Effects of endogenous and exogenous prostaglandins on glycoprotein synthesis and secretion in isolated rabbit gastric mucosa

GASTROENTEROLOGY

1988;95:945-51

Effects of Endogenous and Exogenous Prostaglandins on Glycoprotein Synthesis and Secretion in Isolated Rabbit Gastric Mucosa U. SEIDLER, Abteilungen Hochschule,

K. KNAFLA,

Allgemeine Hannover,

R. KOWNATZKI,

Pharmakologie und Medizinische Federal Republic of German\,

We studied gastric glycoprotein synthesis and secretion in organ culture before and during cyclooxygenase inhibition and replacement with exogenous prostaglandins (16,16-dimethyl prostaglandin E2 and prostaglandin Fz,). Isolated rabbit antral and fundic mucosal explants incorporated [14C]iVacetylglucosamine and [3H]leucine in a linear fashion and steadily secreted labeled proteins and glycoproteins during the 24-h incubation period. On sepharose 4B, >90% of the secreted protein-bound [‘*C]N-acetylglucosamine was found in the high molecular weight peak. Incorporation of tracer was not influenced by cyclooxygenase inhibition with indomethacin or the addition of exogenous prostaglandins. Secretion of newly formed glycoprotein, however, was significantly inhibited by indomethacin and stimulated by both tested prostaglandins in a concentration-dependent manner. 16,16Dimethyl prostaglandin E2 caused significant stimulation in concentrations that are well in the physiologic range for endogenous prostaglandin E2, whereas prostaglandin Faa stimulated in 100 times higher concentrations. We conclude that in the isolated gastric mucosa both endogenous and exogenous prostaglandins stimulate mucus secretion. For prostaglandin E2, but not prostaglandin Fzu, a role in the physiologic regulation of gastric mucus secretion is probable.

T

he gastric mucus coat is thought to be important in both preventing damage to (1,~) and facilitating repair of (3) the gastric epithelium. Therefore, it has been hypothesized that one of the mucosa-protecting mechanisms of prostaglandins (PGs) could be their stimulatory effect on gastric mucus output. When applied intragastrically in micromolar concentrations, some PGs stimulate glycoprotein output into the gastric juice and enhance the thickness of

and K-Fr. SEWING Messgeraete,

Medizinische

the mucus gel adherent to the gastric wall (4-10). Cyclooxygenase inhibitors seem to reduce glycoprotein concentrations in the gastric juice (11,‘lZ). Whether PG concentrations in the physiologic (nanomolar) range directly modulate gastric mucus synthesis or secretion has never been studied in detail. To accurately determine the influence of PGs on mucus biosynthesis and secretion independently of their many effects on other components of the secretory process, the availability of an in vitro model would be desirable in which (a) known drug concentrations can be applied to the mucus cells, (b) endogenous PG synthesis can be blocked, and, most importantly, (c) secretion can be assessed independent of synthesis. MacDermott et al. (13) described an organ culture technique that enabled them to determine rabbit colonic glycoprotein synthesis and secretion rates independently. Gastric mucosal biopsy specimens have also been successfully cultured (14-16). We therefore developed an organ culture technique for rabbit gastric mucosa that was suitable for the measurement of gastric mucus formation. In this model we investigated (a) the influence of cyclooxygenase inhibition and (b) the influence of PGFz,, and 16,16dimethyl PGE2 on gastric mucus synthesis and secretion. Materials

and Methods

Materials Media

and

fetal

Gibco (Grand Island,

calf

serum

were

obtained

N.Y.); PGF,,,, penicillin,

from

streptomy-

Abbreviations used in this paper: PTA, phosphotungstic acid; TCA, trichloroacetic acid. c
946

GASTROENTEROLOGY Vol. 95, No. 4

SEIDLER ET AL.

tin, leucine, N-acetylglucosamine, arachidonic acid, dithiothreitol, and isobutylmethylxanthine from Sigma (St. Louis, MO.); trichloroacetic acid (TCA) and phosphotungstic acid (PTA) from Merck (Darmstadt, F.R.G.); HEPES from Serva (Heidelberg, F.R.G.); and Sepharose 4B from Pharmacia (Piscataway, N.J.). Indomethacin-sodiumtrihydrate was kindly donated by Merck, Sharp and Dohme and 16,16-dimethyl PGEz by Upjohn (Kalamazoo, Mich.). [3H]leucine (sp act, 120-190 Ciimmol) and [14C]Nacetylglucosamine (sp act, 50-60 Ci/mmol) were purchased from Amersham (Buckinghamshire, U.K.).

Culture

Technique

Male rabbits weighing 2-3 kg were killed by blunt head trauma, and the stomach was excised immediately, washed, and placed in ice-cold oxygenated culture medium consisting of 80% Trowell’s T8,10% NCTC-135, 10% heat-inactivated fetal calf serum, 12 mmol/L HEPES, 100 III/ml penicillin G, and 100 pg/ml streptomycin. The mucosa was dissected from the submucosa and small pieces (l-2-mm diameter for 24 h-incubations, 3-mm diameter for pulse-chase studies) were placed on glassfiber filters into organ culture dishes. Cultures were incubated in a specially designed water-jacketed pot at 36”C, in a humid atmosphere under 95% O-5% COZ, on a slowly rotating platform (Heidolph Instruments).

Protein Secretion

and

Glycoprotein Synthesis and 24-h Organ Culture

During

Protein and glycoprotein synthesis rates were determined by measuring the incorporation rates of [3H]leucine and [‘4C]N-acetylglucosamine, respectively, into [3H]TCA- and [‘4C]TCA-PTA-precipitable intracellular and secreted proteins, and secretion rates were assessed as the percentage of total protein-bound tracer that had been secreted into the culture medium. In detail, we added 2 &i/ml of [3H]leucine or [‘4C]N-acetylglucosamine to the culture medium, which was changed every 8 h. Cultured explants were homogenized in 0.15 mol/L NaCl, 10 mmol/L ethylenediaminetetraacetic acid, 5 mmoln N-acetylglucosamine, and 5 mmol/L leucine. The homogenate and collected culture medium were precipitated with ice-cold 10% TCA and 1% PTA, and left at 4°C overnight. To determine the amount of tracer bound to fetal calf serum, media were incubated without explants in each experiment and for each incubation time and processed as described. The samples were then centrifuged at 15,000 rpm and the pellets were resuspended by sonication and washed three times with ice-cold TCA-PTA and twice with ether/ethanol (3:1 volivol). The pellets were dissolved in 0.6 mol/L KOH and counted in liquid scintillation fluid after an aliquot was taken for protein determination (17). Counting efficiency was determined by the external standard-channel ratio method. Results were expressed as disintegrations per minute per milligram of explant protein after subtraction of nonspecific binding of tracer to fetal calf serum. The PGs (lO-y-lO~” mol/L), diluted from alcoholic stock solutions immediately before use, indomethacin, or vehicle were added to the culture medium at the beginning of the incubation period and

then with every medium change. The alcohol concentration in both controls and PG-treated samples was 0.5%. In the PG experiments, explants were preincubated in tracerfree medium with indomethacin (1O-5 mol/L) for 30 min to inhibit endogenous PG production. The medium was then changed and the drugs were added to the organ culture medium together with the tracer. Indomethacin (lo-” mol/L) was present in the medium throughout the experiment.

Pulse-Chase

Experiments

To obtain precise concentration-response relationships for PG-stimulated glycoprotein secretion, pulsechase studies were performed. Explants were pulselabeled for 1 h with 5 &i/ml of [14C]N-acetylglucosamine, rinsed several times, and cultured over isotope-free medium containing 5 mmol/L of cold N-acetylglucosamine and 10 PmoliL of indomethacin. Spontaneous secretion of protein-incorporated tracer over the next 15 h was determined. Drug influence on glycoprotein secretion was determined by performing three consecutive chase periods with fresh medium, each lasting 3 h. At the end of each chase period the explants were gently rinsed and the medium was collected, shaken for 30 min with 10 mmol/L of dithiothreitol to bring visible mucus strands into solution, and centrifuged at 15,000 rpm to remove cellular debris. Debris consisted of exfoliated cells and some insoluble mucus strands, carried 5%-12% of the proteinbound radioactivity in the medium, and was discarded. The medium was then precipitated, washed as described above, and counted. Alternatively, it was subjected to gel chromatography on sepharose 4B and the column fractions were precipitated, washed, and counted. There was good agreement between the protein-incorporated radioactivity in the high molecular weight peak obtained by column fractionation and the total protein-incorporated radioactivity in the chase media after removal of cell debris. Drugs were added to the medium of the third chase period. The ratio of protein-incorporated tracer in the medium of the third chase period (stimulated secretion) and the second chase period (basal secretion) was determined for each set of explants. Results were expressed as percentage of the value observed for the controls to which only the vehicle was added.

Characterization Radioactivity

of Secreted

Protein-Bound

The media of the third chase period from five pulse-chase experiments from either control or PC-treated antral and fundic explants were homogenized, centrifuged, and subjected to gel chromatography on Sepharose 4B. An aliquot of each fraction was precipitated with TCA-PTA and washed, and the precipitate was counted. To further characterize the radiolabeled high molecular weight peak, fractions were concentrated by ultrafiltration (in a IO-ml Amicon stirred cell, under 0.7-bar nitrogen, equipped with a XM-100 membrane) and the concentrate was separated by density gradient centrifugation (continuous self-generating CsCl gradient; starting density, 1.48 g/ml) at 4”C, 150,000 g, for 72 h. The sugar composition of

October

GASTRIC

1988

MUCUS

SECRETION

IN VITRO

947

the labeled glycoproteins was determined by capillar gas chromatography after methanolysis and trimethylsilylation according to the procedure of Preuss and Thier (18), which is a modification of the procedure described by Clamp (19) for the analysis of carbohydrates.

Determination

of Prostaglandin

Production

Antral and fundic explants were cultured as described, but without tracer in the medium. All explants were preincubated for 30 min in either indomethacin (10 and 50 PmoliL) or control medium and then placed in fresh medium. lndomethacin (10 and 50 PmoliL) was present throughout the incubation except in the controls. An aliquot of the culture medium was assayed after 2 and 8 h for PGE, concentration by radioimmunoassay as described previously (20).

1

1

I

I

I

I

J

0

4

a

12

16

20

24

incubationttme~h)

Histology Explants were fixed in 1.25% glutaraldehyde and then for 2 h in 0s04, dehydrated, and embedded in an araldite-Epon 812 mixture. One-micrometer sections were cut and stained with methylene blue-azure II-basic fuchsin [Humphrey’s stain).

Statistical

Analysis

Values are given as mean k SEM; n is always the number of experiments with different rabbits. In each pulse-chase experiment four sets of explants were used for each drug concentration and a set of control explants was processed with each set of test explants. The Student’s t-test for paired samples was used for statistical evaluation if not otherwise indicated.

Results Viability

of the Explants

The incorporation rates of [3H]leucine and [ 14C]N-acetylglucosamine into antral and fundic explants and secreted proteins were determined (data not shown). After an initial lag period of about 2 h, labeled proteins and glycoproteins were secreted steadily into the medium during the 24-h incubation period. The morphology of the explants, judged by light microscopy, was always very well maintained in antral explants and somewhat more variable in fundic explants, where sometimes there was mild interstitial edema with dilation of glands and some vacuolization and clumping of nuclei. The mucus-producing surface cells, however, were well preserved. Efiect of Cyclooxygenase Inhibition on Prostaglandin and Glycoprotein Synthesis and Secretion average

The antral and fundic explants synthesized on 2.3 t 0.3 and 1.0 k 0.08 ng of PGE, per

I

I

I

1

I

I

0

1

4

8

12

16

20

24

mcubation time ( h )

Figure

1. A. [‘4C]N-acetylglucosamine incorporation rates during indomethacin treatment into antral [solid symbols) and fundic (open symbols) intracellular plus secreted glycoproteins [controls (circles), 10 PmoliL of indomethacin (triangles), 50 PmoliL of indomethacin (squares]. n = 5; SEM = 5’#1-11U%, not shown]. B. I’%] N-acetylglucosamine incorporation rates during PC; treatment into antral (solid symbols) and fundic (open symbols) intracellular plus secreted glycoproteins [COIItrols (circles], 10 FmoliL of PGF,,, (triangles). 10 FmoliL of PGE, (squares). n = 5: SEM = 2”/o-7°h,, not shownl.

milligram protein per 2 h, respectively, during the first 2 h of incubation, after which PGE, levels in the culture medium remained stable. After preincubation with indomethacin for 30 min, PGE2 synthesis (as an estimate for cyclooxygenase activity) in the indomethacin-treated explants was suppressed by 93% 2 6% (10 FmoliL) and 97% * 4% (50 PmollL). Indomethacin had no influence on tracer incorporation rates in concentrations of 10 and 50 PmoliL (Figure 1A) but at 1 mmol/L caused a very marked suppression of both [“Hlleucine and 1’4C]N-acetylglucosamine incorporation. The latter could not be influenced by the addition of 1 PmoliL of 16,16dimethyl PGE, (data not shown) and was interpreted as a toxic effect of indomethacin unrelated to its ability to suppress cyclooxygenase activity.

948

SEIDLER

ET AL.

GASTROENTEROLOGY

Vol. 95, No. 4

glycoproteins found in the culture medium was higher in PG-treated explants (Figure 2B). When isobutylmethylxanthine (100 pmol/L) was added together with 16,16-dimethyl PGE2, [‘4C]N-acetylglucosamine incorporation was not stimulated but secretion of labeled glycoprotein was significantly enhanced compared with both controls (see p values in Figure 3) and samples treated with 16,16dimethyl PGE, alone (p < 0.01 for fundic and 0.05 for antral explants) (Figure 3). Figure

2. A. Secreted glycoproteins during indomethacin treatment of the explants in percentage of newly synthesized glycoproteins at 16 h of incubation (4-24 h were measured). *p < 0.05,***p < 0.01. in comparison with the control. indo 10 and 50: 10 and 50 FmoliL of indomethacin is present in the incubation medium (n = 7); indo 50 + 16,16-dm. PGE,: 50 Fmol/L of indomethacin and 2 x lo-’ M 16.16-dimethyl PGE, are present in the incubation medium (n = 4). B. Secreted glycoproteins during PG treatment in percentage of newly synthesized at 16 h of incubation. n = 7. -p < 0.05, **p < 0.01, compared with the control. 16.16-dm. PGE2: 10 FmoliL of 16,16-dimethyl PGE, is present in the incubation medium: PGF,,,: 10 PmoliL of PGF,, is present.

We found a reduction of the release formed glycoprotein with 10 and 50 pmol/L methacin [Figure 2A), more pronounced than in fundic mucosa. This reduction prevented by 16,16-dimethyl PGE, (Figure Efiect of 16,16-Dimethyl Prostaglandin and Prostaglandin FLol on Glycoprotein Synthesis and Secretion

of newly of indoin antral could be 2A).

3

Glycoprotein synthesis (left) and secretion (right) rates in antral and fundic explants during 24-h organ culture. Open bars, controls; shaded bars, 1 FmoliL of 16,16dimethyl PGE?; black bars, 1 FmoliL of 16,16-dimethyl PGE2 and 100 +mol/L of isobutylmethylxanthine (IBMX). n = 6, ***p < 0.001, **p < 0.01, *p < 0.05. compared with the control.

Experiments

Figure 4 shows the secretion rates of labeled glycoproteins after pulse labeling in unstimulated fundic and antral explants. Antral explants not only produce more labeled glycoprotein (see Figure l), but the passage time of precursor [from uptake to secretion of the labeled product) is also shorter in antral than in fundic explants. The second and third chase periods for antral explants were therefore started earlier than those for fundic explants. As seen in Figures 5A and 5B, both PGs stimulated glycoprotein secretion in a concentration-dependent manner. Significant stimulation by 16,16-dimethyl PGEz occurred at 25 nmol/L and by PGFzn at ~500 nmol/L in fundic explants and 210 nmol/L and 21 PmoliL, respectively, in antral explants. Characterization Glycoproteins

EZ

Neither in fundic nor in antral explants could any statistically significant difference of tracer incorporation rates be detected between control and PGtreated explants at any tested PG concentration and incubation time (shown for 10 PmoliL of PGs in Figure 1B). However, the percentage of labeled

Figure

Pulse-Chase

of Secreted

Labeled

On sepharose 4B, most of the secreted proteinbound tracer was found in a single peak that coeluted with blue dextran and therefore has a molecular weight of -2 million daltons. Very little incorporated radioactivity was found in the low molecular weight peaks, although the major portion of secreted proteins eluted in these fractions. Although only a rather small percentage of all secreted proteins have a high molecular weight in the range of

Figure

4. Spontaneous secretion rates of protein-bound [“C]Nacetylglucosamine during a pulse-chase experiment. The arrows indicate the second and third Lhase periods for evaluation of drug effects. Output of protein-bound tracer of each explant was assayed for every 2-h period.

CASTKIC

October 1988

B?'9

Figure

8

7

, 6 5 l-logmo, II

BY

9

c a

I,, 7 5 i-l&lolm

5. Glyc.oprotein secretion. measured in pulse-chase experiments, in response to 16,l&dimethyl PGk& (A] and P(;F,,, (m). Left punel, antral mucosa; right panel, frlndic mucosa. n = 5

the molecular weight for gastric mucus, this percentage is highly labeled (data not shown). Figure 6 shows the elution profile of protein-incorporated tracer (disintegrations per minute per milligram of explant protein) secreted by the explants, indicating that the secreted high molecular weight proteins carry almost all of the label. Figure 6 also shows that, in the pulse-chase experiments, only the secretion of high molecular weight labeled glycoproteins that eluted in the \roid volume was stimulated by the PCs. The sugar composition of this high molecular weight peak for antral secretion was as follows: fucose 10% + 3%,,mannose 3% t 0.4% galactose 32% ? So/,, galactosamine 35% + 5%, glucosamine IS%, + 4%, and neuraminic acid 4% -+ 0.3%; the composition was similar for fundic secretion. We then further fractionated this peak eluate by equilibrium density gradient centrifugation and usually obtained two to three labeled fractions. Most of the label was in a fraction with a density of 1.52-1.58 g/ml, whose sugar content suggested it to be predominan tly gastric mucus glycoprotein (2 l-23), because the amount of fucose, galactose, and galactosamine was high and that of mannose and neuraminic acid low. A variable amount (10%30%) of tracer, however, was incorporated into material with a higher content of aminosugars, mannose, and neuraminic acid than described for gastric mucoprotein and therefore thought to be other glycoproteins attached to the high molecular weight mucin. This material was variably distributed in the low-density top fraction of the tube and a high-density fraction (1.64-1.7 g/ml).

hibition

In the present study both cyclooxygenase inand the addition of exogenous PGs failed to

SECRETION

949

IN VITRO

influence precursor incorporation into intracellular explants and secreted glycoproteins in isolated antral and fundic explants. Even when isobutylmethylxanthine was added to prevent the reported rapid fall in PG-stimulated cyclic adenosine monophosphate increase in gastric mucosal cells (241, incorporation was not stimulated. Cyclooxygenase inhibition consistently reduced gastric glycoprotein output, an effect that was prevented by exogenous PGs. After suppression of endogenous PG production, both tested PGs stimulated glycoprotein secretion in a concentration-dependent manner. The EC,,, values for 16,16-dimethyl PGEz are well in the physiologic concentration range of its endogenous analogue PGE, and similar to the ICsI, values for 16,16dimethyl PGE, and PGF,-mediated inhibition of acid secretion (25-28). Therefore, a role for endogenous PGEz in the regulation of gastric glycoprotein secretion is probable. The potency of PGF2,, was -100 times lower. and a role for this PG in the regulation of mucus secretion seems unlikely because the high concentrations needed will never be available in the stomach. Separation and characterization of secreted glycoproteins revealed that only the secretion of a high molecular weight glycoprotein was stimulated. This glycoprotein appeared to be a mixture of mucus glycoprotein and other glycoproteins with a higher content of aminosugars, mannose, and neuraminic

,200

antra,

?.eCretlOn

lb

b

2b

i0 fraction

600

l;igure

Discussion

MUCUS

fUrldlC

number

secret,on

6. Chromatography of glycoproteins secreted from antral [top panel) and fundic (bottom panel) explants during the third chase period [controls (solid fine): 1 WmoliL of PGF’,,, (dashed line): 1 FmoliL of 16,16-dimethyl PGE, (dotted line)].

950

SEIDLER

ET AL.

acid, and both constituents of the glycoprotein mixture were labeled. Prostaglandins or arachidonic acid have been found to stimulate glycoprotein output in the duodenum (30) and in tracheobronchial (31,32) and gallbladder explants (33). Stimulation by arachidonic acid was blocked by indomethacin, a finding that was taken as evidence for the involvement of endogenously synthesized PGs. Interestingly, the goblet cell mucus secretion of colonic explants was not stimulated by PGs (34). We suggest that PGs act primarily as a secretory stimulus on the gastric mucus cell and have no direct effect on glycoprotein synthesis. The increase in mucosal thickness and increased mucus content of surface cells when PGs are applied for weeks (35) could be secondary to their effects on cell proliferation and could even be mediated by hypergastrinemia after prolonged PG application. The question remains as to why we did not see enhancement of synthesis after the cells had rapidly released a part of their stored mucus granules. The same lack of feedback enhancement of synthesis was found in acetyl(I 3) choline-stimulated colonic mucus secretion during a 24-h incubation, whereas sustained stimulation of pepsinogen secretion resulted in enhanced synthesis (36). Mucus synthesis is, of course, a much slower process than pepsinogen synthesis, and it is possible that within the 24 h of culture, feedback enhancement of synthesis has not yet occurred. We also do not know if, in fact, sustained stimulation of mucus secretion occurs in the presence of the agonist. McQueen et al. (7) showed that, after stimulation of mucus secretion by acetylcholine or PG, no further stimulation could be obtained with a second application of the agonist unless a period of several hours elapsed between the first and second application. Terano et al. (24) observed an initial rise in cyclic adenosine monophosphate levels after PG stimulation of cultured gastric epithelial cells, with a return to near-baseline levels within 1 h in the presence of the PG. It has been described that prolonged exposure to PGs causes both an apparent decrease in the number of PG receptors on the cell surface (down-regulation] and a partial loss of postreceptor function (desensitization) in vitro and in vivo (37,383). This desensitized state is maintained for some time after PC removal. The same process may occur in our organ culture, and this could offer an explanation why PG-stimulated mucus secretion was more pronounced in the pulse-chase studies than in the 24-h cultures. By the time the first labeled glycoproteins appeared in the medium and secretion could be monitored, the PGs had been present in the medium for several hours and the mucus cells could be in a state less responsive to PG stimulation. In

GASTROENTEROLOGY

Vol. 95, No. 4

summary, extremely little is known about the regulation of mucus synthesis, and long-term cultures of gastric mucus cells will be helpful in further clarifying the role of stimulants in mucus synthesis. Our data contrast with the findings by Tao and Wilson (39), who found an up to threefold increased incorporation of radiolabeled precursor into mucosal strips within 3 h when incubated with PGE2, 16,16-dimethyl PGE2, or U46619, a stable thromboxane analogue. Terano et al. (241, however, found in a gastric epithelial cell culture that 1 pmol of 16,16dimethyl PGE, barely stimulated synthesis by i’s, whereas 0.1 and 10 kmol/L did not stimulate synthesis during a 24-h incubation period. Scheiman et al. (40) found no stimulation of precursor incorporation in a canine mucus cell culture by PGE,. In the vascularly perfused rat stomach, Jentjens et al. (35) found a fourfold increase in precursor uptake into mucus cells and an increased radioactivity in secreted mucus within 1 h of application of 16,16dimethyl PGE, and within 30 min of perfusion with the tracer. We and others (41-43) have found the first luminal appearance of tracer truly incorporated into mucus glycoprotein as late as 2-4 h after application of the precursor. The findings of Jentjens et al. might therefore reflect the influence of the PG on vascular and cell permeability and precursor uptake processes rather than mucus synthesis. Unless more is known about the intracellular events following stimulation of the mucus cell by PGs and about the regulation of mucus synthesis in general, it will be difficult to resolve these discrepancies.

References 1. Heatley 2.

3.

4.

5.

6.

7.

8.

NG. Mucosubstances as a barrier to diffusion. Gastroenteroiogy 1959;37:313-18. Allen A, Garner A. Gastric mucus and bicarbonate secretion and their possible role in mucosal protection. Cut 1986;21: 249-62. Wallace JL. Whittle BJR. Role of mucus in the repair of gastric epithelial damage in the rat. Gastroenterology 1986;91:60311. Bickel M, Kauffman GL Jr. Gastric gel mucus thickness: effect of distension, 16,16-dimethyl prostaglandin E,, and carbenoxotone. Gastroenterology 1981;80:770-5. Bolton JP, Palmer D, Cohen MM. Stimulation of mucus and nonparietal secretion by the E, prostaglandins. Am J Dig Dis 1978:23:359-64. Domschke W. Domschke S, Hornig D, Demling L. Prostaglandin-stimulated gastric mucus secretion in man. Acta Hepatogastroenterol 1978;25:292-4. McQreen S. Allen A, Garner A. Measurement of gastric and duodenal mucus gel thickness. In: Allen A, et al., ed., Mechanisms of mucosal protection in the upper gastrointestinal tract. New York: Raven, 1984. Takeuchi K, Magee D, Critchlow J, Matthews J. Silen W. Studies of the pH gradient and thickness of frog gastric mucus gel. Gastroenterology 1983;84:331-40.

October 1988

Y. LaMont 17‘. Ventola AS. Maul1 EA. Szaho S. Cysteamine and prostaglandin Fzij stimulate rat gastric mucin release. (iastroenterology 1983;84:306-11. 10. Johansson C. Kollherg B. Stimulation by intragastrally aclmiliistered E, prostaglandins of human gastric mucus output. Eur J Clin Invest 1979:9:229-32. 11. Menguy K, Masters YY. Effrwts of aspirin on gastric mucus output. Surg Gynecol Ohstet 1965;1200:92-8. I?. El-Bayar H, Steel 1.. Montcalm E, Danquechin-Dor\~al E, IIubois A, Shea-Donohue T. The role of endogenous prostaglandins in the regulation of gastric secretion in rhesus monkeys. Prostaglandins 1~85;30:401-17. 13. MacDermott KP, Donaldson KM, Trier JS. Glycoprotein syllthesis alld secretion by mucosal biopsies of rabbit colon and humarl rectum. 1Clin Invest 1974;54:545-54. 14. Sutton I1K, Donaldson KM Ir. Synthesis and secretion of protein and pepsinogen by rabbit gastric: mucnsa in organ c.ulturct. Gastroenterology 1975:69:166-74. 15. Kapadia CK, Donaldson KM Jr. Macromolecular secretion by isolated gastric: mur:osa. Gastroenterology 1978:74:535-39. 16. Yromans ND, Miller SJ, Synthesis and secretion of glycoproteins by mucosa of rat gastric antrum in organ culture. IXg Dis Sci 398(3;25:295-303. 17. Lobvr!. OH, Kosehrough NJ. Parr LA. Protein measurement with thus lpolin phenol reagent. J Biol Chem 1954;1Y3:265-75. 18. l’reuss A. Thier H-P. Quantitative Analyse natuerlicher Dickungsmittel durch Methanolyse und (:apillaraaschroln;Itgraphic. % Lelxmsm IJnters Borsch 1982;175:93-100. 19. Clamp JK. A gas-liquid c.hromatographic approach to the analysis of cartmhgdrates. Biochem Sot Trans lY77;5:1693--5. 20. Hell M. Sewing K-Pr. Effec:t of omeprazole on eicosanoid formation in and release from guinea-pig gastric mucosal cells. Br 1Pharmac.01 1987;91:69-75. 21. Pouad FM. Waldron-Edward D. Isolation and characterization ot human and Lanine gastric mucosal glycoproteins and their degradation by proteases and hydrolases. Hoppr-Seylers Z Physiol Chem lY80:361:703-13, 22. Pearson IP. Allen A. Parry S. A 70,000 mol. weight protein isolated from purified pig gastric mucus glycoprotein 11) reduction of disulphide bridges and its implication in the polymer structure. Biochem 1 1981:197:155-62. 23. Spee-Brand K. Strous CJA, Kramer MF. Isolation and purificatitru 11t rat gastric. mucous ~lycoprotein. Biochim Biophys A[.ta lYH0;621:104-16. 24. ‘I’erann A, Ivey KJ. Stachura J. et al. Cell culture of rat gastric fundic. mucosa. (;astroenterology 1982;83:1280-91. 2.5. Sol1 AH, Specitic. inhibition by prostaglandins EL and I1 of histarrilnc-stimulatt~(l “C-aminopyrine accumulation and cyc.lic ,~dt:nosint~ monophosphate generation by isolated canine paric:tal cells. ) Clin Invest 1980;65:1222-9. 26. Sol1 .\kl. Whittle BJR. Interac.tion between prostaglandins and cvclic. AMP in the gastric. mucosa. Prostaglandins 1980; Pl(Suppl):39-45. 27. Levine KA, (Cohen KR. Schwartzel EH Jr, Kamsay CE. Prostaglandin E,-histamine interactions on CAMP. cGMP and acid production in isolated fundic glands. Am J Physiol 1982; 242:(:2 l-6, 28. Seidler li, Beinhorn M. Sewing K-Fr. The inhibitory effect of prostaxlandins on acid secretion is mediated via the PGE,receptor (abstr). Gastroenterology 1987:92:1631. 29. Konturek Sj, Piastucki 1, Brozowski T, et al. Kale of prostaglandins in the formation of aspirin-induced gastric: ulcers. (;,lstrotinterology 1981;80:4-9. 30. kosmala M, Carter SK, Konturek SJ, Slomiany A, Slomiany RI,. Mur:us glpcoprotein secretion by duodenal mucosa in

(;ASTKI(:

response 31.

32.

33.

34.

35.

36. 37.

38.

3Y.

40.

41.

42.

4’i. I

to luminal

MllCIIS

arachidonir:

SE(:KF:‘I’ION IL\: VITKO

xitl.

Hiochim

Biophys

951

Acta

1986;884:41’1-28. Marom %, Shelhamer IH. Kaliner hi. Effects of arachidonic. acid, molioh~dro,uyt:icosatetraonnil. ac:icl and prostaglandins on the release of mu(:ous glycol)rotc:ins from hum,rn airways in vitro. J Clin Invest 1981:67:1695-702 Barsigian C. Barbieri t
Received August 27. 1987. Acwpted May 11. 1988. Address requests for reprints to: Dr. Ilrsula Seidler. Il. Medizinische Klinik der Technischen llniversitat Miinchen. lsmaninger Strasse 22. D-8000 Milnchen, F.K.G. This work was supported by BMFI’ grant No. 03 8507 5. This paper was presented in part at the 88th Annual Meeting of the American Gastroenterological Assoc.i