Clinical aspects of gastrointestinal mucus

Analysis of the proteome of human airway epithelial secretions

BACKGROUND

Airway surface liquid, often referred to as mucus, is a thin layer of fluid covering the luminal surface that plays an important defensive role against foreign particles and chemicals entering the lungs. Airway mucus contains various macromolecules, the most abundant being mucin glycoproteins, which contribute to its defensive function. Airway epithelial cells cultured in vitro secrete mucins and nonmucin proteins from their apical surface that mimics mucus production in vivo. The current study was undertaken to identify the polypeptide constituents of human airway epithelial cell secretions to gain a better understanding of the protein composition of respiratory mucus.

RESULTS

Fifty-five proteins were identified in the high molecular weight fraction of apical secretions collected from in vitro cultures of well-differentiated primary human airway epithelial cells and isolated under physiological conditions. Among these were MUC1, MUC4, MUC5B, and MUC16 mucins. By proteomic analysis, the nonmucin proteins could be classified as inflammatory, anti-inflammatory, anti-oxidative, and/or anti-microbial.

CONCLUSIONS

Because the majority of the nonmucin proteins possess molecular weights less than that selected for analysis, it is theoretically possible that they may associate with the high molecular weight and negatively charged mucins to form a highly ordered structural organization that is likely to be important for maintaining the proper defensive function of airway mucus.

Biosynthesis of mucin type O-glycans: lack of correlation between glycosyltransferase and sulfotransferase activities and CFTR expression

Structural differences have been reported in the glycosylation patterns of cystic fibrosis glycoproteins. Although the gene mutated in cystic fibrosis (CFTR) has been cloned and characterized as a chloride channel, its relationship to the highly viscous mucus and structural glycoprotein and mucin abnormalities in cystic fibrosis still remains to be defined. We have evaluated O-glycan biosynthesis in CHO and BHK cells that express CFTR and DeltaF508 CFTR as in vitro models, and utilized the cftr knockout mouse as an in vivo model of CFTR dysfunction. Activities of glycosyltransferases and sulfotransferases synthesizing mucin type O-glycan chains were determined in these models. Differences in transferase activity levels were found between tissues and cell types and during mouse development. No specific patterns of activities were associated with the lack of CFTR or with DeltaF508CFTR expression. This suggests that it is not the presence or absence of normal CFTR, or the presence of mutant CFTR alone, but rather cell specific additional factors or pathophysiological consequences that determine the changes in mucin glycosylation in cystic fibrosis.

Echinostoma caproni and E. trivolvis alter the binding of glycoconjugates in the intestinal mucosa of C3H mice as determined by lectin histochemistry

Mouse (C3H) mucosal glycoconjugates were examined in normal small intestines and intestines infected with Echinostoma caproni or E. trivolvis using six different fluorescein-conjugated lectins: Triticum vulgaris agglutinin (WGA), Ulex europaeus agglutinin I (UEA-I), Ricinus communis agglutinin I (RCA-I), Glycine max soybean agglutinin (SBA), Dolichos biflorus agglutinin (DBA), and Arachis hypogaea peanut agglutinin (PNA). The expression of lectin-binding sites and the intensity of the binding of lectins in the mouse small intestines were changed by infection with the echinostomes. Specific differences in the reaction to glycoproteins were clearly observed between the mouse intestines infected with E. caproni and those infected with E. trivolvis. In E. caproni infection, binding of most of the lectins to the villi was remarkably reduced in accord with the villous atrophy and loss of goblet cells. In contrast, in E. trivolvis infection, the binding of WGA, RCA-I and DBA was reduced in the microvillar surfaces, but binding of UEA-I and SBA were unchanged compared to the control intestines. The lectin binding to goblet cells in E. trivolvis-infected mice mostly increased. These observations may reflect the marked increase in goblet cells and the less severe damage in the villi of E. trivolvis infection compared to E. caproni infection. Most of the glycoconjugates were slightly reduced in the hyperplastic crypts except for N-acetyl glucosamine. It is possible that glucose metabolism in the host intestines infected with E. trivolvis was activated, resulting in an increase in the rate of mucin synthesis as well as qualitative changes in mucus, thereby mediating the expulsion of the worms.

Glycosylation and sulphation of colonic mucus glycoproteins in patients with ulcerative colitis and in healthy subjects

Studies have been made of mucus glycoprotein biosynthesis in different regions of the lower gastrointestinal tract in normal patients and those with ulcerative colitis (UC), active or inactive, by means of 3H-glucosamine (3H-GlcNH2)--35S-sulphate double labelling of epithelial biopsy specimens under culture conditions. The time based rate of 3H-GlcNH2 labelling of mucus in rectal tissue was similar to that in active or inactive UC whereas the rate of 35SO4(2) labelling was significantly increased in active disease. The 3H specific activities measuring the amount of isotopic incorporation into surface and tissue mucus glycoproteins were increased in patients with active UC compared with normal or inactive subjects. The 35S specific activities did not differ significantly between patients with active UC and those in remission. In the rectum, glycosylation of mucus glycoproteins decreases with the increasing age of the patient. Regional differences in 3H-labelling of mucus components are reported for ascending colon, transverse colon, sigmoid colon, and rectum. Sulphation (35S-labelling) was higher in all parts of the colon in left sided UC. Results point to accelerated glycosylation of core proteins in the active phase of UC.

Lectin histochemistry of adenomatous polyps. Not a predictor of metachronous lesions

The binding characteristics of fluorescein isothiocyanate conjugated lectins to normal colonic mucosa, and 43 adenomatous polyps were studied by fluorescence microscopy. The lectin, Dolichos biflorus agglutinin (DBA) stained intensely to upper crypt cells of the sigmoid colon and rectum but to a lesser degree to proximal colonic crypts or lower crypt cells distally. Peanut agglutinin (PNA) and Ulex europaeus agglutinin (UEA) did not bind to the theca of proximal or distal colonic crypts. The lectin Griffonia simplicafolia agglutinin (GSA1) bound intensely to upper and lower crypt cells of both regions. PNA binding was noted in 56% of adenomatous polyps, occurred more often in polyps of the distal colorectum, and increased with polyp size and villous histology. UEA bound to 26% of adenomatous polyps, 42% of proximal polyps, and 17% of distal polyps. DBA staining was noted in 72% of polyps without regional preference. GSA1 stained all polyp specimens. To determine if the lectin binding characteristics of an index (initial) polyp might serve as a predictor of metachronous lesions, 20 patients (29 polyps) without a history of polyps or cancer and who had at least one surveillance colonoscopy 1 to 3 years after the initial polypectomy were studied. The presence or absence of PNA, UEA, or DBA binding in an index polyp did not predict the occurrence of metachronous lesions. Five of the six patients with more than one index polyp had metachronous polyps at follow-up surveillance colonoscopy.

Operative therapy of gallbladder disease in patients with cystic fibrosis

We retrospectively studied 20 patients with cystic fibrosis who underwent surgery for gallbladder disease from 1973 to 1986. A long delay between the onset of symptoms and the diagnosis was noted (mean 7.4 months). This delay was attributed to masking of the symptoms of biliary disease by the malabsorption and pulmonary symptoms seen in this patient population. There was a notable lack of common bile duct disease in our patients as well as in those reported in the literature. This may have been due to a combination of factors: the increased viscosity of the mucus, the small caliber of the gallbladder and ductal system, and the hypotonicity of the gallbladder. We do not recommend routine intraoperative cholangiography in patients with cystic fibrosis and gallbladder disease. Cystic fibrosis is a disease with progressive pulmonary deterioration. Cholecystectomy can be performed in these patients with relative safety if careful preoperative and postoperative care is provided. We recommend early operative intervention in the patient with gallbladder disease and cystic fibrosis.

Cholera enterotoxin-induced mucus secretion and increase in the mucus blanket of the rabbit ileum in vivo

The in vivo rabbit ileum was used to study the relationship of cholera enterotoxin-induced water and electrolyte secretion and mucus secretion and to determine whether the enterotoxin influenced the intestinal mucus blanket. In experiments in which luminal fluid viscosity was used to assess mucus secretion, it was found that while cholera enterotoxin induced a sustained secretion of water and electrolytes, the toxin-induced mucus hypersecretion was short lived (3 to 5 h) and subsequent exposure of the mucosa to cholera enterotoxin or prostaglandin E1 did not stimulate mucus secretion further. Dibutyryl cyclic AMP and theophylline caused a modest mucus secretion in ileal loops which differed from that of cholera enterotoxin in both magnitude and in the fact that the mucus secretion occurred 2 to 3 h after the onset of water and electrolyte secretion. An oral replacement solution was used in the ileum to reduce the enterotoxin-induced loss of water and electrolytes into the lumen. While such a solution slowed the appearance of acidic glycoprotein in the intestinal lumen, it did not change the amount of glycoprotein secreted over a 7-h period, suggesting that toxin-induced mucus secretion was not simply due to a flushing action of the experimentally caused diarrhea. To assess mucus blanket thickness, neutral glycoprotein was recovered from the blanket of rabbit ileal loops 7 h after exposure to cholera enterotoxin and the thickness of the mucus blanket was measured directly 4 and 18 h after toxin exposure. Both methods indicated that even though cholera enterotoxin-induced mucus hypersecretion had subsided and there was histological evidence of goblet cell mucin depletion, there was a sustained increase in mucus blanket thickness that was detectable for at least 18 h after mucosal enterotoxin exposure.

Aberrant expression of monoclonal antibody-defined colonic mucosal antigens in inflammatory bowel disease

Human proximal colon from patients with inflammatory bowel disease and from controls was studied by two techniques to detect tumor-associated antigen expression. A panel of four murine monoclonal antibodies that recognize tumor-associated antigens was used to test purified colonic mucins for epitope expression by radioimmunoassay and to test formalin-fixed, deparaffinized sections of colon by the immunoperoxidase technique. The panel included monoclonal antibodies 19-9, B72.3, DU-PAN-2, and CSLEX1. Colonic mucins were purified from uninvolved surgical specimens by gel filtration with Sepharose 4B and cesium chloride-guanidine hydrochloride density gradient ultracentrifugation. Purified mucins from uninvolved colonic mucosal specimens from 4 of 7 patients with ulcerative colitis expressed one or more of these epitopes by radioimmunoassay, whereas mucins from 6 disease controls did not. Reactivity patterns were heterogeneous. Immunoperoxidase testing demonstrated staining with two or more antibodies in 14 of 18 involved inflammatory bowel disease segments, whereas control sections rarely stained with these antibodies, with the exception of 19-9. Sections of uninvolved mucosa from 4 of 9 patients with ulcerative colitis stained with two or more antibodies. Staining patterns were heterogeneous. The results demonstrate that colonic expression of tumor-associated epitopes occurs frequently in involved segments from both patients with ulcerative colitis and with Crohn's disease, whereas only patients with ulcerative colitis frequently expressed these epitopes in uninvolved segments.

Biosynthesis and secretion of human colonic mucin glycoproteins

Synthesis and secretion of colonic mucin glycoprotein species were assessed during in vitro culture of colonic mucosal explants. DEAE-cellulose chromatography of endogenously labeled mucin glycoproteins from explant tissue demonstrated the presence of six mucin species (I-VI) similar to those identified earlier in surgical specimens of human colonic tissue. The relative proportions of mucin species I-VI in tissue explants remained constant throughout a 30-h culture period. However, the proportional representation of the various mucin species in media was significantly different from that found in tissue, which suggests that some mucin species (I, II, and III) are differentially secreted, whereas others (IV and V) are retained within intracellular pools. Radiolabeled precursors were incorporated into mucin species I, II, and III at a 2.0-2.6-fold greater rate than their concentration in tissue, supporting the concept that these glycoproteins were both synthesized and secreted at a greater rate than species IV and V. Colonic mucosal explants from patients with ulcerative colitis showed greater than 90% reduction of species IV. However, the amount of species IV recovered from culture media of ulcerative colitis explants was comparable to normal controls. It appears that mucin species IV is differentially secreted rather than retained within intracellular pools in mucosa of patients with ulcerative colitis.

Giardia lamblia: stimulation of growth by human intestinal mucus and epithelial cells in serumfree medium

Giardia lamblia trophozoites specifically colonize the upper human small intestine which is normally serumfree but have been grown in vitro only in medium supplemented with serum or serum fractions. Recently, we demonstrated that biliary lipids will support the growth of G. lamblia without added serum. Now, we report that human duodenal jejunal mucus stimulates growth of Giardia in medium with biliary lipids. Stimulation by mucus was enhanced by inclusion of chymotrypsin or crude pancreatic proteases. Coculture of trophozoites with human intestinal epithelial cells also promoted growth, especially in the presence of mucus and/or biliary lipids. With biliary lipids alone, the mean increase in cell number was 3.2 fold and in the presence of mucus 8 fold (P less than 0.01) in 24 serial subcultures. Our demonstration that human intestinal mucus and epithelial cells promote serumfree growth of G. lamblia may help to explain specific colonization of the small intestine by G. lamblia.

Development of a monoclonal antibody specifically reactive to gastrointestinal goblet cells

A mouse monoclonal antibody (7E6A5) of IgG isotype, reacting specifically with mucin-producing goblet cells of the human gastrointestinal tract, has been developed. 7E6A5 reacts by an ELISA with colonic protein eluted from a DEAE column. A screening by immunoperoxidase assay of 76 specimens from 19 different human tissues showed that the immunoreactivity of 7E6A5 was confined exclusively in the globules of goblet cells in the colon, the appendix, and the small intestine. Nongoblet small and large intestinal epithelial cells did not react. Immunoelectron microscopy demonstrated the reactivity with mucin droplets in a homogeneous granular pattern inside the globules of goblet cells. Mucus-secreting cells from remaining parts of the gastrointestinal tract and other mucus-secreting organs such as respiratory, genitourinary tracts, salivary and mammary glands did not show any reactivity to 7E6A5. These findings indicate that the antigen recognized by 7E6A5 is shared by the goblet cells of both the small and large intestines and is unique to them. The monoclonal antibody may be useful in the study of function of mucus-secreting goblet cells and may represent an important tool in the evaluation of diseases such as ulcerative colitis, colon cancer, and intestinal metaplasia in gastric mucosa that are associated with quantitative changes in goblet cell numbers or with qualitative differences in mucin secretion.

Gastrointestinal mucus, a medium for survival and for elimination of parasitic nematodes and protozoa

Mucus is a sticky visco-elastic material which coats all mucosal surfaces. Florey, in 1955, noted the following three functions for gastrointestinal mucus: protection of the underlying mucosa from chemical and physical injury, lubrication of the mucosal surface to facilitate passage of luminal contents, and removal of parasites by binding and entrapment. In the 31 years since Florey's review, detailed analyses of the composition of mucus and of the biochemistry of mucin glycoproteins, as well as measurements of the physical properties of mucus from different organs and sites have yielded information at the molecular level which provide additional support for his views on its function (Allen, 1981; Forstner, Wesley & Forstner, 1982).

A membrane-associated neuraminidase in Entamoeba histolytica trophozoites

Trophozoites of the parasitic amoeba Entamoeba histolytica HM-1:IMSS possess a surface neuraminidase capable of liberating N-acetylneuraminic acid (NANA) from N-acetylneuramin-lactose (alpha 2----3 or alpha 2----6) or mucin in their medium. The neuraminidase was found to be membrane associated, with more than 50% of the yield being recovered in the plasma membrane fraction. The neuraminidase specific activity of the plasma membrane fraction was six times that of internal membrane fraction enzyme. The optimum pH and temperature for this enzyme were 6.7 and 37 degrees C, respectively. Neuraminidase activity was inhibited by ethylene glycol-bis(beta-aminoethyl ether)-N,N,N',N'-tetraacetic acid, and the optimum Ca2+ concentration was 2 mM. The microfilament disruptor cytochalasin D (30 micrograms/ml) inhibited motility and neuraminidase activity of intact Entamoeba trophozoites. The cytochalasin D-induced loss of surface neuraminidase activity was explained in part by a redistribution of enzyme with a loss of plasma membrane enzyme and an increase in intracellular membrane enzyme. A qualitatively similar cytochalasin D effect was observed with two other membrane-associated enzymes, calcium-regulated ATPase and acid phosphatase. Membrane-associated enzyme was minimally affected by Triton X-100 and saponin. An N-acetylneuraminic acid aldolase, optimum pH, 7.4, was found in trophozoite homogenate supernatant fractions. NANA and NANA-containing compounds stimulated trophozoite-directed motility. This motility stimulation by NANA-containing compounds did not apparently require prior release of free NANA by the trophozoite surface neuraminidase. Entamoeba neuraminidase is one of a series of enzymes that may modify the mucus blanket and target cell surface and thereby play a role in the pathogenesis of amebiasis.

Development of anti-human colonic mucin monoclonal antibodies. Characterization of multiple colonic mucin species

Structural relationships between colonic mucin species were assessed using a library of monoclonal antibodies (MAbs) directed against purified human colonic mucin (HCM). After immunization of mice with purified mucin from normal human colonic mucosa, 14% of 1,920 fusion products screened were positive for anti-HCM activity in a solid-phase assay. Patterns of selective binding by hybridomas to six discrete HCM species (I-VI) separated by DEAE-cellulose chromatography suggested the presence of both shared and species-specific antigenic determinants among HCM species I-VI. 23 anti-HCMs MAbs (7 IgM, 7 IgG1, and 9 IgG2) demonstrating a range of anti-HCM species specificities, were produced and used to study structural relationships between mucin species. Binding of various mucin species by individual anti-HCM MAbs was shown by competitive solid-phase radioimmunoassay to reflect the presence of identical epitopes on the different species. Adsorption of HCM species on a variety of affinity resins prepared with anti-HCM MAbs demonstrated that binding to multiple mucin species by a single MAb was related to intrinsic structural determinants. Four anti-HCM MAbs recognized protease-sensitive antigenic structures, which suggests that they may be directed to core HCM proteins. 12 of the anti-HCM MAbs were shown by solid-phase assay to recognize either complete (n = 5) or partial (n = 7) isolated colonic mucin oligosaccharide side chains of defined structure. Collectively, these data show the presence of both shared and unique antigenic structural determinants among colonic mucin species. Chromatographic heterogeneity of mucin glycoproteins seems to be related to the existence of biologically significant subclasses in the normal human colon.

Human colonic goblet cells. Demonstration of distinct subpopulations defined by mucin-specific monoclonal antibodies

We studied glycoprotein content of human colonic goblet cells, using a library of monoclonal antibodies (MAbs) directed against purified human colonic mucin (HCM). Using indirect immunofluorescence (IIF), we found that 17 of 23 anti-HCM MAbs stained some or all goblet cells of normal human colonic mucosa. We observed a variety of cellular staining patterns, including (a) diffuse (homogeneous) staining of intracellular mucin, (b) speckled (inhomogeneous) staining of mucin droplets, (c) peripheral staining of intracellular droplets, (d) cytoplasmic staining of goblet cells, and (e) apical (luminal) surface staining. Staining patterns were not associated with particular HCM species. In addition to variable patterns of IIF within individual cells, anti-HCM MAbs varied in the proportion of goblet cells stained. Some MAbs stained all goblet cells, while others stained a limited number of goblet cells. Although each goblet cell contained more than one type mucin, HCM species III, and IV and V appeared to exist in mutually exclusive goblet cell populations and it was possible to define at least seven subpopulations of goblet cells in colonic mucosa by their content of various combinations of HCM species. Anti-HCM MAbs stained goblet cells from other sites within the gastrointestinal tract to a varying extent. Anti-HCM MAbs also showed extensive cross-reactivity with rodent, rabbit, and monkey colonic mucosa. However, several anti-HCM MAbs stained only human colonic mucosa. These data show that human colonic mucosa contains discrete subpopulations of goblet cells that produce distinctive combinations of specific mucin glycoprotein species.

Effect of guinea pig or monkey colonic mucus on Shigella aggregation and invasion of HeLa cells by Shigella flexneri 1b and 2a

The effects of guinea pig and rhesus monkey colonic mucus preparations on Shigella aggregation and invasion of HeLa cell monolayers by Shigella flexneri serotype 1b, 2a, and 5 strains were investigated. Guinea pig mucus caused agglutination of S. flexneri serotype 1b but not of S. flexneri serotype 2a or 5. Guinea pig mucus also inhibited HeLa cell invasion by S. flexneri serotypes 1b and 2a. Monkey mucus neither agglutinated any Shigella strain nor inhibited HeLa cell invasion.

Interactions of Giardia lamblia with human intestinal mucus: enhancement of trophozoite attachment to glass

Giardia lamblia trophozoites frequently are associated with mucus in vivo. We investigated the effects of human intestinal mucus on parasite attachment and survival in vitro. All samples of mucus from the duodenum and ileum (from four humans and two rabbits) enhanced attachment at 100 micrograms/ml. Attachment increased with mucus concentrations from 1 to 1000 micrograms/ml but declined toward the unstimulated level at concentrations above 1000 micrograms/ml. Mucus from the small intestine also promoted the survival of the parasites during the 2-h incubation. In contrast, colonic mucus promoted survival, but inhibited attachment. Fractionation of mucus from the human small intestine by cesium chloride equilibrium density gradient ultracentrifugation revealed that both attachment- and survival-promoting activities were in the low density, protein-rich fraction. The high density fractions containing the mucins were devoid of activity. Thus, a non-mucin fraction of mucus from the human small intestine may promote colonization by G. lamblia.

Acute and chronic models for hypersecretion of intestinal mucin

Mucin secretion was examined in three functional models relevant to human disease, using rat small intestinal rings or in situ loops, [3H]glucosamine precursor labelling, gel chromatography and a specific radioimmunoassay for mucin. As a model for acute bacterial secretory diarrhoea, tissues were exposed to cholera toxin for up to 4 h. Both stored and newly synthesized radioactive glycoproteins were secreted in amounts twofold to threefold above control levels. Immunoreactive mucin secretion increased fivefold to eightfold. Other agents known to raise cAMP levels did not stimulate mucin secretion, suggesting that cholera may release mucin by a non-cAMP-dependent mechanism. Sepharose 2B chromatography indicated that secreted mucin was smaller in size than intracellular mucin and had compositional differences suggestive of 'immaturity' or protein contamination. In chronically (seven days) reserpinized rats, used as a model of glycoprotein abnormalities relevant to cystic fibrosis, mucin secretion increased twofold to threefold, but the most prominent abnormality was a marked increase in [3H]glucosamine incorporation into all tissue glycoproteins. On purification, the intracellular mucin of reserpine-treated rats had the same composition as mucin from control rats, but the former was smaller in size and had a higher specific radioactivity. Mucin hypersecretion in reserpinized rats may therefore be secondary to a primary and chronic hyperstimulation of mucin biosynthesis. A model of intestinal 'anaphylaxis' or immune-mediated diarrhoea was created in Hooded Lister rats by immunizing with egg albumin (10 micrograms) and challenging with the same antigen in intestinal loops 14 days later. After 4 h, total protein, DNA and brush border sucrase were increased in the lumen. Enhancement of mucin secretion did not occur, however, and therefore does not seem to be a particular feature of the pathophysiology of this model.

Composition of human colonic mucin. Selective alteration in inflammatory bowel disease

Human colonic mucin has been isolated from mucosal scrapings of fresh surgical specimens of normal controls as well as patients with Crohn's colitis and ulcerative colitis. Following sonication and ultracentrifugation, mucin fractions were separated from other soluble colonic glycoproteins by Sepharose 4B chromatography. After nuclease digestion, cesium chloride gradient centrifugation of the excluded material yielded colonic mucin with an average buoyant density of 1.52 g/ml. Subsequent chromatography of the apparently homogeneous colonic mucin on DEAE-cellulose revealed the presence of at least six distinct mucin species (mucin I-VI). Each mucin species was found to have a distinctive hexose, hexosamine, sialic acid, and sulfate content as well as blood group substance activities. Mucin from five patients with Crohn's colitis was found to represent a mixture of at least six discrete species comparable to those isolated from normal colonic specimens. However, in mucin from eight patients with ulcerative colitis there was a marked and selective reduction of one component mucin subclass, designated species IV. Normal mucin and mucin from patients with Crohn's disease contained 48 +/- 17 and 42 +/- 12 mg of species IV/g, while mucin from patients with ulcerative colitis had 5 +/- 3 mg/g solubilized glycoprotein. The selective absence of species IV was found in preparations from both sigmoid (n = 7) and ascending (n = 4) colon and could not be accounted for by an overall decrease in total mucin content. The selective reduction of species IV was also found in mucin isolated from relatively noninflamed colonic mucosa of patients with ulcerative colitis. The carbohydrate composition and blood group activities of the remaining five mucin species were similar to their normal counterparts. Based on the results to date, there appears to be an underlying selective decrease of one colonic mucin subclass in ulcerative colitis.