peanut lectin binding assay for the diagnosis of pancreatic cancer.

BACKGROUND: Mucus glycoproteins are often present in the sera of patients with pancreatic cancer, and their detection and quantification can be used in serologic diagnosis. METHODS: A novel enzyme-linked "sandwich" assay (CAM 17.1/WGA) has been developed in which a lectin, wheat germ agglutinin (WGA), is bound to the solid phase to capture serum glycoproteins, and after addition of test sera, a monoclonal antimucin antibody (CAM 17.1) and peroxidase-tagged second antibody are used as a detection system. RESULTS: The test has been applied to sera from 79 patients with pancreatic cancer and 120 controls. The CAM 17.1/WGA assay alone had a sensitivity of 78% and specificity of 76% in the diagnosis of pancreatic cancer. Combination of the CAM 17.1/WGA test with a previously described peanut lectin binding assay (PNA/ELLA) provided a sensitivity of 92% and specificity of 70%, whereas combination of the CAM 17.1/WGA assay with the CA 19-9 radioimmunoassay had a sensitivity of 85% and specificity of 76%. Combination of all three tests had a sensitivity of 94% and specificity of 66%. In nonjaundiced patients, the combination of CAM 17.1/WGA and PNA/ELLA had a sensitivity of 93% and specificity of 79% in the diagnosis of pancreatic cancer. CONCLUSIONS: This new test adds significantly to the armamentarium of serologic tests for pancreatic cancer. These tests are particularly effective when used in combination to detect different mucin-borne carbohydrate antigens. They deserve more widespread use, particularly in examining nonjaundiced patients with unexplained abdominal pain or weight loss.

Ion exchange chromatography of purified colonic mucus glycoproteins in inflammatory bowel disease: absence of a selective subclass defect.

Previous reports of a selective mucin subclass defect in ulcerative colitis have been reassessed using high performance chromatography (Superose 6 and Mono Q) for mucin purification and fractionation coupled with analysis of the fractions obtained using a combination of enzyme linked lectin and mucin antibody assays. Mucin samples purified from snap frozen rectal biopsy specimens obtained from patients with ulcerative colitis (n = 12), Crohn's disease (n = 5), and non-inflammatory bowel disease control subjects (n = 9) were subject to ion exchange chromatography using a continuous 0-0.35 mol/l NaCl salt gradient with a final 2.5 mol/l NaCl step. In all samples the major proportion (mean (SD) 86.7 (8.9)%) of the mucin detectable by wheat germ agglutinin binding eluted between 0.15 and 0.35 mol/l NaCl with no significant difference in elution profile between ulcerative colitis and control subjects. Significant elution of glycoprotein at less than 0.15 mol/l NaCl did occur, however, when a lower molecular weight mucin containing fraction which contained concanavalin A positive (glucose or mannose containing) material was analysed similarly. Similar ion exchange profiles were obtained when (3H)N-acetylglucosamine labelled mucins were studied after tissue culture of rectal biopsy specimens. No significant alteration in the ion exchange profile of purified mucins in ulcerative colitis has been shown in these studies. It is possible that the previously reported relative depletion of mucin subclass IV (eluting with 0.20 mol/l NaCl) may simply have reflected mucin depletion.