Elizabeth J. Dial

NSAID injury to the gastrointestinal tract: evidence that NSAIDs interact with phospholipids to weaken the hydrophobic surface barrier and induce the formation of unstable pores in membranes

In this review, we have discussed our current understanding of the barrier properties that are in place to protect the upper gastrointestinal mucosa from luminal acid, and the pathogenic mechanism by which nonsteroidal anti‐inflammatory drugs (NSAIDs) induce injury to the gastrointestinal tract. The changes in our view of the importance of NSAID‐induced cyclo‐oxygenase (COX) inhibition on the pathogenesis and prevention of NSAID‐induced gastrointestinal injury is presented. The focus of this paper has been placed on the effects of NSAIDs on the mucosal surface, and specifically the effect of these powerful drugs in inducing changes in the hydrophobicity, fluidity, biomechanical and permeability properties of extracellular and membrane phospholipids. Lastly, recent evidence is presented that salicylic acid and related NSAIDs may alter the stability of membranes, inducing the formation of unstable pores that may lead to back‐diffusion of luminal acid and membrane rupture. This understanding of the interaction of NSAIDs with membrane phos‐pholipids may prove valuable in the design of novel NSAID formulations with reduced gastrointestinal side‐effects.

Role of luminal ammonia in the development of gastropathy and hypergastrinemia in the rat

BACKGROUND/AIMSnChronic infection with Helicobacter pylori causes persistent elevations in gastric juice ammonia levels. Thus, we studied the effects of experimentally induced increases in gastric juice ammonia levels on gastric structure and function and gastrin homeostasis.nnnMETHODSnRats were fed either normal chow or the diet supplemented (20 g/dL) with ammonium or sodium acetate.nnnRESULTSnLong-term dietary ammonium loading for 2 weeks or longer resulted in a 1.5-2-fold increase in the weight and mucosal thickness of the stomach and proximal duodenum with evidence of mild gastritis and enterochromaffinlike cell hyperplasia. The ammonium-containing diet also induced a significant 2-3-fold increase in both circulating gastrin levels of fed rats and an increase in the postprandial gastrin responses over control values. Antral gastrin levels were also markedly elevated by long-term ingestion of the test diet, which was increased 3-4-fold over control values in fasted animals and less so after meal stimulation. Consistent with these findings, gastrin-specific messenger RNA was increased 2.5-3-fold in the antrum of ammonium fed rats, whereas actin-specific messenger RNA was not affected or decreased. Animals fed a diet supplemented with 20 g/dL sodium acetate sustained modest increases in mucosal thickness and serum and antral gastrin concentration, suggesting that nonspecific gastric injury and inflammation is also a factor that influences G-cell function.nnnCONCLUSIONSnLong-term exposure of the antral mucosa to elevated levels of ammonia in the gastric juice in the presence of gastritis, conditions similar to that occurring in subjects infected with H. pylori, seem to be causative factors in the development of G-cell hyperfunction.

Attenuation of hydrophobic phospholipid barrier is an early event in Helicobacter felis-induced gastritis in mice

Helicobacter pylori infection has been linked tothe development of gastritis which can then progress toa number of disease entities including peptic ulcerdisease and gastric cancer. Since the pathogenic mechanism by which the bacteria causesgastritis is unresolved, we employed a model system, theH. felis-infected mouse to investigate the temporalrelationship between bacterially-induced alterations in the hydrophobic phospholipid barrier of thestomach and the development of gastritis. In the presentstudy, C57BL/6 mice were inoculated with 109CFU of H. felis and the changes in gastric wet weight, histology, surface hydrophobicity,phospholipid/phosphatidylcholine concentration,phospholipase A2 activity, and the pH ofcollected gastric juice were measured 0.5-2 monthspostinoculation. In related experiments, we investigated the effects oftreating H. felis infected mice with antibiotic/bismuththerapy on the above gastric properties. It wasdetermined that both gastric surface hydrophobicity and phospholipid composition were significantlyattenuated as early as 2-4 weeks postinfection,preceding signs of mucosal inflammation and glandularatrophy as indicated by increases in gastric wet weight, pH and a disappearance in parietal cells. Theseearly H. felis-induced changes in gastric surfacehydrophobicity and phospholipid concentration werereversed by antibiotic/bismuth therapy. Based on these results we conclude that H. felis infectioninduces an early transformation of the stomach from ahydrophobic to an acid-sensitive hydrophilic state thatmay trigger the subsequent development ofgastritis.

Gastric protective activity of mixtures of saturated polar and neutral lipids in rats.

It has been shown that intragastric treatment of rats with a suspension of dipalmitoylphosphatidylcholine and tripalmitin at a 1:4 ratio (5 mg lipid/mL per rat) provided rats with highly efficaceous and consistent protection against a variety of ulcerogenic agents and conditions. The gastric protective activity of this mixture was of long duration (t 1/2 approximately 9 hours. In an attempt to understand the mechanism of protection, it was determined that the ulcerogen-induced reduction in gastric surface hydrophobicity was reversed in rats pretreated with the mixture. However, the lipid mixture did not affect the gastric emptying rate and maintained its cytoprotective activity in indomethacin-treated rats. These results indicate that the mixtures protective effect was not mediated by alterations in either gastrointestinal motility or the gastric accumulation of lipids or cytoprotective metabolites (prostaglandins). The mixture also appreciably reduced gastric lesion score in response to acid if one or both the lipids was substituted for a metabolically inert ether analogue, suggesting that lipid metabolism makes a negligible contribution to the protective response. Electron microscopic observation indicated that the predominent structure in the mixture is a microemulsion in which a dipalmitoylphosphatidylcholine monolayer encapsulates a tripalmitin core. Last, the improved gastric protective activity of the mixture in comparison to dipalmitoylphosphatidylcholine liposomes is discussed regarding marked differences in the physical structure of the two suspensions and the rate at which lipids in these states adsorb to a surface to enhance its hydrophobic properties.

Isolated rat gastric parietal cells: Cholinergic response and pharmacology

Abstract Isolated, partially purified or enriched rat gastric muscosal parietal cells were shown to respond to carbamycholine (EC 50 = 2 μ M) and other muscarinic cholinergic agonists as measured by an increased accumulation of 14 C-aminopyrine, an indirect measure of acid secretion. The secretory response to carbamylcholine was shown to be inhibited stereoselectively and reversibly by nanomolar concentrations of muscarinic cholinergic antagonists. Non-muscarinic antagonists, including cimetidine, were either ineffective or very weak inhibitors. The affinity constants calculated for cholinergic antagonist inhibition of 14 C-aminopyrine accumulation induced by carbamylcholine were similar to those previously calculated from direct binding studies on purified parietal cell particulate fractions using 3 H-QNB (1). These studies support the existence of specific parietal cell muscarinic cholinergic receptors with which the natural secretagogue acetylcholine interacts to regulate gastric acid secretion.

Milk protection against experimental ulcerogenesis in rats.

The antiulcer activity in pasteurized/homogenized bovine milk and a lipid extract of this milk was tested in an attempt to isolate and identify the active component. Using 0.6 N HCl as a damaging agent in pylorus-ligated rats, the protective property of milk appeared to be related to its phospholipid content and not its protein constituents. With intact (non-pylorus-ligated) rats, milk had demonstrable protective activity against 0.6 N HCl, as well as 100% ethanol and 160 mM taurocholic acid. The increasing phospholipid concentrations in skim, whole, and buttermilk paralleled their antiulcer efficacy. A lipid extract of whole milk showed significant protection against 0.6 N HCl when given alone or following indomethacin treatment. Measurements of the contact angle (hydrophobicity) of the gastric surface showed that it was maintained near control levels in the presence of 0.6 N HCl, if rats were first pretreated with milk. These results are consistent with the possibility that surface-active lipids in dairy milk, such as phospholipids, may account for a significant protion of milks antiulcer activity by maintaining the hydrophobicity of the luminal surface of the gastric mucosa in the presence of a damaging agent.

Monoamine oxidase: An important intracellular regulator of gastrin release in the rat*

The role of monoamine oxidase (MAO) in the meal-induced or amino acid-induced release of gastrin was investigated. Rats that were pretreated with the nonspecific MAO inhibitor nialamide (200 mg/kg) showed a greater rise in meal-induced serum gastrin than did untreated controls. In vitro experiments demonstrated that gastrin secretion from dispersed antral G cells in response to a stimulatory dose of phenylalanine or methylbenzylamine (10 mM) was markedly enhanced if the cells were treated with nialamide. Studies with the more specific MAO inhibitors clorgyline and deprenyl indicated that antral mucosa contained predominantly type A activity. Inhibition of MAO type A with clorgyline, both in vivo and in vitro, resulted in a greater release of gastrin after stimulation by a meal or phenylalanine. It is concluded that MAO may play an important role in the regulation of gastrin release from the G cell by partially controlling the level of amines within the cell.

Phosphatidylcholine-associated nonsteroidal anti-inflammatory drugs (NSAIDs) inhibit DNA synthesis and the growth of colon cancer cells in vitro

The use of NSAIDs or COX-2 inhibitors for chemoprevention of colorectal cancer has been suggested for patients at high risk for this disease. However, the gastrointestinal side effects of traditional NSAIDs which consist of bleeding and ulceration, and the cardiovascular effects of COX-2 inhibitors may limit their usefulness. In preclinical studies, our laboratory has shown that the addition of phosphatidylcholine (PC) to the NSAIDs aspirin (ASA) or ibuprofen (IBU) results in a NSAID–PC with fewer GI side effects and also maintained or enhanced analgesic, anti-pyretic and anti-inflammatory efficacy over the unmodified NSAID. Because NSAID–PCs have not been tested for anti-cancer activity, in the present study, ASA–PC and IBU–PC were tested on the SW-480 human colon cancer cell line. SW-480 cells were incubated in media containing 1–5xa0mM NSAID or NSAID–PC for 2xa0days. Measurements were made of cell number, cell proliferation (DNA synthesis), and manner of cell death (necrosis and apoptosis). ASA and IBU reduced cell number in a dose-dependent manner with IBU showing a greater potency than ASA. The association of PC to the NSAID resulted in greater reductions of cell number for both NSAIDs. Furthermore, the NSAID–PC formulation had significantly greater efficacy and potency to inhibit cellular DNA synthesis than the unmodified NSAID. PC alone at the doses and times used had no effect on cell number in this cell line, but did have a small effect to reduce DNA synthesis. None of the drugs had a clear effect on cell death by necrosis. Only IBU and IBU–PC caused cell death by apoptosis in SW-480 cells. We conclude that NSAID–PCs have activity to impede the growth of colon cancer cells in vitro, which is due, in major part, to a marked reduction in DNA synthetic activity of these cells. This growth inhibitory effect appears to be independent of COX-2 activity, since it is known that SW-480 cells do not have this inducible COX isoform. Due to its greater efficacy in this model system, IBU–PC should be further evaluated as a chemopreventive agent that is safer for the GI tract than unmodified NSAID.

Surface hydrophobicity of the gastric mucosa in the developing rat: Effects of corticosteroids, thyroxine, and prostaglandin E2

Surface hydrophobicity of the gastric mucosa and its variation in response to treatments with corticosteroids, thyroxine, and 16,16-dimethyl prostaglandin E2 were measured in developing rats. A developmental increase in the hydrophobicity of the luminal surface of the gastric mucosa was recorded between the first and third weeks of life. The hydrophobicity of the stomach was not consistently influenced by an acute administration of 16,16-dimethyl prostaglandin E2 (5 micrograms/kg, 30 min before examination) until the end of the third week of life, at which time a significant 40% increase was recorded. Similarly, the decrease in surface hydrophobicity that resulted from luminal administration of an ulcerogenic dose of HCl (0.6 N, 6 ml/kg) was blocked by 16,16-dimethyl prostaglandin E2 only in 3-wk-old rats and not in rats 1 and 2 wk of age. Neither the normal developmental increase nor the 16,16-dimethyl prostaglandin E2-induced enhancement in gastric surface hydrophobicity was induced precociously by corticosterone or thyroxine. The possible importance of these findings on the development of gastric surface hydrophobicity to the ontogeny of both gastric barrier function and prostaglandin-induced gastric protection is discussed.

Effects of milk, prostaglandin, and antacid on experimentally induced duodenitis in the rat: use of myeloperoxidase as an index of inflammation.

Ulcerogenesis of the duodenal mucosa frequently involves an inflammatory reaction with infiltration of leukocytes. Measurement of neutrophil myeloperoxidase activity might thus be a sensitive indicator of damage, before visible lesions occur. To test this possibility, a rat model for duodenal injury was used where fasted animals were treated with indomethacin and histamine-diHCl. Twenty-four hours after indomethacin treatment, duodenal tissues were collected for histochemical staining and biochemical assay for myeloperoxidase activity. Indomethacin- and histamine-challenged rats had significantly elevated myeloperoxidase activity compared to unchallenged controls (P < 0.05) for both histochemistry and biochemistry. There was also a significant correlation between these two parameters (r=0.68, P < 0.001). The duodenal injury model then was used to test the effectiveness of known gastric protective agents. Results indicated that milk and buttermilk did not aggravate or protect against duodenal injury, while antacid and prostaglandin did significantly protect against inflammation (P < 0.02). We concluded that measurement of myeloperoxidase activity is a sensitive and potentially useful estimate of duodenal injury that can be valuable in assessing ulcerogenesis and healing.