Inhibitor and ion binding sites on the gastric H,K-ATPase

The gastric H,K-ATPase catalyzes electroneutral exchange of H(+) for K(+) as a function of enzyme phosphorylation and dephosphorylation during transition between E(1)/E(1)-P (ion site in) and E(2)-P/E(2) (ion site out) conformations. Here we present homology modeling of the H,K-ATPase in the E(2)-P conformation as a means of predicting the interaction of the enzyme with two known classes of specific inhibitors. All known proton pump inhibitors, PPIs, form a disulfide bond with cysteine 813 that is accessible from the luminal surface. This allows allocation of the binding site to a luminal vestibule adjacent to Cys813 enclosed by part of TM4 and the loop between TM5 and TM6. K(+) competitive imidazo-1,2alpha-pyridines also bind to the luminal surface of the E(2)-P conformation, and their binding excludes PPI reaction. This overlap of the binding sites of the two classes of inhibitors combined with the results of site-directed mutagenesis and cysteine cross-linking allowed preliminary assignment of a docking mode for these reversible compounds in a position close to Glu795 that accounts for the detailed structure/activity relationships known for these compounds. The new E(2)-P model is able to assign a possible mechanism for acid secretion by this P(2)-type ATPase. Several ion binding side chains identified in the sr Ca-ATPase by crystallography are conserved in the Na,K- and H,K-ATPases. Poised in the middle of these, the H,K-ATPase substitutes lysine in place of a serine implicated in K(+) binding in the Na,K-ATPase. Molecular models for hydronium binding to E(1) versus E(2)-P predict outward displacement of the hydronium bound between Asp824, Glu820, and Glu795 by the R-NH(3)(+) of Lys791 during the conformational transition from E(1)P and E(2)P. The site for luminal K(+) binding at low pH is proposed to be between carbonyl oxygens in the nonhelical part of the fourth membrane span and carboxyl oxygens of Glu795 and Glu820. This site of K(+) binding is predicted to destabilize hydrogen bonds between these carboxylates and the -NH(3)(+) group of Lys791, allowing the Lys791 side chain to return to its E(1) position.     

Comparison of covalent with reversible inhibitor binding sites of the gastric H,K-ATPase by site-directed mutagenesis

The gastric H,K-ATPase is covalently inhibited by substituted pyridyl-methylsulfinyl-benzimidazoles, such as omeprazole, that convert to thiophilic probes of luminally accessible cysteines in the acid space. The K(+) competitive inhibitor, SCH28080, prevented inhibition of acid transport by omeprazole. In stably expressing HEK293 cells, the benzimidazole-reactive cysteines, Cys-321 (transmembrane helix (TM) 3), Cys-813 and Cys-822 (TM5/6), and Cys-892 (TM7/8) were mutated to the amino acids found in the SCH28080-resistant Na,K-ATPase and kinetic parameters of H,K-ATPase activity analyzed. Mutations of Cys-822 and Cys-892 had insignificant effects on the K(i(app)), K(m(app)) or V(max), but mutations of Cys-813 to threonine and Cys-321 to alanine decreased the affinity for SCH28080. Mutation of Cys-321 to alanine produced mixed kinetics of inhibition, still with higher affinity for the cation-free form of phosphoenzyme. Since the phenylmethoxy ring of the imidazo-pyridine inhibitors binds to TM1/2, as shown by earlier photoaffinity studies, and the mutations in TM6 (Cys-813 --> Thr) as well as the end of TM3 (Cys-321 --> Ala) decrease the affinity for SCH28080, the TM1/2, TM3, and TM6 helices lie within approximately 16 A of each other based on the size of the active, extended conformation of SCH28080.     

The gastric [H,K]ATPase:H+/ATP stoichiometry

An H+/ATP ratio of 2 for H+ transport was determined from initial rate measurements at pH 6.1 in a purified gastric microsomal fraction containing the [H,K]ATPase. This ratio was independent of external KCl, though the apparent K0.5 for ATP was increased from 10.78 +/- 0.51 (n = 3) to 64.6 +/- 11.9 (n = 3) microM ATP and from 5.13 +/- 0.64 (n = 3) to 65.2 +/- 0.64 (n = 3) microM ATP for H+ transport and the K+-stimulated ATPase, respectively, as K+external was increased from 12 to 150 mM. The H+/ATP ratio was also relatively independent of ATP concentration. Maximum initial rates obtained in KCl-equilibrated vesicles were independent of added valinomycin, though net H+ transport was increased 29.3 +/- 1.03% (n = 6) by the addition of ionophore. Maximum net H+ transport in this vesicle preparation was 185 +/- 2.1 (n = 14) nmol mg-1 of protein. Initial rate measurements of ATPase represent a burst of K+-dependent activity of approximately 10-15 s duration. The H+/ATP stoichiometry was calculated based on the K+-stimulated component of hydrolysis. Under most conditions, the Mg2+-dependent component of hydrolysis was less than 10% of the (Mg2+ + K+) component of hydrolysis.     

SCH28080, a K+-competitive inhibitor of the gastric H,K-ATPase,binds near the M5-6 luminal loop,preventing K+ access to the ion binding domain

Inhibition of the gastric H,K-ATPase by the imidazo[1,2-alpha]pyridine, SCH28080, is strictly competitive with respect to K+ or its surrogate, NH4+. The inhibitory kinetics [V(max), K(m,app)(NH4+), K(i)(SCH28080), and competitive, mixed, or noncompetitive] of mutants can define the inhibitor binding domain and the route to the ion binding region within M4-6. While mutations Y799F, Y802F, I803L, S806N, V807I (M5), L811V (M5-6), Y928H (M8), and Q905N (M7-8) had no effect on inhibitor kinetics, mutations P798C, Y802L, P810A, P810G, C813A or -S, I814V or -F, F818C, T823V (M5, M5-6, and M6), E914Q, F917Y, G918E, T929L, and F932L (M7-8 and M8) reduced the affinity for SCH28080 up to 10-fold without affecting the nature of the kinetics. In contrast, the L809F substitution in the loop between M5 and M6 resulted in an approximately 100-fold decrease in inhibitor affinity, and substitutions L809V, I816L, Y925F, and M937V (M5-6, M6, and M8) reduced the inhibitor affinity by 10-fold, all resulting in noncompetitive kinetics. The mutants L811F, Y922I, and I940A also reduced the inhibitor affinity up to 10-fold but resulted in mixed inhibition. The mutations I819L, Q923V, and Y925A also gave mixed inhibition but without a change in inhibitor affinity. These data, and the 9-fold loss of SCH28080 affinity in the C813T mutant, suggest that the binding domain for SCH28080 contains the surface between L809 in the M5-6 loop and C813 at the luminal end of M6, approximately two helical turns down from the ion binding region, where it blocks the normal ion access pathway. On the basis of a model of the Ca-ATPase in the E2 conformation (PDB entry 1kju), the mutants that change the nature of the kinetics are arranged on one side of M8 and on the adjacent side of the M5-6 loop and M6 itself. This suggests that mutations in this region modify the enzyme structure so that K+ can access the ion binding domain even with SCH28080 bound.     

Cation transport by gastric H+:K+ ATPase.

A vesicular microsomal fraction isolated from hog fundic mucosa demonstrates the capacity to take up equal amounts of RB+ and Cl-. The amount of the Rb+ uptake is sensitive to the extravesicular osmolarity, and rate of uptake is sensitive to temperature. 86Rb+ efflux is dependent upon the cation composition of the diluting solution. ATP, but not beta-gamma methylene ATP, induces a reversible efflux of 86Rb+ from loaded vesicles, and this is dependent upon a functional K+-ATPase. The ATP induced efflux is not affected by CCCP (carbonyl cyanide m-chlorophenylhydrazone) or TCS (tetrachlorosalicylanilide) nor by lipid soluble ions or valinomycin. Nigericin inhibits the efflux by 40%. Uptake of the lipid soluble ion 14C-SCN- has been demonstrated and is enhanced by ATP only in the presence of valinomycin. The results are consistent with a neutral or isopotential exchange of H+ for Rb+ mediated by K+-ATPase.     

Solubilization, purification, and characterization of (H+, K+)ATPase from hog gastric microsomes

The (H+,K+)ATPase-enriched microsomal fraction prepared from hog gastric mucosa by sucrose density gradient centrifugation was effectively solubilized with Emulgen, with apparent preservation of the enzyme activity, and then the ATPase was highly purified by polyethylene glycol fractionation, and Blue Sepharose CL-6B and amino-hexyl Sepharose chromatographies. The purified enzyme showed a single band, with an apparent molecular mass of approximately 94 kDa, on SDS-PAGE, and exhibited both K+-ATPase and K+-stimulated-p-nitrophenyl phosphatase (pNPPase) activities. The optimum pH for the ATPase activity was 7.0. Amino acid analysis of the purified enzyme showed that it contains a large amount of hydrophobic amino acid (42%) and a small amount of glucosamine and galactosamine. The rabbit antibody monospecific for the ATPase, in the Ouchterlony double immunodiffusion and Western blotting tests, markedly inhibited both the K+-ATPase and K+-pNPPase activities.     

Modulation of gastric H+,K+-transporting ATPase function by sodium

Gastric H+,K+-ATPase activity is not affected by Na+ at pH 7.0 but is significantly stimulated by Na+ at pH 8.5. For the stimulation at the latter pH, the presence of both Na+ and K+ were essential. Contrary the H+,K+-ATPase, the associated K+-pNPPase was inhibited by Na+ at both pH values. Sodium competes with K+ for the K+-pNPPase reaction. Also, unlike the H+, K+-ATPase activity the ATPase-mediated transport of H+ within the gastric microsomal vesicles was inhibited by Na+. For the latter event only the extravesicular and not the intravesicular Na+ was effective. The data suggest that the K+-pNPPase activity does not represent the phosphatase step of the H+,K+-ATPase reaction. In addition, the observed inhibition of vesicular H+ uptake by Na+ appears to be due to the displacement by Na+ of a cytosolic (extravesicular) H+ site responsible for the vectorial translocation of H+.     

Acid secretion and the H,K ATPase of stomach.

The regulation of acid secretion was clarified by the development of H2-receptor antagonists in the 1970s. It appears that gastrin and acetylcholine exert their effects on acid secretion mainly by stimulation of histamine release from the enterochromaffin-like (ECL) cell of the fundic gastric mucosa. The isolated ECL cell of rat gastric mucosa responds to gastrin/cholecystokinin (CCK), acetylcholine, and epinephrine with histamine release and to somatostatin and R-alpha-methyl histamine by inhibition of histamine release. Histamine and acetylcholine stimulate the parietal cell by elevation of cAMP or [Ca]i by activation of H2 or M3 receptors, respectively. These independent pathways converge to activate the gastric acid pump, the H+,K+ ATPase. Activation is a function of the association of the ATPase with a potassium chloride transport pathway that occurs in the membrane of the secretory canaliculus of the parietal cell. Hence the secretory canaliculus is the site of acid secretion, the acid being pumped into the lumen of the canaliculus. The pump is composed of two subunits, a large catalytic and a smaller glycosylated protein. This final step of acid secretion has become the target of drugs also designed to inhibit acid secretion. The target domain of the benzimidazole class of acid pump inhibitors is the extracytoplasmic domain of the pump that is secreting acid, and the target amino acids are the cysteines present in this domain. The secondary structure of the pump can be analyzed by determining trypsin-sensitive bonds in intact, cytoplasmic-side-out vesicles of the ATPase, and it has been shown that the alpha subunit has at least eight membrane-spanning segments. Omeprazole, the first acid pump inhibitor, forms a disulfide bond with cysteines in the extracytoplasmic loop between the fifth and sixth membrane-spanning segment and to a cysteine in the extracytoplasmic loop between the seventh and eight segments, preventing phosphorylation of the pump by ATP. As a result of the effective and long-lasting inhibition of acid secretion by the acid pump inhibitor, superior clinical results have been found in all forms of acid-related disease.     

Effect of calcium on the H+/K+ ATPase of hog gastric microsomes

Summary The K⁺-stimulated, ouabain-insensitive ATPase activity present in vesicles of microsomal fractions from hog gastric mucosa can be demonstrated in fresh preparations by adding Ca²⁺ (M range) to the incubation medium. Ca²⁺ effect is similar but not additive to the effect of gramicidin or freezing. High Ca²⁺ concentrations (1 mM) produce an inhibotory effect on the K⁺-stimulated ATPase activity. This effect, is not seen in the presence of gramicidin. Calcium increases the magnitude of ATP-driven H⁺ uptake in vesicles exposed to K⁺ for periods of time up to 60 min. At longer times of exposure (120 min) the response does not differ from controls. It is concluded that Ca²⁺ at low concentrations (m range) enhances the K⁺ permeability of the vesicular membrane. At higher concentrations (mm range), Ca²⁺ becomes inhibitory to the K⁺ permeability. A role for Ca²⁺ as a second messenger in stimulus-secretion coupling in the parietal cell is discussed.     

Mutational analysis of the K+-competitive inhibitor site of gastric H,K-ATPase.

The gastric H,K-ATPase is inhibited selectively and K(+)-competitively from its luminal surface by protonated imidazo[1,2alpha]pyridines (e.g., SCH28080). Identification of the amino acids in the membrane domain that affect SCH28080 inhibition should provide a template for modeling a luminally directed vestibule in this enzyme, based on the crystal structure of the sr Ca-ATPase. Five conserved carboxylic residues, Glu343, Glu795, Glu820, Asp824, Glu936, and unique Lys791 in the H,K-ATPase were mutated, and the effects of mutations on the K(i) for SCH28080, V(max), and K(m,app)[NH(4)(+)] were measured. A kinetic analysis of the ATP hydrolysis data indicated that all of these residues significantly affect the interaction of NH(4)(+) ions with the protein but only three of them, Glu795, Glu936, and Lys791, greatly affected SCH28080 inhibition. A Glu795Asp mutation increased the K(i) from 64 +/- 11 to 700 +/- 110 nM. Since, however, the mutation Glu795Gln did not change the K(i) (86 +/- 31 nM), this site has a significant spatial effect on inhibitor kinetics. A Glu936Asp mutation resulted in noncompetitive kinetics while Gln substitution had no effect either on inhibitor affinity or on the nature of the kinetics, suggesting that the length of the Glu936 side chain is critical for the exclusive binding of the ion and SCH28080. Mutation of Lys791 to Ser, the residue present in the SCH28080-insensitive Na,K-ATPase, resulted in a 20-fold decrease in SCH28080 affinity, suggesting an important role of this residue in SCH28080 selectivity of the H,K-ATPase versus Na,K-ATPase. Mutations of Asp824, Glu343, and Glu820 increased the K(i) 2-3-fold, implying a relatively minor role for these residues in SCH28080 inhibition. It appears that the imidazopyridine moiety of SCH28080 in the protonated state interacts with residues near the negatively charged residues of the empty ion site from the luminal side (TM4, -5, -6, and -8) while the hydrophobic phenyl ring interacts with TM1 or TM2 (the latter conclusion based on previous data from photoaffinity labeling). The integrity of the SCH28080 binding site depends on the presence of Lys791, Glu936, and Glu795 in H,K-ATPase. A computer-generated model of this region illustrates the possible involvement of the residues previously shown to affect SCH28080 inhibition (Cys813, Ile816, Thr823, Met334, Val337) and may predict other residues that line the SCH28080 binding vestibule in the E(2) conformation of the pump.     

Proton binding sites and conformational analysis of H+K(+)-ATPase

It is proposed that the hydronium ion, H3O+, binds to the E1 conformation of the alpha-subunit of gastric proton pump. The H3O+ binding cavities are characterized parametrically based on valence, sequence, geometry, and size considerations from comparative modeling. The cavities have scope for accommodating monovalent cations of different ionic radii. The H3O+ transport is proposed to be aided by arenes which are arranged regularly along the pump starting from N-domain through the transmembrane region. Step-by-step structural changes accompanying H3O+ occlusion are studied in detail. The observations corroborate well with earlier experimental studies.     

125I]azidosalicylyl melittin binding domains: evidence for a polypeptide receptor on the gastric (H+ + K+)ATPase

We have previously shown that melittin, a bee venom peptide, potently inhibited the catalytic and transport functions of rabbit gastric (H+ + K+)ATPase. A radioactive photoaffinity analog of melittin, ([125I]azidosalicylyl melittin), labeled the (H+ + K+)ATPase. These results suggested that melittin exerted inhibitory effects through direct interaction with the (H+ + K+)ATPase. In this study we attempt to define the melittin-binding domain of the (H+ + K+)ATPase using conformation-dependent proteolytic fragmentation of [125I]azidosalicylyl melittin-labeled hog gastric (H+ + K+)ATPase. In the presence of KCl (E2 form) the 95,000-Da [125I]-azidosalicylyl melittin-labeled (H+ + K+)ATPase was cleaved by trypsin to a 40,000-Da NH2-terminal tryptic fragment and a 56,000-Da COOH-terminal fragment through cleavage at Arg 454 of the (H+ + K+)ATPase. The 40,000-Da fragment was labeled by [125I]-azidosalicylyl melittin. The 56,000-Da fragment was not labeled. When unmodified (H+ + K+)ATPase was trypsinized in the presence of KCl, and the fragments were then reacted with [125I]azidosalicylyl melittin, similar tryptic fragmentation results were obtained. In the absence of KCl (E1 form), the 56,000- and 40,000-Da fragments did not accumulate. Chymotryptic hydrolysis of [125I]azidosalicylyl melittin-labeled (H+ + K+)-ATPase was very slow in the presence of KCl (E2 form). In the absence of KCl (E1 form), chymotryptic hydrolysis was more rapid, with accumulation of a major 42,000-Da fragment which was radiolabeled. The melittin-binding region on the (H+ + K+)ATPase is N-terminal to Arg 454 of the (H+ + K+)ATPase. This region is known to contain the aspartyl phosphate residue (Asp 385), the site of phosphoenzyme formation on the (H+ + K+)ATPase. Melittin is also known to bind to calmodulin and other proteins. Another known calmodulin-binding peptide with a different sequence but similar structure, Trp-3, (Leu-Lys-Trp-Lys-Lys-Leu-Leu-Lys-Leu-Leu-Lys-Lys-Leu-Leu-Lys-Leu-Gly) also inhibited the (H+ + K+)ATPase and label incorporation by [125I]azidosalicylyl melittin. These Trp-3 results suggested that the (H+ + K+)ATPase contains a peptide-binding domain which is similar to the peptide-binding domains found on other melittin-binding proteins.     

Ultrastructural and cytochemical analysis of Na+, K+, ATPase and H+, K+, ATPase in parietal cells of gastric mucosa in the rabbit

Summary Rabbit gastric secretion has the physiological peculiarity of being continuous and uninfluenced by food intake. In this respect, ultrastructural analysis of rabbit parietal cells has revealed morphofunctional features situated between states of rest and very active acid secretion. Our cytochemical study shows that Mg²⁺ ATPase and ADPase activities vary from cell to cell and can even be totally absent. These activities concern either microcanaliculi or laterobasal folds or both, but never tubulovesicles. Application of the technique of Mayahara to K⁺ pNPP, associated or not with inhibitors (ouabain, vanadate, N-ethyl-maleimide, sodium fluoride), enabled us to confirm the coexistence of H⁺, K⁺, ATPase and Na⁺, K⁺, ATPase activities in the rabbit and to determine that these activities concern basolateral folds, microcanaliculi, hyaloplasm and tubulovesicles. The global activity of K⁺, pNPPase varied considerably in intensity. The results of using inhibitors suggest that proton transport ceases completely in certain cells. The signs of functional alternation found in this study are in agreement with physiological data relative to this animal.     

Ultrastructural and cytochemical analysis of Na+, K+, ATPase and H+, K+, ATPase in parietal cells of gastric mucosa in the rabbit.

Rabbit gastric secretion has the physiological peculiarity of being continuous and uninfluenced by food intake. In this respect, ultrastructural analysis of rabbit parietal cells has revealed morphofunctional features situated between states of rest and very active acid secretion. Our cytochemical study shows that Mg2+ ATPase and ADPase activities vary from cell to cell and can even be totally absent. These activities concern either microcanaliculi or laterobasal folds or both, but never tubulovesicles. Application of the technique of Mayahara to K+ pNPP, associated or not with inhibitors (ouabain, vanadate, N-ethyl-maleimide, sodium fluoride), enabled us to confirm the coexistence of H+, K+, ATPase and Na+, K+, ATPase activities in the rabbit and to determine that these activities concern basolateral folds, microcanaliculi, hyaloplasm and tubulovesicles. The global activity of K+, pNPPase varied considerably in intensity. The results of using inhibitors suggest that proton transport ceases completely in certain cells. The signs of functional alternation found in this study are in agreement with physiological data relative to this animal.     

Investigation of ion binding to the cytoplasmic binding sites of the Na,K-pump

A dual-wavelength fluorimeter was constructed, which used two light emitting diodes (LEDs) to excite the fluorescence dye RH 421 alternately with two different wavelengths. The ratio of the emissions at the two excitation wavelengths provided a drift-insensitive signal, which allowed detection of very small changes of the fluorescence intensity. Those small changes were induced by ion binding and release in conformation E₁ of the Na,K-ATPase. Titration experiments were performed to determine equilibrium dissociation constants (± standard deviation) for each step in the complete binding and release sequence: 0.12 ± 0.01 mM (E₂(K₂) KE₁), 0.08 ± 0.01 mM (KE₁ E₁), 3.0 ± 0.2 mM (NaE₁ E₁), 5.2 ± 0.4 mM (Na₂E₁ NaE₁) and 6.5 ± 0.4 m_M (Na₃E₁ Na₂E₁) at pH 7.2 and T=16°C. These numbers show that the affinities of the binding sites exposed to the cytoplasm, are higher for K⁺ than for Na⁺ ions, similar to what was found on the extracellular side. The physiological requirement for extrusion of Na⁺ from the cytoplasm, and for import of K⁺ from the extracellular medium seems to be facilitated not by favorable binding affinities in state E₁ but by the two ATP-driven reaction steps of the cycle, E₂(K₂) + ATP K₂E₁ · ATP and Na₃E₁ · ATP (Na₃) E_l-P, which border the ion exchange reactions at the binding sites in conformation E₁. Correspondence to: H.-J. Apell     

Potential cation and H+ binding sites in acid sensing ion channel-1

Acid sensing ion channels (ASICs) are cation-selective membrane channels activated by H⁺ binding upon decrease in extracellular pH. It is known that Ca²⁺ plays an important modulatory role in ASIC gating, competing with the ligand (H⁺) for its binding site(s). However, the H⁺ or Ca²⁺ binding sites involved in gating and the gating mechanism are not fully known. We carried out a computational study to investigate potential cation and H⁺ binding sites for ASIC1 via all-atom molecular dynamics simulations on five systems. The systems were designed to test the candidacy of some acid sensing residues proposed from experiment and to determine yet unknown ligand binding sites. The ion binding patterns reveal sites of cation (Na⁺ and Ca²⁺) localization where they may compete with protons and influence channel gating. The highest incidence of Ca²⁺ and Na⁺ binding is observed at a highly acidic pocket on the protein surface. Also, Na⁺ ions fill in an inner chamber that contains a ring of acidic residues and that is near the channel entrance; this site could possibly be a temporary reservoir involved in ion permeation. Some acidic residues were observed to orient and move significantly close together to bind Ca²⁺, indicating the structural consequences of Ca²⁺ release from these sites. Local structural changes in the protein due to cation binding or ligand binding (protonation) are examined at the binding sites and discussed. This study provides structural and dynamic details to test hypotheses for the role of Ca²⁺ and Na⁺ ions in the channel gating mechanism.     

Binding of intrinsic ATPase inhibitor to mitochondrial ATPase--stoichiometry of binding of nucleotides, inhibitor, and enzyme

F1-ATPase inhibitor was purified from yeast, Saccharomyces cerevisiae. The purified inhibitor blocked ATPase activity in the presence of ATP and Mg2+ by forming a latent equimolar enzyme-inhibitor complex with ATP and ADP newly bound to loose sites on the enzyme. A small portion of externally added ATP was hydrolyzed before the latent complex was formed but the hydrolysis was not directly related to the complex formation. Newly bound ATP tended to be converted to ADP when the ATP concentration of the medium was low. ATP tightly bound to the enzyme was not directly involved in formation of the complex. The complex was fairly stable in the presence of excess inhibitor and ATP but at a high concentration of the enzyme (10(-5) M), the inhibition was not complete, although only about 0.03% of the original activity remained unblocked.     

Species-specific distribution of alpha-galactosyl epitopes on the gastric H/K ATPase beta-subunit: relevance to the binding of human anti-parietal cell autoantibodies.

The gastric H/K ATPase beta-subunit, an abundant glycoprotein of the secretory membranes of gastric parietal cells, is the major autoantigen recognized by human parietal cell autoantibodies in gastric autoimmunity. Our previous studies demonstrated that the human autoantibodies recognize the H/K ATPase beta-subunit from a number of species and that glycosylation of the beta-subunit with complex N-glycans is required for autoantibody binding. The N-glycans of the beta-subunit contain polylactosamine chains. The lactosamine chains of the rabbit beta-subunit are terminated with alpha-linked galactosyl residues (alpha-galactosyl epitope) (Tyagarajan et al., Biochemistry, 1996, 35, 3238-3246). Here we have investigated the expression of alpha-galactosyl epitopes on the H/K ATPase beta-subunit from a number of species. Using the alpha-galactosyl binding lectin, BS1-IB4, and naturally occurring anti-alpha-galactosyl antibodies, we have demonstrated that the rat H/K ATPase beta-subunit also contains terminal alpha-galactosyl residues, but not the beta-subunit from pig, dog, and mouse, indicating species-specific differences in the terminal saccharide sequences of the beta-subunit. We also investigated the potential contribution of the alpha-galactosyl epitopes to the binding by human sera. The reactivity of human pernicious anemia serum with gastric parietal cells could not be inhibited with saccharide inhibitors and, in addition, no binding was observed with normal human sera. We conclude that the H/K ATPase beta-subunit oligosaccharides from rabbit and rat are terminated with alpha-galactosyl epitopes, and although the presence of this epitope does not contribute to binding by human parietal cell autoantibodies at the concentrations routinely used, it is recommended that neither rat or rabbit stomachs be used for screening human sera.     

A conformation-specific interhelical salt bridge in the K+ binding site of gastric H,K-ATPase

Homology modeling of gastric H,K-ATPase based on the E2 model of sarcoplasmic reticulum Ca2+-ATPase (Toyoshima, C., and Nomura, H. (2002) Nature 392, 835-839) revealed the presence of a single high-affinity binding site for K+ and an E2 form-specific salt bridge between Glu820 (M6) and Lys791 (M5). In the E820Q mutant this salt bridge is no longer possible, and the head group of Lys791, together with a water molecule, fills the position of the K+ ion and apparently mimics the K+-filled cation binding pocket. This gives an explanation for the K+-independent ATPase activity and dephosphorylation step of the E820Q mutant (Swarts, H. G. P., Hermsen, H. P. H., Koenderink, J. B., Schuurmans Stekhoven, F. M. A. H., and De Pont, J. J. H. H. M. (1998) EMBO J. 17, 3029-3035) and, indirectly, for its E1 preference. The model is strongly supported by a series of reported mutagenesis studies on charged and polar amino acid residues in the membrane domain. To further test this model, Lys791 was mutated alone and in combination with other crucial residues. In the K791A mutant, the K+ affinity was markedly reduced without altering the E2 preference of the enzyme. The K791A mutation prevented, in contrast to the K791R mutation, the spontaneous dephosphorylation of the E820Q mutant as well as its conformational equilibrium change toward E1. This indicates that the salt bridge is essential for high-affinity K+ binding and the E2 preference of H,K-ATPase. Moreover, its breakage (E820Q) can generate a K+-insensitive activity and an E1 preference. In addition, the study gives a molecular explanation for the electroneutrality of H,K-ATPases.