Poly-N-acetyllactosamine-specific tomato lectin interacts with gastric parietal cells. Identification of a tomato-lectin binding 60–90 X 10(3) Mr membrane glycoprotein of tubulovesicles

The cytoplasmic tubulovesicular and canalicular membranes of gastric parietal cells are intimately involved in hydrochloric acid secretion. To characterise the glycoproteins of these membranes, we examined a panel of lectins for reactivity with parietal cells in paraffin sections of rat, dog and pig stomach. The poly-N-acetyllactosamine-specific lectin from Lycopersicon esculentum (tomato) and from Solanum tuberosum (potato), and the galactose-specific lectin Ricinus communis agglutinin (RCA120), showed strong cytoplasmic binding of parietal cells of all three species, with a pattern indicative of an intracellular membrane network. Binding to parietal cells was confirmed by double-labelling studies with parietal cell auto-antibodies from patients with autoimmune gastritis. Mucous cells and mucin also bound these lectins strongly. Other gastric cell types did not stain with either tomato or potato lectin, but stained weakly with RCA120. Electron-microscopic examination of lectin binding sites using biotinylated tomato lectin or RCA120 and streptavidin-gold, revealed specific binding to the luminal face of parietal cell tubulovesicular and canalicular membranes as well as the contents of mucous cell secretory granules. Tomato lectin and RCA120 reacted by lectin blotting with a major species of apparent molecular weight 60–90 X 10(3) Mr from rat, dog and pig gastric membranes. A tubulovesicular membrane fraction, enriched 10-fold for K(+)-dependent phosphatase activity, was also enriched three-fold for tomato lectin binding as assessed by a solid-phase lectin assay. The 60–90K (K = 10(3) Mr) component, in 125I-labelled detergent extracts of dog tubulovesicular membranes, bound to an affinity support of tomato lectin-Sepharose and was specifically eluted with N,N',N'-triacetylchitotriose. Digestion with N-glycanase collapsed the 60–90K component into a sharp 35K band. We conclude that: (1) a 60–90K membrane glycoprotein localised on the luminal face of tubulovesicles and canaliculi of parietal cells interacts strongly with tomato lectin and RCA120; and (2) the glycoprotein is composed of a 35K core protein glycosylated with N-glycans probably containing poly-N-acetyllactosamine sequences with terminal galactosyl residues. The properties of this 60–90K glycoprotein are identical to a major parietal cell autoantigen recognised by sera of patients with autoimmune gastritis.(ABSTRACT TRUNCATED AT 400 WORDS)

Cell digestion and genotoxicity assessment in gastric mucosa

It has been claimed that the in vivo incubation of rat gastric mucosa with a test drug, followed by labelling with tritiated thymidine and selective in vitro digestion is a suitable method for evaluating genotoxic potential. However, the method rests on the assumptions that the radiolabel released on digestion has been incorporated into damaged DNA which has undergone repair (i.e. unscheduled DNA synthesis, UDS) and that the results are not confounded by the presence of dividing cells (scheduled DNA synthesis). This study examined the method by repeating the digestion procedure on rat gastric mucosa which had been labelled with bromodeoxyuridine in vivo, to mimic the 3H-TdR experiments previously reported. Results showed that the cell digest contained a mixture of cells, including parietal cells, and that there was a 6.9 +/- 0.69% (n = 43) contamination with dividing cells over a 2-hour period of labelling. This represents a 10-fold enrichment of dividing cells as compared to intact tissue. It is concluded that measurements of total radioactivity in DNA made using this digestion method cannot be related solely to UDS, but are related to a variety of unrecognized artifacts.

Interaction of calcium channel antagonists with parietal cell acid production, adenylate cyclase, intracellular-free Ca2+ and H+/K+-ATPase

The calcium channel antagonists verapamil, gallopamil and nifedipine were tested for their effects on acid secretion stimulated by histamine and dibutyryl-cyclic AMP in isolated and enriched guinea pig parietal cells, on adenylate cyclase activity mediated by histamine H2 receptors, on histamine-stimulated increase in cytosolic-free Ca2+ concentration [Ca2+], on gastric H+/K(+)-ATPase activity and on H+/K(+)-ATPase-mediated proton uptake in intact gastric membrane vesicles. Verapamil and gallopamil impaired all cellular and enzymatic test systems studied. Both drugs affected with highest potency the acid secretion in the parietal cell preparation (IC50: 1-2 mumol/l) and the H+/K(+)-ATPase-mediated H+ uptake in gastric membrane vesicles, whereas their inhibitory action was less pronounced on adenylate cyclase and on histamine-induced increase in cytosolic-free [Ca2+]. The type of interaction found in the gastric membrane vesicle preparation indicates that both drugs act as protonophores. Nifedipine was less effective as an inhibitor of acid secretion in the parietal cell preparation and in reducing proton concentration in isolated gastric membrane vesicles. The drug failed to block adenylate cyclase and H+/K+-ATPase activity. Since nifedipine is a more effective calcium channel blocking agent but a less lipophilic drug than verapamil and gallopamil, we conclude that the antisecretory activity of calcium channel antagonists in vitro is mediated by a nonspecific, i.e. a protonophoric, action. We suggest that verapamil exhibits its antisecretory activity in vivo partially by its protonophoric action at the secretory membrane of the parietal cell, whereas the decrease in acid secretion by nifedipine is not mediated by this mechanism.

Regulation of CA II and H+,K(+)-ATPase gene expression in canine gastric parietal cells

Acid secretagogue-specific receptor activation in parietal cells triggers rapid and coordinate gene expression of gastric H+,K(+)-ATPase and of CA II. The rapid rise in steady-state levels of CA II mRNA is due to new transcription of the CA II gene in stimulated cells. Although the presumed function of CA II in activated parietal cells is to catalyze the generation of HCO3- from OH-, regulation of CA II gene expression appears to be independent of the generation of H+ (and OH-) through the action of H+,K(+)-ATPase.

Melittin inhibition of the gastric (H+ + K+) ATPase and photoaffinity labeling with [125I]azidosalicylyl melittin

Melittin is a 26-amino acid amphipathic polypeptide toxin from bee venom which forms anion-selective ion channels in bilayers and biological membranes under the influence of membrane potential. Melittin has been shown to interact with a number of membrane proteins. We found that melittin inhibited K+-stimulated ATP hydrolysis by the (H+ + K+) ATPase in parietal cell apical membrane vesicles derived from histamine-stimulated rabbit gastric mucosa with a KIapp of 0.5 micron. Melittin also inhibited K+-stimulated p-nitrophenyl hydrolysis activity which is associated with the gastric (H+ + K+) ATPase in a dose-dependent manner with a KIapp of 0.95 micron. ATP-driven, K+-dependent H+ transport was inhibited over this same concentration range, even in the absence of a membrane potential. Melittin did not appear to increase the H+ leak from vesicle with preformed H+ gradients when the H+ pump was arrested by Mg2+ chelation, but all possible membrane perturbation effects were difficult to rule out. However, the data suggest that melittin exerts its inhibitory effect through interaction with the (H+ + K+) ATPase. In order to determine whether direct interactions between the (H+ + K+) ATPase and melittin occurred, a radioactive derivative of melittin, [125I]azidosalicylyl melittin, was prepared and photoreacted with sealed rabbit gastric membranes and highly purified hog gastric membrane containing the (H+ + K+) ATPase. In the purified hog preparation only a 95,000-Da band, the (H+ + K+) ATPase was labeled, while in the rabbit preparation a 95,000-Da band and one other membrane protein of 70,000 Da were labeled with this reagent. Label incorporation into the (H+ + K+) ATPase and the 70,000-Da band was greatly reduced by addition of excess unlabeled melittin, suggesting specificity of the interaction. Label incorporation occurred in the absence of ATP or added salts and was not reduced by SCH28080 (a K+ site inhibitor) suggesting that the melittin binding site was distinct from the luminal K+ site of action of SCH28080.

Pancreastatin: a novel peptide inhibitor of parietal cell signal transduction

The direct inhibition of secretion by pancreastatin was investigated in rabbit isolated parietal cells. Pancreastatin exerted no influence on basal aminopyrine uptake. Pancreastatin inhibited histamine stimulated aminopyrine uptake through a decrease in intracellular cAMP. Pancreastatin inhibition of histamine stimulated uptake was blocked in the presence of pertussis toxin. Pancreastatin also inhibited the carbachol stimulated increase in aminopyrine accumulation. However, the effects of pancreastatin on carbachol stimulation were not reversed by pertussis toxin. Pancreastatin did not alter the carbachol induced increase in cytosolic free calcium. Thus, pancreastatin appears to inhibit parietal cell signal transduction at multiple points along the second messenger pathways.

Histamine methyltransferase in dispersed cells from rabbit fundic mucosa

The cellular distribution of histamine N-methyltransferase was studied in rabbit gastric mucosa. The fundic mucosa was dispersed by collagenase treatment in Hanks' or calcium-free medium. In calcium-free medium, the number of dispersed cells/g wet tissue, as well as their viability was increased; histamine N-methyltransferase recovery was up to three-fold larger than in cells prepared in Hanks' medium. Furthermore, the calcium-free medium led to a greater acid secretory response, whereas the cellular pepsinogen content tended to be lower. Histamine N-methyltransferase activity was found in all cell fractions but was higher in the larger cell types. The enzyme activity showed only a partial correlation with either oxyntic or chief cells. These results indicate that the use of calcium-free medium to disperse and isolate rabbit mucosal cells improves cell quality. Histamine N-methyltransferase in the rabbit fundic mucosa, is found in more than one cell type, primarily the oxyntic and chief cells.

Carbonic anhydrase in the normal rat stomach and duodenum and after treatment with omeprazole and ranitidine

A low pH in the lumen of the stomach and duodenum stimulates gastroduodenal mucosal secretion of bicarbonate, particularly in the duodenum. Long-term deprivation of this acid stimulus might affect the ability of the mucosa to secrete bicarbonate, with a consequent decrease in mucosal protection against the acid. This could occur by 'down-regulation' of carbonic anhydrase (CA) activity in the bicarbonate-transporting cells. Levels of CA activity and amounts of CA isoenzymes in rat gastric and duodenal mucosa were determined by biochemical assay and histochemical and immunohistochemical staining. Control animals and animals pre-treated for 4-6 weeks with the histamine H2-receptor antagonist ranitidine (600 mg kg-1 daily) or the H+,K+-ATPase inhibitor omeprazole (28 mg kg-1 daily) were examined. Both drugs are potent inhibitors of gastric secretion of acid. Both gastric and duodenal mucosal total CA activity and the distribution of isoenzymes were very similar in control animals and animals treated with these drugs. In the stomach, CA II was found in the surface epithelial and parietal cells. In the duodenum both CA I and CA II were observed. The staining for CA I was restricted to a small number of villus cells which looked like ordinary duodenal enterocytes. CA II in the duodenum was found in all villus cells, except the goblet cells. The staining decreased gradually from the top to the bottom of the villi and was absent in the crypts. Duodenal bicarbonate secretion is dependent on mucosal CA activity, and the distribution of CA II thus suggests that this alkaline secretion is of villous rather than cryptal origin.

Carbonic anhydrase II gene expression in isolated canine gastric parietal cells

The participation of carbonic anhydrase (CA II) in the gastric acid secretory process has been the subject of considerable controversy. We utilized a cDNA probe for CA II to measure CA II gene expression in canine gastric parietal cells that had been stimulated with the three principal acid secretagogues, carbachol, gastrin, and histamine. Hybridization to dot blots of total parietal cell RNA showed a significant increase in specific CA II mRNA content within minutes of secretagogue addition: carbachol stimulation led to an increase of 52 +/- 8.9%, reaching a maximum within 20 min; gastrin stimulation led to an increase within 60 min of 104 +/- 10.6%; stimulation with histamine was followed within 20 min by an increase of 30 +/- 7.2%. We also measured transcription rates for the CA II gene in cells stimulated by each agent and found an increase within 15 min. Our results show that CA II gene expression is regulated by agents that stimulate the parietal cell to secrete acid and that the accumulation of CA II mRNA subsequent to the initial interaction of stimulant appears to result from new transcription of the CA II gene. These data suggest the participation of CA II in the acid secretory response of the parietal cell to secretagogues.

Histamine metabolism of the human gastric mucosa--a study on the regional distribution of the amine and enzyme activities

Histamine content, histidine decarboxylase activity and histamine methyl transferase activity of the human gastric mucosa have been studied in healthy subjects. Histamine content was evenly distributed over the gastric mucosa and no significant difference could be observed between the oxyntic and pyloric gland area. A 'specific' histidine decarboxylase was found in the human gastric mucosa and the enzyme activity was significantly higher in the corpus compared to the antral mucosa. We found no difference in enzyme activity of the mucosa obtained from three defined areas of the corpus region of the stomach. Histamine methyl transferase activity was of the same magnitude in the corpus as in the antral mucosa. In addition, with regard to histamine methyl transferase activity, we were unable to demonstrate any difference in activity in the mucosal specimens obtained from the defined areas of the stomach. The presented technique to obtain gastric mucosal tissue specimens during routine fibreoptic endoscopy of the stomach, allows studies over gastric mucosal histamine metabolism in health as well as in disease states.

Inhibition of H+,K+-ATPase by methyl(E)-2-(3,4-dimethoxystyryl)-benzimidazole-4-carboxylate (ALE-36)

ALE-36, as well as omeprazole and SCH 28080, markedly inhibited the [14C]aminopyrine (AP) accumulation induced by dibutyryl cyclic AMP (dbcAMP) and H+,K+-ATPase activity in a concentration-dependent manner. The inhibitory effect of omeprazole on the dbcAMP-induced [14C]AP accumulation was reversed by treatment with beta-mercaptoethanol, but those of ALE-36 and SCH 28080 were not. ALE-36 and SCH 28080 did not inhibit dog renal Na+,K+-ATPase activity, while omeprazole and ouabain did inhibit this enzyme activity. These results suggest that the inhibitory action of ALE-36 on acid secretion is due to the specific inhibition of gastric H+,K+-ATPase, the manner being different from in the case of omeprazole.

Omeprazole, SCH 28080 and doxepin differ in their characteristics to inhibit H+/K+-ATPase driven proton accumulation by parietal cell membrane vesicles

The effects of omeprazole, SCH 28080 and doxepin were studied on H+/K+-ATPase mediated H+ accumulation in parietal cell membrane vesicles. Omeprazole had no effect on the initial rate of H+ accumulation and the initial steady state concentration of H+; an inhibition was found after the vesicles were acidified. This inhibition was counteracted by the SH reducing agent dithioerythritol. SCH 28080 inhibited the initial rate of H+ accumulation and the steady state H+ concentration. The inhibitory effect of SCH 28080 was counteracted by KCl. Doxepin (3-100 microM) reduced the initial steady state H+ concentration. Doxepin concentrations lower than 0.5 microM had no such effect but dissipated the proton gradient after the vesicles were fully acidified. This doxepin effect was partially counteracted by KCl and was also obtained in vesicles in which the pump reaction was stopped by EDTA. These data show that (i) omeprazole is an acid-activated compound which interferes with SH groups of the H+/K+-ATPase localized inside the vesicles; (ii) SCH 28080 interferes with the K+ site of the H+/K+-ATPase; and (iii) doxepin interacts by a K+ antagonistic activity at the H+/K+-ATPase site and in addition by intravesicular neutralization and/or a protonophoric mechanism with the process of H+ formation.

Effects of starvation, feeding, and time of day on the activity of proton transport adenosine triphosphatase in the parietal cells of the mouse gastric glands

The effects of starvation, feeding, and time of day on mouse gastric glands were studied by means of an enzyme histochemical method for K+-dependent p-nitrophenyl phosphatase (K+-NPPase), a partial reaction of the proton pump ATPase which drives gastric acid secretion. The stomachs of mice starved for 24 h showed very low levels of parietal cell K+-NPPase histochemical reaction. However, a brief meal following such a period of starvation produced an abrupt increase in K+-NPPase reaction within most of the parietal cell-containing glands though not all parietal cells were equally susceptible to stimulation. The number of glands containing K+-NPPase-reactive parietal cells fell slowly in the hours following a feeding stimulus. These changes were shown to be caused by feeding rather than by general arousal and to follow the feeding cycle in ad libitum fed animals. The reasons that parietal cells in the basal parts of mouse gastric glands cannot be induced to show K+-NPPase reactivity by a feeding stimulus are not understood.

Kinetic properties of potassium stimulated ATPase purified from gastric mucosa

In this study, partially purified K+-H+ ATPase from frog gastric mucosa were obtained by using differential and density gradient centrifugation. Optimum activity of K+-H+ ATPase (Vmax), Michealis-Menten constant (Km) and Hill coefficient (h) were found as 83.3 mumol Pi.mg prot-1.h-1, 95.2 microM and 0.91, respectively. Enzyme preparations were more stable in glycerol solutions stored at -40 degrees C. Minimum activity lost was determined for samples stored in 40% (v/v) glycerol solution at -40 degrees C for two months.

Amiloride inhibits rat mucosal ornithine decarboxylase activity and DNA synthesis

Refeeding fasted rats induces a dramatic trophic response in gastrointestinal mucosa and is associated with elevations in both the rate of DNA synthesis and ornithine decarboxylase (ODC) activity. The signal for these increases is unknown. Amiloride prevents cell alkalinization by blocking Na+-H+ exchange at apical epithelial cell membranes. In study 1, rats were fasted 48 h, treated with amiloride (0.5 to 500 mg/kg), and refed for 4 h. Refeeding increased ODC activities in the jejunal mucosa (X8) and liver (X19) but not in the oxyntic gland mucosa. In the jejunum, but not the liver, the activation of ODC was completely abolished by 100 mg/kg amiloride. In study 2, the rate of DNA synthesis was determined by measuring the rate of [3H]thymidine incorporation 16 h after refeeding. Refeeding resulted in significantly increased rates of DNA synthesis (dpm.microgram DNA-1.30 min-1) over fasted levels, and amiloride at 100 mg/kg significantly reduced the elevations in the jejunum and liver. In conclusion, amiloride inhibits the postprandial increases in jejunal ODC activity and DNA synthesis in the jejunum and liver. The results indicate that 1) the Na+-H+ antiport is essential to the increased ODC activity in the jejunum and the stimulation of DNA synthesis in the jejunum and liver after a meal and 2) increases in DNA synthesis and their suppression by amiloride are not necessarily linked to ODC activity.

Function and structure of parietal cells after H+-K+-ATPase blockade

Omeprazole was administered to rabbits as a single dose or daily for 1 wk. H+-K+-ATPase and isolated gastric glands were prepared from the oxyntic corpus mucosa and used for functional and quantitative morphological studies. Both 10 and 100 mumol omeprazole/kg increased the pH of the gastric content when measured at death. The stimulated oxygen uptake and the rate of aminopyrine (AP) uptake were both inhibited in the isolated gastric gland preparations. Morphometric studies of biopsy specimens taken from the corpus mucosa and isolated gastric glands showed that omeprazole treatment increased the volume density of the acid compartments. This expansion provides an increased accumulation space for AP. Therefore, an increased AP uptake might be seen in glands isolated from omeprazole-treated animals. Blockade of the H2-receptor by ranitidine transformed the morphology of the cell into a more resting type and, furthermore, reduced the omeprazole-induced increase in the volume density of the acid compartments in the parietal cell. The H+-K+-ATPase activity measured in membrane fractions from the omeprazole-treated animals was decreased dose dependently and inhibited by 95% after 100 mumol omeprazole/kg. However, the concentration of the enzyme in these fractions did not change. These results indicate a specific inhibitory action of omeprazole on the H+-K+-ATPase.

Interaction of antidepressants and neuroleptics with histamine stimulated parietal cell adenylate cyclase and H+ secretion

The tricyclic antidepressants trimipramine and doxepin, and the neuroleptic agents trifluoperazine and haloperidol were tested for their effect on histamine H2-receptor-mediated adenylate cyclase activity and H+ secretion in guinea-pig parietal cells. All compounds inhibited histamine-stimulated adenylate cyclase and H+ secretion in a concentration-dependent manner. The antisecretory potency was 1-2 orders of magnitude higher than that for adenylate cyclase inhibition. All drugs caused a rightward shift in the concentration-response curves of histamine-induced adenylate cyclase activation with Schild-plot lines having a slope significantly different from unity. Histamine-stimulated H+ secretion was inhibited by the drugs in a noncompetitive fashion. These results demonstrate that antidepressants and neuroleptics interfere noncompetitively with the parietal cell histamine H2-receptor and that this receptor blocking activity is not related to the antisecretory activity of the drugs.

Effects of ethanol on parietal cell membrane phospholipids and proton pump function

The effects of ethanol on membrane phospholipids of isolated rabbit parietal cells (PC) and fundic glands (FG) were evaluated by using proton nuclear magnetic resonance spectroscopy (NMR). The most conspicuous resonance bands in PC and FG spectra at 3.2 and 0.9 ppm were due to the protons of +N(CH3)3 groups of phosphatidylcholine and, to a lesser degree, other phospholipid derivatives and terminal -CH3 groups for all phospholipids, respectively. Ethanol treatment (1 min) of PC and FG at 37 degrees C significantly increased the peak intensity of -CH3 groups in a dose-dependent manner, while no intensity change in PC and a lesser change in FG were observed for the +N(CH3)3 groups, suggesting a greater disorganizing effect of ethanol in the inner hydrophobic (-CH3) compared with the outer hydrophilic [+N(CH3)3] membrane moieties of phospholipids. Ethanol at concentrations greater than 5% (vol/vol) inhibited K+-stimulated paranitrophenyl phosphatase activity of intact PC, which correlated with inhibition of basal and histamine-stimulated aminopyrine uptake ratio by PC and FG. NMR is a sensitive, noninvasive probe to show ethanol-induced changes in intact PC lipid membrane organization without altering ultrastructural morphology.

Protein kinase C and parietal cell function

Ca2+-phospholipid dependent protein kinase activity and the effect of 12-O-tetradecanoylphorbol-13-acetate (TPA) on H+ secretion was studied in guinea-pig parietal cells. In resting parietal cells PK-C activity was distributed almost equally in both the cytosolic and particulate cell fraction. Exposure of the cells to TPA resulted in a loss of cytosolic PK-C activity which was not accompanied by a concomitant increase of particulate activity. Furthermore TPA inhibits the K+/H+-ATPase and dissipates H+ gradient in intact gastric vesicles by a protonophoric action. The role of these biochemical events for the antisecretory action of TPA found in the intact parietal cell are discussed.

Oncofetal expression of glutathione S-transferase placental form in human stomach carcinomas

The content of glutathione S-transferase placental form (GST-pi) was immunohistochemically analyzed in 100 cases of human stomach carcinoma of various histological types and compared with the content in the stomach mucosa of the human fetus, infant and adult. High levels of GST-pi content were demonstrated in all stomach carcinomas except for signet ring cell carcinomas. High levels were also present in surface mucous cells and glandular cells of the stomach from a fetus aged 18 weeks, whereas only the parietal cells of fundic glands in the adult stomach showed slight staining for GST-pi. These results indicate that the phenotypic expression of GST-pi in human stomach carcinoma is oncofetal in character.

[Morphofunctional changes in the parietal cells of the stomach after adrenalectomy]

As demonstrated experiments performed in 150 white male rats by means of light and electron microscopy, cytochemistry with successive morpho-, cytophotometric and biochemical analysis, after bilateral adrenalectomy, inhibition of proliferative processes of cervical epitheliocytes and their differentiation into parietal cells take place. This results in decreasing amount of the latter, in appearance among them not completely differentiated forms. Dystrophic processes develop, manifesting as a decreasing balance of the secretory membranes and destruction of the mitochondrial apparatus, accompanied with a sharp drop in activity of the main oxidoreductases and ATP-ase. The gastric juice acidity decreases. These changes depend mainly on deficiency of glucocorticoid hormones, but not mineralocorticoids.

[Effect of alpha-tocopherol (vitamin E acetate) on the metabolism of stomach secretory cells in hyperthyroid rats]

Alpha-tocopherol has been shown to cause less marked structural and metabolic changes in parietal stomach cells of hyperthyroid rats.

Prostanoid inhibition of canine parietal cells

Cellular mechanisms underlying the anti-secretory actions of the prostaglandin E2 analogue enprostil were studied using enzyme-dispersed, elutriator-enriched canine parietal cells and the accumulation of the weak base 14C-labeled aminopyrine as a functional index. Enprostil inhibited the accumulation of aminopyrine stimulated by histamine and the phosphodiesterase inhibitor isobutylmethyl, but not by carbachol, gastrin, or dibutyryl cyclic adenosine monophosphate. Inhibition by enprostil was dose-dependent (0.1 nM to 1 microM), with maximal inhibition ranging from 65 to 95 percent. Over the same concentration range, enprostil inhibited the histamine-stimulated generation of cyclic adenosine monophosphate. This selective inhibition of histamine activation of parietal cell function was comparable to that found for prostaglandin E2. Forskolin, a diterpene that directly activates the catalytic subunit of adenylate cyclase, was also markedly inhibited by nanomolar concentrations of prostaglandin E2 and enprostil. We conclude that at least a component of the secretory inhibition by enprostil reflects direct interference with histamine stimulation of parietal cell adenylate cyclase.