Hyperosmotic xylitol, prostaglandins and gastric mucosal barrier

Prostaglandins & Medicine 7:

63-70,

1981

HYPEROSMOTIC XYLITOL, PROSTAGLANDINS AND GASTRIC MUCOSAL BARRIER Georges Assouline and Abraham Danon, Clinical Pharmacology Unit, Faculty of Health Sciences, Ben-Gurion University of the Negev and Soroka Medical Center, Beer-Sheva, Israel (reprint requests to AD). ABSTRACT We have previously reported that hyperosmotic solutions of sodium chloride or of xylitol possess potent anti-ulcer activity and reduce gastric acidity in the rat. They also stimulate gastric prostaglandin (PG) biosynthesis, which may bear a causal relationship to the above effects. In the present investigation we studied the effect of intragastric hyperosmolarity on the transmucosal potential difference (PD) and on the permeability to H+ ions in the rat stomach. We also studied the effect of the prostaglandin synthetase inhibitors, indomethacin and flufenamic acid, on these parameters. Rat stomach was perfused in vivo, under urethane anesthesia, by xylitol solutions made up in 0.01 N HCl. While moderately hyperosmotic (13%) xylitol was without effect, the perfusion of intensely hyperosmotic xylitol (34.5%) resulted in a long lasting reduction of the transmucosal PD from a mean (?SEM) of -6324 mV to a trough value of -40+3 mV. This depression of transmucosal PD was inhibited in a dose-related fashion by prior treatment with the PGsynthetase inhibitors. Acid recovery in the effluent was significantly reduced by the 34.5% xylitol solution,and indomethacin pretreatment aid not modify the effect of hyperosmotic xylitol. It is concluded that, although intensely hyperosmotic xylitol produces some of the characteristic effects of a barrier breaker, i.e. depression of transmucosal PD and acid back diffusion, these two phenomena probably involve different mechanisms, as indicated by their differential response to indomethacin. INTRODUCTION Recently we reported that hyperosmotic solutions of sodium chloride or of xylitol or sorbitol are potent anti-ulcer agents in the rat. We also observed a concomitant decrease of gastric acidity following the instillation of hyperosmotic NaCl into the rat stomach (l-2). The latter finding could indicate antisecretory activity of the hyperosmotic solutions, as we have indeed observed (unpublished data). However, some degree of acid back diffusion with subsequent loss of acidity from the stomach may also be a contributing factor. Thus, Altamirano suggested

63

that very hyperosmotic,solutions of urea, glucose or sucrose caused increasedionic permeabilityofthe gastricmucosa (3). On the other hand, as the conceptof mucosal barrier breakage carries the implication of damage to the mucosa, it seems to be incompatiblewith the obvious anti-ulcereffects of the named hyperosmoticsolutions.The present investigationwas thereforeundertakento ascertainthe effects of intragastricinstillationof hyperosmoticsolutionsof xylitol on the mucosal barrier.Moreover,since we have observedthat such hyperosmotic solutionsaugment gastricprostaglandin(PG) biosynthesis(2,4) it was of interestto study whether cyclooxygenaseinhibitorscould alter any of the effect of hyperosmoticsolutionson the mucosal barrier. MATERIALSAND METHODS Gastric Lumen Perfusion. Female Charles-Riverrats weighing 180-230 g were starved for 24 hours, during which time they were housed on grid bottom cages. Free access to water was provided during the fasting period. Anesthesiawas effected with urethane, 1.5 g/kg intraperitoneally, while rectal temperaturewas maintainedat 37?l°C by means of a heating lamp. The gastric lumen was perfusedby a techniquesimilar to that describedby Ghosh and Schild (5). Briefly, a polyethylenecatheterwas introducedin the stomach through the esophagusand connectedto an infusion/withdrawal pump (HarvardApparatus,Millis, Mass.) at a flow rate of 0.093 ml/min. The perfusatewas collectedin 5 min fractionsthrough a pyloric catheter. Allowancewas made for a lag of some five min that elapsedbefore the fluid reached the outer tip of the pyloric cannula..Transmucosal potential difference(PD) was measuredby electrodesconsistingof polyethylene tubing (1.6 mm i.d.) filled with 4% w/v agar in saturatedpotassium chloride.The detectingelectrodewas introducedin the gastric lumen via the pyloric cannula,while the referenceelectrodelaid on the external side of the gastricwall inside the peritonealcavity. The free ends of each electrodewere introducedin a 30 ml glass container filled with saturatedKCl, in which one of a pair of balanced calomel cells (ModelEA404, Metrohm,Herisau, Switzerland)was placed. The transmucosalPD was continuouslymonitoredby means of a digital voltmeter (8000 A Digital Multimeter,Flucke, Seattle,Wa.) and recorded at 5 min intervals. ExperimentalProcedure. Each experimentconsistedof three periods: a basal period, during which 0.01 N HCl was perfused, lasted from 60 to 90 min, until both PD and the fractionvolume stabilizedfor at least 30 min. A 30 min test period followed,during which an isoosmotic(4.6%)or hyperosmoticsolutions (13% or 34.5%) of xylitol (Sigma,St. Louis, MO.) in acid (0.01 N HCl) were perfused.The osmolaritiesof these solutionswere determinedto be 327+19 (n=S),925+19 (n=8) and 2307?26 mOs/l (n=ll)respectively(measured with a Fiske OM osmometer.Uxbridge,Mass.). Finally, a 90 min recovery period was allowed,during whi~h~0.01N-HCl was again perfused. The acid contentof each fractionthat was collectedwas determinedby titrationwith 0.001 N NaOH using phenolphthalein as indicator.

64

In some experiments an identical procedure was applied in either indomethacin or flufenamic acid pretreated rats. Indomethacin (Sigma, St. at a concenLouis, MO.) was freshly dissolved in 2% sodium bicarbonate intraperitoneally at tration of 2, 4 and 8 mg/ml, pH68.0, and injected 5, 10 and 20 mg/kg respectively 45 min prior to surgery. Flufenamic acid (kindly donated by Rafa Labs. Ltd, Jerusalem, Israel) was freshly suspended in distilled water by means of a drop of tween 80 to make suspensubcutaneously at 25, 50 and sions of 10, 20 and 40 mg/ml, and injected 100 mg/kg respectively 45 min prior to surgery. Calculations. The values reported for PD (mV) and osmolarities (mOs/l) are means ? standard errors of the mean, for each point, after making the individual time scale coincide in the same treatment group, so that the peak osmolarities that had been measured occurred at the same time. Although acid recovery (uEq) was determined for each 5 min fraction, it was calculated for 30 min periods relative to the peak intragastric osmolarity and reported as loss of acid recovery relative to the 27.9 PEq that had been perfused. Comparisons were made using Student t test. RESULTS The osmolarity of the outflow increased gradually during perfusion of the xylitol solutions and peaked at the end of each 30 min test period. The maximal osmolarities in the outflow were 310 + 17 (n=5), 774 + 73 (n=8) and 1790 f 81 mOs/l (n=ll) after perfusion of 4.6, 13.0 and 34.5% xylitol, respectively. Effect

on PD.

Perfusion of isoosmotic (4.6%) as well as 13% xylitol solution did not significantly alter the base,line PD of -64 + 3 mV (n=5, fig. 1) and solution was -69 5 4 mV (n=4) respectively. However, when 34.5% xylitol perfused, the PD dropped sharply to a minimum of -40 ? 3 mV (n=7, p< 0.0005 compared with control rats at the same time. Fig. 1). The trough in PD coincided with the peak osmolarity in the stomach. After cessation of the hyperosmotic perfusion, the PD returned slowly towards the baseline, but recovery was not complete until the end of the experiment. In 4 out of 7 rats, the drop in PD was preceded by a transient increase up to a mean value of -8OmV. Pretreatment with indomethacin (20 mg/kg) resulted in a significantly lower basal PD : -57 * 2 mV (n=8), compared with -67 f 2 mV (n=16) in non-pretreated rats (p
65

peak

osmolarity

PO (-mV)

I

5%”

60

0

5

10

1

H+deficit

(pEq/30

min)

??

, n=7) or with (o---o, Fig. 1. Effect of 34.5% xylitol without (-, 8, n=4) indomethacin pretreatment, compared with 4.6% xylitol f$-x, n=5) on PD (upper panel) and acid deficit (lower panel). The hypere& osmotic xylitol was perfused for 30 min, and the time scale is relative to the peak intragastric osmolarity, as indicated. PD was determined at 5 min intervals, while acid deficit is the integrated value for 30 min periods. Each point is the mean + SEM for the number of experiments indicated. * ~~0.05 compared with corresponding control value (4.6% xylitol).

Effect on acid recovery. perfusion of isoosmotic (4.6%, fig. 1) as we11 as moderately hyperosmotic (13%) xylitol did not alter the acid recovery in the effluent. However intensely hyperosmotic (34.5%) xylitol brought about a 23% reduction in acid recovery of the effluent during the first half hour following the peak gastric osmolarity (fig. 1). Indomethacin pretreatment did not modify the reduction in acid recovery that was produced by 34.5% xylitol.

66

AV mV 40-

(3)

30-

20-

10-

ono pretreatment

Indomethacln 5

10 mU

FI ufenamic 20

kg

25

acid

50 mg/

100 kg

Fig. 2. Effects of pretreatment with various doses of PG-synthetase inhibitors on the drop of PD that was induced by 34.5% xylitol. AV: mean difference (+ SEM) between minimal PD after xylitol perfusion and basal level recorded for each animal. The number of experiments is indicated in parenthesis. * p < 0.02 ,**p < 0.005 compared with non-pretreated rats. DISCUSSION The results indicate that the intensely hyperosmotic xylitol (34.5%) caused marked reductions in both PD across the rat gastric mucosa and acid recovery, both parameters being indicative of presumed mucosal barrier breakage. On the other hand, the moderately hyperosmotic xylitol (13%) was without effect on the same. This is in agreement with data reported by others. Altamirano (3,6) showed that the output of Na+, K+ and Cl- to the luminal side of the dog gastric mucosa, in vitro, remained practically unchanged until the solute concentration on the luminal side was raised with various nonelectrolyte solutes up to 0.5 - 0.8 M. However, above this critical level, the ionic outputs increased with the hyperosmolarity of the solution. Sernka and colleagues (7,8) also showed that hyperosmotic solutions of various solutes, at osmolarities of 1000 mOs/l or higher, lowered the PD and at the same time increased the mucosal permeability to urea as well as to Na+ and Cl-. Such ionic fluxes across the mucosa may explain the decreased PD that was presently observed with intensely hyperosmotic xylitol. Hyperosmotic xylitol solutions of a similar intensity have been shown to increase PGE output from the rat stomach (2). Indomethacin and flufenamic acid inhibited the drop in PD that was produced by 34.5% xylitol in a

67

dose related fashion. These data seem to suggest that the effect of xylitol on PD could be ascribed to a PG related mechanism. However, positive implication of PGs as mediators of this phenomenon requires that the administration of exogenous PGs also produce a lowering of PD. Unfortunately, the relevant data available in the literature vary, depending on species, experimental design, nature, concentration and route of administration of the PG derivatives used. Under certain experimental conditions, PD was reported to decrease (9-11) following the administration of a PG, while in others it was augmented (12-14). Although the great variability in experimental design precludes definite conclusions, it appears, however, that the dose of PG administered is a crucial determinant of the outcome. Thus Miller and Tepperman (15) found that lower doses of 16-16 dimethyl PGE2 increased PD, while on the contrary, higher doses depressed it under the same experimental conditions. Results obtained in the present investigation and also reported by others indicate that indomethacin, flufenamic acid (results not shown) and aspirin (10, 13), all decrease the PD. Whether this effect is related to inhibition of PG synthetase or not, cannot be concluded. However, if we assume that the low physiological concentrations of endogenous PGs are such that they increase the resting PD, this action may be expected from the PG synthetase inhibitors, Intensely hyperosmotic xylitol (34.5%) induced a loss of acid recovery which was not affected by indomethacin pretreatment. Sernka and Jackson (7) reported that hyperosmotic solutions depressed the spontaneous acid secretion of rats which had been selected as vigorous spontaneous secretors. We have been able to confirm the same with rats that had a high rate of acid secretion, either spontaneously or after histamine infusion, but not with low secretors, to which neutral hyperosmotic xylitol was In the present study the stomachs were instilled (unpublished results). perfused with acid solutions, under which conditions the secretory rate was obviously negligible, thus back diffusion may represent a major component of the loss of acidity, The concurrence of decreased PD and reduced acid recovery is conventionally reffered to as “mucosal barrier breakage”. However the present data, showing differential effect of the PG synthetase inhibitor indomethacin on these two components of “barrier breakage”, seems to indicate that they may not necessarily be interrelated. Moreover, while a hyperosmotic solution of xylitol exerts some degree of barrier breakage, it also protects the gastric mucosa against ulcerogenic agents (2). Indeed, the generally admitted correlation between “barrier breakage” and mucosal damage has been questioned (16). Examples of such an apparent discordance have been recently described. Deregnaucourt and Code (17) reported that the first exposure of the rat gastric mucosa either to taurocholic acid or to ethanol at such concentrations which break the mucosal barrier reduced the breaking effects of second exposure to the same agent. Robert and associates (18, 19) observed that prior exposure of the rat stomach to 5 mMtaurocholate, 20% ethanol or 70°C water prevented the ulcerogenic effects of 80mM taurocholate, absolute alcohol or boiling water respectively given 15 minutes later, and this protective effect was inhibited by indomethacin pretreatment. These authors suggested that lower concentrations of damaging agents may function as “mild irritants” and induce the formation of endogenous PG, which in turn protect the gastric mucosa against further

68

damage. The same mechanism could apply to the gastric protective action of hyperosmotic solutions. While 25% NaCl severely damaged the rat gastric mucosa (19)) we reported that 10% NaCl protected it against and at the same time greatly increased various experimental ulcers, the PGE contents of the gastric mucosa and juice (2). If the ulcerogenic effect on the one hand and the increased formation of endogenous PG in response to exposure to such agressive agents on the other hand are dose related, there may be a critical point where the ulcerogenic effect surpasses the cytoprotective action of the PG that is released. Alternatively, if PGs exert a biphasic effect, as suggested the mucosal barrier at by some experimental results (15)) tightening lower doses, weakening it at higher doses and even damaging the gastric mucosa under extreme conditions (20), then excessive concentrations of “mild irritants” may induce the release of such amounts of endogenous PGs which may no longer be cytoprotective. More quantitative information is needed for a better understanding of this complex interrelationship. ACKNOWLEDGEMENT This work was supported in part by a grant Bureau, Ministry of Health, Israel.

from the Chief

Scientist’s

REFERENCES 1.

: antiulcer Assouline G, Danon A. Gastric prostaglandins of hypertonic solutions. Isr. J Med Sci 14: 493, 1978.

2.

Danon A, Assouline G. Antiulcer activity of hypertonic solutions in the rat : possible role of prostaglandins. Eur J Pharmacol 58: 425-431, 1979.

3.

Altamirano M. Action on the dog gastric

4.

Assouline G, Leibson V, Danon A. Stimulation of prostaglandin output from rat stomach by hypertonic solutions. Eur J Pharmacol 44: 271-273, 1977.

5.

Ghosh MN, Schild HO. Continuous recording in the rat. Brit J Pharmacol 13: 54-61,

6.

Altamirano M. Action on the dog gastric

7.

Sernka TJ, Jackson AF. Hyperosmotic Life Sci 17: 435-442, 1975.

8.

Rollin RE, Jacobson ED, Sernka TJ. Gastric mucosal transport and metabolism in hyperosmotic solutions. Nutr Rep Int 20: 787-797, 1979 *

9.

Bolton JP, Cohen MM. Permeability effects of E2 prostaglandins on canine gastric mucosa. Gastroenterology 70: 865, 1976.

10.

Dajani EZ, Callison on canine gastric 442, 1978.

activity

of concentrated solutions of non electrolytes mucosa. Am J. Physiol 216: 33-40, 1969.

of acid 1958.

gastric

secretion

of solutions of reduced osmotic concentration mucosa. Amer J Physiol 216: 25-32, 1969. instillation

of rat stomach.

DA, Bertermann RE. Effects of E prostaglandins potential difference. Amer J Dig Dis 23: 436-

69

11.

TeppermanBL, TeppermanFS, Fang WF, JacobsonED, Effects of 16, 16 - dimethylprostaglandinE2 on ion transportby isolated rabbit gastricmucosa and rat intestinalepithelialcells. Can J Physiol Pharmacol56: 834-839,1978.

12.

Whittle BJR. Mechanismsunderlyinggastricmucosal damage induced by indomethacinand bile salts and the action of prostaglandins. Brit J Pharmacol60: 455-460, 1977.

13.

ChaudhuryTK, JacobsonED. Prostaglandincytoprotectionof gastric mucosa. Gastroenterology74: 59-63, 1978.

14.

Colton DG, CallisonDA, Dajani EZ. Effects of a prostaglandinEl derivative,SC-29333and aspirinon gastric ionic fluxes and potentialdifferencein dogs. J PharmacolExp Ther 210: 283-288, 1979.

15.

Miller TA, TeppermanBL. Effect of prostaglandinE2 on aspirininduced gastricmucosal injury.J Surg Res 26: 10-17, 1979.

16.

Fromm D. Gastricmucosal "barrier".Gastroenterology77: 396-398, 1979.

17.

DeregnaucourtJ, Code CF. Increasedresistanceof the gastricmucosal barrier to barrier breakers in the rat. Gastroenterology 77: 309-312, 1979.

18.

ChaudhuryTK, Robert A. Preventionby mild irritantsof gastric necrosisproduced in rat by Na-taurocholate.Fed Proc 37: 303, 1978.

19.

Robert A, LancasterC, Hanchar AJ, Nezamis JE. Mild irritantsprevent gastricnecrosis throughprostaglandinformation.Histological study. Gastroenterology74: 1086, 1978.

20.

Kenyon GS, Ansell IF, Carter DC. Methylatedanaloguesof prostaglandin E2 and the gastricmucosal barrier. Prostaglandins15: 779-794,1978.

70