Omeprazole inhibits acetylsalicylic acid ~ modified histamine stimulation of acid secretion in rabbit gastric glands

OT BROSSEUK, IGM CLEA TOR, A) RAE, G W ANKLING. Omeprazole inhibits acetylsalicylic acid-modified histamine stimulation of acid secretion in rabbit gastric glands. Can J Gastroenterol 1994;8( 1 ): 15-20. The effects of misoprostol and omeprazole on basal-, histamineand acetylsalicylic acid (ASA)-induced gastric acid secretion by isolated rabbit gastric glands were studied. The authors found that ASA at a concentration of 2.4X10'3 M significantly inhibited acid secretion in the isolated gastric glands to 65% of basal levels, and that ASA at a concentration of 2.4Xl0'2 M significantly inhibited the histamine stimulation of acid secretion to 78% of maximal. Misoprostol inhibited acid secretion to 76% of basal acid secretion, while omeprazole inhibited secretion to 58% of basal values. Misoprostol inhibited the ASA-modified histamine stimulation to 82% of maximal stimulation. In contrast, omeprazole was able to inhibit the ASA-modified histamine stimulation to 48% of maximal. This omeprazole inhibition of secretagogue-induced acid production reduced acid secretion to levels below basal secretion, indicating that neither histamine nor ASA ( at the concentrations used), alone or in combination, had any stimulatory effect in the presence of omeprazole. Misoprostol is the recommended drug of choice for prevention and treatment of nonsteroidal anti-inflammatory drug (NSAID),induced gastrointestinal mucosal injury. In vitro results suggest that omeprazole appears to treat this condition more effectively if gastric acid secretion is a necessary prerequisite for NSAID-induced mucosa! injury.

N ONSTERO!DAL ANTl-!NFLAMMA- tory drug (NSA!D )-induced gastro• intesti nal mucosa!injury can range from mild gastritis to frank peptic ulceration.NSA!Ds in general, and acerylsalicylic acid (ASA) in particular, are extremely commonly ingested drugs used to treat conditions ranging from arthritis to vascular occlusive diseases, in add ition to their use as analgesics.T his widespread use of NSAlDs and the high associated incidence of NSAJD-induced mucosa!damage ( 1) means that this condition is frequently encountered by an yone deali ng with gastrointestinal disorders.Misoprostol, a prostaglandin Et analogue, is the present drug of choice for prevention of this NSA!D-induced damage.Misoprostol has been shown to exert its action on the G i-subunit of the adenylate cyclase coupled to histamine-H2 receptor in rabbit parietal cells (2).Omeprazole is a newer drug in the armamentarium of treatment modalities for gastrointestinal mucosa!injury; it exerts its effect at the level of the final common pathway for gastric acid secretion -the proton pump (3 ).Previous studies in vivo have indicated chat omeprazole may be protective against NSA!D-induced gastric mucosa!damage (4,5).We hypothesized that if ASA-induced gastric au niveau des glandes gastriques de lapins ont ete etuP,ies.Les auteurs ont decouvert que l'AAS, a une concentration de 2,4 x 10-M inhibait de fai;on significative la secretion acide au niveau des glandes gastriqu~ isolees pour donner65% des taux de base et qu'a une concentration de 2,4 x 10-M, il inhibait de fa~on significative la stimulation par histamine de la secretion acide pour donner 78% des valeurs maximales.Le misoprostol a inhibe la secretion acide pour donner 75% de la secretion acide de base, alors que l'omeprazole a inhibe la secretion pour donner 58% des valeurs de base.Le misoprostol a inhibe la stimulation par !'histamine modifiee par I' AAS pour donner 82 % de la stimulation ma.ximale.Par ailleurs, l'omeprazole a pu inhiber la stimulation par !'histamine modifiee par l'AAS pour donner 48% des valeurs maximales.Cet effet inhibiteur de l'omeprawle sur la production d'acide par le secretagogue a reduit la secretion d'acide pour donner des taux inferieurs a la secretion basale indiquant ainsi que ni !'histamine, ni l'AAS (aux concentrations utilisees) seuls ou en association, n'ont pu exercer un effet stimulant en presence de l'omeprazole.Le misoprostol est le medicament de choix, recommande pour la prevention et le traitement des lesions de la muqueuse digestive associees aux anti-inflammatoires non steroidiens (AINS).Selon les resultats in vitro, l'omeprazole semble traiter cette affection plus efficacement si la secretion d'acide gastrique est une composante de la lesion muqueuse induite par les AINS.mucosal injury requires an acidic gastric environment, omeprazole should be able to inhibit this injury through its blockade of the proton pump, and we should be able to show inhibition of acid secretion in the presence of ASA in vitro.We chose isolated rabbit gastric glands as our model to study this hypothesis.

MATERIALS AND METHODS
Gastric glands were obtained in a fashion similar to the method established by Berglindh and Obrink (6).The study protocol was approved by the University of British Columbia Animal Care Committee.New Zealand white rabbits were heparinized and anesthetized with ketamine and xylazine.The stomach was perfused via a catheter inserted through the abdominal aorta with clamping of the supradiaphragmatic aorta to isolate perfusion to the upper abdominal organs.The perfusate was a phosphate-buffered saline solution (149.6 mM sodium chloride, 3.0 mM c.lipotassium hydrogen phosphate, 0.64 mM sodium biphosphate monohydrate, pH 7.4) which caused edema of the gastric wall to enhance separation of the gastric mucosa.The antrum was then excised from the stomach and gastric mucosa was harvested through blunt dissection.After mechanical mincing, mucosa was digested in an oxygenated collagenase solution ( LJO.O mM sodium chloride, 12.0 mM sodium bicarbonate, 3.0 mM sodium biphosphate monohyc.lrate,3.0 mM sodium phosphate dibasic, J .O mM dipotassium hydrogen phosphate, 2.0 mM magnesium sulphate, 1.0 mM calcium chloride, 5.0 mg/L phenol red, 1.0 mg/10 mL collagenase [Sigma type 1, Missouri].1.5 mg/10 mL albumin, 1.5 mg/10 mL glucose) for 60 mins.During this incubation and subsequent incubations, water bath temperatures remained at J7°C and a pH of 7.4 was maintained.Following collagenase digestion, the collagenase-enzyme solution and gastric glands were filtered through 1 70 µM pores to remove large debris.Glands were rinsed in respiratory medium (132.4 mM sodium chloride, 5.4 mM potassium chloride, 5.0 mM sodium phosphate dibasic, 1.0 mM sodium biphosphate monohydrate, 1.2 mM magnesium sulphate, 1.0 mM calcium chloride, 5 mg/L phenol red, 2.0 mg/10 mL albumin, 2.0 mg/10 mL glucose).At this stage, glands were permitted to settle, supernatant was poured off and glands were resuspended in JO mL of respiratory medium containing 0.11 mL (0.2 MBq) of C-14 aminopyrine (Amersham) in solution.
A second incubation of these glands in the combined aminopyrine/respiratory medium took place over 30 mins.During the two incubations, a 96-well tray was prepared for the final step in the experiment.Each well was filled with 170 µL of respiratory medium and a 15 µ L solution of each secretagogue or inhibitor, or combination of these substances, assigned to the particular well.In this series of experiments, l.Oxl0--4 M histamine (Sigma), l.8xl0-6 M misoprostol (Searle Canada), l.9x10--4 M omeprazole (Astra Phanna) and 2.4x10--4 M ASA were used.
In a secondary experiment, ASA in concentrations of 2.4x10-3 M and 2.4X] 0-2 M were usec.1 to analyze the effect of increasing ASA dosing.Upon completion of a second incubation, 100 µL of cell solution was adc.led to each well, according to Adrian's micromethod (7).A third and final incubation involving placing the incubation chamber housing the microtitre plate into a water bath was performed for 60 mins.Flow rate of oxygen through the chamber was 3 L/min.Following incubation, the microtitre plate was placed on a vacuum filtration unit for J to 4 mins to remove the supernatant.Any wells that did not adequately filter through due to an excess of cellular debris were discarded and not included in any subsequent analysis.Filters in each well were dried and then punched into scintillation vials.Ten millilitres of Ecolite( +) (ICN Biochemicals) scintillant was added to each vial and beta counting was performed for 5 mins (Beckman LS7500).
In the second set of experiments to analyze the effects of increasing concentrations of ASA, a separate batch of scintillant was used and beta counting was performed on a separate beta counter.Uptake of aminopyrine into the parietal cell estimates acid secretion (6) and, in tum, indicates thesecretory effect of secretagogues and acidinhibiting drugs upon these cells.

STATISTICS
Differences between drug effects were determined by calculating weighted comparisons between the mean response across animals.The average value for each animal was estimated as the mean of all wells from that subject (typically seven to nine wells).These values were weighted to the between well variation for each animal.The weighted estimates were used to compute 95 to 99% confidence intervals about the mean response.Tighter confidence intervals were obtained to correct for multiple comparisons.Significant differences were assumed when confidence intervals did not overlap (P<0.05).

RESULTS
The wells belonging to the control group produced a mean count of 1206.6±32.1 decays per minute (dpm).These decay values were counted for 5 mins and then averaged.Figure 1 shows that addition of histamine significantly increased the beta count to 1873.4±60.4dpm, representing a 55% increase in acid secretion above baseline.Addition of ASA alone had an inhibitory effect on the wells that was not significant at a concentration of 2.4xl0-4 M, but that, in a subsequent experiment to analyze the effect of increasing concentrations of ASA, was a significant inhibitor at a concentration of 2.4xlo-3 M (Figure 2).It is of note that this second set of experiments used a separate batch of scintillant and a separate beta counter which resulted in across-the-board higher beta counts in the second set of experiments compared with the first (cg, control counts of2391.8versus 1206.6).This does not, however, affect the comparisons being made within each set of experiments.
When ASA and histamine were added to the cells, there was an inhibition of the histamine stimulation with beta counts of 1465.1±47.5 dpm; this represents a decreased acid secretion of 12% below stimulation of histamine alone (Figure 1) which is not statistically significant, but which, once again, in subsequent experiments was found to be significant inhibition with a concentration of 2.4x10-2 M (Figure 2).Misoprostol added to the cells produced a significant inhibitory effect with beta counts of913.9±14.1 dpm, a 24% decrease in acid secretion below baseline.Omeprazole significantly in-hibited acid secretion by 42% below basal levels, producing beta counts of 696.3± 16.9 dpm (Figure l ).The inhibition by omeprazole is a statistically significant better inhibition than that shown by misoprostol.When misoprostol was added to the wells containing both ASA and histamine, it produced beta counts of 1204.2±30.55dpm, a significant 18% reduction in acid secretion below the maximal levels.Omeprazole was able to inhibit the maximal secretion by 52%, with beta counts of 702.2±20.4dpm (Figure 1 ).Once again the increase in inhibition by omeprazole compared with misoprostol attained statistical significance.This inhibition by omeprazole on wells containing both A SA and histamine produced virtually the same result as when omeprazole was added to cells being stimulated by histamine alone, which resulted in beta counts of 731.6±24.5 dpm (Figure 3 ).Thus, the greatest inhibition of unmodified, and of particular interest to our study - Omeprazole and ASA-Induced mucosal damage ASA-modified acid secretion, was exhibited by omeprazole at the concentrations that we tested.Regardless of additives, omeprazole consistently inhibited acid secretion to about 700 dpm readings.

DISCUSSION
Misoprostol has been shown in a double-blind, placebo controlled trial to prevent NSAID-induced gastric ulcer (8).Misoprostol is also noted co be effective in the prevention of NSAID-induced duodenal ulceration (9).Although H2recepcor antagonists have been shown to have some role in the prevention of NSA ID-induced duodenal ulceration ( 10), no other agent has been proven to be as effective as misoprostol in the prevention of NSAID-induced gastric and duodenal ulceration.It has been suggested that the pathogenesis of gastroduodenal lesions by ASA and other NSA!Ds involves the back diffusion of hydrogen ions (11); if this is correct, any agent that maximally inhibits the production of hydrogen ions should be a very useful agent for the prevention and treatment of these lesions.
Omeprazole, a substituted benzimidazole, is a potent inhibitor of both histaminic and cholinergic stimulation of acid secretion by gastric parietal cells (3).Romano an<l colleagues (12) have demonstrated that omeprazole may also have a protective effect on gastric cells independent of inhibition of gastric acid secretion.Konturek ( 13) showed that this cytoprotective effect of omeprazole extends to gastric lesions induced specifically by ASA.It has not been previously demonstrated that omeprazole can inhibit acid secretion in the presence of ASA, in add ition to any direct cytoprotection it affords against the ASA effect.
The literature contains contradictory reports on the effect that ASA has on gastric acid secretion, with d ifferent investigators suggesting there is no effect, stimulation and inhibition of acid secretion by ASA.
Levine et al ( 14) have suggested that ASA does not affect basal acid secretion, but may poten tiate the secretagogue stimulation of acid secretion by gastric parietal cells.In our experiment, using a refined model of isolated rabbit gastric glands, we have confirmed the finding that a low concentration of ASA does not alter baseline acid secretion; in addition, we fo und that ASA at higher concentrations actually inhibits basal acid secretion.These results are in agreement with the findings of Shea-Donahue and co-workers ( 15) who showed in vivo that ASA inhibits H + secretion in rhesus monkeys.These investigators speculated that this inhibition in v ivo may be due to a combination of the back diffusion of H+ into the cells fo llowing ASA-induced mucosa!injury and a direct inhibitory action of ASA on the parietal cells.
Levine and colleagues (14) demonstrated ASA potentiation of histaminestimulated acid secretion in isolated parietal cells.Under the same conditions, they were not able to demonstrate potentiation in isolated gastric glands except after a 'prolonged' incubation of 30 to 45 mins.
In our experiments using isolated gastric glands with incubation times of 60 mins, we did not reproduce the potentiation demonstrated by Levine and, in fact, we showed inhibition with increasing ASA concentration.O ne possible explanation for this inconsistency in results between Levine's isolated parietal cells and our isolated gastric glands is that perhaps the integrity of the gastric gland as a functional unit with its chief cells, mucous cells and endocrine cells, together with the parietal cells, is necessary to show the inhi bitory effect of ASA which we have demonstrated.This, however, does not reconcile the inconsistency between the inhibition that we found with glands incubated for 60 mins compared with the potentiation which Levine found with glands incubated for 30 to 45 mins.Levine's hypothesis t hat ASA potentiatcs secretion of H+ through a mechanism of mobilization of intracellular Ca 2 + may be correct, but the level of intraluminal H+ depends on both the rate of H+ secretion and on the rate of back d iffusion of the ions into the cells from the lumen.The rate at which ASA facilitates back diffusion of the ion into CAN J 0ASTR0ENTEROL VOL 8 No I JANUARY/FEBRUARY 1994 the gastric cells may be sufficient to more than compensate fo r Levine's hypothesized increased secretion of the ion, resulting in a net decrease in the intraluminal H+ in response to the ASA.
If the theory of back d iffusion of H+ into the gastric mucosa!cells is correct, it suggests that mucosa!damage by ASA is mediated not so much by changes in intraluminal pH, bur perhaps more so by changes in the intracellular pH seco ndary to chis ionic back d iffus ion.The mucosa! injury still depends on the presence of some intraluminal acid which must provide the source of H+ for back diffusion.

CONCLUSIONS
We have demonstrated that high concentrations of ASA inhibit basal and stimulated levels of acid secretion by isolated gastric glands.We have also demonstrated that omeprazole is able co inhibit completely the secrecagogue effect of histamine on the gastric glands, and that the presence of ASA does not interfere with t he ability of omeprazole to inhibit the histamine stimulation of acid secretion.Misoprostol was able to inhibit the histamine stimulation of acid secretion both with and without the presence of ASA to a much lesser degree than the omeprazole.l n fact, we have shown that omeprazole inhibits both stimulated and unstimulated levels of acid secretion to a constant level of secretion which is not altered significantly by any histamine or ASA effect.
These results in vitro indicate that omeprazole may be a very effective agent for the prevention and treatment of ASA-induced gastrointestinal mucosa!injury if an acidic gastric environment is a prerequisite to this injury as the source of the injurious back d iffusing H+ ions.