Defects in TCIRG1 subunit of the vacuolar proton pump are responsible for a subset of human autosomal recessive osteopetrosis.

Osteopetrosis includes a group of inherited diseases in which inadequate bone resorption is caused by osteoclast dysfunction. Although molecular defects have been described for many animal models of osteopetrosis, the gene responsible for most cases of the severe human form of the disease (infantile malignant osteopetrosis) is unknown. Infantile malignant autosomal recessive osteopetrosis (MIM 259700) is a severe bone disease with a fatal outcome, generally within the first decade of life. Osteoclasts are present in normal or elevated numbers in individuals affected by autosomal recessive osteopetrosis, suggesting that the defect is not in osteoclast differentiation, but in a gene involved in the functional capacity of mature osteoclasts. Some of the mouse mutants have a decreased number of osteoclasts, which suggests that the defect directly interferes with osteoclast differentiation. In other mutants, it is the function of the osteoclast that seems to be affected, as they show normal or elevated numbers of non-functioning osteoclasts. Here we show that TCIRG1, encoding the osteoclast-specific 116-kD subunit of the vacuolar proton pump, is mutated in five of nine patients with a diagnosis of infantile malignant osteopetrosis. Our data indicate that mutations in TCIRG1 are a frequent cause of autosomal recessive osteopetrosis in humans.

[3H]Bafilomycin as a probe for the transmembrane proton channel of the osteoclast vacuolar H(+)-ATPase.

Bone resorption by the osteoclast is dependent on acidification of the bone surface by a vacuolar type H+-ATPase (V-ATPase) present in the ruffled membrane of the actively resorbing cell. V-ATPases are a highly conserved family of proton pumps consisting of two functional complexes: a cytoplasmic catalytic sector (VC) and a membrane bound proton channel (VB). Bafilomycin A1, a macrolide antibiotic, is a highly potent inhibitor of V-ATPases, and inhibits bone resorption in vitro in isolated rat calvariae. In order to investigate the binding of bafilomycin to the osteoclast V-ATPase, we used a tritiated bafilomycin which had been prepared by acetylating the 21-hydroxyl group of bafilomycin A1. Osteoclast ruffled membrane vesicles were prepared from purified chicken osteoclasts by differential centrifugation and proton transport in these vesicles was shown to be inhibited by [3H]bafilomycin (IC50 approximately 2 nM). Control membrane vesicles or membrane vesicles partially inhibited with [3H]bafilomycin were solubilized and separated by centrifugation on 15-30% glycerol gradients. V-ATPase activity and reconstitutable proton transport activity could be recovered in high density fractions of the gradient. However, the peak of [3H]bafilomycin radioactivity (>70% of total radioactivity in the gradient) was present in a single peak at lower density. Antibodies against subunits of VC (70, 56 and 40 kDa) reacted only in fractions containing the peak V-ATPase activity whereas an antibody to the 39 kDa subunit of VB reacted both with fractions containing the peak V-ATPase activity but also, and more strongly, in fractions containing the peak [3H]bafilomycin. The fractions in the control gradient corresponding to the peak of [3H]bafilomycin were reconstituted into liposomes and shown to mediate passive bafilomycin A1-inhibitable proton conductance. SDS-PAGE followed by autoradiography, indicated that the bafilomycin was not covalently bound to the V-ATPase or the proton channel. Quantification of VB by [3H]bafilomycin binding or by antibody staining suggested an excess of the free proton channel to that of the intact holoenzyme. A corresponding amount of free catalytic sector could not be found in any fraction throughout the isolation procedure of the V-ATPase from the initial homogenate. Thus, in conclusion, bafilomycin inhibits the V-ATPase by binding tightly but non-covalently to the proton channel region of the V-ATPase which appears to be present in excess over the intact holoenzyme in the osteoclast. The possible role of an excess of the proton channel subcomplex in the osteoclast is discussed.

Omeprazole and bafilomycin, two proton pump inhibitors: differentiation of their effects on gastric, kidney and bone H(+)-translocating ATPases.

The effects of omeprazole and bafilomycin on processes dependent on two different types of H(+)-translocating ATPases were compared. A H(+)-ATPase of the E1E2-type, the H+,K(+)-ATPase, was purified from gastric mucosa. Vacuolar type H(+)-ATPases were prepared both from kidney medulla and from osteoclast-containing medullary bone. H+,K(+)-ATPase-mediated proton transport in gastric vesicles was selectively inhibited by omeprazole with a high potency (inhibitory concentrations greater than or equal to 3 microM) and in time- and pH-dependent manner. This result is consistent with the mechanism of action of omeprazole, which is dependent on acid-induced transformation of the drug into an active inhibitor reacting with luminally accessible sulfhydryl groups of the enzyme. Accordingly, the presence of the membrane-impermeable mercaptane glutathione did not affect the inhibitory action of omeprazole on the H+,K(+)-ATPase. Proton transport in kidney- and bone-derived membrane vesicles was also inhibited by omeprazole, but with a lower potency (inhibitory concentrations greater than or equal to 100 microM). Furthermore, the presence of glutathione totally abolished this inhibition, indicating that cytosolic, rather than luminal, SH-groups of the respective vacuolar H(+)-ATPase were interacting with omeprazole at high concentrations. In line with these results, it was found that omeprazole was much more potent in inhibiting acid production in isolated gastric glands (IC50 approximately 0.25 microM) than in inhibiting osteoclast-mediated 45Ca-release in isolated mouse calvaria (IC50 approximately 200 microM). Bafilomycin, on the other hand, was much more effective in inhibiting proton transport mediated by the vacuolar H(+)-ATPases in the kidney- and bone-derived membrane vesicles (IC50 approximately 2 nM) than in inhibiting H+,K(+)-ATPase-mediated proton transport in gastric membrane vesicles (IC50 approximately 50 microM). Thus, approximately 10(4) times higher concentrations of bafilomycin were needed to inhibit the H+,K(+)-ATPase to the same extent as the vacuolar H(+)-ATPase. A similar difference in potency of bafilomycin was found when its inhibitory effect was determined in isolated mouse calvaria (IC50 approximately 2.5 nM) and in isolated gastric glands (IC50 approximately 5 microM). Hence, omeprazole was found to be a specific inhibitor of the H+,K(+)-ATPase under physiological conditions, i.e. in the presence of glutathione, while bafilomycin was found to be selective towards vacuolar H(+)-ATPases.

Characterization of proton transport in bone-derived membrane vesicles.

ATP-dependent proton transport in membrane vesicles prepared from the medullary bone of egg-laying hens, a source rich in osteoclasts, was characterized. Proton transport was abolished by bafilomycin A1 (10 nM) and N-ethylmalemide (50 microM), but not by oligomycin (15 micrograms/ml), vanadate (100 microM) or SCH 28080 (100 microM), thereby differentiating this H(+)-ATPase from the F1F0- and phosphorylated-type of ATPases. Preincubation of the membrane vesicles at 0 degrees C for 1 h in the presence of KCl (0.3 M) and Mg-ATP (5 mM) resulted in almost complete loss of H(+)-transport activity (cold-inactivation). Preventing the formation of a membrane potential by voltage clamp (Kin+ = Kout+ + valinomycin) increased both the rate of H(+)-transport and the equilibrium delta pH, suggesting an electronic proton transport mechanism. Thus, the H(+)-ATPase in this bone-derived membrane vesicle preparation shows the characteristics of a vacuolar H(+)-ATPase in its inhibitor- and cold-sensitivity and its electrogenic mechanism. The anion sensitivity of the H(+)-ATPase was investigated by varying the intra- and/or extra-vesicular salt composition. The H(+)-ATPase had no absolute requirement for any specific anion, but membrane permeable anions were found to stimulate proton transport activity, presumably by acting as charge compensators for the electrogenic hydrogen ion transport. However, some anions, such as sulfate, acetate and nitrate were directly inhibitory to the ATPase. The results are in agreement with the recently proposed mechanism of osteoclast acidification: a vacuolar H(+)-ATPase working in parallel with a Cl(-)-channel resulting in electroneutral HCl secretion.

Characterization and cellular distribution of the osteoclast ruffled membrane vacuolar H+-ATPase B-subunit using isoform-specific antibodies.

Acidification of the bone surface, leading to bone resorption, is accomplished by a vacuolar-type H+-ATPase present in a specialized domain of the plasma membrane of the osteoclast known as the ruffled membrane. Structure and function appears to be highly conserved within this class of multisubunit enzymes. However, cloning and sequencing of complementary DNA has shown that one of the subunits in the catalytic domain, the B-subunit, exists in at least two forms, B1 and B2. B1 messenger RNA has been found almost exclusively in the kidney, whereas messenger RNA for B2 has been found in all tissues studied, including the kidney. It has been speculated that the B1 isoform might be involved in targeting to the plasma membrane. In the present study, we have characterized the B-subunit of the chicken osteoclast H+-ATPase using antibodies directed against peptides with isoform-specific or conserved sequences of the B-subunit. Western analysis was performed on chicken osteoclast membrane vesicles and on partially purified chicken osteoclast H+-ATPase and was compared with similar analysis of H+-ATPase isolated from bovine kidney and brain. The B1-specific antibody reacted with a polypeptide of approximately 56 kD on immunoblots of the renal H+-ATPase, whereas no reaction could be detected against the osteoclast H+-ATPase or the osteoclast membrane vesicle preparation. In contrast, the antibody against a B2-specific sequence reacted with a peptide of approximately 56 kD on immunoblots of the osteoclast H+-ATPase, the renal H+-ATPase, and the clathrin-coated vesicle H+-ATPase. The antibody against a conserved region of the B-subunit did not generate any evidence for the presence of isoforms other than B2 in the osteoclast. Immunocytochemistry of rat osteoclasts on bovine bone slices using the B2 antibody showed intense polarized staining along the plasma membrane facing the bone surface in actively resorbing osteoclasts whereas nonresorbing osteoclasts were diffusely stained throughout the cytoplasm. By confocal microscopy, the B2 staining was located to the level of the ruffled membrane and appeared to be concentrated to the peripheral areas of the membrane adjacent to the sealing zone. We conclude that the osteoclast vacuolar H+-ATPase contains the B2 isoform and suggest that upon initiation of resorption the pump is translocated from the cell interior to a special domain of the ruffled membrane close to the sealing zone.

Different in vitro and in vivo profiles of substituted 3-aminopropylphosphinate and 3-aminopropyl(methyl)phosphinate GABA(B) receptor agonists as inhibitors of transient lower oesophageal sphincter relaxation.

BACKGROUND AND PURPOSE: Gastro-oesophageal reflux is predominantly caused by transient lower oesophageal sphincter relaxation (TLOSR) and GABA(B) receptor stimulation inhibits TLOSR. Lesogaberan produces fewer CNS side effects than baclofen, which has been attributed to its affinity for the GABA transporter (GAT), the action of which limits stimulation of central GABA(B) receptors. To understand the structure-activity relationship for analogues of lesogaberan (3-aminopropylphosphinic acids), and corresponding 3-aminopropyl(methyl)phosphinic acids, we have compared representatives of these classes in different in vitro and in vivo models. EXPERIMENTAL APPROACH: The compounds were characterized in terms of GABA(B) agonism in vitro. Binding to GATs and cellular uptake was done using rat brain membranes and slices respectively. TLOSR was measured in dogs, and CNS side effects were evaluated as hypothermia in mice and rats. KEY RESULTS: 3-Aminopropylphosphinic acids inhibited TLOSR with a superior therapeutic index compared to 3-aminopropyl(methyl)phosphinic acids. This difference was most likely due to differential GAT-mediated uptake into brain cells of the former but not latter. In agreement, 3-aminopropyl(methyl)phosphinic acids were much more potent in producing hypothermia in rats even when administered i.c.v. CONCLUSIONS AND IMPLICATIONS: An enhanced therapeutic window for 3-aminopropylphosphinic acids compared with 3-aminopropyl(methyl)phosphinic acids with respect to inhibition of TLOSR was observed and is probably mechanistically linked to neural cell uptake of the former but not latter group of compounds. These findings offer a platform for discovery of new GABA(B) receptor agonists for the treatment of reflux disease and other conditions where selective peripheral GABA(B) receptor agonism may afford therapeutic effects.

Osteoclastic acid transport: mechanism and implications for physiological and pharmacological regulation.

In order to solubilize bone mineral and degrade the organic matrix of bone osteoclasts must secrete 1-2 protons for every Ca2+ liberated. This transport is a major metabolic activity of osteoclasts requiring an electrogenic H(+)-ATPase, a conductive chloride channel, a chloride-bicarbonate exchanger, carbonic anhydrase, and functional/morphological polarization of the cell. The osteoclast H(+)-ATPase is electrically coupled to a chloride channel in the ruffled membrane as are similar transport activities found in acidic intracellular vesicles, but the vanadate sensitivity of the osteoclast proton pump is intermediated between that of the E- and v-type proton pumps. The carbonic anhydrase and chloride-bicarbonate exchange provide an interface with pH regulation and integrate bone resorption into systemic acid-base balance. With the molecular mediators of bone resorption being known we may consider the control of bone resorption with an eye to mechanism and specificity that has not previously been possible. The effects of systemic acidosis to increase bone resorption and the effects of carbonic anhydrase deficiency are consistent with our mechanism of osteoclast ion transport.

Characterization of proton transport in bone-derived membrane vesicles.

ATP-dependent proton transport was characterized in membrane vesicles, prepared by differential centrifugation from medullary bone of egg laying hens, a source rich in osteoclasts. The H(+)-ATPase present in this preparation showed the characteristics of a vacuolar H(+)-ATPase in its sensitivity to inhibitors, including bafilomycin A1, its sensitivity to cold treatment and its electrogenic mechanism. There was no evidence for a direct activation of the H(+)-ATPase by anions, including Cl-. These results are consistent with the view that the osteoclast, the cell responsible for bone resorption, secrets acid by means of a vacuolar H(+)-ATPase.

Effects of chloride transport inhibitors on intestinal fluid and ion transport in vivo and in vitro.

The hypothesis that intestinal fluid secretion is driven by Cl- has been tested by investigating the effects of NPPB (5-nitro-2-(3-phenyl-propylamino)-benzoate), a blocker of Cl- channels in nephrons, and the loop diuretic bumetanide, an inhibitor of the Na+, K+, 2Cl(-)-co-transporter. Both NPPB (IC50 (inhibitory concentration) approximately 100 microM) and bumetanide (IC50 approximately 2 microM) inhibited stimulated short-circuit current (Isc) in monolayers of a colonic cell line T84. NPPB also inhibited 36Cl- uptake by these cells, indicating that NPPB acts as a Cl- channel blocker in the T84 cells. NPPB (300 microM) and bumetanide (10 microM) abolished both stimulated Isc and Cl- secretion in isolated rat colonic mucosa. As judged by autoradiography, [3H]NPPB was found both in the crypts and at the surface after exposure of either side to the compound. In line with these results, NPPB and bumetanide reduced stimulated fluid secretion in everted colon sacs from the rat. In the anaesthetized rat model, neither bumetanide nor NPPB affected the net fluid transport. After luminal administration of [3H]NPPB to the rat, radioactivity was found mainly in the villus tips, whereas no labelling was found in the crypts. NPPB was bound to plasma protein (99%), and the inhibitory effects of both NPPB and bumetanide on Isc in T84 cells and fluid secretion in the colonic sacs decreased in the presence of albumin, again indicating that the compounds might not reach the in vivo target, or that the mechanism for fluid secretion in vivo may not be explained solely by the secretion of Cl-.

Isolation and reconstitution of a vacuolar-type proton pump of osteoclast membranes.

A vacuolar-type proton-translocating ATPase was extracted from ruffled membranes of chicken osteoclasts with 1% polyoxyethylene 9-lauryl ether (C12E9) and was purified 13-fold by glycerol gradient centrifugation. The isolated pump appears by sodium dodecyl sulfate-polyacrylamide gel electrophoresis to have a subunit composition similar to that of the clathrin-coated vesicle proton pump, in that subunits of apparent molecular masses of 116, 71, 57, 40, 39, 33, and 17 kDa are present in the osteoclast pump preparation. In addition, the 116-, 71-, 57-, and 40-kDa components were shown to cross-react with specific antisera generated against the homologous subunits of the clathrin-coated vesicle proton pump. The isolated osteoclast H(+)-ATPase was reconstituted into liposomes prepared from purified lipids (phosphatidylcholine, phosphatidylethanolamine, phosphatidylserine, and cholesterol) by a cholate-dilution, freeze-thaw method. Proton transport catalyzed by the reconstituted pump was inhibited by bafilomycin A1 (10 nM) and N-ethylmaleimide (1 mM) but was insensitive to vanadate. We propose that osteoclast-mediated bone resorption is effected by a vacuolar-type proton pump with functional and structural similarities to that isolated from clathrin-coated vesicles.

Properties of three isoforms of the 116-kDa subunit of vacuolar H+-ATPase from a single vertebrate species. Cloning, gene expression and protein characterization of functionally distinct isoforms in Gallus gallus.

Vacuolar H+-ATPases (V-ATPases) are involved in a wide variety of essential cellular processes. An unresolved question is how the cell regulates the activity of these proton pumps and their targeting to distinct cellular compartments. There is growing evidence for the presence of subunit diversity amongst V-pumps, particularly regarding the 116-kDa subunit (called the a subunit). We have cloned and characterized three isoforms (a1, a2 and a3) of this subunit from chicken. The amino-acid sequences of these homologues are approximately 50% similar and their nucleotide differences indicate that they are products of distinct genes. The levels of mRNA expression of these isoforms was quantified by ribonuclease protection analysis. The a1 and a2 isoforms have a similar tissue distribution, with the highest level of mRNA expression in brain, an intermediate level in kidney and relatively low levels in liver and bone. In contrast, the highest level of expression of the a3 isoform is in bone and liver, with a moderate level in kidney, and the lowest level in brain. An antibody against the a1 isoform reacted with a 116 kDa protein in a brain V-ATPase preparation that was not detected in bone or liver V-ATPase preparations, whereas an antibody against the a3 isoform reacted with a 116-kDa peptide in bone and liver, but not brain V-ATPases preparations. The bone and brain V-ATPases showed differential sensitivity to the inhibitors bafilomycin and (2Z,4E)-5-(5,6-dichloro-2-indolyl)-2-methoxy-N-[4-(2, 2,6,6-tetramethyl)piperidinyl]-2,4-pentadienamide. Thus, this work demonstrates the presence of structurally and functionally distinct V-ATPases in a single vertebrate species.

Solid phase synthesis of diamides as potential bone resorption inhibitors.

Unsymmetrical diamide libraries have been prepared by a general and versatile solid phase route, using diacid templates in combination with aromatic and aliphatic amines chosen with the help of statistical experimental design. The compounds were tested as potential inhibitors of osteoclast vacuolar ATPase.

Aminomethyl-2,6-difluorophenols as a novel class of increased lipophilicity GABA(C) receptor antagonists.

3- and 4-(Aminomethyl)-2,6-difuorophenols were tested for activity against the three major classes of GABA receptors. 4-(Amninomethyl)-2,6difluorophenol was shown to be a competitive and somewhat selective antagonist at p1 GABA(C) receptors expressed in Xenopus oocytes (K(B) = 75.5 microM with a 95% Confidence Interval range of 75.2 microM to 75.8 microM). This is the first in a novel class of increased lipophilicity GABA(C) receptor antagonists with little activity at alpha1beta2gamma2 GABA(A) and GABA(B) receptors.

Subunit G of the vacuolar proton pump. Molecular characterization and functional expression.

The vacuolar type proton pump of clathrin-coated vesicles has a multisubunit ATP hydrolytic center that is peripheral to the membrane. Polypeptides present in this domain include the well characterized subunits A, B, C, D, E, and F; SFD, a dimer composed of 50- and 57-kDa polypeptides; and polypeptides termed G and H. Of these, subunits A, B, C, and E have been shown to be necessary but not sufficient for significant ATPase activity; in addition, either polypeptide G or H is also required for ATP hydrolysis (Xie, X.-S. (1996) J. Biol. Chem. 271, 30980-30985). In this study, the polypeptides G and H were purified and directly sequenced. Subsequent molecular analysis has revealed that these proteins are isoforms, which we designate G1 and G2. The cDNAs encoding the rat and bovine brain and chicken osteoclast forms of G1 have been cloned. The open reading frames of the rat and bovine clones encode hydrophilic proteins of 118 amino acids that differ at only five residues; bovine G1 has 36% identity with VMA10, a component of the proton channel of yeast. Northern blot analysis revealed a 1. 0-kilobase pair transcript encoding G1 in bovine brain, kidney, heart, and spleen. The cDNA encoding bovine polypeptide H was cloned and sequenced, revealing this protein to be 64% identical to G1, constituting isoform G2. In Northern blot analysis, a single 1. 7-kilobase pair transcript hybridized with a probe to G2 in brain, but not in heart, kidney, or spleen. An antibody against a bovine G1-specific domain reacts with V pump from bovine brain, kidney, and chromaffin granule, whereas an anti-G2 antibody reacts only with proton pump from brain. The bovine forms of G1 and G2 were subsequently expressed in Escherichia coli and Sf9 cells, respectively, and purified to homogeneity. Reconstitution of ATP hydrolysis was achieved by combination of recombinant subunits A, B, C, and E with either recombinant G1 or G2, demonstrating the role of these isoforms in pump function.