Potential role of molecular mimicry between Helicobacter pylori lipopolysaccharide and host Lewis blood group antigens in autoimmunity.

Helicobacter pylori is involved in gastritis, gastric and duodenal ulcers, gastric adenocarcinoma, and mucosa-associated lymphoid tissue lymphoma. Earlier studies already suggested a role for autoimmune phenomena in H. pylori-linked disease. We now report that lipopolysaccharides (LPS) of H. pylori express Lewis y, Lewis x, and H type I blood group structures similar to those commonly occurring in gastric mucosa. Immunization of mice and rabbits with H. pylori cells or purified LPS induced an anti-Lewis x or y or anti-H type I response, yielding antibodies that bound human and murine gastric glandular tissue, granulocytes, adenocarcinoma, and mucosa-associated lymphoid tissue lymphoma cells. Experimental oral infections in mice or natural infection in humans yielded anti-Lewis antibodies also. The beta chain of gastric (H+,K+)-ATPase, the parietal cell proton pump involved in acid secretion, contained Lewis y epitopes; gastric mucin contained Lewis x and y antigenic determinants. Growth in mice of a hybridoma that secretes H. pylori-induced anti-Lewis y monoclonal antibodies resulted in histopathological evidence of gastritis, which indicates a direct pathogenic role for anti-Lewis antibodies. In conclusion, our observations demonstrate that molecular mimicry between H. pylori LPS and the host, based on Lewis antigens, and provide understanding of an autoimmune mechanism for H. pylori-associated type B gastritis.

In vivo trafficking of nascent H(+)-K(+)-ATPase in rabbit parietal cells.

Protein metabolic labeling in vivo was used to determine a time course for trafficking of nascent H(+)-K(+)-adenosinetriphosphatase (H(+)-K(+)-ATPase) from endoplasmic reticulum (ER) to mature tubulovesicles in parietal cells. Stomachs of cimetidine-treated rabbits were taken 15-90 min after injection of [35S]methionine/cysteine, and mucosal microsomes were fractionated on sucrose gradients for analysis by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, Western blot, and autoradiography. After 15 min, labeled alpha-subunit peaked at approximately 1.14 g/ml, matching the distribution of the high-mannose beta-subunit precursor, "pre-beta." After 30 min, most labeled alpha-subunit was in a peak at approximately 1.10 g/ml, considered to be Golgi. By 90 min, most labeled alpha-subunit was in a light peak, at approximately 1.07 g/ml, aligned with the major peak of total H(+)-K(+)-ATPase previously characterized as mature tubulovesicles. From material enriched in pre-beta, alpha-subunit was coprecipitated with pre-beta by a terminal mannose-specific lectin, Galanthus nivalis agglutinin, in the same ratio as the mature alpha:beta ratio. Thus alpha- and beta-subunits associated early in the ER. This is the first use of protein metabolic labeling to study early trafficking of the H(+)-K(+)-ATPase in vivo. The techniques may be usefully applied to examining changes in H(+)-K(+)-ATPase synthetic rate in response to various pharmacological treatments and studying the divergent pathways for nascent H(+)-K(+)- and Na(+)-K(+)-ATPases.

Direct measurement of extracellular proton flux from isolated gastric glands.

We used the microphysiometer, a sensitive extracellular pH sensor, to resolve luminal (or apical) H+ secretion and basolateral release of OH- as well as liberation of acidic metabolites in rabbit gastric glands. Stimulation of glands via the adenosine 3',5'-cyclic monophosphate pathway produced a biphasic change in the extracellular acidification rate (EAR): after an initial transient decrease below the unstimulated baseline (-40.9 +/- 3.4%), the EAR increased to a steady-state maximal plateau (+98.1 +/- 5.3%) within 30 min (n = 37). We interpret the biphasic EAR profile as an initial excess of basolaterally released OH- followed by delayed luminal efflux of simultaneously produced H+. The elevated EAR at steady state reflected liberation of metabolic acid attributed to H(+)-K(+)-ATPase enzymatic activity. The presence of H2-4,4'-diisothiocyanostilbene-2,2'-disulfonic acid prevented OH- release and reduced steady-state EAR. Basolateral OH- release and steady-state EAR were also inhibited by the H(+)-K(+)-ATPase inactivators omeprazole and SCH-28080. Inhibition of Na+/H+ exchange did not reduce steady-state EAR and did not affect apical H+ production, as judged by the accumulation of the weak base aminopyrine. Sodium thiocyanate (1 mM), which short circuits intraluminal H+ accumulation, blocked OH- release, demonstrating its dependence on H(+)-OH- separation at the apical membrane. A computerized model was developed to illustrate how the observed biphasic EAR profile would result from a delayed luminal efflux of H+ due to transitory intraluminal compartmentalization.

Omeprazole decreases H(+)-K(+)-ATPase protein and increases permeability of oxyntic secretory membranes in rabbits.

Amounts and fractional distributions of gastric H(+)-K(+)-adenosinetriphosphatase (ATPase) activity and H(+)-K(+)-ATPase protein as well as properties of H(+)-K(+)-ATPase-containing membranes were studied in rabbits injected with omeprazole (OM; 1 mg/kg sc twice daily for 5 days). Total H(+)-K(+)-ATPase activity decreased to 22 +/- 2% of control (n = 4). Densitometry of sodium dodecyl sulfate-polyacrylamide gel electrophoresis and Western blots showed H(+)-K(+)-ATPase protein was decreased to 60-70% of control. In vitro reduction of the enzyme-OM disulfide bond with 0.1 M 2-mercaptoethanol increased microsomal H(+)-K(+)-ATPase activity to 56 +/- 7% of control (n = 3), consistent with a substantial decrease in enzyme protein. Incorporation of 35S-labeled methionine for 30 min before death resulted in 2.2-fold more label per unit of microsomal alpha-subunit protein (5 days OM vs. control). Thus the decrease in enzyme protein resulted from increased breakdown rather than decreased synthesis. A striking shift in distribution of H(+)-K(+)-ATPase-containing microsomes (tubulovesicles) on sucrose gradients reflected slow equilibration of most control vesicles with the gradient medium and faster equilibration after 5 days OM, indicating increased permeability. After 5 days OM, microsomal vesicle acidification (by acridine orange uptake assay) was negligible, even with 2-mercaptoethanol treatment, and H+ leakage on sudden delta pH was faster than control. We conclude that extended OM treatment not only inhibits H(+)-K(+)-ATPase but accelerates its breakdown and renders H(+)-K(+)-ATPase-containing membranes more permeable. It is thus possible that increased backward H+ flux contributes to profound inhibition of acid secretion during extended omeprazole treatment. In parallel experiments, H(+)-K(+)-ATPase activity and density gradient sedimentation of tubulovesicles returned to near normal 3 days after OM withdrawal.

Ontogeny of gastric H(+)-K(+)-ATPase in suckling rabbits.

Gastric mucosal homogenates were prepared from resting and stimulated stomachs of rabbits, age 3-57 days postnatal, and fractionated by differential centrifugation. Total H(+)-K(+)-adenosinetriphosphatase (ATPase) (assayed as K(+)-dependent ouabain-insensitive hydrolysis of p-nitrophenyl phosphate) was low in the first 3 wk but rapidly accumulated between days 20 and 43. Specific activity rose eightfold from day 3 to a typically adult level of 2 mumol.mg-1.h-1 by day 43. The microsomal fraction (P3) was subfractionated on sucrose gradients (20, 27, and 33% steps or 10-40% continuous gradient). H(+)-K(+)-ATPase from P3 of resting stomachs was distributed bimodally on the continuous gradients, with activity mainly in the denser peak (or on the 33% sucrose step) before day 20, but accumulating mainly in the lighter peak (or in the lighter step-gradient fractions) after day 20. Throughout the age range tested, in vivo stimulation with histamine just before removal of the stomach caused a loss of most H(+)-K(+)-ATPase from P3 and an increase in H(+)-K(+)-ATPase in a lower-speed fraction P1. Thus, even in the 1st postnatal wk, when H(+)-K(+)-ATPase is low, most of the enzyme occurs in cells with histamine H2 receptors and all the intracellular mechanisms for fusion of oxyntic cell tubulovesicles (enriched in P3) with the apical membrane (enriched in P1). These studies delineate a 3-wk period of sharply accelerated synthesis of H(+)-K(+)-ATPase before weaning. Age-related changes in distribution of H(+)-K(+)-ATPase among microsomal density subfractions suggest maturational changes either in the intracellular partitioning of the enzyme or in properties of the membranes containing the enzyme.