Bile acid-induced alterations of mucin production in differentiated human colon cancer cell lines.

Damage to the gastrointestinal tract mucous layer may render underlying cells susceptible to intraluminal toxins or carcinogens. Our aim was to determine the effect of bile acids on mucin, the primary constituent of mucous. Differentiated Caco-2 and HT29 cells were used as models of human colonic epithelial cells. Mucin was measured by [3H]-glucosamine labeling. Short term (30 min) incubations with 1-5 mM unconjugated bile acids or taurodeoxycholic acid induced mucin release relative to bile acid hydrophobicity. Longer incubations were cytotoxic. Long term (7 days) incubation at nontoxic concentrations (0.1 mM) of deoxycholic acid (DC) decreased total mucin by 36 +/- 2% (SEM, P = 0.0003) in differentiated HT29 cells and by 57.2 +/- 2% (P < 0.05) in Caco-2 cells. Tauroursodeoxycholic acid (TUDC) or ursodeoxycholic acid (0.1-0.5 mM) did not alter mucin levels. Simultaneous incubation of 0.1 mM DC and 0.1-0.5 mM TUDC or 2.5 mM TDC and TUDC did not change mucin levels. Differentiated HT29 and Caco-2 cells contained high levels of intestinal mucin MUC3 mRNA while undifferentiated HT29 cells did not possess a MUC3 message. Deoxycholic acid (0.1 mM) did not alter the MUC3 mRNA level. Neither cell type showed detectable expression of intestinal MUC2 or gastric MUC6. Thus, cytotoxic concentrations of bile acids induce mucin release, presumably due to detergent effects. Nontoxic concentrations of DC reduce mucin levels in differentiated enterocyte-like cells, which can be prevented by coincubation with TUDC. The bile acid-induced alterations in mucin production by enterocytes observed in vitro may influence intestinal cytoprotection in vivo.

Mouse gastric mucin: cloning and chromosomal localization.

Mucins protect gastric epithelium by maintaining a favourable pH gradient and preventing autodigestion. The purpose of this study was to clone a mouse gastric mucin which would provide a foundation for analysis of mucin gene regulation. Mucin was purified from the glandular portion of gastric specimens and deglycosylated by HF solvolysis. Antibodies against native and deglycosylated mouse gastric mucin (MGM) were raised in chickens. Screening of a mouse stomach cDNA library with the anti-(deglycosylated MGM) antibody yielded partial clones containing a 48 bp tandem repeat and 768 bp of non-repetitive sequence. The 16-amino-acid tandem repeat has a consensus sequence of QTSSPNTGKTSTISTT with 25% serine and 38% threonine. The MGM tandem repeat sequence bears no similarity to previously identified mucins. The MGM non-repetitive region shares sequence similarity with human MUC5AC and, to a lesser extent, human MUC2 and rat intestinal mucin. Northern blot analysis reveals a polydisperse message beginning at 13.5 kb in mouse stomach with no expression in oesophagus, trachea, small intestine, large intestine, caecum, lung or kidney. Immunoreactivity of antibodies against deglycosylated MGM and against a synthetic MGM tandem repeat peptide was restricted to superficial mucous cells, antral glands and Brunner's glands in the pyloric-duodenal region. DNA analysis shows that MGM recognizes mouse and rat DNA but not hamster, rabbit or human DNA. The MGM gene maps to a site on mouse chromosome 7 homologous to the location of a human secretory mucin gene cluster on human chromosome 11p15. Due to sequence similarity and predominant expression in the stomach, the MGM gene may be considered a MUC5AC homologue and named Muc5ac.

Expression cloning of gastric mucin complementary DNA and localization of mucin gene expression.

BACKGROUND & AIMS: Secretory mucins play an important role in gastric cytoprotection and are derived from a heterogeneous family of genes. The aim of this study was to determine the specific type and location of mucin gene expression in the human stomach. METHODS: Expression cloning was performed by screening a human gastric complementary DNA expression library with antisera against deglycosylated gastric mucin. RNA analysis and immunohistochemistry were used to quantify and localize mucin gene expression. RESULTS: Sequencing of positive clones revealed two clones containing tandem repeats. The first contained a 169-amino acid repeat and was named MUC6 (as previously described). The second contained the same 8-amino acid repeat consensus sequence (APTTSTTS) as complementary DNAs previously isolated from a tracheobronchial complementary DNA library and was labeled MUC5 (or MUC5AC). RNA analysis indicated that the gastric epithelium contains high levels of MUC5 and MUC6 messenger RNA with little or no MUC2, MUC3, and MUC4 messenger RNA. Immunohistochemical analysis showed that surface mucous cells of the cardia, fundus, and antrum expressed MUC5 peptide. In contrast, MUC6 peptide expression was limited to mucous neck cells of the fundus, antral-type glands of the antrum and cardia, and Brunner's glands of the duodenum. CONCLUSIONS: MUC5 and MUC6 represent major secretory mucins in the stomach and are localized to distinct cell types.

Mucin gene expression in normal, preneoplastic, and neoplastic human gastric epithelium.

Mucins synthesized by malignant cells may contribute (via decreased cellular adhesion and immune recognition) to cancer invasion and metastases. Human mucins are derived from a heterogeneous family of genes, labeled MUC1-6. Our aim was to determine the pattern of mucin gene expression in normal, preneoplastic (intestinal metaplasia), and malignant gastric specimens. Probes and antibodies for specific mucin tandem repeat sequences were used for RNA and immunohistochemical analysis. Normal stomach mucosa was characterized by expression of MUC1, MUC5, and MUC6 mRNA and immunoreactive protein, without MUC2, MUC3, and MUC4 gene expression. In contrast, high levels of MUC2 and MUC3 mucin mRNA and immunoreactive protein were found in specimens with intestinal metaplasia. Gastric cancers exhibited markedly altered secretory mucin mRNA levels compared with adjacent normal mucosa, with decreased levels of MUC5 and MUC6 mRNA and increased levels of MUC3 and MUC4 mRNA. Overall, immunoreactive MUC1 mucin was detected in 72% of 33 gastric cancers, and secretory mucin core peptides were expressed in 34% (MUC2), 45% (MUC3), 19% (MUC5), and 57% (MUC6) of these specimens. Coexpression of multiple (three or more) mucin core proteins occurred in 15 of 25 (60%) advanced (stages III and IV) cancers compared with 1 of 8 (12.5%) early (stages I and II) cancers (P < 0.048). We conclude that human gastric epithelium has a unique mucin gene pattern, which becomes markedly altered in preneoplastic and neoplastic specimens. Increased mucin gene heterogeneity in gastric adenocarcinomas is associated with advanced cancer stage.

Experimental model of upper intestinal adenocarcinoma induced by N-methyl-N'-nitro-N-nitrosoguanidine in C57BL/6 mice.

The purpose of this study was to develop a model of gastrointestinal carcinogenesis using C57BL/6 mice. Treatment regimens consisted of one control group and 2 groups which received N-methyl-N'-nitro-N-nitrosoguanidine (MNNG) in drinking water: 50 micrograms/ml x 52 weeks and 100 micrograms/ml x 27 weeks. In addition, 2 protocols using adjuvant agents intended to increase tumor formation were used: MNNG (100 micrograms/ml) x 27 weeks + 0.2% taurocholic acid added to the diet from weeks 13-52, and MNNG (50 micrograms/ml) x 33 weeks+caerulein (10 micrograms/kg) subcutaneously 3 times/week from weeks 21-52. High-grade dysplasia was observed in the duodenum of 1/13 mice treated with MNNG (50 micrograms/ml). The combination of the latter and caerulein did not augment tumorigenesis. Mice treated with MNNG (100 micrograms/ml) frequently developed neoplasia in the duodenum and upper jejunum. Foci of low-grade and high-grade dysplasia alone were found in 3/12 (25%) mice; and intramucosal and invasive adenocarcinoma were found in 7/12 (58.3%) mice. The addition of taurocholic acid significantly increased the number and histological stages of the tumors (adenocarcinoma occurred in 100%, P = 0.03) and decreased the time for tumor formation.

Heterogeneity of mucin gene expression in normal and neoplastic tissues.

To determine the relative expression of distinct mucin genes in normal and neoplastic tissue, antibodies and cDNA probes that recognize the core tandem repeat sequences of membrane-bound (MUC1) and secreted (MUC2 and MUC3) mucins were used for immunohistochemical and RNA Northern and slot-blot analysis. MUC1 mRNA was detected in all epithelial tissues tested. MUC1 core peptide, recognized by monoclonal antibodies 139H2 and DF3, was highly expressed on apical membranes of bronchus, breast, salivary gland, pancreas, prostate, and uterus, and was sparsely expressed in gastric surface cells, gallbladder, small intestine, and colonic epithelium. In contrast, MUC2 and MUC3 gene expression was primarily restricted to the intestinal tract. MUC2 mRNA was highly expressed in normal jejunum, ileum, and colon, compared with very low levels in normal bronchus and gallbladder. MUC3 mRNA was highly expressed in normal jejunum, ileum, colon, and gallbladder. Immunohistochemical studies using antibodies against synthetic MUC2 (anti-MRP) and MUC3 (anti-M3P) peptides indicate that MUC2- and MUC3-producing cells in the gastrointestinal tract are distinct. Goblet cells of the small intestine and colon reacted strongly with anti-MRP, whereas M3P reactivity was restricted to columnar cells of small intestinal villi, surface colonic epithelium, and gallbladder. Mucin protein epitopes and mRNA levels were frequently altered in adenocarcinomas compared to corresponding normal tissues. Alterations included increased expression, aberrant expression, and, less frequently, loss of expression. Increased MUC1 immunoreactivity was observed in most adenocarcinomas of the breast, lung, stomach, pancreas, prostate, and ovary. In addition, with the exception of prostate cancer, focal aberrant expression of MUC2 and MUC3 epitopes was frequently observed. Increased MUC1, MUC2, and MUC3 epitopes were present in colon adenocarcinomas of all histological subtypes, with the greatest increase of MUC2 epitopes observed in colloid (mucinous) colon cancers. MUC2 or MUC3 mRNA levels were increased in colloid colon cancer compared with normal colon, however in well- and moderately well-differentiated colon cancers MUC1, 2 and 3 mRNA levels were decreased. Compared with corresponding normal tissue, MUC1 mRNA levels were increased in breast cancer and well-differentiated lung cancers, and MUC3 mRNA was increased in gastric adenocarcinomas. Normal stomach lacked both MUC2 and MUC3 immunoreactivity and mRNA, however, MUC2 and MUC3 proteins and mRNA were highly expressed in gastric intestinal metaplasia. In conclusion, mucin genes are independently regulated and their expression is organ- and cell type-specific. Furthermore, neoplastic transformation is associated with dys-regulated expression of both membrane-bound and secreted mucin core protein epitopes and may be due to altered mucin mRNA levels and/or altered mucin glycosylation.

Valproic acid induction of coma in rats: synergism with NH4+ and pentobarbital.

Dose-response curves for the incidence of coma after intraperitoneal injections of various doses of valproic acid (VP) and octanoic acid (OA) showed that, mole for mole, valproic acid was less toxic than octanoic acid. However, a simultaneous subcoma dose of pentobarbital (PB) enhanced the toxicity of VP more than that of OA. The dose-response curve for NH4Cl was affected by simultaneous subcoma doses of VP and OA but not by PB. VP enhanced the toxicity of the NH4+ by 52%; OA enhanced the toxicity by 12%. PB added significantly to the toxicity of VP and NH4+ when the three were given simultaneously. Doses of 0.7 mmol NH4+ and 0.5 mmol VP given separately had little or no encephalopathic effect, with blood ammonias of 250-1250 micrograms/dl. When given simultaneously they induced a deep coma and raised the blood ammonia threefold, to about 3600 micrograms/dl. Similar doses of OA and NH4+, induced a similar deep coma, but blood levels of ammonia were not as high. Simultaneous injections of 250 mg glucose did not alter the results. Thus VP toxicity is enhanced substantially by its synergistic interactions with PB and the NH4+.

Studies of ammonia loading: effects of rate of delivery and enhanced removal of NH4 on blood levels of ammonia and coma induction.

Using dose-response curves, the dose of NH4Ac inducing coma in one-half of the animals was increased by 60 to 80% after 1 mmol of arginine. The larger increase occurred in larger rats but was not proportional to the increase in weight. Incremental subcoma doses of NH4 raised the amount of NH4 required for inducing coma and the brain level of ammonia at the point of coma. After a portacaval shunt the results were similar, although lower doses of NH4 were required from the beginning. Blood ammonias after a loading dose (1.25 mmol) of NH4 were influenced by the duration of a preinfusion of NH4 and by the preinjection of various amino acids involved in the disposal of NH4 in the urea cycle. The amount of reduction in blood ammonia by ornithine and arginine compounds was less the longer the preinfusion of NH4. Blood ammonia was not lowered by glutamate at any time but was increased with longer preinfusion periods. Hepatectomy (Hx) reduced the removal of an NH4 load. After a modest load (0.85 mmol) of NH4, blood ammonia increased 5-fold, over that of sham-operated rats, with 70% Hx and 15-fold with 90% Hx. Ornithine reduced these blood ammonias by about 50%. Arginine had no effect. These studies indicate ways of reducing toxicity of NH4 and factors that predispose to or enhance toxicity.

Effect of endotoxin and immunoglobulin on the course of experimental hepatic encephalopathy.

A single intraperitoneal dose of endotoxin (500 micrograms) shortened the time for development of hepatic coma by 27% in 300-g rats that had an end-to-side portacaval shunt followed within 48 hr by hepatic artery ligation. The body temperature of the rats was maintained at 37 degrees C, and the endotoxin was injected just after the hepatic artery was ligated. Controls were injected similarly with saline. The time to death was also shortened by 27%. A single intravenous dose of immunoglobulin (150 mg) delayed the time from the massive hepatic ischemia to the onset of hepatic coma by 19%. The immunoglobulin was injected just after the portacaval shunt was completed. Controls were injected similarly with 0.6 ml of 25% human serum albumin. While not large, these opposite effects of endotoxin and immunoglobulin were highly significant statistically. These observations complement the findings in human liver failure.

Hepatic regenerative enzyme activity after pericentral and periportal lobular toxic injury.

Pericentral and periportal liver injuries involving less than 50% of the parenchyma were produced with acetaminophen and allyl alcohol, respectively. Doses were selected to produce comparable peak serum malate dehydrogenase, sorbitol dehydrogenase, and SGPT activities. The regenerative response was assessed by serial measurements of hepatic thymidine kinase (TK) activity and ornithine decarboxylase (ODC) activity. The initial responses reflected in ODC activity were more or less similar. However, the ultimate regenerative response reflected by TK activity was almost three times as great after periportal injury as after pericentral injury, after allowing for differences in the extent of necrosis. Histologic examination also showed greater mitotic and tissue reparative responses after periportal injury. These results suggest that the concept of hepatocellular heterogeneity applies to the regenerative response of liver cells as well as the metabolic functions previously identified.

Ammonia toxicity: comparative protective effect of various arginine and ornithine derivatives, aspartate, benzoate, and carbamyl glutamate.

Ornithine and arginine compounds were highly effective in preventing an increase in blood ammonia and in preventing or minimizing encephalopathy after acute subcoma, coma-inducing, or lethal doses of NH4+. Similar protection was seen after subacute loading with glycine. Ornithine ketoacid derivatives were no more effective than ornithine alone or ornithine glutamate. Ornithine appeared to be a little more effective than arginine, but the differences were slight. Aspartate and glutamate alone were ineffective. Carbamyl glutamate was much less effective than either ornithine glutamate or arginine glutamate. Orotic acid excretion was markedly increased in the presence of excess NH4+. This increment was eliminated with ornithine or arginine, although the reduction with arginine was unpredictably erratic. Aspartate increased the orotic acid excretion and the amount of urea formed. Sodium benzoate was borderline in its effect on the blood ammonia and on orotic acid excretion.

Acetaminophen liver injury: sequential changes in two biochemical indices of regeneration and their relationship to histologic alterations.

Massive liver injury was produced in fasting male Sprague-Dawley rats weighing 200 +/- 25 gm each by gastric administration of 1400 mg/kg acetaminophen. The time sequence of changes in liver ornithine decarboxylase (ODC) activity, which reflects the earliest phases of cell multiplication, liver thymidine kinase (TK) activity, which reflects DNA synthesis, and liver histology (necrosis, mitosis, and repair processes) was recorded. ODC showed the usual biphasic response. By 12 hours, it reached its first peak, a six- to eightfold increase. At this time there was no histologic evidence of necrosis, and serum malate dehydrogenase (MDH), sorbitol dehydrogenase (SDH), and alanine aminotransferase (SGPT) were normal. During the next 12 hours ODC decreased by 60% to 70% and cellular necrosis became evident, and reached a peak at 24 to 36 hours, as did serum MDH, SDH, and SGPT. The serum enzymes fell precipitously at 48 hours, but the histologic evidence of necrosis subsided gradually over 60 hours. The secondary ODC peak, a fourfold increase, coincided with rising activity of TK, which increased 25- to 35-fold over 54 to 72 hours, and then subsided. At 54 hours, when DNA synthesis had already peaked, there was no histologic evidence of repair other than mitoses. However, within the next 6 hours, evidences of repair became prominent, and remained so for another 36 hours before subsiding. Thus, with acetaminophen injury, the initial phases in preparation for cell multiplication occurred before histologic evidence of injury was apparent, and DNA synthesis peaked before other evidence of tissue repair became evident.(ABSTRACT TRUNCATED AT 250 WORDS)

Ammonia, octanoate and a mercaptan depress regeneration of normal rat liver after partial hepatectomy.

Four injections of subcoma doses of ammonium acetate, octanoic acid or dimethyl disulfide during the first 24 hr after two-lobe hepatectomy in normal rats markedly depressed DNA synthesis as reflected by liver thymidine kinase activity or the incorporation of tritiated thymidine into hepatic DNA. Recovery from the depressant effects of the three toxins took 16 to 28 hr. Similar doses of the same toxins injected hourly for 3 or 5 hr after the two-lobe hepatectomy had similar depressant effects on the early peak of ornithine decarboxylase activity measured at 4 or 6 hr. Recovery occurred within 3 hr perhaps because of the very short half-life of ornithine decarboxylase and its rapid regeneration time. These observations may have implications for the lack of regeneration observed in many patients with fulminant hepatic failure who have accumulated sufficient ammonia, methanethiol and fatty acids over periods of days or weeks to become encephalopathic.

Brain methanethiol and ammonia concentrations in experimental hepatic coma and coma induced by injections of various combinations of these substances.

In normal rats in a coma induced by NH+4 alone or by methanethiol alone, the brain and blood levels of ammonia or methanethiol are much higher than those observed in rats in experimental hepatic coma. When various smaller dosage combinations of NH+4, methanethiol, and octanoic acid were injected simultaneously, coma occurred at lower brain and blood concentrations of ammonia and methanethiol. Brain ammonia and methanethiol concentrations in normal rats receiving 0.75 mmol NH+4 plus 0.15 mmol octanoic acid plus 18 mumol methanethiol were comparable with those observed in 24 rats in hepatic coma after fulminant hepatic failure caused by acute massive ischemic liver necrosis. The normal rats became comatose. In these rats and in the rats in hepatic coma, the ammonia level in the brain was increased threefold and the methanethiol level in the brain was increased fivefold. Because these levels of ammonia and methanethiol were sufficient to induce coma in normal rats, they should also have been sufficient to induce coma in rats with damaged livers. Therefore, the accumulation of ammonia and methanethiol in the central nervous system after the acute massive ischemic necrosis may have been sufficient to account for the coma that ensued, without the involvement of other factors.

Urea excretion, galactose elimination, and aminopyrine disappearance during normal liver regeneration after partial hepatectomy. Comparison with other function tests reported.

After removal of 70% of normal rat liver, liver weight was 37% of control at 3 1/2 hours and 49% at 24 hours. Urea formation per gram liver after an NH4+ load and galactose elimination per gram liver were well-preserved during this early posthepatectomy period. At 46 hours after removal, galactose elimination was transiently less than would be expected from liver weight. The aminopyrine disappearance rate fell to 50% to 63% of control values until 72 hours, when it decreased further to 43%, disproportionately less than liver weight. The disappearance rate per gram liver at this time was 63% of control. Serum bile acid concentration increased 17-fold by 24 hours, and was five times that of control concentration at 72 hours. Comparisons of these four measures and six other in vivo measures of function previously reported permitted a segregation of the livers into two groups according to liver function after hepatectomy: those that were disproportionately preserved or were proportional to the amount of liver present, and those that were disproportionately reduced in relation to liver weight. Four additional measurements reported in isolated perfused livers after hepatectomy were similarly segregated. It appears that those liver functions that most closely reflect cellular growth or associated changes in liver blood flow are preserved, whereas those that are possibly more specialized and differentiated are depressed during regeneration.

Toxicity of a fatty acid and ammonia: interactions with hypoglycemia and Krebs cycle inhibition.

In rats, hypoglycemia induced with insulin and Krebs cycle inhibition produced by fluoroacetate poisoning augmented the encephalopathic effects of octanoic acid and NH4Cl. Blood sugars in the range of 25 to 45 mg/dl resulted in a decrease of approximately 30% in the doses of NH+4 and of octanoate required to induce coma in otherwise normal rats. Mild fluoroacetate poisoning resulted in a corresponding decrease of 25% in the dose of NH+4 and 10% in the dose of octanoate. A single coma-inducing dose of octanoate or several subcoma doses resulted in blood sugar decrements of 50% or more within 1 to 3 hr. A combination of subcoma doses of octanoate and NH+4 had the same effect. Pretreatment with glucose prevented the hypoglycemia and blunted the encephalopathic response to octanoate. A single coma-inducing dose of NH+4 or several subcoma doses resulted in blood sugar increments of 50% or more within 1/2 to 5 hr.

Encephalopathic effect of phenol in rats.

The coma-inducing effect of phenol was studied in normal 300 +/- 50 gm Sprague-Dawley rats. Dose-response curves were developed which showed that one-half the animals became deeply comatose with 540 mumol of intraperitoneal phenol and 100% with 600 mumol. Five stages of encephalopathy were readily distinguished. In stage I, spontaneous activity was noticeably decreased, posture was upright, back-leg control normal, muscle tonus slightly increased, and response to stimuli normal. In Stage V, activity was gone, the rats were deeply unconscious, righting reflex and back-leg control were gone, muscles were completely relaxed, and there was no response to stimuli. With a coma-inducing dose of phenol, spontaneous activity decreased. After 1 min a body tremor developed interspersed with unpredictable jumping, and all four limbs began to shake. Leg control and then the righting reflex were lost by 2 min. Within 5 min rats were deeply unconscious and muscles completely relaxed while shaking of limbs continued until recovery 20 to 60 min later or death. With 480 mumol or less, none of the rats became comatose, but the shaking of limbs was present. The coma-inducing dose of phenol was reduced by 10% to 20% with simultaneous injection of subcoma doses of NH4+ or OA and by 20% to 30% with simultaneous DMDS leads to 2 methanethiol. Conversely, a subcoma dose of phenol reduced the coma-inducing doses of NH4+, OA, and DMDS by approximately 20% to 25%. Thus phenol, which accumulates in human hepatic failure, induces coma by itself in rats and acts synergistically with other hepatic failure toxins.