Experimental ulceration of rabbit antral mucosa

GASTROENTEROLOGY

1981;88:77-83

Experimental Ulceration of Rabbit Antral Mucosa E. KIVILAAKSO, and W. SILEN

A. BARZILAI,

Departments of Surgery, Massachusetts

Harvard

Medical

The mechanism of acute ulceration and the nature of mucosal protective mechanisms were investigated in rabbit antrum, which is more resistant to injury than fundus of the same species, with particular reference to the relationship between H’ back diflusion, tissue acidification, and occurrence of ulceration. Exposure of antral pouches to increasing JuminaJ [H+] during short-term hemorrhagic shock caused progressive acidification of the mucosa as measured by a microelectrode in the Jamina propria, but ulceration of the mucosa did not occur unless an unphysiologically high [H+] (225 mM) was used. Addition of exogenous pepsin to a JuminaJ [H+] of 80 mM increased the rate of H’ back diffusion and the degree of mucosal acidification, but ulcerations did not develop, suggesting that endogenous pepsin does not account for the ulceration of fundic mucosa under the same experimental situation. In contrast, addition of sodium taurocholate (5 mM) to the same Juminal [H’] (80 mM) produced ulcerations and increased the rate of H’ back diffusion. Systemic administration of acetazolamide, an inhibitor of carbonic anhydrase, enhanced the development of uJceration without increasing the rate of back diffusion or the degree of tissue acidification. The results suggest that rabbit antral mucosa is protected from excessive acidification and ulceration by its relative impermeability to H’. When the rate of H’

Received October 9, 19%. Accepted July 31, 1980. Address requests for reprints to: W. Silen, M.D., Department of Surgery, Beth Israel Hospital, 330 Brookline Avenue, Boston, Massachusetts 02215. Dr. Kivilaakso’s present address is: Second Department of Surgery, University Central Hospital, Helsinki, Finland. This work was supported by U.S. Public Health Service grants AM 15881 and AM 17917, and a grant from the S&rid Juselius Foundation, Helsinki, Finland. Dr. Fromm is the recipient of Career Development Award AM ooO53. 0 1981 by the American Gastroenterological Association 0016-5085/81/010077-07$02.50

R. SCHIESSEL, School

D. FROMM,

and Beth Israel Hospital,

Boston,

back diffusion is artificially augmented to a level encountered in the fundus in an ulcerogenic situation, profound acidification with ulceration also occurs in the antral mucosa. The data also suggest that carbonic anhydrase, an enzyme abundantly present in gastric mucosa, may have a protective function in the mucosa, possibly by contributing to the regulation of intracellular pH and/or to the maintenance of mucosal HCO,- secretion. It is well established that the antral portion of the stomach is less susceptible to acute ulceration than the fundus, but the mechanism underlying this difference remains obscure. In the rabbit, fundic mucosa readily ulcerates when exposed to physiologic concentrations of luminal acid during short-term hemorrhagic shock (1) or even under normotensive conditions (2). Previous studies have shown that ulceration under these circumstances is preceded by substantial acidification of the fundic mucosa from influxing luminal H’ (1,2). In the antrum, the degree of acidification is significantly less (1,~) suggesting that the factors contributing to the protection of antral mucosa might include lesser entry and/or better disposal of the influxing luminal H’. Whether the antral mucosa is, in addition, inherently more tolerant to the harmful effects of intramural H’ remains to be elucidated. The present study investigates in greater detail the role of back diffusion of luminal H’ in the pathogenesis of antral ulceration in the rabbit. The degree of acidification of the lamina propria of the antrum in various potentially ulcerogenic situations was assessed, and the requirements for ulceration were determined in the presence and absence of hemorrhagic shock. In addition, potective mechanisms of the antral mucosa were studied with special emphasis on the possible role of mucosal carbonic anhydrase.

76

KIVILAAKSO

GASTROENTEROLOGY

ET AL.

Material and Methods While under chloralose anesthesia (70 mg/kg), New Zealand white rabbits were prepared with innervated antral pouches. A gastrotomy was made slightly proximal to the fundoantral junction, and the stomach was emptied of its contents. The fundic portion of the stomach proximal to the gastrotomy was closed with a silk ligature, carefully preserving the vasculature, and the antral lumen was cleansed meticulously with warm saline. The anterior wall of the antrum was fixed with sutures to an intraluminal wire grid support measuring 1.5 cm in diameter. Both ends of the antral pouch were closed with a silk ligature. The shaft of the wire grid consisted of a metal tube which also served as the filling/emptying conduit of the pouch. In order to measure net fluxes of H+ during the test condition under study, the pouch was filled with 6 ml of the luminal test solution, which contained Y-labeled polyethylene glycol (PEG) with unlabeled PEG, 1 mg/ml, as a volume marker. After 15 min, the contents of the pouch were recovered and analyzed for titratable H’ or OH- by electrometric titration to pH 7.6 (Radiometer, Copenhagen, Denmark). The net fluxes of H’ were calculated as described before (1). During each collection period, the contents of the pouch were mixed twice by aspiration and reinfusion with a syringe. Usually, H’ fluxes for two successive 15-min normotensive periods were determined. Intramural pH of the mucosa was measured using pHsensitive glass (Microelectrodes Inc., Londonderry, N.H.) or antimony microelectrodes (Transidyne General, Ann Arbor, Mich.). The pouch was filled with the luminal test solution, and a small avascular area of the gastric wall (2 3 mm in diameter) fixed to the support was denuded of its seromuscular coat. The pH-microelectrode was advanced into the mucosa underlying the denuded area to a depth of 566 pm by using a micromanipulator. At this depth, the tip of the electrode is located in the midportion of the mucosa (2). The electrode was kept in position throughout the entire experiment, and the area was coated with a drop of sesame oil to prevent excessive evaporation and loss of CO, from the denuded area. A calomel reference electrode was connected to the serosal surface of the pouch via a saline-agar bridge. The millivoltage was recorded continuously by a high-input impedance pH/electrometer (Corning 112) connected to chart recorder. As both the pHelectrode and the reference electrode were located at the same (serosal) side of the epithelial lining, the millivoltage was not affected by transmural PD. The electrodes were calibrated before and after the measurements in nonphosphate buffers at 37%. Table

1.

Group Control

Acetazol

Arterial Blood Gases and Acid-Base Acetazolamide Period” 1 2 1 2

Balance

Arterial blood pH 7.39 7.17 7.33 7.16

a Period 1 = before shock: period 2 = after shock.

f f f f

0.01 0.03 0.02 0.04

Vol. 60, No. 1

In experiments with hemorrhagic shock, the baseline for intramural pH was first measured for 15-30 min. Whole blood, 16 ml/kg was then removed over 3-4 min from a cannula inserted in the femoral artery, producing hypotension to a mean pressure of 20-30 mmHg. Mean arterial pressure was monitored with a mercury manometer connected via a three-way stopcock to the same arterial line. After 30 min of shock, the shed blood, which had been collected in a heparinized syringe, was returned over 5-7 min into an ear vein. Samples of arterial blood were taken before and at the end of the shock period and analyzed for pH, Pco,, and PO, (Radiometer, Copenhagen, Denmark). After the experiment the test animal was killed, and the antral pouch was removed. The mucosa was inspected for the presence of macroscopic lesions and biopsied for microscopic study. Statistical analyses were performed using Student’s ttest for paired and unpaired variates.

Results Luminal

Acidity

The requisites for ulceration were first assessed in untreated antral pouches by determining the rate of diffusion of luminal H’ into the mucosa and the degree of acidification of the tissue (lamina propria) during exposure of the luminal surface to varying [H’]. The measurements of H’ fluxes were made during the normotensive period before shock. Previous studies have shown that the rate of luminal loss of H’ in antral pouches in the present model does not change significantly during hemorrhagic shock (1). In the presence of neutral luminal solution (150 mM NaCl), a minimal but constant secretion of alkali (0.90 + 0.1 pmol/lS min) was observed in all pouches (Figure 1). When acidic solutions were instilled into the pouches, a net loss of H’ from the luminal solution occurred, and the rate of this loss (H’ backflux) was linearly dependent upon the concentration of luminal H’ (y = -1.98 + 0.56x; r = 0.84). In pouches containing neutral luminal solution, only a slight and insignificant (p < 0.2) decrease in intramural pH from 7.22 + 0.02 to 7.15 f 0.06 (n = 5) occurred during hemorrhagic shock (Figure 2). (In this and all subsequent experiments, mean pH after the experimental manipulation refers to the mean

in Rabbits Before and After Shock With and Without PCOz (mmHg) 37.6 41.0 49.2 57.2

f 1.2 f 6.1 f 0.6 z!z2.6

[HCOI (mEq/L) 22.1 14.0 24.6 19.9

f f f f

1.2 1.6 0.2 1.6

POz (mm%) 93.2 106.2 92.6 101.6

f f f f

2.2 1.3 1.2 1.9

January

EXPERIMENTALULCERATIONOF RABBITMUCOSA

1981

I

2 SEM

N=5

-100

-

T .c

E

r %

the degree of acidification and the presence of lesions. In the three pouches without lesions, the average intramural pH was 6.58 (range 6.50-6.70), whereas in the seven pouches containing lesions it was significantly lower (p < 0.01, Wilcoxon’s test), 6.18 (range 6.00-6.40).

-

Pepsin

E”_ S50

-

Fl

S iL + I t; 2

79

I

II

oOmM WI

Figure 1.

I

60 mM HCI

: 150 mM I-ICI

225 mM HCI

LUMINAL

SOLUTION

: 60mM HCI + pepsin

60 mM HCI + TC

H+ fluxes in rabbit antral pouches exposed to different concentrations of luminal H+, sodium taurocholate, and pepsin. The measurements were performed in a normotensive state before measurement of intramural pH. In each experiment, the average flux for two successive 15-min periods was determined.

Net

pH at the end of the shock period.) In contrast to pouches containing neutral lumina! solution, acidic luminal solution caused more profound acidificatiqn of the mucosa (Figure 2). The degree of acidification was dependent upon the concentration of luminal H’ and hence on the rate of back flux of luminal H’. In pouches containing 80 mM HCl, the preshock rate of luminal H’ loss was 45.0 f 6.0 pmo1/15 min (n = 5). Intramural pH decreased from 7.27 & 0.02 to 6.89 f 0.07 (p < 0.01) during the subsequent hemorrhagic shock. In pouches containing 150 mM HCl, the rate of luminal H‘ loss was 84.2 +- 11.7 pmol/l5 min (n = 5), and intramural pH decreased from 7.12 + 0.03 to 6.57 f 0.02 (p < O.Ol), whereas in pouches containing 225 mM HCl, the rate of luminal H’ loss was 126.9 rt_ 15.1 pmol/l5 min (n = 5) and intramural pH decreased from 7.16 f 0.06 to 6.30 + 0.08 (n = 10, p < 0.01) during shock, The relationship between the rates of net-H+ fluxes in the normotensive preshock period and the intramural pH during the subsequent shock period in individual experiments is shown in Figure 3. A highly significant (p < 0.001) correlation (r = 0.92) between the rate of backflux of luminal H’ and the degree of acidification of the mucosa was observed. None of the pouches containing [H’] of 150 mM or less developed mucosal lesions during shock. Of the 10 pouches containing 225 mM, seven had lesions consisting of small erosions or ulcerations and/or areas of mucosal necrosis with preservation of the basic mucosal architecture. Furthermore, a significant correlation was observed in this group between

Addition of hog pepsin (obtained from Sigma Company LVith a potency of 5000 units/mg), 5000 U/ ml, a concentration which is equivalent to that in normal rabbit gastric juice, to the luminal solution containing 80 mM HCl increased the rate of luminal H’ loss from 42.3 + 9.6 to 61.4 Tf: 10.7 pmol/l5 min (p < 0.05, Figure 1). During shock, the intramural pH decreased from 7.18 f 0.04 to 6.56 + 0.07 (p < 0.01, Figure 4). This decrease approximates that observed in pouches containing a higher concentration of acid (150 mM) alone. In order to exclude the possibility that a buffering effect of pepsin interfered with titration, we performed experiments in which we found that, in fact, an overestimation in titration of 2.6% occurred in pepsin-containing solutions. However, none of the pouches containing pepsin had mucosal lesions. Taurocholate Addition of sodium taurocholate, 5 mM, to the luminal solution containing 80 mM HCl increased the rate of luminal H’ loss from 46.1 + 6.2 to 88.0 & 8.6 pmol/l5 min (p ( 0.01, Figure 1). During hemorrhagic shock, profound acidification

1.5 -

Hemorrhage +

o--o a-d H

6.0 0

Figure 2.

10

16 ml/kg w

0 mM HCI 60 mM HCI 150 mM HCI 225 mM HCI

20

30 40 TIME (minutes)

50

60

pH in antral pouches exposed to different concentrations of luminal H+ during short-term hemorrhagic shock.

Intramural

KIVILAAKSO

I30

GASTROENTEROLOGY

ET AL.

Vol. 80, No. 1

7.5 y = 7.15 + o.oos5x (p
r : 0.92 N= 20

E7.0

-

< 5 I d 5 -0.5

_ _

Figure 3. Relationship between the rate of net H+ flux before the shock and intramural pH during the subsequent hemorrhagic shock in rabbit antral pouches.

.

-

/A’

.i

I

I

-200

I

I

-50 -100 NET Ii+ FLUX (pmoles/lSmins.)

-150

of the mucosa occurred. In pouches containing 80 mM HCl with 5 mM sodium taurocholate, intramural pH decreased from 7.23k 0.07to 6.50f 0.06(p This decrease was of the same magnitude as < 0.01). the decrease in intramural pH observed in the nonulcerating pouches exposed to a higher concentration of HCl(l50) alone, but was less than that of the ulcerating pouches containing 225 mM HCl (Figure 4).All five pouches containing taurocholate developed mucosal lesions during shock. These lesions ranged from small areas of mucosal necrosis to extensive destruction of the mucosa.

Hemorrhage *

57.0

16 ml/kg 4

-

I

0

Acetazolamide The effect of systemic administration of acetazolamide (100 mg/kg) on the rate of backflux of H’ was studied in a series of experiments with normotensive animals. The luminal solution contained 150 mM HCl. The rate of luminal H’ loss was somewhat less in the animals receiving acetazolamide than in control animals (Figure 5), but the difference between the two groups was not statistically significant (p > 0.4). During hemorrhagic shock, intramural pH decreased from 7.17f 0.01to 6.54k 0.07when the pouches contained 150 mM HCl and 100 mg/kg of acetazolamide was given (Figure 6). This decrease closely approximates the decrease in intramural pH observed in animals not treated with acetazolamide under the same conditions. However, in contrast to untreated animals, all five rabbits receiving acetazolamide developed antral mucosal lesions. These lesions consisted of multiple small punctuate ulcerations measuring l-2 mm in diameter.

2 -100 [

5

_

&atazolamide

z $ ? z 6.5 -

1-,.-.i 12SEM N=5 e--e . ..... P--O

6.0 II 0 Figure

60 mM 60 mM 60 mM 150 mM

11 10

I 20

HCI HCI + pepsin HCI + TC HCI ’

1

’

1

30 40 TIME (minutes)

1 50

’

1 60

pH in antral pouches containing pepsin, 5000 U/ml, or sodium taurocholate (TC), 5 m&l, with luminal acid (80 mM HCl) of rabbits subjected to hemorrhagic shock. For comparison, values for pouches containing 80 or 150 mM HCl alone are also depicted.

4. Intramural

+ -

8

’

t

0

-

Acetazolamide Control I

0

I

15

I

30

I

45

I

1

60 75 TIME (minutes)

I

I

I

90

105

120

Figure 5. Effect of acetazolamide (100 mg/kg i.v.) on the rate of H+ fluxes in antral pouches containing 150 mM HCl. The measurements were performed in normotensive rabbits.

EXPERIMENTAL

A more dramatic response was observed when a larger dose, 200 mg/kg, of acetazolamide was used. Intramural pH decreased profoundly, even during the early phase of shock. It continued to decrease in a nearly linear fashion throughout the entire shock period (Figure 6). The average intramural pH of the five pouches at the end of the shock period was 5.35 + 0.16.In addition, whereas the intramural pH in other groups usually showed varying degrees of recovery during the retransfusion of the shed blood intramural pH continued to decrease even after the end of the shock in animals receiving the high dose of acetazolamide. This extreme acidification of the tissue may be secondary to breakdown of the mucosal lining, as all five pouches in this group showed practically complete destruction of the mucosa at the end of the experiment, both macroscopically and microscopically.

is an index of intracellular pH. Wheth r the increased tolerance against ulceration in antrum is due to the added availability of ATP in the antral mucosa, as suggested by Menguy and Masters (4),remains to be elucidated. We also considered the possibility that the absence of endogenous pepsin might account for the greater resistance of antrum to luminal H’ since pepsin is necessary for gross ulceration to develop in Addition of pepsin to the luminal solution vivo (5-7). significantly increased the rate of luminal H’ loss and the degree of mucosal acidification in the present studies, but ulceration of the mucosa did not occur, even though the concentration of pepsin (5000 U/ml) was considerably higher than the concentration of endogenous pepsin in fundic pouches in the present model (200-500 U/ml, unpublished data). The possibility that acidification of the mucosa caused by back diffusion of H’ might provoke autocatalytic activation of pepsinogen in the tissue itself (8) has not been excluded by the present experiments, however, if indeed pepsinogen is present in the antrum. Bile salts are known to increase the permeability of gastric mucosa to H’, and the ulcerogenic action of these compounds are generally attributed to this effect. The present experiments are in accord with

Discussion The present experiments demonstrate that rabbit antral mucosa is relatively resistant to ulceration. During short-term hemorrhagic shock, ulceration does not develop unless the mucosa is exposed to unphysiologically high concentrations of luminal H’. This is in sharp contrast to the fundic mucosa of the same species, which regularly ulcerates when exposed to only 80 mM of luminal H’ during identical experimental conditions (1). Previous investigations have shown that the permeability of rabbit fundic mucosa to H’ is 2-3 times greater than that of the antral mucosa (1,2). It seems plausible that the “tightness” of the antral mucosa contributes to its protection by impeding the access of luminal H’ to the tissue. The present studies support this view because we demonstrated that if the rate of diffusion of luminal H’ into the mucosa is artificially augmented to levels associated with ulcerogenesis in fundic pouches (1,2), ulceration also develops in antral mucosa. However, as the intramural pH associated with the development of ulcers during shock in the absence of luminal bile salts or systemic acetazolamide was -6.3 in the antrum and -6.6 in fundus, it seems possible that antral mucosa may also be inherently more resistant to the harmful effects of influxing H’. We recognize that the comparison of the intramural pH in the antrum and the fundus may not be ideal because the studies were not carried out simultaneously in the same animal, but the experiments were done in vivo under identical conditions. We believe that the intramural pH measurement reflects the pH of the interstitial tissue of the lamina propria. As we have not simultaneously measured intracellular pH, no assumptions can be made in regard to whether the intramural pH

ULCERATION

OF RABBIT MUCOSA

81

January 1981

Hemorrhage c

16 ml/kg e

I2SEM

I -

0.

< 6.55

-

2

L7

F

_ z 6.0 -

N=5

5.5 -

n--a

I-ICI. 150 mM

o--o

HCI, 150mM + Acetazolamide 100 mg/kg

-

+%I I-ICI, 150 mM + Acetazolamide 200 mg/kg

5.0 -

I

0

I

I 10

I

I 20

II

I

I

30 40 TIME (minutes)

11

11

50

60

Figure 6. Effect of acetazolamide on intramural pH in antral pouches of rabbits subjected to hemorrhagic shock. The pouches contained 150 mM HCl. For comparison, values for untreated animals are also depicted.

82

KIVILAAKSO ET AL.

this view as taurocholate did enhance acidification of the tissue as measured by the intramural microelectrode. However, the data suggest that taurocholate may also promote ulceration by increasing the sensitivity of the mucosa to the detrimental influences of the influxing luminal H’. Evidence for an increased sensitivity is that the rate of luminal H’ loss and the degree of mucosal acidification in pouches exposed to 80 mM HCl and taurocholate is roughly equal to those in pouches exposed to 150 mM HCl alone, but ulceration occurred only in the presence of taurocholate. There are several possible explanations for these observations. Bile salts are known to interfere with energy metabolism in the gastric mucosa by uncoupling oxidative phosphorylation (9) and by decreasing mucosal ATP content (10). Bile salts also may damage the integrity of intracellular organelles with the subsequent release of lysosomal enzymes, especially acid hydrolases, which could provoke autodigestion of tissue (11).Bile salts cause loss of functional integrity of cellular membrane-bound proteins such as Na+-K+-ATPase (12).If bile salts also affect other enzymes within the cell or cell membrane, especially those contributing to the maintenance of normal intracellular pH, the resultant disruption of H’ ion gradient across the tissue could further impair the capacity of the mucosa to withstand acidification. Inhibition of carbonic anhydrase has been shown to promote gastric ulceration in various experimental models (3,13,19, but the mechanism for the enhanced ulceration remains largely unknown. In vivo, the systemic effects of carbonic anhydrase inhibitors could play a role in ulcerogenesis, as transient respiratory acidosis (15)and retention of CO, in the tissues (16)are known to occur after administration of acetazolamide. Although the increased arterial Pcoa in our experiments indicates some retention of CO,, accentuation of systemic acidosis by acetazolamide was not observed during shock. This, together with studies in isolated frog gastric sacs showing that acetazolamide enhances ulcerogenesis in vitro (3) suggests that local influences of acetazolamide play a more important role in ulcerogenesis than the systemic effects, especially since the gastric mucosa is exceptionally rich in carbonic anhydrase (17,lB). The gastric mucosa is known to secrete small amounts of HCO,- (19-21), and it has been speculated that this alkaline secretion may serve as a protective barrier to impede the access of luminal H+ into the mucosa. The secretion of HCO,- seems to require the presence of carbonic anhydrase (22), and, consequently, the ulcerogenic action of acetazolamide might be mediated by inhibition of mucosal secretion of HCO,-. However, the finding that

GASTROENTEROLOGY

Vol. 80, No. 1

acetazolamide (100 mg/kg) promotes ulceration without increasing the rate of H’ back diffusion or the degree of mucosal acidification suggests that acetazolamide, like taurocholate, interferes in some as yet unexplained manner with the mechanisms that regulate the disposal of influxing H’. These mechanisms may include a carbonic anhydrase-requiring proton pump to extrude the influxing luminal H’ from epithelial cells, such as has been described in several other tissues (23-26). The possibility that acetazolamide may have an effect independent of its action on carbonic anhydrase has not been excluded in our experiments. The present studies reemphasize the role of luminal acid and its diffusion into the mucosa in the pathogenesis of acute gastric ulceration. In an ulcerogenic situation, luminal H’ accumulates excessively in the mucosa, and when a critical concentration is achieved, breakdown of the tissue occurs. In the antrum in the absence of bile salts or acetazolamide, the acidity of the tissue as reflected by the intramural pH bears a definite relationship to the development of ulceration, but is not the sole determining factor. For example, certain ulcerogenic agents, such as acetazolamide and taurocholate, can enhance the sensitivity of the mucosa to the influxing luminal H’ and thus render it more susceptible to ulceration.

References 1. Kivilaakso E, Fromm D, Silen W. Relationship between ulceration and intramural pH of gastric mucosa during hemorrhagic shock. Surgery 1978;84(1):70-7. 2. Kivilaakso E, Fromm E, Silen W. Effect of the acid secretory state on intramural pH of rabbit gastric mucosa. Gastroenterology 1978;75:641-8. 3. Kivilaakso E, Barzilai A, Schiessel R, et al. Ulceration of isolated amphibian gastric mucosa. Gastroenterology 1979;77:317. 4. Menguy R, Masters YF. Gastric mucosal energy metabolism and “stress ulceration.” Ann Surg 1974;180:538-46. 5. Rassers JFR. Uber die pathogenese des ulcus digestivum. Klin Wochenschr 1925;4:644-5. 6. Schiffrin MJ, Warren AA. Some factors concerned in the production of experimental ulceration at the GI tract of cats. Am J Dig Dis 1942;9:205-9. 7. Alphin RS, Vokac VA, Gregory RL, et al. Role of intragastric pressure, pH and pepsin in gastric ulceration in the rat. Gastroenterology 1977;73:495-569. 8. Anderson W. Activation of pepsinogen by sulphated glycosaminoglycans: a possible role in peptic ulcerogenesis. J Pharm Pharmacol 1969;21:264-6. 9. Menguy R, Masters YF. Mechanism of stress ulcer. Influence of sodium taurocholate on gastric mucosal energy metabolism during hemorrhagic shock and on mitochondrial respiration and ATPase in gastric mucosa. Am J Dig Dis 1976:21:1961-7. 10. Kuo Y-J, Shanbour LL. Inhibition of ion transport by bile salts

January 1981

in canine gastric mimosa. Am J Physiol 1976;231:1433-7. 11. Waldron-Edward D, Boutros MIR, Himal HS. Effect of bile on lysosmal stability in the mucosa of the canine gastric antrum. Gastroenterology 1977:73:986-4. 12. Little MPG, McCarthy CF, Mooney PA, et al. The relationship between gastritis and the effect of bile salts on Na, K-ATPase in human gastric mucosa. J Physiol 1977;273:61. 13. Davies R, Edelman J. The function of carbonic anhydrase in the stomach. Biochem 1951;50:190-4. 14. Werther JL, Hollander F, Altamirano M. Effect of acetazolamide on gastric mucosa in canine vivo-vitro preparations. Am J Physiol 1965;269:127-33. 15. Cain SM. Respiratory effects at carbonic anhydrase inhibition. US Air Force, Sch Aviat Med 1966;66-48:1-52. 16. Mithoefer JC, Davis JS. Inhibition of carbonic anhydrase: effect on tissue gas tensions in the rat. Proc Sot Exp Biol Med 1958;98:797-861. 17. Boass A, Wilson TH. Cellular localization of gastric intrinsic factor in the rat. Am J Physiol 1964;296:783-6. 18. O’Brien P, Rosen S, Trencis-Buck L, et al. Distribution of car-

EXPERIMENTAL

19. 20.

21. 22. 23. 24.

25. 26.

ULCERATION OF RABBIT MUCOSA

83

bonic anhydrase within the gastric mucosa. Gastroenterology 1977;72:870-4. Flemstrbm G, Sachs G. Ion transport by amphibian antrum in vitro. I. General characteristics. Am J Physiol1975;228:118-98. Fromm D, Schwartz JH, Robertson R, et al. Ion transport across isolated antral mucosa of the rabbit. Am J Physiol 1976;231:1783-9. Garner A, Flemstrom G. Gastric HCO,- secretion in the guinea pig. Am J Physiol 1978;234:E535-41. Flemstrbm G. Active alkalinization by amphibian gastric fundic mucosa in vitro. Am J Physiol 1977;233:El-12. Thomas RC. Ionic mechanism of the H+ pump in a snail neurone. Nature (Lond) 1976;262:54-5. Thomas RC. The effect of carbon dioxide on the intracellular pH and buffering power of snail neurones. J Physiol 1976; 255:715-35. Russell JM, Boron WF. Role of chloride transport in regulation of intracellular pH. Nature (Lond) 1976;264:78-4. Ellis D. Thomas RC. Direct measurement of intracellular pH of mammalian cardiac muscle. J Physiol 1976;262:755-71.