Cytological transformations associated with parietal cell stimulation: critical steps in the activation cascade

Cultured rabbit parietal cells were used to evaluate morphological responses to activators and inhibitors of HCl secretion. Immunofluorescence was used to localize the proton pump protein, H, K-ATPase, and the apical membrane-cytoskeletal linker protein, ezrin; fluorescent-labeled phalloidin was used as a marker of F-actin. Treatment of healthy control parietal cells with secretagogues resulted in exaggerated swelling of apical membrane vacuoles, presumably with the accumulation of HCl and water. Thus stimulation-associated swelling of apical vacuoles was blocked by inhibitors that work at various steps in the secretion-activation cascade. When secretion was blocked by agents that prevent the translocation of H,K-ATPase-rich tubulovesicles to apical membrane vacuoles (such as H2-receptor antagonists and protein kinase A inhibitors), the general resting morphology was maintained. ME-3407 (a functional analogue of wortmannin) was unique in preventing H, K-ATPase redistribution and effecting the delocalization of ezrin from apical membrane vacuoles. When secretion was blocked by agents that inhibit the H+ pump or induce H+ backflux, the translocation of H,K-ATPase to apical membrane vacuoles occurred but the large vacuolar swelling associated with HCl and H2O accumulation was greatly diminished. These data support the membrane recycling/recruitment hypothesis of HCl secretion in which H, K-ATPase-rich tubulovesicles are recruited from a cytoplasmic domain to the apical surface, and they are inconsistent with models proposing that the tubulovesicles, regardless of shape, are contiguous with the apical plasma membrane. These studies also demonstrate the utility of the parietal cell culture model in distinguishing a general site of action for various inhibitors and antisecretory agents.

Carbachol activates IkappaB kinase in isolated canine gastric parietal cells

IkappaB kinase (IKK) is a recently discovered kinase complex composed of the kinases IKKalpha and beta, which plays a crucial role in the activation of NF-kappaB. In this study we examined the regulation of IKK by carbachol in isolated gastric parietal cells. IKKalpha and beta activities were measured by immune complex kinase assay. Carbachol induced both IKK alpha and beta in a time-dependent fashion, with a maximal stimulatory effect detected after 5 min of incubation. The action of carbachol was inhibited by the intracellular Ca(++) chelator BAPTA-AM, the PKC inhibitor GF109203X, and the NF-kappaB inhibitor PDTC. Carbachol also induced degradation of IkappaBalpha, which was reversed by addition of both GF109203X and PDTC and stimulated the activity of a NF-kappaB-luciferase reporter gene plasmid in COS-7 cells stably expressing the human M3 muscarinic receptor. In conclusion, carbachol induces IKK in the parietal cells via intracellular Ca(++)- and PKC-dependent signaling pathways. This observation represents a novel mechanism for the regulation of NF-kappaB through the activation of seven transmembrane G-protein-coupled receptors.

Up-regulation of H2 receptor and adenylate cyclase in rabbit parietal cells during prolonged treatment with H2-receptor antagonists

Intragastric hyperacidity occurs after abrupt withdrawal of histamine H2-receptor antagonists, and the prolonged administration of these agents induces tachyphylaxis of the inhibitory effects on gastric acid secretion. We examined the effect of the prolonged administration of H2-receptor antagonists on the H2-receptor signaling system in parietal cells isolated from rabbits that had received H2-receptor antagonists for 14 days. [125I]aminopotentidine (APT) binding sites to H2 receptors in parietal cell membranes were increased without any significant change in the affinity for [125I]APT. The expression of Gs(alpha), guanosine triphosphate (GTP)-binding protein coupled to H2 receptor, was slightly increased. Basal as well as GTP- or histamine-stimulated cAMP production was increased, but no significant change was observed in the presence of an H2-receptor antagonist. The up-regulation of the H2 receptor and adenylate cyclase appeared to cause hypersecretion of acid after withdrawal of H2-receptor blockade.

Kex2 family endoprotease furin is expressed specifically in pit-region parietal cells of the rat gastric mucosa

The proprotein-processing endoprotease furin is localized in the gastric epithelial cells of the pit region in the rat gastric gland. The gastric pit is composed of several cell types, including gastric surface mucosal (GSM) cells and parietal cells. Furin converts many growth- or differentiation-related proproteins to their active forms. We examined identification of furin-positive cells by immunostaining of rat gastric mucosa and regulators of the furin expression by measuring the furin promoter activity by luciferase assay. Furin-positive cells were stained for H(+)-K(+)-ATPase, indicating that they are parietal cells. Furin-positive parietal cells were not stained for transforming growth factor-alpha (TGF-alpha) but were surrounded by TGF-alpha-positive GSM cells. In contrast, parietal cells below the proliferative zone were positive for TGF-alpha but not for furin. Furin-positive parietal cells expressed a high level of epidermal growth factor receptor (EGFR). TGF-alpha stimulated the furin promoter activity highly in a mouse GSM cell line GSM06. Thus we suggest that the parietal cells of the pit region have furin-mediated functions that can be stimulated by EGFR signaling.

Carbachol activates ERK2 in isolated gastric parietal cells via multiple signaling pathways

We previously reported that both carbachol and epidermal growth factor (EGF) are potent inducers of the extracellular signal-regulated protein kinases (ERKs) in isolated gastric canine parietal cells and that induction of these kinases leads to acute inhibitory and chronic stimulatory effects on gastric acid secretion. In this study we investigated the molecular mechanisms responsible for these effects. Both carbachol (100 microM) and EGF (10 nM) induced Ras activation. The role of Ras in ERK2 induction was examined by transfecting parietal cells with a vector expressing hemoagglutinin (HA)-tagged ERK2 (HA-ERK2) together with a dominantly expressed mutant (inactive) ras gene. HA-ERK2 activity was quantitated by in-gel kinase assays. Dominant negative Ras reduced carbachol induction of HA-ERK2 activity by 60% and completely inhibited the stimulatory effect of EGF. Since Ras activation requires the assembly of a multiprotein complex, we examined the effect of carbachol and EGF on tyrosyl phosphorylation of Shc and its association with Grb2 and the guanine nucleotide exchange factor Sos. Western blot analysis of anti-Shc immunoprecipitates with an anti-phosphotyrosine antibody demonstrated that both carbachol and EGF induced tyrosyl phosphorylation of a major 52-kDa shc isoform. Grb2 association with Shc was demonstrated by blotting Grb2 immunoprecipitates with an anti-Shc antibody. Probing of anti-Sos immunoprecipitates with an anti-Grb2 antibody revealed that Sos was constitutively bound to Grb2. To examine the functional role of Sos in ERK2 activation, we transfected parietal cells with the HA-ERK2 vector together with a dominantly expressed mutant (inactive) sos gene. Dominant negative Sos did not affect carbachol stimulation of HA-ERK2 but inhibited the stimulatory effect of EGF by 60%. We then investigated the role of betagamma-subunits in carbachol induction of HA-ERK2. Parietal cells were transfected with the HA-ERK2 vector together with a vector expressing the carboxy terminus of the beta-adrenergic receptor kinase 1, known to block signaling mediated by betagamma-subunits. In the presence of this vector, carbachol induction of HA-ERK2 was inhibited by 40%. Together these data suggest that, in the gastric parietal cells, carbachol activates the ERKs through Ras- and betagamma-dependent mechanisms that require guanine nucleotide exchange factors other than Sos.

Expression and characterization of protein kinase C in isolated rabbit parietal cells

We investigated the expression, characterization and distribution of protein kinase C (PKC) isozymes in isolated rabbit parietal cells (PC). Cellular extracts of PC were analyzed by Western blot using isozyme-specific antibodies. The Ca2+-independent PKC-epsilon was detected in cytosolic, membrane and cytoskeletal fractions of basal and histamine-stimulated PC, whereas the Ca2+-dependent PKC-alpha was confined to the cytosolic and membrane fractions. Cytosolic and membrane fractions were partially purified by DEAE cellulose column chromatography with elution of increasing NaCl concentration. Eluates of 0.15 M and 0.3 M NaCl PC fractions were identified as PKC-alpha and -epsilon isoforms, respectively. Phorbol 12-myristate 13-acetate (TPA) treatment of PC for 15, 30 and 60 sec decreased significantly cytosolic PKC-alpha and increased membrane-associated PKC-alpha. In contrast to the distribution of PKC-alpha, TPA did not alter membrane or cytosolic level of PKC-epsilon. Comparison of the dose-response curves between TPA-induced hydrogen (H+) secretion, as measured by aminopyrine (AP) uptake, and the membrane-associated PKC-alpha suggests that translocation of PKC-alpha is not involved in the H+ secretory process in PC. Furthermore, a PKC inhibitor, staurosporine, produced a concentration-dependent enhancement of histamine-stimulated H+ secretion. These findings suggest that PKC-alpha plays a negative modulatory role, rather than an obligatory role, in H+ secretion. The localization and distribution of PKC-epsilon into the cytoskeletal fraction of PC also suggests that this isozyme may be involved in the cellular regulation of reversible morphological transformation during stimulation.

Modulation by nitric oxide of gastric acid secretion in toads

Nitric oxide (NO) is a novel chemical messenger that mediates a variety of biological actions. This study was undertaken to investigate the effects of NO on parietal cell function. The rate of [3H]arginine conversion to [3H]citrulline, a parameter of NO synthase activity, and NO formation (as NO2-), were inhibited by the NO synthase inhibitor, NG-nitro-L-arginine methyl ester (L-NAME), in a concentration-dependent manner in the non-stimulated toad gastric mucosa. This range of concentrations of L-NAME provoked stimulation of H+ secretion in a similar fashion, which was blocked by L-arginine but not by D-arginine. Pre-treatment with carbachol plus ethylene glycol-bis(beta-aminoethyl ether)-N,N,N',N'-tetra-acetic acid (EGTA) prevented the effect of L-NAME on H+ secretion and drastically reduced NO synthase activity. L-arginine had an inhibitory effect on H+ secretion in non-stimulated and carbachol-stimulated gastric mucosa, which was reversed by L-NAME. Carbachol and pentagastrin, but not histamine, significantly increased NO formation in the toad gastric mucosa. The results suggest that changes in NO synthesis in the gastric mucosa may modulate parietal cell function and that a calcium-dependent mechanism may be involved.

Regulation of c-Jun NH2-terminal kinases in isolated canine gastric parietal cells

c-Jun NH2-terminal kinases (JNKs) are protein kinases that are activated by a wide variety of extracellular signals. This study investigated the expression and regulation of JNKs in isolated gastric canine parietal cells. Western blot analysis of cell lysates from highly purified (>95%) parietal cells with an antibody recognizing JNK1 and to a lesser degree JNK2 revealed the presence of two bands of 46 and 54 kDa, respectively. JNK1 activity was quantitated by immunoprecipitation and in-gel kinase assays. Of the different agents tested, carbachol was the most potent inducer of JNK1 activity, whereas histamine and epidermal growth factor induced weaker responses. The proinflammatory cytokine tumor necrosis factor-alpha stimulated JNK1 but had no effect on extracellular signal-regulated kinase (ERK2) induction, suggesting that activation of JNK1 might represent an important event in mediation of the inflammatory response in the stomach. The action of carbachol was dose (0.1-100 microM) and time dependent, with a maximal stimulatory effect (fourfold) detected after 30 min of incubation and sustained for 2 h. Addition of the specific protein kinase C (PKC) inhibitor GF109203X did not affect the stimulatory action of carbachol. The intracellular Ca2+ chelator 1,2-bis(2-aminophenoxy)ethane-N,N,N', N'-tetraacetic acid-AM inhibited carbachol induction of JNK1 activity by 60%. Thapsigargin (1 microM), an intracellular Ca2+-rising agent, induced JNK1 activity more than threefold. Carbachol activation of JNK1 resulted in induction of c-Jun (protein) transcriptional activity and in stimulation of parietal cell mRNA content of c-jun. In conclusion, our data indicate that carbachol induces JNK activity in gastric parietal cells via intracellular Ca2+-dependent, PKC-independent pathways, leading to induction of c-jun gene expression via phosphorylation and transcriptional activation of c-Jun.

Inhibition of gastric mucosal prostaglandin synthetase activity by mercaptomethylimidazole, an inducer of gastric acid secretion--plausible involvement of endogenous H2O2

We have reported earlier that mercaptomethylimidazole (MMI), an antithyroid drug of thionamide group, induces gastric acid secretion at least partially through the liberation of histamine, sensitive to cimetidine. Now, we show that the drug has a significant inhibitory effect on the cyclooxygenase and peroxidase activity of the prostaglandin (PG) synthetase of the gastric mucosal microsomal preparation. The effect can also be mimicked by low concentrations of H2O2. While studying the possible intracellular effect of MMI on acid secretion, a cell fraction (F3) enriched in parietal cell was isolated by controlled digestion of the mucosa with protease. This cell fraction is activated by MMI as measured by increased O2 consumption. The activation is sensitive to omeprazole, a proton-pump inhibitor, indicating that the activation is due to increased acid secretion by MMI. MMI was also found to directly inhibit the peroxidase activity of the F3 cell fraction and may thus increase the intracellular level of H2O2. The cyclooxygenase activity of the PG synthetase of the F3 cell fraction is also inhibited by MMI and the effect can be reproduced by low concentrations of H2O2. Both MMI and H2O2 can also inhibit the peroxidase activity of the PG synthetase. We suggest that in addition to the activation of the parietal cell by MMI possibly through endogenous H2O2, MMI induces acid secretion in vivo by inactivating the PG synthetase thereby inhibiting the biosynthesis of PG and removing its inhibitory influence on acid secretion so that the histamine released by MMI can stimulate acid secretion with maximum efficiency.

Demonstration and immunolocalization of ATP diphosphohydrolase in the pig digestive system

Two isoforms of ATP diphosphohydrolase (ATPDase; EC 3.6.1.5) have been previously characterized, purified, and identified. This enzyme is an ectonucleotidase that catalyzes the sequential release of gamma- and beta-phosphate groups of triphospho- and diphosphonucleosides. One of its putative roles is to modulate the extracellular concentrations of purines in different physiological systems. The purpose of this study was to define, identify, and localize these two isoforms of ATPDase in the pig digestive system. ATPDase activity was measured in pig stomach, duodenum, pancreas, and parotid gland. Enzyme assays, electrophoretograms, and Western blots with a polyclonal antibody that recognizes both isoforms demonstrate the presence of ATPDase in these organs. Immunolocalization showed intense reactions with gastric glands (parietal and chief cells), intestine (columnar epithelial cells), parotid gland, and pancreas. Smooth muscle cells all along the digestive tract were also highly reactive. Considering the variety of purinoceptors associated with the digestive system, the ATPDase is strategically positioned to modulate purine-mediated actions such as electrolyte secretion, glandular secretion, smooth muscle contraction, and blood flow.

Protein kinase-dependent phosphorylation of H,K ATPase-containing membranes from rat and pig stomachs

Previously H,K ATPase preparations from pig stomach were shown to contain intrinsic protein kinase activities which phosphorylated specific tyrosine and serine residues in the N-terminal of the alpha-chain of H,K ATPase (Togawa et al. 1996). In the present investigation, pig H,K ATPase-containing membrane preparations were compared with rat preparations. In contrast to results obtained with the alpha-subunit of H,K ATPase from pig, phosphorylation was not observed in the rat enzyme. Addition of rat preparations to the pig preparations resulted in decreased phosphorylation in pig preparations. To follow the phosphorylation of membrane proteins in vivo, 32P-loaded gastric cells prepared from rat were stimulated with several secretagogues. Proteins with molecular weights of about 120 and 80 kDa were markedly phosphorylated upon stimulation, but the alpha-subunit of H,K ATPase was not. These results suggest that phosphorylation of tyrosine or serine residues of H,K ATPase found in pig H,K ATPase preparations may not be involved in the acid secretion pathway.

Structural aspects of the gastric H,K ATPase

The gastric H,K ATPase is an alpha beta heterodimeric member of the eukaryotic alkali-cation P-type ion-motive ATPase family. The alpha subunit is composed of 1033 amino acids and the beta subunit of 291 amino acids with 6 or 7 potential N-linked glycosylation sites. Much effort has been expended to define the membrane domain of P-type ATPases. A membrane domain of the large subunit consisting of 10 membrane-spanning sequences is suggested by a combination of methods such as (1) tryptic digestion, separation, and sequencing of membrane peptides, (2) labeling with extracytoplasmic reagents, and (3) in vitro translation of hydrophobic segments. The beta subunit has a single transmembrane segment with strong hydrophobic interactions with the alpha subunit. Blue native gel electrophoresis shows that the enzyme is an (alpha-beta)2 dimer. Cross-linking with Cu-phenanthroline provides evidence that association is between the alpha subunits, and the potential SH groups that are Cu sensitive are at cysteine 565 and cysteine 615, in the region of the large cytoplasmic loop between the fourth and fifth transmembrane segments. No cross-linking is observed in the membrane domain. ATP prevents cross-linking because of a conformational change at the surface of the protein induced by ATP or by direct binding of the nucleotide at the site of cross-linking. The WGA binding properties of the beta subunit allow investigation of the region of interaction with the alpha subunit. Thus, digestion of the enzyme by trypsin followed by SDS solubilization and selective elution from a WGA column resulted in coelution of the membrane fragment containing TM7 and TM8. This result demonstrates major hydrophobic interaction between the seventh and eighth transmembrane segments and the beta subunit. An antibody generated against rat parietal cells also recognized shared epitopes in the same region of both the alpha and beta subunits. Biochemical investigation of the arrangement of the transmembrane segments has been hindered by the lack of effective cross-linking reagents probably because of the compact arrangement of this domain, preventing even Cu access. A series of antiulcer drugs has been developed that have a unique chemistry related to their inhibition of the gastric H,K ATPase. They are 2-(substituted pyridyl methylsulfinyl) benzimidazoles, weak bases with a pKa of 4.0. After accumulation in the acidic space generated by the H,K ATPase either in vivo or in vitro, they undergo acid-catalyzed conversion to a tetracyclic sulfenamide which reacts with luminally accessible SH residues to form stable disulfide derivatives. In the particular case of pantoprazole, 2-(3,4-dimethoxy-2-pyridyl-methylsulfinyl)-5-difluoromethoxy benzimidazole, reaction is confined largely to cysteine 813, placed between the fifth and sixth transmembrane segments. The 5 azido analog of pantoprazole provided acid transport-dependent inhibition of the isolated transporting ATPase by this photoactivatable covalent SH reagent. The inhibited enzyme was then photolyzed, cleaved with trypsin, and the membrane fragments compared before and after photolysis. Disappearance of the segment corresponding to TM3,4 and a relative loss of the segment corresponding to TM7,8 suggests close proximity of these two membrane pairs to the loop joining the fifth and sixth transmembrane segments, in particular TM3,4. Use of this type of covalent, photoactivatable site-specific reagent to determine loop proximity can be extended to other acid transporters.

Functional impairment of the individual rat stomach ECL cell in response to sustained hypergastrinemia

ECL cells in the oxyntic mucosa secrete histamine and pancreastatin in response to gastrin. The present study examined gastrin-evoked ECL-cell responses over a 10-week time span in terms of individual ECL cells and unit ECL cell volume. Rats were treated with omeprazole (400 micromol/kg per day orally). The concentrations of gastrin and pancreastatin in serum and of histamine and pancreastatin in the oxyntic mucosa were measured as was the activity of the oxyntic mucosal histidine decarboxylase (HDC). The ECL cells were visualized by immunostaining of histamine and examined by electron microscopy. The total ECL cell number and volume, and the mean ECL cell diameter and volume were determined. The HDC, chromogranin A (CGA) and cholecystokinin-B (CCK-B) receptor mRNA concentrations were determined. In terms of individual ECL cells and unit ECL cell volume, the serum pancreastatin concentration, the oxyntic mucosal histamine content, HDC activity, and HDC, CGA and CCK-B receptor mRNA contents increased slowly at first and then leveled off or started to decline after 2 weeks. After 10 weeks all ECL-cell parameters (expressed per unit ECL cell volume) were back to or approaching the starting value. In conclusion, sustained hypergastrinemia first activates each individual ECL cell (with a peak after 1-2 weeks) and then causes gradual functional impairment, the activity returning towards the pre-stimulation level.

Neonatal injection of native proton pump antigens induces autoimmune gastritis in mice

BACKGROUND & AIMS

Autoimmune gastritis is associated with gastric H+, K(+)-adenosine triphosphatase (ATPase)-specific autoantibodies (HKAb). The (auto) antigen that triggers disease and the pathogenic role of the autoantibodies are unknown. The aim of this study was to analyze when these autoantibodies are produced during autoimmune gastritis in neonatally thymectomized mice and whether a native H+, K(+)-ATPase antigen preparation can induce disease in mice.

METHODS

Autoantibodies were characterized by a novel assay based on immunoprecipitation of a functional H+, K(+)-ATPase expressed in Xenopus oocytes. Normal mice were injected intraperitoneally with H+, K(+)-ATPase-enriched gastric membranes in the absence of adjuvant.

RESULTS

Conformational autoantibodies recognizing both H+, K(+)-ATPase subunits appeared simultaneously with the gastric lesions 1 month after thymectomy. Immunization of neonates, but not adults, induced a persistent autoimmune gastritis in the body mucosa, characterized by lymphocytic infiltrations, loss of parietal and chief cells, metaplasia, and H+, K(+)-ATPase-specific autoantibodies. The histopathological lesions of this new model are similar to those in humans and thymectomized mice.

CONCLUSIONS

The onset of gastritis and autoantibody production parallels the expression of the H+, K(+)-ATPase during ontogeny. Exposure of the neonatal immune system to organ-specific antigens expressed late after birth induces autoimmune gastritis in adult mice.

Glucose suppresses the activity of rat oxyntic histidine decarboxylase without affecting gastrin levels

Glucose suppressed the activity of oxyntic mucosal histidine decarboxylase within 2 h when given either intragastrically or intraperitoneally to rats fasted for 24 h. Serum levels of gastrin, secretin, glucagon, and somatostatin and oxyntic mucosal levels of gastrin, histamine, and somatostatin showed no significant changes after glucose. Glucose suppressed the aspirin-induced histidine decarboxylase activity without changing serum gastrin. It also suppressed the pentagastrin-induced histidine decarboxylase activity. Neither fructose nor mannitol had such an effect. These results suggest that glucose acts directly on the enterochromaffin-like cells in rat oxyntic mucosa to suppress histidine decarboxylase activation.

Immunization with gastric H+/K(+)-ATPase induces a reversible autoimmune gastritis

The gastric H+/K(+)-ATPase has been implicated as a major autoantigen in pernicious anaemia in humans and in thymectomy-induced autoimmune gastritis in mice. Here we have shown that autoimmune gastritis can be generated by direct immunization of non-thymectomized BALB/c mice with mouse gastric H+/K(+)-ATPase in complete Freund's adjuvant. The gastritis was characterized by infiltration of the gastric submucosa and mucosa with macrophages, CD4+ and CD8+ T cells, and B cells and by circulating autoantibodies to the H+/K(+)-ATPase. The mononuclear infiltrate within the gastric mucosa was accompanied by loss of parietal and zymogenic cells and accumulation of small immature epithelial cells. Splenocytes from gastritic mice adoptively transferred gastritis to naive recipients. Cessation of immunization resulted in decrease in autoantibody titre and regeneration of parietal and zymogenic cells. The results directly confirm that the gastric H+/K(+)-ATPase is the causative autoantigen in the genesis of autoimmune gastritis. Recovery of the lesion following cessation of immunization suggests that homeostatic mechanisms can reverse a destructive autoimmune process.

A tyrosine-based signal targets H/K-ATPase to a regulated compartment and is required for the cessation of gastric acid secretion

Gastric acid secretion is mediated by the H/K-ATPase of parietal cells. Activation of acid secretion involves insertion of H/K-ATPase into the parietal cell plasmalemma, while its cessation is associated with reinternalization of the H/K-ATPase into an intracellular storage compartment. The cytoplasmic tail of the H/K-ATPase beta subunit includes a four residue sequence homologous to tyrosine-based endocytosis signals. We generated transgenic mice expressing H/K-ATPase beta subunit in which this motif's tyrosine residue is mutated to alanine. Gastric glands from animals expressing mutant beta subunit constitutively secrete acid and continuously express H/K-ATPase at their cell surfaces. Thus, the beta subunit's tyrosine-based signal is required for the internalization of H/K-ATPase and for the termination of acid secretion. As a consequence of chronic hyperacidity, the mice develop gastric ulcers and a hypertrophic gastropathy resembling Menetrier's disease.

Two Rab proteins, vesicle-associated membrane protein 2 (VAMP-2) and secretory carrier membrane proteins (SCAMPs), are present on immunoisolated parietal cell tubulovesicles

The tubulovesicles of gastric parietal cells sequester H+/K+-ATPase molecules within resting parietal cells. Stimulation of parietal cell secretion elicits delivery of intracellular H+/K+-ATPase to the apically oriented secretory canaliculus. Previous investigations have suggested that this process requires the regulated fusion of intracellular tubulovesicles with the canalicular target membrane. We have sought to investigate the presence of critical putative regulators of vesicle fusion on immunoisolated gastric parietal cell tubulovesicles. Highly purified tubulovesicles were prepared by gradient fractionation and immunoisolation on magnetic beads coated with monoclonal antibodies against the alpha subunit of H+/K+-ATPase. Western blot analysis revealed the presence of Rab11, Rab25, vesicle-associated membrane protein 2 (VAMP-2) and secretory carrier membrane proteins (SCAMPs) on immunoisolated vesicles. The same cohort of proteins was recovered on vesicles immunoisolated with monoclonal antibodies against SCAMPs and VAMP-2. In contrast, whereas immunoreactivities for syntaxin 1A/1B and synaptosome-associated protein (SNAP-25) were present in gradient-isolated vesicles, none of the immunoreactivity was associated with immunoisolated vesicles. The observation of VAMP-2 and two Rab proteins on immunoisolated H+/K+-ATPase-containing tubulovesicles supports the role for tubulovesicles in a regulated vesicle fusion process. In addition, the presence of SCAMPs along with Rab11 and Rab25 implicates the tubulovesicles as a critical apical recycling vesicle population.

Activation of parietal cell by mercaptomethylimidazole: a novel inducer of gastric acid secretion

Mercaptomethylimidazole (2-Mercapto-1-methylimidazole, MMI), an antithyroid drug of thionamide group, significantly activated the parietal cell for acid secretion, as evidenced by increased O2 consumption by more than 2.5-fold over the basal level. When compared, MMI-induced activation was similar to that of histamine but significantly higher than that of isobutylmethylxanthine or carbachol. Activation by MMI was not prevented by receptor blockers of the parietal cell, indicating that its effect was not mediated through the cell surface histamine-H2 receptor or the muscarinic receptor. However, the activation was almost completely blocked only by omeprazole, an established inhibitor of the terminal proton-pumping H+-K+-ATPase of the parietal cell. That MMI-induced activation was coupled with the H+ transport was further confirmed by significant increase in [14C]-aminopyrine uptake by MMI in rabbit gastric gland preparation. MMI-dependent activation of the parietal cell correlated well with the inhibition of the endogenous peroxidase activity. In vitro studies indicated that MMI irreversibly inactivated both peroxidase and catalase activity of the parietal cell in presence of H2O2. As inactivation of these H2O2-scavenging enzymes should increase accumulation of intracellular H2O2, the effect of latter was studied on acid secretion. H2O2 at a low concentration, stimulated acid secretion by sevenfold in isolated gastric mucosa, which was sensitive to omeprazole. It also significantly stimulated [14C]-aminopyrine uptake in gastric gland preparation. We suggest that MMI activated parietal cells to stimulate acid secretion by endogenous accumulation of H2O2 through inactivation of the peroxidase-catalase system.

Bombesin inhibits histamine release from the rat oxyntic mucosa by a somatostatin-dependent mechanism

BACKGROUND AND METHODS

This study examines the effect of bombesin on endogenous somatostatin and the histamine-synthesizing enterochromaffin-like cells. Somatostatin and histamine were measured in the venous effluent of isolated/antrectomized vascularly perfused rat stomachs after administration of bombesin and gastrin alone or combined. Histidine decarboxylase (HDC) enzyme activity and mRNA abundance were measured in the gastric corpus after intravenous administration of bombesin to conscious rats.

RESULTS

Bombesin released somatostatin from the isolated stomachs and reduced basal and gastrin-stimulated venous histamine. Somatostatin antiserum partially reversed the effect of bombesin on basal and gastrin-stimulated histamine release. In conscious fed rats, intravenous bombesin doubled serum gastrin concentrations and increased HDC activity.

CONCLUSION

We conclude that endogenous (paracrine) somatostatin inhibits basal and gastrin-stimulated histamine release from the ECL cell. In intact animals this effect is surmountable by simultaneously released gastrin, suggesting that a balance between the effects of gastrin and somatostatin determines the activation of the ECL cell.

Transcriptional activation of H+/K+-ATPase genes by gastric GATA binding proteins

H+/K+-ATPase (composed of alpha and beta subunits) and histamine H2 receptor are specifically expressed in gastric parietal cells. The GATA binding proteins (GATA-GT1 and GATA-GT2, also called GATA-6 and GATA-4, respectively) originally found in the gastric mucosa recognized a sequence motif [gastric motif, (G/C)PuPu(G/C)NGAT(A/T)PuPy] in the upstream regions of the ATPase genes [Tamura, S., Wang, X.-H., Maeda, M., and Futai, M. (1993) Proc. Natl. Acad. Sci. USA 90, 10876-10880]. These proteins activated the transcription of the reporter gene ligated downstream of the control region of the rat ATPase alpha or beta subunit gene but had no effect on the same reporter ligated downstream of the H2 receptor gene. Deletion analyses suggested that the upstream 249 (alpha gene) and 323 (beta gene) base pair sequences from the first letter of the initiation codon are sufficient for activation by the GATA proteins. Interestingly, two and three gastric motifs are located near the TATA-boxes of the alpha and beta genes, respectively. Mutagenesis studies demonstrated that the two motifs proximal to the TATA-box sequences of the ATPase alpha and beta subunit genes were essential for the activation. These results suggest that both the alpha and beta subunit genes are regulated similarly by the GATA binding proteins. The expression system established in this study is a useful system for analyzing the roles of GATA proteins in transcriptional regulation of the H+/K+-ATPase gene.

Association of syntaxin 3 and vesicle-associated membrane protein (VAMP) with H+/K(+)-ATPase-containing tubulovesicles in gastric parietal cells

H+/K(+)-ATPase is the proton pump in the gastric parietal cell that is responsible for gastric acid secretion. Stimulation of acid secretion is associated with a reorganization of the parietal cells resulting in the incorporation of H+/K(+)-ATPase from a cytoplasmic membrane pool, the tubulovesicle compartment, into the apical canalicular membrane. To better characterize the role of membrane trafficking events in the morphological and physiological changes associated with acid secretion from parietal cells, we have characterized the expression and localization of soluble N-ethylmaleimide-sensitive factor attachment protein receptors (SNAREs) in these cells. Each of the six different SNARE proteins examined [syntaxins 1 through 4 of 25-kDa synaptosome-associated protein, and vesicle-associated membrane protein] were found to be expressed in parietal cells. Furthermore, two of these SNAREs, vesicle-associated membrane protein and syntaxin 3, were associated with H+/K(+)-ATPase-containing tubulovesicles while the remainder were excluded from this compartment. The expression of syntaxin 1 and synaptosome-associated protein of 25 kDa in parietal cells, two SNAREs previously thought to be restricted to neuroendocrine tissues, suggests that parietal cells may utilize membrane trafficking machinery that is similar to that utilized for regulated exocytosis in neurons. Furthermore, the localization of syntaxin 3, a putative target membrane SNARE, to the tubulovesicle compartment indicates that syntaxin 3 may have an alternative function. These observations support a role for intracellular membrane trafficking events in the regulated recruitment of H+/K(+)-ATPase to the plasma membrane after parietal cell stimulation.

Evidence that rat stomach ECL cells represent the main source of circulating pancreastatin

Recently, we showed that the ECL cells in the oxyntic mucosa of the rat stomach are an important source of circulating pancreastatin, a fragment of chromogranin A. The present study examined how much the ECL cells contribute to the circulating levels of pancreastatin during omeprazole-evoked hypergastrinemia. Rats received omeprazole (400 mumol kg-1 day-1) by the oral route for 3 weeks. Two weeks after the start of the treatment, the rats were subjected to a sham operation or fundectomy. The concentrations of gastrin and pancreastatin in serum were monitored before and after the operations. The ECL cells were visualized by pancreastatin immunostaining and their number was determined. The activity of oxyntic mucosal histidine decarboxylase (HDC) was measured before and after 2 weeks of omeprazole treatment. Omeprazole-induced hypergastrinemia resulted in elevated serum pancreastatin and increased oxyntic mucosal HDC activity. Pancreastatin-immunoreactive cells were equally numerous before and after 2 weeks of omeprazole treatment. After surgical removal of the ECL cells by fundectomy, the serum gastrin concentration remained high whereas the serum pancreastatin concentration decreased by 90%. We conclude that the ECL cells in omeprazole-treated rats are responsible for 90% of circulating pancreastatin.

A cyclic GMP-dependent housekeeping Cl- channel in rabbit gastric parietal cells activated by a vasodilator ecabapide

1. The membrane potential of rabbit gastric parietal cells is dominated by a Cl- channel with a subpicosiemens single channel conductance in the basolateral membrane. The effects of 3-[[[2-(3,4-dimethoxyphenyl)ethyl]carbamoyl]amino-N-methylbenzamide++ + (DQ-2511: ecabapide), a vasodilator, on the opening of this Cl-1 channel, the cyclic GMP content and the intracellular free Ca2+ concentration ([Ca2+]i) of parietal cells were investigated by whole-cell patch-clamp technique, enzyme immunoassay and Fura 2-fluorescence measurement. 2. Ecabapide stimulated the opening of the Cl-1 channel as determined by the reversal potential. This stimulation was concentration-dependent, and its EC50 value was 0.2 microM. Both the basal and ecabapide-induced openings of the channel were inhibited by 5-nitro-2-(3-phenylpropylamino)-benzoate (NPPB, 500 microM), a Cl- channel blocker. Another Cl- channel blocker, niflumic acid (500 microM) was much less effective. 3. The power spectra of the currents before and after the addition of ecabapide (10 microM) were analysed. Both spectra contained only one Lorentzian (1/f2) component. 4. 6-Anilino-5,8-quinolinedione (LY83583; 5 microM) which prevents activation of soluble guanylate cyclase, significantly inhibited both the basal and ecabapide (10 microM)-induced openings of the Cl- channel. 5. Ecabapide (0.01-100 microM) concentration-dependently elevated the cyclic GMP content in the parietal cell-rich suspension. The EC50 value was 0.2 microM. 6. In single Fura 2-loaded parietal cells, ecabapide (10-100 microM) did not increase [Ca2+]i. 7. These results indicate that ecabapide stimulates an intracellular production of cyclic GMP in the parietal cell without increasing [Ca2+]i, and leads to an activation of the housekeeping Cl- channel.

Pathogenesis of post-thymectomy autoimmune gastritis. Identification of anti-H/K adenosine triphosphatase-reactive T cells

Autoimmune gastritis spontaneously develops following thymectomy of 3-day-old BALB/c mice (d3Tx). These mice develop autoantibodies to the gastric parietal cell proton pump, H/K ATPase, and aberrant expression of the H/K ATPase in the neonatal thymus prevents the induction of disease post-thymectomy. To characterize the effector cells mediating autoimmune gastritis, we isolated H/K ATPase-enriched preparations of parietal cell microsomes and further purified the enzyme by lectin affinity chromatography. Both preparations induced significant proliferative responses of gastric lymph node cells, which were mediated by CD4+, MHC class II-restricted T cells. Surprisingly, T cells reactive to the Ag could only be demonstrated in lymph nodes in the immediate proximity of the stomach; little or no response was seen when mesenteric or peripheral lymph nodes were tested. It is likely that the H/K ATPase-reactive T cells are actually the effector cells in this disease, as they could only be detected in mice that developed gastritis, as indicated by anti-parietal cell Ab, gastric inflammation, and the presence of cells capable of transferring disease into nu/nu mice. H/K ATPase-specific T cell proliferative responses could first be detected 5 wk post-thymectomy and were accompanied by high background responses at this time point. These latter responses may represent enhanced syngeneic MLRs, which we have previously shown to be elevated in d3Tx mice. Characterizations of the H/K ATPase-reactive and self-reactive T cell populations may reveal the factors that break peripheral T cell tolerance and lead to the development of organ-specific autoimmune disease.