Ca2+ Transport by Reconstituted Synaptosomal ATPase Is Associated with H+ Countertransport and Net Charge Displacement*

The synaptosomal plasma membrane Ca2+-ATPase (PMCA) purified from pig brain was reconstituted with liposomes prepared by reverse phase evaporation at a lipid to protein ratio of 150/1 (w/w). ATP-dependent Ca2+ uptake and H+ ejection by the reconstituted proteoliposomes were demonstrated by following light absorption and fluorescence changes undergone by arsenazo III and 8-hydroxy-1,3,6-pyrene trisulfonate, respectively. Ca2+uptake was increased up to 2–3-fold by the H+ ionophore carbonyl cyanide p-trifluoromethoxyphenylhydrazone, consistent with relief of an inhibitory transmembrane pH gradient (i.e. lumenal alkalinization) generated by H+countertransport. The stoichiometric ratio of Ca2+/H+ countertransport was 1.0/0.6, and the ATP/Ca2+ coupling stoichiometry was 1/1 at 25° C. The electrogenic character of the Ca2+/H+countertransport was demonstrated by measuring light absorption changes undergone by oxonol VI. It was shown that a 20 mV steady state potential (positive on the lumenal side) was formed as a consequence of net charge transfer associated with the 1/1 Ca2+/H+ countertransport. Calmodulin stimulated ATPase activity, Ca2+ uptake, and H+ ejection, demonstrating that these parameters are linked by the same mechanism of PMCA regulation.

Regulation of the intracellular free Ca 2ϩ concentration in nerve terminals is a key factor in the mechanism of synaptic transmission (1,2). Reduction of cytosolic Ca 2ϩ in nerve terminals is dependent on removal by the plasmalemmal Ca 2ϩ -ATPase (PMCA) 1 and sequestration by the intracellular (SERCA) Ca 2ϩ -ATPase (3). Coexistence of both ATPases, as well as the presence of various transport and permeability channels in synaptosomal preparations (4), has rendered difficult a detailed characterization of the synaptosomal PMCA with regards to Ca 2ϩ and H ϩ transport and net charge trans-fer, as done with other Ca 2ϩ transport systems (7)(8)(9)(10)(11)(12)(13).
We purified the PMCA from pig synaptosomes by calmodulin column chromatography (5,6) and obtained proteoliposomal reconstitution by the method originally developed by Rigaud and coworkers (12, 14 -18). The main advantages of this procedure were: (i) optimal protein incorporation and orientation, (ii) very low permeability of the proteoliposomal membrane to electrolytes, and (iii) large intravesicular volume which delays back inhibition by high concentrations of lumenal Ca 2ϩ (8). We then studied the functional properties of the reconstituted synaptosomal PMCA, measuring ATP-dependent Ca 2ϩ -uptake activity, H ϩ ejection, and formations of transmembrane electrical potential.

MATERIALS AND METHODS
Egg phosphatidylcholine (EPC) and egg phosphatidic acid (EPA) were obtained from Avanti Polar Lipids Inc. Calmodulin, calmodulinagarose, C 12 E 8 , FCCP, valinomycin and A23187 were obtained from Sigma. Oxonol VI and 8-hydroxy-1,3,6-pyrene trisulfonate (pyranine) were from Molecular Probes. Bio-Beads SM-2 and the anion exchange chromatography resin AG1X8 were purchased from Bio-Rad. All other reagents were of the highest purity available.
Purification of Plasma Membrane Ca 2ϩ -ATPase-The pig brain plasma membrane Ca 2ϩ -ATPase was purified as described in Salvador and Mata (5). Briefly, fresh pig brain was homogenized in 10 volumes of 10 mM Hepes/KOH, pH 7.4, 0.32 M sucrose, 0.5 mM MgSO 4 , 0.1 mM phenylmethylsulfonyl fluoride and 2 mM 2-mercaptoethanol. After two centrifugations at low and high speed, respectively, the pellet was layered in a discontinuous 20 -40% (w/v) sucrose gradient and centrifuged at 63,000 ϫ g. The synaptosomes obtained in the interface were lysed in a hypotonic medium containing 1 mM EDTA and centrifugated at 20,000 ϫ g. The synaptic plasma membrane vesicles were obtained in the pellet and solubilized in a medium containing 20 mM Hepes/KOH, pH 7.4, 130 mM KCl, 0.5 mM MgCl 2 , 50 M CaCl 2 , 15% glycerol, 2 mM 2-mercaptoethanol, and 0.6% (w/v) Triton X-100. The solubilized fraction was applied onto a calmodulin-agarose column previously equilibrated with the same buffer except that 1 mM MgCl 2 and 100 M CaCl 2 was used. Finally the Ca 2ϩ -ATPase was eluted in the above buffer containing 0.06% (w/v) Triton X-100 and 2 mM EDTA instead of calcium. The samples were stored at Ϫ80°C until use.
Preparation of Liposomes and Reconstitution of the Ca 2ϩ -ATPase-EPC/EPA (9/1, w/w) liposomes were prepared by reverse phase evaporation as described previously (19) in a buffer containing 20 mM Pipes-K ϩ , pH 7.1, 130 mM KCl (Pipes buffer, pH 7.1). After sequential extrusion through 0.8, 0.4, and 0.2 m nucleopore filters, the liposomes were resuspended at 4 mg/ml in Pipes buffer, pH 7.1, and solubilized by 8 mg/ml C 12 E 8 . The purified Ca 2ϩ -ATPase (0.11 mg/ml) was added to the liposomes at a lipid/protein ratio of 150/1 (w/w) and incubated for 2 min. The detergent was then removed in 3 h by subsequent additions of Bio-Beads SM-2 (80 mg/ml/h) under continuous stirring. After complete detergent removal, the proteoliposomes were pipetted off. When measurements of H ϩ ejection were done, the fluorescent pH indicator pyranine (200 M) was added in the incubation medium before Bio-Beads addition. After detergent removal, the external pyranine was removed by passing the proteoliposomes through an anion exchange chromatog-* This work was supported in part by Grant PB95-0144 from Dirección General de Investigación Científica y Técnica and Grant EIA94 -29 from Junta de Extremadura, Spain (to A. M. M) and by National Institutes of Health Grant P01-HL-27867 (to G. I.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Ca 2ϩ -ATPase Activity-The enzymatic activity was measured at 25°C by using a coupled enzyme assay following spectrophotometrically the NADH absorption change at 340 nm. The proteoliposomal protein (3 g/ml) was incubated in 2 ml of a medium containing Pipes buffer, pH 7.1 plus 5 mM MgCl 2 , 10 M CaCl 2 , 0.22 mM NADH, 0.42 mM phosphoenolpyruvate, 10 units of pyruvate kinase, and 28 units of lactate dehydrogenase. Additions of 0.2 mM ATP or 1 mM ATP were done to assay the activity of the reconstituted or solubilized protein, respectively.
Ca 2ϩ Uptake-Ca 2ϩ uptake by reconstituted proteoliposomes was measured by following the differential absorption of the metallochromic indicator arsenazo III (50 M) as described previously (20). A 687-660-nm wavelength pair was used in a dual wavelength spectrophotometer (DW-2000, SLM-Aminco) to monitor changes in external Ca 2ϩ concentration. The proteoliposomal protein (6 g) was incubated at 25°C in 2 ml of a medium containing Pipes buffer, pH 7.1, plus 0.42 g/ml calmodulin, 5 mM MgCl 2 , and 50 M arsenazo III. The absorbance signal uptake was calibrated by the addition of 10 M final Ca 2ϩ , and the reaction was initiated by adding 0.2 mM ATP.
ATP-dependent H ϩ Ejection-Lumenal H ϩ ejection of the proteoliposomes was measured by following the changes in the fluorescence intensity of the pH indicator pyranine that is entrapped during the reconstitution procedure (9). Fluorescence intensity was measured using a exc 460 nm and em 510 nm in a spectrofluorometer (Jasco FP-777). Calibration of lumenal pH changes was done by subsequent additions of 0.5 mN H 2 SO 4 to 2 ml of medium containing 6 g of proteoliposomal protein in Pipes buffer, pH 7.1, plus 5 mM MgCl 2 , 10 M CaCl 2 , and 4 M FCCP.
The stoichiometry of lumenal H ϩ was done by taking into account the lumenal volume, as described in Levy et al., 1992 (17). The internal volume of the proteoliposomes (400 l/mg of protein) was taken from Hao et al. (12).
Electrical Potential Assay-The transmembrane electrical potential was measured by following the differential absorption (625-603 nm) of 2 M oxonol VI, using dual wavelength spectrophotometry (12). The reaction mixture was as described for Ca 2ϩ uptake except that oxonol VI was used instead of arsenazo III. The calibration was done by several additions of 130 mM KCl to the medium in the presence of 1 M of the K ϩ ionophore valinomycin and in the absence of ATP. Absorption changes were standardized with electrical potential by using the Nernst equation.
Protein Determination-The protein content was measured by the Bradford method (21). The interference of Triton X-100 was avoided by a previous treatment of the samples with Bio-Beads SM-2.
SDS-Polyacrylamide Gel Electrophoresis-Electrophoresis was performed by the Laemmli procedure (22) in a 7.5% polyacrylamide gel.

Purification and Reconstitution of Synaptosomal PMCA-
The pig brain PMCA was purified from solubilized synaptic vesicles by calmodulin affinity chromatography. Electrophoretic analysis of the synaptic vesicles (Fig. 1, lane 1) demonstrates the presence of several protein bands in addition to the PMCA which is hardly visible because of its very low quantity. In contrast, the purified protein eluted from the calmodulin column appears as a single band of apparent 140-kDa molecular weight (Fig. 1, lane 2). We found that the purified enzyme had an ATPase activity of 1.10 Ϯ 0.10 IU in the presence of pure EPC, as compared with 0.063 Ϯ 0.011 IU in the solubilized synaptosomal vesicles.
ATP-dependent Ca 2ϩ Uptake-The pig brain purified PMCA was reconstituted into EPC/EPA liposomes at a lipid/protein ratio of 150/1 (w/w). The reconstitution procedure was performed by detergent removal from lipid-protein-detergent micellar solutions using Bio-Beads SM2 as detergent removing agent. The reconstitution procedure and Ca 2ϩ transport activity were optimized by using different detergents and ionic conditions. The highest ATP-dependent Ca 2ϩ uptake was obtained when C 12 E 8 was used as a detergent in the reconstitution procedure, and Cl Ϫ (rather than SO 4 2or isothiocyanate) was used as the prevailing anion in the reaction medium. Similar findings were reported by Hao et al. (12) for reconstitution of the erythrocyte PMCA.
The reconstituted PMCA sustained high rates of ATP-dependent Ca 2ϩ uptake (Fig. 2), which increase approximately 2-fold upon addition of calmodulin (Table I). In the presence of calmodulin and in the absence of FCCP ( Fig. 2A), Ca 2ϩ transport into the lumen of the vesicles occurred with an initial velocity of 0.40 Ϯ 0.01 mol of Ca 2ϩ /mg of protein⅐min Ϫ1 . This value was calculated 30 s after triggering the reaction with ATP. Asymptotic levels of about 1 mol of Ca 2ϩ /mg of protein were reached in a few minutes, at which point addition of FCCP stimulated Ca 2ϩ uptake activity to yield up to 4-fold Ca 2ϩ accumulation. The activity was not affected by valinomycin (not shown) and was completely inhibited by 50 M vanadate. Addition of the Ca 2ϩ ionophore A23187 produced immediate release of the Ca 2ϩ accumulated into the vesicles. When FCCP was present in the reaction medium before ATP addition (Fig. 2B), Ca 2ϩ uptake reached Ca 2ϩ accumulation levels of 3.25 Ϯ 0.2 mol Ca 2ϩ /mg of protein in 10 min with no need for further additions. The initial rates of Ca 2ϩ uptake determined 30 s after ATP addition, however, were not affected by FCCP. These observations suggest that a transmembrane H ϩ gradient is formed during Ca 2ϩ accumulation by pig brain PMCA proteoliposomes and that FCCP collapses the H ϩ gradient and relieves its inhibition of Ca 2ϩ transport.
ATPase Activity-We have studied under similar experimen- tal conditions the effects of calmodulin, H ϩ -and Ca 2ϩ -ionophores, and vanadate upon the Ca 2ϩ -ATPase hydrolytic activity of the reconstituted enzyme (Fig. 3). In the presence of 0.2 mM ATP, we observed hydrolytic rates of 0.21 Ϯ 0.02 mol of P i /mg of protein⅐min Ϫ1 . This activity was stimulated by addition of calmodulin, FCCP, and A23187. The initial rate was increased 2-fold by calmodulin (Table I). FCCP did not affect the initial rate, but increased the ATPase activity only after an initial period of activity, i.e. only when a pH gradient was present. The addition of vanadate totally inhibited the ATPase activity. A comparison of the initial rates of Ca 2ϩ transport and ATPase activity yields a coupling stoichiometry of 1/1 at 25°C, independently of the presence of calmodulin and/or FCCP.
ATP-dependent H ϩ Ejection-H ϩ ejection from the lumen of the proteoliposomes, associated with ATP-dependent Ca 2ϩ uptake was demonstrated directly by measuring changes in lumenal pH. This was accomplished by monitoring the fluorescence intensity of the pH indicator pyranine, which was trapped in the lumen of the proteoliposomes during the reconstitution procedure. Upon ATP addition (Fig. 4A), the initial rate of H ϩ extrusion calculated after 30 s was approximately 0.1 mol of H ϩ /mg of protein⅐min Ϫ1 . Analogous to Ca 2ϩ uptake, H ϩ ejection was stimulated by calmodulin (see Table I and Fig.  4B) to a rate of approximately 0.26 Ϯ 0.05 mol of H ϩ /mg of protein⅐min Ϫ1 , reaching in 5 min a maximal amount of 0.36 Ϯ 0.04 mol of H ϩ ejected/mg of protein. It is noteworthy that although H ϩ ejection was not significantly affected by valino-mycin in the presence of K ϩ , FCCP decreased the fluorescence signal to the original level. Therefore, H ϩ ejection and formation of pH gradient are not driven by any electrical gradient, but are primary events linked to ATP-dependent Ca 2ϩ uptake in the form of a Ca 2ϩ /H ϩ countertransport. Our measurements indicate that the coupling stoichiometry of Ca 2ϩ uptake and H ϩ ejection is 1.0/0.6.
Electrical Potential-The net charge displacement associated with the observed stoichiometry of Ca 2ϩ /H ϩ countertransport suggests that a transmembrane electrical potential is generated across the proteoliposomal membrane as a consequence of this process. In fact, we observed the development of a transmembrane electrical potential by measuring light absorption undergone by oxonol VI. It is shown in Fig. 5 that activation of the Ca 2ϩ pump by addition of ATP under optimal conditions produces a steady-state transmembrane potential of 19 Ϯ 2.1 mV. Addition of valinomycin, a K ϩ ionophore, collapses the electrical potential by rendering the membrane permeable to K ϩ and thereby compensating for the uneven charge displacement.

DISCUSSION
The aim of this study was a characterization of Ca 2ϩ transport by the plasma membrane Ca 2ϩ -ATPase from pig brain in terms of countertransport and electrogenic properties. The study of ion fluxes was rendered possible by the use of PMCA reconstituted in proteoliposomes with a low electrolyte perme- ability. The reconstituted synaptosomal PMCA sustains a Ca 2ϩ transport rate up to 0.4 mol of Ca 2ϩ /mg of protein⅐min Ϫ1 , similar to those obtained by Niggli et al. (13) and Hao et al. (12) with reconstituted erythrocyte PMCA. In addition, the low lipid/protein ratio used for reconstitution has provided a large lumenal volume per ATPase molecule, thereby permitting accumulation of large amounts of Ca 2ϩ and relatively long experimental times before establishment of an inhibitory Ca 2ϩ concentration in the lumen of the proteoliposomes. The 1/1 stoichiometric ratio of Ca 2ϩ uptake and ATP hydrolysis obtained by measurements of initial rates (Table I) is in the same range as that reported with the reconstituted erythrocyte PMCA (12,13), but lower than the ratio of 2/1 obtained with SERCA (9). This difference may be related to intrinsic characteristics of the two families of Ca 2ϩ pumps (23)(24)(25).
The use of the proteoliposomal reconstituted system turned out to be very advantageous in demonstrating that the synaptosomal PMCA pump operates Ca 2ϩ /H ϩ countertransport. Collapse of the related H ϩ gradient by the H ϩ ionophore FCCP, and not by the K ϩ ionophore valinomycin, is consistent with direct Ca 2ϩ /H ϩ countertransport through the pump, rather than H ϩ extrusion secondary to electrical potential. The Ca 2ϩ /H ϩ stoichiometric ratio of 0.6 found in the synaptosomal plasma membrane Ca 2ϩ -ATPase, is similar to the ratios reported for the other Ca 2ϩ -ATPases. For example, Hao et al. (12) obtained a stoichiometric Ca 2ϩ /H ϩ ratio of 1 with the reconstituted erythrocyte PMCA at 12°C temperature but a lower ratio at 30°C. Although the Ca 2ϩ /H ϩ ratio may be primarily determined by the stoichiometry and pKs of acidic residues operating the Ca 2ϩ /H ϩ exchange (18), it is possible that higher temperatures and/or other experimental variable may produce H ϩ slippage (26). The uneven charge exchange produced by the observed Ca 2ϩ /H ϩ ratios is consistent with the development of transmembrane electrical potential following addition of ATP (Fig. 5).
The Ca 2ϩ /H ϩ countertransport and the net charge transfer demonstrated with purified and reconstituted brain PMCA will be useful in understanding Ca 2ϩ regulation in synaptosomal function. Furthermore, demonstration of Ca 2ϩ /H ϩ countertransport in several Ca 2ϩ transport ATPases indicates that a shift in the affinity of protein carboxylic functions for H ϩ and Ca 2ϩ occurs upon formation and cleavage of phosphorylated enzyme intermediate and that this is a common mechanistic feature in the coupling of ATP utilization and Ca 2ϩ transport.