Hydrogen sulphide synthesis in the rat and mouse gastrointestinal tract

AIMS

Hydrogen sulphide (H2S) exerts several anti-inflammatory effects, accelerates the healing of experimental gastric ulcers, and can stimulate intestinal secretion. Little is known about H2S synthesis in the gastrointestinal tract. The aim of this study was to characterize H2S synthesis throughout the gastrointestinal tract.

METHODS

H2S synthesis in various gastrointestinal tissues of rats and mice was determined. The effects and selectivity of inhibitors of two key enzymes for H2S synthesis, cystathionine-gamma-lyase and cystathionine-beta-synthase, were examined. Cystathionine-gamma-lyase and cystathionine-beta-synthase expression was evaluated by Western blotting and immunohistochemistry. Cystathionine-gamma-lyase and cystathionine-beta-synthase expression in biopsies of human colon was also examined.

RESULTS

H2S synthesis was variable throughout the gastrointestinal tract in parallel with variations in cystathionine-gamma-lyase and cystathionine-beta-synthase expression. The efficacy of cystathionine-beta-synthase and cystathionine-gamma-lyase inhibitors to reduce H2S synthesis in these tissues was also variable. Cystathionine-beta-synthase is the predominant source of H2S synthesis in the colon of rodents. Cystathionine-gamma-lyase and cystathionine-beta-synthase were also expressed in healthy human colon biopsies.

CONCLUSIONS

The capacity for H2S synthesis varies throughout the rodent gastrointestinal tract, as does the distribution and contribution of the two key enzymes. Investigation of additional enzymatic sources of H2S and the development of more selective inhibitors are suggested.

Reduction of gastrointestinal injury in acute endotoxic shock by flurbiprofen nitroxybutylester

Nitric oxide has been reported to have paradoxical effects in experimental endotoxic shock, contributing to the hemodynamic consequences of endotoxin administration, but apparently protecting the gastrointestinal mucosa. A novel class of nitric oxide-releasing nonsteroidal anti-inflammatory drug (NSAID) derivatives has recently been described which exert anti-inflammatory activities but produce significantly less gastrointestinal injury than the parent nonsteroidal anti-inflammatory drugs from which they are derived. Thus, the present study was performed to determine the effects of one of these derivatives, flurbiprofen 4-nitroxybutylester, compared to the native nonsteroidal anti-inflammatory drug, flurbiprofen, in an experimental model of endotoxic shock. Intravenous administration of endotoxin from Salmonella typhosa to rats pretreated with flurbiprofen produced a profound decrease in systemic arterial blood pressure, an increase in hematocrit and extensive gastric and small intestinal damage. In rats pretreated with flurbiprofen 4-nitroxybutylester, endotoxin produced comparable changes in blood pressure and hematocrit to those seen in rats treated with flurbiprofen; however, the severity of gastrointestinal damage was significantly reduced. Gastric blood flow was profoundly decreased following endotoxin administration, but was significantly higher in rats pretreated with flurbiprofen 4-nitroxybutylester than in rats pretreated with flurbiprofen. These results demonstrate that despite not affecting the acute systemic effects of endotoxin administration, flurbiprofen 4-nitroxybutylester is capable of protecting the gastrointestinal mucosa from injury, possibly through preservation of mucosal blood flow.

Adaptation of rat gastric mucosa to aspirin requires mucosal contact

Adaptation of the gastric mucosa to repeated administration of aspirin is a well-documented phenomenon, but the underlying mechanism is not fully understood. In this study, we tested the hypothesis that adaptation of the rat stomach to chronic aspirin administration required contact between the aspirin and the gastric mucosa. Rats were orally treated twice daily with either aspirin (100 mg/kg) or the vehicle. After various periods of treatment (< or = 20 days), the rats were given a higher dose of aspirin (250 mg/kg po), and the extent of gastric damage was assessed 3 h later. Rats receiving chronic aspirin demonstrated the development, in a time-dependent manner, of resistance to the damaging effects of aspirin. Chronic aspirin administration also significantly decreased the susceptibility of the rat stomach to damage induced by indomethacin or naproxen. The adaptation phenomenon was associated with a parallel increase in inflammatory infiltration of the mucosa, as measured by tissue myeloperoxidase activity and histology. Prostaglandin synthesis was markedly suppressed (> 80%) in all rats treated with aspirin. Gastric mucosal ornithine decarboxylase activity was not affected by chronic aspirin administration. If aspirin was administered subcutaneously or intrajejunally for 20 days, neither adaptation nor inflammation of the gastric mucosa was observed. These studies demonstrate that the rat stomach adapts to chronic oral administration of aspirin, but not to aspirin administration via other routes. Adaptation of the gastric mucosa occurred in parallel to infiltration of granulocytes. Whether these two phenomena are mechanistically or causally linked is not yet clear.

Hepatic and gastric cytoprotective effects of long-term prostaglandin E1 administration in cirrhotic rats

BACKGROUND

Acute administration of prostaglandin E (PGE) may be cytoprotective for hepatocytes in acute hepatitis and for gastric mucosa in cirrhotic rats. We examined the effects of long-term PGE treatment on liver and stomach in cirrhotic rats.

METHODS

Cirrhosis was induced by bile duct ligation. Controls had a sham operation. Half the rats received a PGE1 analogue, misoprostol (PGE1) (10 micrograms orally, daily) on days 1-29 postsurgery, and the others received vehicle only. On day 31, all rats underwent ex vivo gastric chamber procedures. Liver chemistry, portal pressures, and hepatic and gastric tissue levels of prostaglandin E2, leukotriene B4, myeloperoxidase, and collagen were determined.

RESULTS

PGE1-treated cirrhotic rats had less hepatosplenomegaly, lower serum alanine aminotransferase levels, and portal pressures and higher arterial pressure than vehicle-treated cirrhotic rats. Hepatic and gastric leukotriene B4, myeloperoxidase and collagen levels were significantly lower in the PGE1-treated compared with vehicle-treated cirrhotic rats. Vehicle-treated cirrhotic rats had greater spontaneous and ethanol-induced gastric damage and failed to show a gastric hyperemic response to ethanol, whereas PGE1-pretreated rats did. PGE1 did not significantly affect sham-operated rats.

CONCLUSIONS

Long-term PGE1 administration was cytoprotective for both the liver and gastric mucosa in cirrhotic rats. Clinical trials of PGE in human cirrhosis or portal hypertensive gastropathy may be warranted.

Attenuation of epithelial injury in acute experimental colitis by immunomodulators

Intestinal epithelial permeability can be modulated by the immune system and can be greatly increased by transepithelial migration of neutrophils. Since immunosuppressants have been reported to inhibit the ability of neutrophils to migrate, we assessed the effects of two immunosuppressants on epithelial permeability and granulocyte infiltration in a model of acute colitis. Epithelial permeability was measured at 3 and 6 h after induction of colitis in the rabbit by intracolonic administration of trinitrobenzene sulfonic acid. At these times, blood-to-lumen leakage of 51Cr-EDTA was elevated by approximately 8- and 18-fold, respectively, above levels observed in healthy controls. Pretreatment with either of the immunosuppressants (cyclosporin A and L-683,590) significantly reduced the changes in 51Cr-EDTA leakage observed at the latter time point. These drugs also significantly attenuated granulocyte infiltration of the colon after induction of colitis, as measured by tissue myeloperoxidase activity. Unlike the immunosuppressants, misoprostol, a prostaglandin analogue, attenuated the increases in colonic permeability but had no effect on granulocyte infiltration in this model. These results demonstrate that two structurally unrelated immunosuppressants are capable of markedly reducing neutrophil infiltration and the colonic permeability changes observed in an experimental model of acute colitis, although the mechanisms through which these effects are produced remain unclear.

Characterization of a simple animal model for nonsteroidal anti-inflammatory drug induced antral ulcer

Most animal models of nonsteroidal anti-inflammatory drug (NSAID) induced gastric injury are characterized by acute, superficial erosions in the corpus region, whereas the clinically significant injury in man is the deep, antral ulcer. The purpose of this study was to characterize a model of NSAID-induced antral ulceration that more closely resembles the type of damage seen in man. Rabbits received indomethacin subcutaneously every 12 h. The progression of ulcer formation was followed by killing groups of animals after one to seven doses of indomethacin. The dose dependency of ulcer formation was assessed by giving indomethacin at doses of 1 to 20 mg/kg. Healing of antral ulcers was determined by examining the stomach at various times after administering the seventh dose of indomethacin (20 mg/kg). The effects of prophylactic treatment with misoprostol or ranitidine on ulcer formation were assessed. Indomethacin administration initially produced superficial erosions in the corpus and antrum, but with time, ulcers became apparent in the antrum. The formation of these ulcers was dependent upon the number of times indomethacin was administered and the dose. Similar ulcers could be induced with a second NSAID, diclofenac. Misoprostol treatment resulted in a significant reduction in the extent of indomethacin-induced antral ulceration, but ranitidine had no effect. Antral ulcers healed progressively following cessation of indomethacin administration and were almost completely resolved by 108 h after the final dose of indomethacin. These results demonstrate that subcutaneous NSAID administration to rabbits is a simple and reproducible method for producing ulcers that bear striking macroscopic resemblance to NSAID-induced antral ulcers in man.(ABSTRACT TRUNCATED AT 250 WORDS)

Protective effects of PF-10040 in experimental NSAID-gastritis: role of leukocytes, leukotrienes and PAF

The effects and mechanism of action of PF-10040, a quinoline derivative, were examined in an experimental model of NSAID-gastritis. Oral pretreatment with PF-10040 dose-dependently reduced the severity of indomethacin-induced gastric damage, with a significant protective effect being observed with doses of 10 mg/kg or greater. The protective effects of this compound persisted for as long as 6 h after administration. PF-10040 also significantly reduced the severity of aspirin- or naproxen-induced gastric damage. At protective doses, PF-10040 did not significantly affect gastric LTB4 synthesis, LTB4-induced granulocyte recruitment to intradermal injection sites, or LTD4-induced changes in gastric blood flow. The free radical scavenging effects of PF-10040 were examined using an in vitro assay, in which it failed to exert significant effects at concentrations of up to 10 mM. PF-10040 had no significant effect on gastric acid secretion. In rat mesenteric venules, PF-10040 inhibited PAF-, but not LTB4-induced leukocyte adherence and emigration. These results suggest that the protective effects of PF-10040 are not attributable to scavenging of free radicals, inhibition of leukotriene synthesis, blockade of LTB4 or LTD4 receptors or inhibition of LTB4-mediated leukocyte endothelial cell adhesion.

Characterization of spontaneous and ethanol-induced gastric damage in cirrhotic rats

Cirrhosis was induced in Sprague-Dawley rats via ligation of the common bile duct. Changes in gastric blood flow and mucosal architecture were examined. Using an ex vivo gastric chamber preparation, the susceptibility of the cirrhotic gastric mucosa to injury by 20% ethanol was also examined. The gastric mucosa of cirrhotic animals was abnormal, even before ethanol administration. The macroscopically visible damage in these animals ranged from superficial hyperemia to epithelial sloughing. These gastric lesions were similar in appearance to the gastropathy described in cirrhotic patients, including "cherry-red spots" and areas of generalized erythema. Cirrhotic rats had a lower resting gastric transmucosal potential difference than control rats, and their gastric mucosa was also significantly more susceptible to damage by topical ethanol application. Ethanol administration caused a significant increase in gastric blood flow in control rats, whereas it significantly decreased gastric blood flow in cirrhotic rats. This lack of a reactive hyperemic response in cirrhotic rats may be responsible for the increased susceptibility of the gastric mucosa to ethanol-induced damage.

Capsaicin-induced hyperemia in the stomach: possible contribution of mast cells

Topical application of capsaicin to the gastric mucosa results in marked hyperemia as a consequence of the release of vasoactive neuropeptides from sensory afferent neurons. Because many of these neuropeptides have the capacity to induce mast cell degranulation, we investigated the possible contribution of mast cells to capsaicin-induced hyperemia. Application of capsaicin to the gastric mucosa of normal rats resulted in a concentration-dependent increase in blood flow. In rats in which mastocytosis was induced by prior infection with Nippostrongylus brasiliensis, the hyperemic responses to capsaicin were significantly greater than in control rats. This augmented hyperemic response could be significantly attenuated by pretreatment with a histamine H1-receptor antagonist (pyrilamine) or with a mast cell stabilizer (doxantrazole). Depletion of mucosal mast cells through treatment with dexamethasone also significantly reduced the hyperemic response to capsaicin. Hyperemic response to capsaicin in normal rats and in rats with mucosal mastocytosis could be completely abolished by pretreatment with ruthenium red or prior ablation of the sensory afferent neurons with capsaicin. These results suggest that in rats with gastric mastocytosis, sensory neuron-dependent activation of mast cells contributes to the hyperemic response to topical capsaicin. These findings are therefore consistent with the hypothesis that there is communication between nerves and mast cells in the gastric mucosa, at least in rats previously infected with N. brasiliensis.

Prevention and reversal of experimental colitis by a monoclonal antibody which inhibits leukocyte adherence

The role of neutrophils in the pathogenesis of acute colitis was investigated using a rabbit model. Colitis was induced by intracolonic administration of trinitrobenzene sulfonic acid in 30% ethanol. Myeloperoxidase activity was measured at various times after induction of colitis as an index of neutrophil infiltration, and this was confirmed by histology. The permeability of the colonic epithelium to [51Cr]EDTA was also measured at various times after induction of colitis. The most marked increase in neutrophil infiltration of the colon occurred during the period 3-6 h after induction of colitis. This was also the period in which the greatest increase in colonic permeability was observed. Pretreatment with a monoclonal antibody (IB-4) directed against the leukocyte adhesion molecule, CD18, markedly suppressed neutrophil infiltration into the colonic tissue after induction of colitis. This pretreatment also significantly reduced the extent of epithelial injury. Administration of IB-4 to rabbits 12 h after induction of colitis resulted in a rapid decline in tissue myeloperoxidase activity. When measured 12 h after IB-4 administration (3 mg/kg), colonic myeloperoxidase activity was reduced by about 80% compared to the control group treated with the vehicle. These results are consistent with the hypothesis that neutrophils contribute significantly to the epithelial dysfunction that characterizes colitis and suggest that antibodies directed against adhesion molecules may represent a novel approach to the treatment of intestinal inflammatory disorders.

A monoclonal antibody against the CD18 leukocyte adhesion molecule prevents indomethacin-induced gastric damage in the rabbit

The role of leukocyte adherence in the mechanism of gastropathy induced by nonsteroidal antiinflammatory drugs was investigated using a rabbit model. Gastric damage was induced by intragastric instillation of indomethacin [5 mg/mL] for a period of 30 minutes. Histologically, this treatment resulted in extensive vascular congestion and leukocyte margination within the mucosa. Pretreatment with a monoclonal antibody [IB-4] directed against the common beta subunit of the CD11/CD18 adhesion glycoprotein complex significantly (P less than 0.05) reduced both the vasocongestion and the prevalence of leukocyte margination. Macroscopically, indomethacin treatment resulted in the formation of numerous hemorrhagic lesions in the corpus region of the stomach. Pretreatment with IB-4 reduced the extent of gastric hemorrhagic damage by approximately 85% (P less than 0.001). Damage in the group pretreated with IB-4 did not differ significantly from that in rabbits that did not receive indomethacin. In separate experiments, the dose of IB-4 used was shown to completely suppress the recruitment of granulocytes in response to two different agonists. These results support the hypothesis that leukocyte adherence to the vascular endothelium is an important event in the pathogenesis of ulceration induced by nonsteroidal antiinflammatory drugs. Leukocytes might contribute to ulceration by occluding microvessels, thereby reducing mucosal blood flow, and by releasing various mediators, proteases, and free radicals that can produce tissue necrosis.

The mucoid cap over superficial gastric damage in the rat. A high-pH microenvironment dissipated by nonsteroidal antiinflammatory drugs and endothelin

The ability of the "mucoid cap" of mucus, fibrin, and cellular debris that forms over sites of superficial gastric damage to provide a relatively high-pH microenvironment was investigated in the rat. Furthermore, the effects of administration of nonsteroidal antiinflammatory drugs or reduction of mucosal blood flow on this microenvironment were investigated. Superficial mucosal damage induced by a 2-minute exposure to hypertonic saline produces a mucoid cap that has a pH of 4-6 despite the presence in the lumen of a solution of acid with a pH of less than 1. This relatively high-pH microenvironment remained even when the exposure to 0.15-mol/L HCl was continued for as long as 1 hour. Intraperitoneal administration of indomethacin (5 mg/kg) or naproxen (10 mg/kg) resulted, within 10-25 minutes, in complete dissipation of this high-pH microenvironment, with the subsequent development of hemorrhagic erosions. These effects of indomethacin and naproxen occurred subsequent to significant inhibition of gastric prostaglandin synthesis by these agents. Rapid dissipation of the pH gradient could also be produced by brief (30-second) clamping of the arterial blood supply to the chambered mucosa or by systemic administration of a vasoconstrictor (endothelin-1), in both cases resulting in the development of hemorrhagic erosions. These results show that the mucoid cap over sites of superficial damage provides a relatively high-pH microenvironment and that dissipation of this microenvironment by inhibitors of prostaglandin synthesis or by reduction of mucosal blood flow can convert sites of superficial mucosal injury to hemorrhagic erosions. Such inhibitory effects might contribute to the ulcerogenic actions of nonsteroidal antiinflammatory drugs.

Platelet-activating factor mediates gastric damage induced by hemorrhagic shock

The role of platelet-activating factor (PAF) as a mediator of the gastric damage associated with hemorrhagic shock was investigated using a rat model. With use of an ex vivo gastric chamber preparation, the gastric mucosa was bathed with 0.1 M HCl for 90 min. At minute 10 the systemic arterial blood pressure (BP) was reduced to 25 mmHg by bleeding from the femoral artery. BP was maintained at this level for 15 min, then the shed blood was reinfused. In control rats subjected to this protocol, extensive gastric damage developed during and after the shock period and involved an average of 50 +/- 8% of the total area of glandular mucosa. A marked decrease in transmucosal potential difference (PD) was observed during shock, with little recovery thereafter. Also, significant appearance of protein and hemoglobin (Hb) in the gastric lumen was detected after induction of shock. Oral pretreatment of the rats with the PAF antagonist WEB 2086 (0.5-20 mg/kg) dose dependently reduced the extent of macroscopically visible gastric damage, the decrease in transmucosal PD, and the appearance in the lumen of protein and Hb. A similar protective effect was observed with another PAF antagonist, BN 52021 (10 mg/kg). With use of laser-Doppler flowmetry, changes in gastric blood flow were determined before, during, and after induction of shock.(ABSTRACT TRUNCATED AT 250 WORDS)

Effects of leukotrienes on susceptibility of the rat stomach to damage and investigation of the mechanism of action

The ability of various leukotrienes to alter the susceptibility of the rat gastric mucosa to injury by 20% ethanol and the possible mechanism of action were examined using an ex vivo gastric chamber preparation. Intraarterial infusions of leukotriene B4 or N-acetyl leukotriene E4 (0.01-1.0 microgram/kg per min for 10 min) had no significant effect on the extent of damage induced by topically applied 20% ethanol. However, infusion of leukotriene C4, D4, or E4 (1.0 micrograms/kg per min) significantly increased ethanol-induced damage, as measured macroscopically, histologically, and functionally. Lower doses of leukotriene C4, D4, or E4 were without significant effect in this model. The increase in damage induced by these three leukotrienes could be blocked by pretreatment with either of two structurally unrelated leukotriene D4 antagonists (L-649,923 or L-660,711). The augmentation of damage by leukotriene C4 was not affected by pretreatment with indomethacin or with a specific thromboxane A2-receptor antagonist (L-670,596). At the dose that increased ethanol-induced damage, none of the leukotrienes tested significantly altered gastric vascular permeability, as measured by Evan's blue leakage. However, using laser-Doppler flowmetry, leukotrienes C4 and D4 were found, when administered intraarterially at doses in the 0.05-1.0 micrograms/kg per min range, to produce dose-dependent reductions of gastric blood flow while N-acetyl leukotriene E4 was without effect and leukotriene B4 induced slight increases. The effects of leukotrienes C4 and D4 on gastric blood flow could be inhibited by the two leukotriene D4 antagonists but not by the thromboxane antagonist. These results demonstrate that although they do not produce damage by themselves, leukotrienes C4, D4, and E4 are capable of augmenting ethanol-induced injury to the gastric mucosa. Changes in vascular permeability do not appear to play a role in the mechanism of action of the leukotrienes, while their effects on gastric blood flow are likely to be important. Under certain condition it is therefore possible that local release of leukotrienes could, at least in part through reducing vascular perfusion, predispose the surrounding tissue necrosis.

Comparison of the damage-promoting effects of leukotrienes derived from eicosapentaenoic acid and arachidonic acid on the rat stomach

The ability of leukotrienes derived from eicosapentaenoic acid were compared with counterpart leukotrienes derived from arachidonic acid in terms of their ability to affect susceptibility of the stomach to injury induced by a topical irritant and their ability to alter gastric blood flow. Intra-arterial infusion of leukotriene C4 (LTC4) and LTD4 (0.1-3 micrograms/kg/min for 5 min) produced dose-dependent increases in gastric mucosal damage induced by topically applied 20% ethanol, as assessed macroscopically, by changes in transmucosal potential difference and by measurement of efflux of protein into the gastric lumen. Similar doses of LTC5 or LTD5 did not produce significant changes in any of these three parameters, when compared with control rats receiving the vehicle. With a higher dose of LTC5 or LTD5 (5 micrograms/kg/min), significant damage was observed. LTC4 and LTD4 were also found to be more potent at reducing gastric blood flow than LTC5 and LTD5. These results demonstrate that the peptido-leukotrienes derived from eicosapentaenoic acid (LTC5 and LTD5) are on the order of five times less potent than the leukotrienes derived from arachidonic acid (LTC4 and LTD4), in terms of increasing the susceptibility of the gastric mucosa to damage and reducing gastric blood flow. These results may have important implications in terms of the hypothesis that fish oil diets may be protective or may accelerate healing in ulcerative diseases of the gastrointestinal tract.

Comparison of the effects of endothelin-1 and endothelin-3 on the rat stomach

The effects of systemic administration of endothelin-1 and endothelin-3 on the susceptibility of the stomach to injury were compared in the anesthetized rat, as were their effects on gastric vascular tone, systemic blood pressure and hematocrit. When infused at concentrations in the 10(-7)-10(-6) M range, endothelin-1 was a far more potent hypertensive agent than endothelin-3. Endothelin-1 caused significant hemoconcentration, while endothelin-3 did not Endothelin-1 was approximately 5- to 10-times more potent as a vasoconstrictor in the stomach and a similar difference in potencies was observed when the ability of these peptides to increase the susceptibility of the stomach to ethanol-induced damage was compared. The two peptides were equipotent in producing gastric mucosal hemorrhage in the absence of any exogenous irritant. These results demonstrate that like endothelin-1, endothelin-3 has ulcerogenic and vasoconstriction actions in the stomach. While there are very large differences in the potencies of the two peptides in terms of producing systemic hypertension and hemoconcentration, the differences in the potency of the gastric ulcerogenic and vasoconstrictor effects are much less marked.

The modulation of gastric mucosal integrity by endothelin-1 and prostacyclin

The interaction of the vasoconstrictor peptide, endothelin-1 (ET-1), and the endothelium-derived vasodilator eicosanoid, prostacyclin, was examined as it pertains to the modulation of gastric mucosal integrity. Using an ex vivo chamber preparation of the rat stomach, the effects of intravenous ET-1 on the susceptibility of the mucosa to damage induced by topical application of an irritant, 20% ethanol, were examined. ET-1 significantly augmented gastric hemorrhagic damage induced by the irritant when administered at concentrations in the 10(-7) to 10(-6) M range. Pretreatment with indomethacin at a dose that inhibited gastric prostacyclin synthesis by over 85% resulted in significant augmentation of the ulcerogenic actions of ET-1. The damaging actions of ET-1 could be significantly reduced by topical pretreatment of the gastric mucosa with prostacyclin (5-50 micrograms/ml). This pretreatment also significantly reduced the hypertension and hemoconcentration observed following ET-1 administration. ET-1 also significantly augmented the susceptibility of the gastric mucosa to injury induced by hydrochloric acid. Oral administration of 150 mM HCl produced little or no gastric damage in control rats. However, a 5-min intravenous infusion of ET-1 produced significant increases in the severity of acid-induced gastric damage in a concentration-dependent manner (10(-7) to 10(-6) M). These results demonstrate that ET-1 is a potent ulcerogenic agent in the rat stomach. The ulcerogenic actions of ET-1 can be significantly reduced by prostacyclin, suggesting that the balance between endothelial cell release of ET-1 and prostacyclin may be an important factor in modulating gastric mucosal integrity.