MUCUS AND ESOPHAGUS

Intestinal metaplasia of the esophagus is characterized by the presence of acidic mucins in goblet cells, and is the hallmark of Barrett's esophagus. Chen et al, from Harvard University (Boston, USA), evaluated the significance of histochemical changes in acidic mucins at the esophagogastric junction in 20 patients with and 19 patients without intestinal metaplasia of the esophago-gastric junction. They observed that positivity to sulfomucin is highly associated with the presence of a metaplastic epithelium. They conclude that sulfomucin staining may be useful to detect metaplastic epithelium at the esophagogastric junction.

Marcinkiewicz et al (Charlotesville and Kansas City, USA) reported that mucin concentrations in the esophagus are lower in patients with reflux esophagitis than in healthy subjects. Furthermore, patients with Barrett's esophagus (BE) are known to experience less symptoms of heartburn. The authors hypothesized that decreased heartburn could be related to increased mucin synthesis by the metaplastic epithelium. They compared glycoconjugate release in 10 patients with BE and in 17 asymptomatic volunteers. Using an esophageal perfusion catheter, they infused saline (0.9% NaCl), then HCl, then HCl plus pepsin, then again saline. Mucins were determined by the periodic acid/Schiff (PAS) method in esophageal perfusates from which salivary secretion was carefully withdrawn. They showed that mucin production was elevated in BE patients. So, they concluded that restitution of the rate of esophageal glycoconjugate release may well contribute to the diminished heartburn symptomatology in these patients.

Synthesis and secretion of mucins

The gastric mucosa synthesizes mucins derived from the MUC5AC gene. Shekels et al (Minneapolis, USA) investigated the mechanisms regulating MUC5AC gene expression in pathophysiological conditions. They examined the expression of this gene in both human and mouse gastric epithelial cells in culture treated with : cytokines (TNFa, EGF, IL-1b, IL-4); mucus secretagogues (prostaglandin (PG) E2; pentagastrin; PMA, phorbol 12-myristate 13-acetate; calcium ionophore A23187; nitric oxide donor sodium nitroprussiate.
Amongst these agents, only PMA and TNFalpha increased MUC5AC gene expression by cultured cells; PGE2, pentagastrin and A23137 stimulate mucin secretion but not MUC5AC gene expression. The data indicate that both phorbol ester and TNFalpha-responsive transcription factor binding sites may contribute to the regulation of MUC5AC gene in gastric epithelial cells.

Turner et al (Harvard, Boston, USA) recalled that mucins are high molecular weight glycoproteins, whose primary function is to protect the gastric epithelium against noxious stimuli, including gastric HCl. Then, the authors used pig gastric mucosal explants to evaluate the effects of dexamethasone (a known ulcerogenic antiinflammatory drug) and retinoids (i.e. trans-retinoic acid and 9-cisretinal, two known transcriptional regulators), on both mucin release and mucin gene expression. They showed that dexamethasone produced a 50% increase and retinoids a 40% decrease in mucus release.
Similarly, dexamethasone caused a 66% decrease in MUC5AC mRNA levels, while both retinoids increased them by 100 and 123% respectively. Because the effects observed on both mucin secretion and gene expression appear to be parallel, it is concluded that, at least in part, these mediators act at the transcriptional level.

Gastric cytoprotection

· Human gastric mucins play a major role in gastroprotection against acid and pepsin. Thus, the primary function of MUC6 could be the protection of epithelial cells from endogenous agents such as proteases, acid and bile salts. Moreover, the cytoprotective efficiency of MUC6 would depend on its ability to withstand breakdown in an hostile environment, a feature which could be determined by the characteristics of the molecule. Toribara et al (San Francisco and Minneapolis, USA) used molecular biology techniques to specify the primary structure of this mucin. From the corresponding cDNA sequence, they deduced that MUC6 is comprised of approximately 5100 amino acids, organized in three regions: an amino-terminal, "non tandem-repeat" region of 1400 amino acids; an intermediate "tandem-repeat" region, where a 169 base-pairs unit is repeated 20 to 24 times; a carboxy terminal "non tandem-repeat" region of 361 amino acids. The authors also showed that MUC6 utilizes a multimerization mechanism, similar to that of the von Willebrand factor, in the formation of the mucus gel: its molecule minimizes the number of regions necessary to form disulfur bonds, thus limiting the exposure of MUC6 to the outside environment.

The "needs" for cytoprotection are different among mucosal tissues, which may explain the existence of different secretory mucins (MUC2 to 8). Batmann et al (Minneapolis and San Francisco, USA; Lille, France; Leeds, UK) investigated the expression of one of these secretory mucins, namely MUC6, a human gastric mucin, in several fetal tissues (18-20 wk gestation) exposed to potential mucosa-threatening endogenous agents. They found MUC6 mRNA in mucous neck cells of the gastric fundus, antral glands, duodenal Brunner's glands, gallbladder mucosa, common bile duct, seminal vesicles... This distribution suggests that MUC6 primary function could be the protection of vulnerable epithelia from endogenous damaging agents such as proteases, acid or biliary salts, respectively.

Watanabe et al (Kanagawa, Japan) compared the stimulatory effects of L-cystein and PGE2 (a known mucin secretagogue), on mucin secretion by cultured gastric epithelial cells. Similarly to PGE2, L-cystein (10 and 100 µM) concentration-dependently stimulates mucin secretion and inhibits intracellular mucin accumulation. However, in contrast to PGE2, L-cystein did not increase intracellular cAMP levels, indicating that the mechanism by which cystein stimulates mucus release is different from that reported for PGE2.

The stimulatory effects of PGE2 on mucin secretion have been largely described in several animal species, particularly on gastric mucosal explants or gastric epithelial cells in short-term cultures. Enss et al (Hannover, Germany) designed a study aimed at determining, first, if PGE2 exerts a similar effect in human gastric cells in primary culture and, second, using a comprehensive panel of specific lectins to determine possible alterations in the carbohydrate composition of the released mucins. They show, indeed, that PGE2 concentration-dependently stimulates mucin secretion by these cells, but also that this increased production was accompanied by changes in their carbohydrate moiety (fucose: -20%; N-acetyl glucosamine: -45%; mannose: +280%; galactose: +60%) compared to mucins released by non-stimulated cells. They conclude that these structural modifications may induce alterations in the cytoprotective capability of the gastric mucus.

Several studies report that gastrin accelerates mucin biosynthesis by the mucosa of the gastric corpus, but its physiological mechanism of action remains obscure. Ishikawa et al (Yokohama, Japan) wanted, first, to ascertain if this effect of gastrin was limited to the mucus cells of the gastric mucosa and, second, to investigate a possible involvement of nitric oxide (NO). They show, indeed, that gastrin increased [3H]glucosamine incorporation into mucins (which is a witness of mucus synthesis) from full-thickness corpus mucosa, but not specifically by the deeper gland mucus cell layer, indicating that only superficial cells are targets for gastrin. In addition, this gastrin-induced enhancement was suppressed by L-NNA (Nw-nitro arginine, an inhibitor of NO synthase), and by a NO scavenger (carboxy-PTIO), which suggest the involvement of NO in this mechanism.

Teprenone (geranyl-geranyl-acetone, GGA) is an antiulcer drug widely used in Japan. Its cytoprotective effects are largely mediated by the induction of heat-shock proteins (HSPs). In addition to this mechanism, Teshima et al (Tokushima and Kyoto, Japan) (3610) have also found, in guinea-pig gastric epithelial cell cultures, that GGA (1µM) activated NFkb, a cell regulatory protein involved, among other, in the mechanisms of NO generation. They observed an enhancement of inducible NO synthase (iNOS) mRNA expression, increased levels of nitrates and nitrites (both derived from NO metabolism), and an up-regulation of NO-dependent mucin release in these cell cultures. The data suggest that, in addition to the trascriptional activation of HSP genes, teprenone may exert its cytoprotective effect through NFkb-mediated pathways leading to iNOS induction.

Mucus and Helicobacter pylori (Hp)

The relative inefficacy of antibiotics in the treatment of Helicobacter pylori may result from the inability of the molecule to reach the bacteria hidden within the viscous mucus layer. Grubel and Cave (Boston, USA) investigated the association of two antibiotics, amoxicillin and clarythromycin, with the antacid magaldrate (Riopan gel or powdered tablets), to modify the pH of a solution of semipurified pig gastric mucins. They observed that the mucus markedly delayed the penetration of both antibiotics, but clarythromycin penetrated mucus at a ten-fold higher rate than amoxicillin. They also showed that both formulations of the antacid magaldrate influence mucus layer permeation to antibiotics; these effects, however, were not univocal: tablets increase more than 3-fold the penetration of amoxiciline, whereas the gel retards the penetration of clarythromycin. The data underline the relevance of antacids for the improvement of therapies aimed at HP eradication.

Kaise et al (Tokyo, Japan) also addressed the issue of HP eradication. They recommend a triple therapy, including omeprazole (a proton pump inhibitor) plus amoxicilline and clarythromycin (treatment 1);
they investigated possible improvements to this therapy by adding either pronase (a proteinase which dissolves gastric mucus, sanctuary of the bactery; treatment 2),
or oxidizing water (which kill various organisms; treatment 3).

All three treatments, applied to 104 patients randomly allocated into three regimen groups, were satisfactory with eradication rates of 74%, 75% and 67% respectively, and acceptable mild side effects. So, it is most probably not necessary to weight down the recommended triple therapy.

Furuta et al (Hamamatsu, Japan) examined the amount of HP present in gastric mucus of patients during and after eradication treatments: lansoprazole alone (n=11); lansoprazole and amoxicilline (n=12). They used different techniques for this determination: cPCR (competition polymerase chain reaction), histology, culture and rapid urease test. They observed that cPCR still revealed the presence of important amounts of HP after the end of the treatment in several patients, although they had turned negative according to the other methods. These patients will return seropositive one month or one year later. In contrast, 8 out of 12 patients in the second group, that were HP seronegative by both cPRC and the other methods, remained negative one month or one year later. Therefore, cPRC determination of HP in gastric mucus seems necessary for the precise assessment of success in HP eradication therapy.

Pathophysiological conditions such as those resulting from the infection by HP or peptic ulcer disease have a negative impact on the synthesis of gastric mucins. These mucins derive mainly from two major genes, MUC5AC and MUC6, which are highly polymorphic. Using gastric biopsies obtained from in 86 patients undergoing endoscopy, Fesenmeyer et al (Minneapolis, USA; Leeds, UK), investigated the effects of HP infection on both gene expression (at mRNA and protein levels) and polymorphisms. HP infection, which they found in 25 patients, was associated with an increase in MUC6 expression, but also to a decrease of both MUC5AC and total mucins, compared to mucins from healthy controls. In contrast, in 16 patients previously treated for HP eradication, the respective levels of the different mucins were identical to those of healthy controls. In addition, they observed that gene polymorphism was apparently not associated with infection status. The authors suggest that this decrease in MUC5AC could well contribute to the pathogenesis of HP-related peptic ulcer disease.

Trefoil peptides (intestinal trefoil peptide, ITF; human spasmolytic peptide, hSP) are HP specific chemotaxins. This is the hypothesis that Turner et al (Nottingham and London, UK) wanted to ascertain. Indeed, these authors previously showed that HP chemotaxin, as well as trefoil peptides and other mucus components, were all resistant to boiling and/or proteolysis; but mucus itself is a well known chemotaxin for other enteric pathogens. Therefore, they compared the chemotaxis of different bacteria (Clostridium difficile, Escherichia coli, Campylobacter pylori and Helicobacter pylori) towards ITF, hSP and purified pig gastrin mucins. All these bacteria presented chemotaxis for the mucus, which except for C. jejuni, was resistant to boiling and proteolysis. Only HP, however, presents chemotaxis for ITF and hSP, suggesting that other components in the boiled mucus are responsible for the chemotaxis of other bacteria. Trefoil peptide localization to the antrum and ulcer margin overlaps the topography of HP infection and may represent the means that HP utilizes to subvert host's mucosa protection mechanisms.

Intestinal metaplasia (IM) is considered to be a precursor of intestinal type gastric carcinoma, with higher risk of cancer development for those secreting incomplete sulphomucins (type III IM), than for those releasing sialomucins (types I-II). Zerbib et al, (Bordeaux, France) described the histochemical features of mucins secreted by HP infection-associated IMs. They examined biopsies from 255 patients with proven HP infection: among them, they identified IM in 40 (15%), who were mainly older patients, of which only 3.5% were of type III, thus, more susceptible to developing cancers.

HP infection may also disturb the integrity of the adherent mucus layer of the gastric epithelium which is the home of the bacteria. The layer's structure would depend on the relative amounts of polymerized mucins (i.e. gel-forming mucins), and mucins of smaller sizes (i.e. non gel-forming). Newton et al (Newcastle, UK) examined in parallel biopsies (secreted gel and intraglandular mucus) and brushings (secreted gel only) taken from patients infected by HP (n=24) or healthy controls (n=24). Their study shows a significant reduction (18%) in the proportion of polymerized gel-forming mucins in the adherent mucus of infected patients. The authors conclude that although this degree of variation is unlikely to impair the efficacy of the mucus barrier, it could sign a local alteration of the gel structure in the immediate vicinity of the HP.

The same authors, Newton et al (Newcastle, UK) extended their investigations by measuring the thickness of the adherent mucus layer in patients infected by HP (HP+, n=20) and in healthy controls (HP-, n=20). Significant differences in mucus gel thickness were not found between HP- and total HP+ subjects. In contrast, in HP+ patients with associated gastric atrophy (n=9/20), the mucus layer was significantly thinner than in HP+ without gastric atrophy (n=11/20), which were not different from healthy controls. They conclude that it is gastric atrophy and not HP infection by itself that leads to the reduction in the thickness of the adherent mucus gel layer.

The pathogenesis of HP associated diseases could well result from an aberrant expression of gastric mucins. Thus, by comparing the mucin composition of antral biopsies obtained by endoscopy from HP+ patients (n=29) and HP- controls (n=32), Byrd et al (Detroit, USA) (3768), observed: that MUC6, which is present in the controls only in the mucosal glands, is also expressed in superficial cells in infected patients; that MUC5, a mucin from superficial cells, is less expressed in infected patients; finally, that MUC1 and trefoil peptides pS2 (superficial epithelium) and hSP (glands) were not different between both groups. The consequences of these variations in terms of mucus barrier efficacy remained to be specified, but could, once again contribute to the pathogenicity of HP.

Very little is known of the structural alterations of the oligosaccharide side-chains of the mucin molecule in response to HP infection. Ota et al (Houston, USA; Matsumoto, Japan) performed biopsies of both gastric antrum and corpus in 21 patients, before and after eradication treatment, and and used monoclonal antibodies directed against specific oligosaccharides. They found that HP infection markedly altered the glycosylation of gastric mucins.

MUCUS AND BILE DUCTS

Mucin synthesis and gallbladder cytoprotection

Klinkspoor and Lee (Seattle, USA) wanted to verify if bile salts could exert their well known stimulating effects on mucin secretion, simply because of their detergent effect on the cell membrane. Thus, they used primary cultures of dog gallbladder epithelial cells, in the presence of either a detergent (Triton X-100 [12.5 - 200 µM]; tween-20 [0.1 - 1.6 mM]) or the bile salt TUDC (tauroursodeoxycholic acid, [10 mM]). Mucin synthesis was assessed by measuring [³H]N-acetyl glucosamine incorporation in secreted glycoproteins; the viability of cells aggressed by these agents was assessed by measuring the leakage of the enzyme lactate dehydrogenase (LDH) in the culture medium. They observed that, as opposed to TUDC, the detergents were unable to stimulate mucin secretion without causing cell damage. As such, there appears to be an additional mechanism to the simple detergent effect, by which bile salts interact with the gallbladder epithelium to stimulate mucus secretion.

Gallstones and gallbladder inflammation

Gallbladder mucin hypersecretion is the most frequently proposed mechanism to explain gallstones formation. However, bacterial infection may also be an important etiologic factor. Choi et al (Seattle, USA; Chung-Buk, South Korea) used dog gallbladder epithelial cells to investigate, first, whether lipopolysaccharide (LPS) from E. coli stimulates mucin secretion, and second, whether TNFa, an inflammatory cytokine, or nitric oxide (NO) were involved in this stimulation. Therefore, they tested LPS (100 and 200 µg/ml) in the presence or absence of either TNFalpha (100 ng/ml), or NO donor (NOR-4, 1 mM), or NO synthase inhibitior (L-NAME, N-w-nitro-L-arginine methylester; 4 mM). At both concentrations, LPS increased (160 and 200% respectively) the secretion rate of mucins by apical membrane cells, without causing cell damage, but remained inactive on basolateral membrane cells.
This effect was not reproduced by TNFalpha or by the NO donor. Similarly, this effect was not prevented by L-NAME, or by an antibody directed against TNFalpha . Thus, the mechanism of LPS-induced mucin stimulation remains to be elucidated.

Cao et al (Boston and Worcester, USA) were interested by the kinetics of the interactions between mucins and cholesterol vesicles, whose aggregation and fusion precedes gallstone formation. They identify two distinct phases. A first step of lipid aggregation is reversible, independent of microviscosity, and requires only glycosylated domains on the mucin molecule. A second step, irreversible, when vesicle fusions depend on non-glycosylated domains of the mucins. The authors conclude that an effective anti-lithogenic therapy may target specific domains on the mucin molecule.

These interactions also awakened the curiosity of del Pozo et al (Munich, Germany) who confirmed the existence of strong interactions between lipid vesicles and mucins in the bile of patients with lithiasis. They underline that these interactions are very likely to be relevant factors for the precipitation of cholesterol crystals and gallstone formation.

The same group, (Muller et al, Munich, Germany) tried to elucidate the mechanism of action of ursodeoxycholic acid (UDCA), which is regarded as effective for the dissolution of cholesterol gallstones and for preventing the risk of recurrence after nonsurgical therapy. These authors administered either UDCA (750 mg/j; n=7) or a placebo (n=8) to 15 patients with gallstones, randomly allocated to two groups. UDCA significantly reduced both the amount of cholesterol crystals in the bile and the concentration of cholesterol in vesicles. Moreover, they observed a significant drop of mucin and protein levels, thus improving bile fluidity and reflecting diminished damage to the gallbladder epithelium in patients treated with UDCA.

The role of the gallbladder epithelium in lipid and protein composition of bile was investigated by Keulemans et al (Amsterdam, The Netherlands). These authors measured the lipid and protein concentrations and compositions of paired bile samples, collected from 14 patients undergoing elective cholecystectomies. They observe that lipid composition (cholesterol, phospholipids, bile salts) did not change between the liver and the gallbladder. This was not true for proteins: some proteins (IgA, IgM) were absorbed, the concentration of other proteins (mucins, albumin, aminopeptidase N) increased more than could be explained by water absorption alone. These data indicate a differential absorption of proteins in the gallbladder; especially the rise of albumin in mucus could by itself explain the increase in protein concentration in patients with gallstones.

A frequently addressed issue in Japan is hepatolithiasis. In this form of lithiasis prevalent in the Far East, most of the calculi are brown (calcium-billirubine) and located in inflamed intrahepatic bile ducts secreting large amounts of mucins. The presence of marked duct inflammation lead Asano et al (Ibaraki and Tokyo, Japan) to investigate the role of PLA2 (a key enzyme of the inflammatory reaction which stimulates arachidonic acid metabolism) in the pathogenesis of this lithiasis. In patients with hepatolithiasis or cholelithiasis, they observed a very clear increase of PLA2 in the bile, reaching much higher concentrations than those measured in patients presenting only gallbladder calculi. The authors suggest that this marked increase in PLA2 could be of pathogenic importance through potentiating bile duct inflammation and the associated increase in the expression of biliary sulfomucins.

Synthesis and regulation

ATP (adenosine triphosphate), as well as other nucleotides (UTP, uridine triphosphate; ADP, adenosine diphosphate; cAMP, cyclic adenosine monophosphate) are major regulators of cellular functions. Several studies have demonstrated that ATP stimulates mucin secretion by airway epithelial cells and intestinal goblet cells. Savard and Lee (Seattle, USA) (1607) studied the response of dog pancreatic duct epithelial cells, to exogenous nucleotides. When the apical surface of cell was exposed to ATP (10, 100 and 1000 µM), mucus secretion was enhanced to 172%, 362% and 444% respectively, compared to controls, without causing any detectable cell damage. Comparatively, UTP was as active as ATP, and much more efficient than ADP for this effect. No effect was obtained with cAMP or adenosine. Finally, ATP was almost devoid of effect when applied on the basolateral membrane, in contrast with what has been reported for other secretagogues such as secretine or VIP (vasoactive intestinal polypeptide). A P2u-type purinoceptor mediated mechanism on the apical surface is suggested by the authors, to explain the specificity of action of these nucleotides which are constitutively present in pancreatic fluid.

MUCUS AND COLON

Synthesis and regulation

N-formylated peptides are a class of chemotactic peptides released by intestinal bacteria, found at high concentrations in colonic secretions. One of these, fMLP (Nformyl-methyonyl-leucyl-phenylalanine), has a specific receptor (fMLPr) involved in both secretion of mucins and the release of cytokines such as interleukin (IL)-8. Campbell et al (Liverpool, UK) used PCR (polymerase chain reaction) to identify the expression of this receptor in several epithelial cell lines: gastric (AGS) and colonic (LS174T, HT29-MTX, HT29). It appeared that only the HT29-MTX cell line expressed this 615 base-pair receptor, which responds concentration-dependently to the stimulation by fMLP by secreting IL-8 and large amounts of mucins. This receptor, whose biological significance is still unknown, could be involved in the pathophysiological protection mechanisms developed by the intestinal mucosa in response to infection.

Van Klinken et al (Amsterdam, The Netherlands) tried to determine if the mucin HGBM (human gallbladder mucin), first described as specific to the gallbladder, could be expressed in other tissues, particularly throughout the intestinal tract. These authors proceeded to perform histochemical studies on biopsies of gallbladder, duodenum, jejunum, and ascending, transverse and sigmoid colon, using antibodies directed against HGBM, BGBM (bovine) and MUC5B. Not surprisingly they detected HGBM in the gallbladder, but also in the colon. HGBM displays a complete homology with MUC5B and with its bovine homologue; HGBM is stored in small granules in the goblet cells of the lower colonic crypts region and can be secreted.

Evidence is accumulating that ulcerative colitis (UC) and Crohn's disease (CD) have polygenic origins. Moreover, a deficiency in some mucin sub-classes which characterizes ulcerative colitis have been found in unaffected monozygotic twins, suggesting a genetic basis to this disease. Parkes et al (Oxford, UK) looked at the presence of chromosomic markers of the MUC2 gene in patients with IBD and their parents. They found a genetic linkage, although modest, but no significant difference between UC and Crohn's disease. For the first time evidence is presented for a linkage of MUC2 gene and IBD susceptibility.

Lucas et al (Boston, USA) investigated the expression of the mucin genes MUC2 and MUC5B, but also pS2, hSP and hITF genes of the trefoil peptides associated with epithelial mucins in patients with UC or CD. Only MUC5B gene expression was found markedly and significantly decreased in Crohn's disease but not in UC. Thus, the identification of the regulatory factors of MUC5B expression would be of paramount interest for the understanding of IBD pathogeny.

Diversion of fecal flow often results in an itrogenic inflammation. Edwards et al (Oxford and Bristol, UK; Galveston, USA) examined mucins in biopsies obtained from both defunctioned colonic segments with either normal or inflammatory mucosa, and functional colonic segments of patients with either UC or Crohn's disease. Both total mucin synthesis and the ratio of mucins stained with the lectin Dolichos biflorus were decreased in defunctioned segments, independently of the underlying pathology. The O-acetylated sialic acid content was identical in all biopsies, whether from defunctioned or functional segments. In contrast, a diminished level of sulphated mucins was specifically observed in defunctioned segments of UC patients.

A technical difficulty inherent to the study of mucins from surgical or biopsy specimens is that it is difficult to separate secreted mucins from submucosal proteoglycans. Saitoh et al (Hirosaki, Japan) used chromtographic techniques to identify mucins released into a colonic lavage solution in patients with UC or CD, and in healthy controls. The total amount of mucins and the high molecular weight fraction were diminished in UC and enhanced in CD. Therefore, the study of mucins in colonic lavage fluids may well constitute a new approach to differential diagnosis of Crohn's disease and UC.

Several lines of evidence support the involvement of autoimmune mechanisms linking colonic mucins and UC. Takaishi et al (Tokyo, Japan) detected antimucin antibodies directed against colonic S (soluble)- and M (membrane)-mucins, in 11 out of 60 patients with UC. In addition, rats immunized with a colonic mucin developed chronic colitis. These rats develop diarrhea, and anatomical and pathological alterations of colonic mucosa comparable to those observed during UC. Thus, according to these authors, the autoantibodies directed against colonic mucins would contribute to the pathogeny of ulcerative colitis.

The trefoil peptides ITF and SP are secreted by mucus cells and are involved in mucosal restoration processes. Moreover, immune cells, especially macrophages, play a major role in mucosal repair. The aim therefore, was to ascertain whether these immune cells expressed ITF and SP. Cook et al (Footscray, Australia) demonstrated that ITF and SP are expressed by tissues normally involved in immune regulation (spleen, bone marrow, thymus). In the spleen, ITF and SP are synthesized and released by macrophages, and their production is enhanced by lipopolysaccharide administration. These data highlight a role for ITF and SP in mucosal repair by macrophages.

The mechanisms regulating mucin synthesis are complex and multifactorial. It is well known that stress stimulates colonic mucin release. It has also recently been shown that stress induces mast cell degranulation. Castagliuolo et al, from Harvard University (Boston, USA) demonstrated that stress was unable to elicit the secretion of colonic mucins in mast cell deficient mice. Thus, mast cells are very likely involved in the modulation of colonic mucin production.

Gastric Cancer

Several mucin genes (MUC1, MUC5AC, MUC6) have been recently identified in the normal gastric mucosa. In contrast, the expression of these genes in gastric tumoral cell lines remains unclear. Wagner et al (Hannover, Germany) observed that gastric cancer cell lines showed a decreased expression of "gastric" type mucins and an aberrant expression of "intestinal" type mucins. In addition, TNFa predominantly stimulates gastric-specific mucin genes. According to these authors, this enhancement may play an important function in gastric inflammation through an alteration of mucin composition.

Cancer of bile ducts

Carbohydrate composition and protein core antigens have been found to be modified in pancreatic carcinoma. A related issue is whether similar changes occur in extrahepatic bile duct cancers. Imamura et al (Kagoshima, Japan) investigated several mucin antigens in extrahepatic bile duct carcinoma. Their results suggest that the antigens sTn and MUC1-AR may play a role in the transition from normal epithelium to carcinoma in extrahepatic bile ducts.

Colon cancer

STn is a carbohydrate antigen whose expression has been found to be associated with colon cancer in humans. The sTN epitope has only been identified in mucin type glycoproteins, leading to the hypothesis that mucins may influence the behavior of colon cancer cells. Ogata et al (New York, USA) were the first to detect STn on a non-mucin type molecule, termed CD44, a metastasis-associated surface protein. These results suggest that the O-glycane residues on the CD44 glycoprotein may be involved in the biology of colon cancer cells.

Apomucin-specific O-glycosylation mechanisms are poorly understood. The same authors as above (Ogata et al, New York, USA) demonstrated that the same MUC1 peptide could be differently glycosylated in different colonic cancer cell lines. They also showed that the STn disaccharide may be present on MUC1 apomucin.

Sialyl-Lex (sLex) belongs to a family of carbohydrate antigens whose expression increases progressively in human rectocolonic carcinomas. The data gathered by Mann et al (Berlin and Kiel, Germany) demonstrate that overexpression of sLex-bearing mucins results from a gradual drop of sLex O-acetylation. This chemical alteration can be considered as a new marker of colonic carcinoma development. Since sLex expression has been associated with the capacity of tumoral cells to metastasize, these results lead to the hypothesis for a role of sialic acid O-acetylation in the metastatic process.

The search for tumoral markers in stools is extremely relevant to the identification of colorectal cancers. Tissue analyses have shown that MUC1 is predominantly expressed in these cancers and very seldom in normal colonic epithelial cells. Limburg et al from the Mayo Clinic (Rochester, USA) reported that fecal concentrations of MUC1 did not allow a differentiation of patients with colorectal cancer from healthy controls. Moreover, no significant correlation was found between the amount of MUC1 and tumor stage or site, patient age or gender. So, it seems that the assay of MUC1 in stools would be of little use in screening colorectal cancers.

MUC1 mucin core peptide is often detected in colorectal cancer tissues and in cancer cell lines. Hiraga et al (Hiroshima and Ehime, Japan) observed that MUC1 was expressed in 43 out of 59 lesions (73%) examined. MUC1 expression was higher in the presence of lymphatic or venous invasion, or in lesions with liver metastasis. In contrast, MUC1 expression was not correlated with histologic grade and depth of invasion. According to these authors, MUC1 expression would be relevant to the prediction of metastatic potential and prognosis of colorectal cancers.

Phorbol esters such as phorbol 12-myristate 13-acetate (PMA) are known to modulate diverse cellular responses through intracellular pathways involving the protein kinase C (PKC). Han et al (San Francisco, USA) have shown that PMA increased MUC2 and MUC3 mucin genes and stimulated cancer cell-associated carbohydrate antigens. These changes were associated with an enhancement of their invasive properties and a reduction of the adhesion properties of colonic cancer cells. These data suggest that PKC signaling is probably involved in the invasive and metastatic properties of these cells.

Index de la page .
coproweb@free.fr