Ca2+/H+ countertransport and electrogenicity in proteoliposomes containing erythrocyte plasma membrane Ca-ATPase and exogenous lipids.

A reconstituted proteoliposomal system was obtained with Ca-ATPase purified from human erythrocyte membrane (plasma membrane, PM ATPase), and liposomes prepared by reverse-phase evaporation. The reconstituted PM ATPase behaved as an electrogenic Ca2+/H+ exchanger and, under optimal conditions, utilization of 1 mol of ATP was accompanied by uptake of one Ca2+ by the vesicles, and ejection of one H+ from the lumen of the vesicles. Ca2+ uptake was greatly (5-fold) stimulated by the addition of calmodulin, and by collapsing the H+ gradient with the ionophore carbonyl cyanide p-trifluoromethoxyphenylhydrazone. In the presence of calmodulin and p-trifluoromethoxyphenylhydrazone, the reconstituted system sustained transport rates of 1.00 +/- 0.12 mumol of Ca2+/mg of protein min-1 (30 degrees C), reaching asymptotic levels of 8.05 +/- 0.41 mumol of Ca2+/mg of protein (i.e. 20 mM lumenal Ca2+). The corresponding net charge transfer produced a maximal electrical gradient of 40.5 +/- 1.8 mV at steady state. Demonstration of the electrogenic behavior of the PM ATPase, obtained for the first time with these experiments, was critically dependent on the detergent used in the reconstitution procedure. The lumenal pH rise had a much greater rate-limiting effect on the pump, than the electrical potential developed by the pump.

* 1.8 mV at steady state. Demonstration of the electrogenic behavior of the PM ATPase, obtained for the first time with these experiments, was critically dependent on the detergent used in the reconstitution procedure. The lumenal pH rise had a much greater rate-limiting effect on the pump, than the electrical potential developed by the pump.
The outward directed Ca2+ pump plays an important role in the long-term maintenance of a steep Ca2+ gradient across the plasma membrane (PM).' Mechanism and regulation of the P M Ca2+ pump have been reviewed in detail (Carafoli, 1991(Carafoli, , 1992. However, disagreement is still found in the literature regarding countertransport and/or electrogenic properties of the pump. There is considerable evidence to support the existence of Ca2+/H+ countertransport. Concerning the H+/Ca2+ stoichiometry, however, a ratio of 1 was observed with erythrocyte membrane resealed inside-out vesicles, while a ratio of 2 was obtained with purified ATPase in reconstituted proteoliposomes (Niggli et al., 1982;Smallwood et al., 1983;Gassner et al., 1988).
A H+/Ca2+ ratio of 1 or 2 would determine whether the pump is or is not electrogenic. In this regard, experiments with inside-out vesicles (Waisman et al., 1981;Rossi et al., 1982; * This work was supported by the National Institutes of Health and the American Heart Association. The costs of publication of this article therefore be hereby marked "aduertisement" in accordance with 18 were defrayed in part by the payment of page charges. This article must U.S.C. Section 1734 solely to indicate this fact.
The abbreviations used are: PM, plasma membrane; FCCP, carbonyl panesulfonic acid; PIPES, 1,4-piperazinediethanesulfonic acid; SR, sar-cyanide p-trifluoromethoxyphenylhydrazone; MOPS, 4-morpholineproticulum Ca2+ATPases. coplasmic reticulum; CaM, calmodulin; SERCA, sarcs-endoplasmic re- Smallwood et al., 1983) provided evidence for an electrogenic behavior of the pump, consistent with a H+/Ca2+ ratio lower than 2. On the contrary, no development of electrical potential was observed using proteoliposomes reconstituted with purified ATPase (Niggli et al., 19821, suggesting that the PM Ca2+ pump operates as an electroneutral H+/Ca2+ exchanger, with a 2:l stoichiometric ratio. Further disagreement is found in the literature on whether the kinetics of the plasma membrane Ca2+ pump are sensitive to H' and/or electrical gradients. Using purified ATPase i n reconstituted proteoliposomes, Niggli et al. (1982) obtained evidence for inhibition of the Ca2+ pump by H' gradients, but not by electrical potentials generated by Na+ or K' gradients. On the one hand, Gassner et al. (1988) reported that Ca2+ extrusion from intact red cells is affected by electrical potentials. Furthermore, Smallwood et al. (1983) noted that Ca2+ accumulation into inside-out vesicles was enhanced by the diffusion of external anions through the band I11 protein.
We describe here the successful reconstitution of purified human erythrocyte PM Ca2+-ATPase with liposomes prepared by reverse-phase evaporation. By the choice of a favorable detergent for the reconstitution procedure, we were then able to obtain parallel measurements of ATP-dependent Ca2+ uptake, H+ ejection, and development of electrical potential, with the same experimental system. We found that the PM operates as an electrogenic Ca2+/H+ exchanger.
MATERIALS AND METHODS Chemicals-Octaethylene glycol-n-dodecyl ether (C,,E,) was obtained from Nikko Chemical Co. (Tokyo, Japan). Purified egg yolk phosphatidylcholine and phosphatidic acid were obtained from Avanti Polar Lipids Inc. Calmodulin-Sepharose 4B, camodulin (CaM), valinomycin, carbonyl cyanidep-trifluoromethoxyphenylhydrazone (FCCP), calcimycin (A231871, and all other chemicals were obtained from Sigma. Preparation of Human Red Cell Membrane-Human red cell membrane ghosts were prepared from outdated human erythrocytes obtained from the blood bank as described by Niggli et al. (1979), except for the use of 10 m M MOPS-K+ instead of Tris-HC1 in the last 4 washings. The entire procedure was performed at 4 "C.
Purification of Erythrocyte Membranes Ca2+-ATPase-The Ca2+-ATPase was purified from red cell membrane ghosts by calmodulin affkity chromatography, an adaptation of the method described by Niggli et al. (1981) and Kosk-Kosicka et al. (1986). Briefly, 300-500 mg of red cell membrane deficient in calmodulin were solubilized with a buffer containing 0.4% C,,E, (w/v), 130 m M KC1, 20 m M PIPES-K', pH 7.2,lOO PM CaCl,, 1 m M MgCl,, 2 m M dithiothreitol, and 0.3 M sucrose to obtain a final concentration of 6-8 mg of proteidml, and kept on ice for 10 min. Nonsolubilized material was removed by centrifugation at 40,000 x g for 40 min. Phosphatidylcholine was added to the supernatant to a concentration of 0.5 mg/ml. After sonicating l min on ice with a micro-ultrasonic cell disrupter (Kontes, model 1440-4000) at power setting 2-3 and tone 3-4, the supernatant was applied to a calmodulin-Sepharose 4B column (1.5 x 10 cm) pre-equilibrated with 0.05% C,,E,, 130 m M KCl, 20 m M PIPES-K', pH 7.2,100 VM CaCl,, 1 m M MgCl,, 2 m M dithiothreitol, 0.3 M sucrose, and 0.05% (w/v) phosphatidylcholine. The column was washed with several bed volumes of the same buffer until the elution medium did not contain anymore protein (monitored by 280 nm absorption). Then the column was further eluted with the same buffer, containing 1 m~ EDTA-K+ instead of CaCl,. The Ca2+-ATPase activity peak eluted by EDTA-K+ was collected, 50 p~ CaCl, and 2 m M MgCI, were added, and the pooled enzyme was stored a t -80 "C for use within 1 month. The purification procedure was carried out a t 4-5 "C.
Preparation ofliposomes-Unilamellar liposomes were prepared by reverse-phase evaporation by the method described by Rigaud et al. (1983) and . For these experiments, the aqueous medium contained 20 mM PIPES-K+, pH 7.1, and 130 m M KCI.
Reconstitution ofCa2+-ATPase with Liposomes-For this purpose the liposomes stock (16 mg of lipid/ml) was diluted with the aqueous medium used in the preparation of the liposomes, to yield a final lipid concentration (including the enzyme added later) of 4 mg/ml. The liposomes were then solubilized by adding incremental amounts of C,,E8, while stirring slowly, until solubilization of lipids resulted in a dramatic decrease of turbidity. At this time, the C,,E&pid molar ratio was approximately 2. If measurements of lumenal alkalinization were planned, the fluorescent pH indicator pyranine was added a t this time, to reach a 200 p~ concentration. Purified plasma membrane Ca2+-ATPase (120-180 pg of proteidml) presolubilized by the addition of 0.5 mg/ml C,,E8 was then added to the liposomes to obtain a lipid/protein ratio (w/w) of 150. The detergent/protein/phospholipid mixture was kept for 1 min under gentle stirring, then the detergent was removed by adding 0.6 g/ml prewashed wet Bio-Bead SM-2. The mixture was gently stirred a t room temperature for 1.5 h, and the Bio-Bead was removed by passing the mixture through empty poly-prep columns (Bio-Rad). Finally, the proteoliposomes were passed through an anion exchange chromatography column (AG 1-X8, Bio-Rad) to eliminate the pyranine in the medium outside the proteoliposome. One-ml bed volume per 4 ml of proteoliposome suspension was sufficient for this purpose. Preparation of liposomes and proteoliposomes was carried out a t room temperature.
Functional Measurements of Proteoliposomes-Ca2+ uptake, electrical potential, and pH changes across the membrane of proteoliposomes were measured by slight modifications of the methods described by Yu et al. (1993). Ca2+ uptake was followed by monitoring the differential (660 uersus 687 nm) absorption changes undergone by Arsenazo I11 (Scarpa, 1979), which were linearly proportional to changes of the Ca2+ concentration in the external medium. In most cases the reaction mixture contained 130 mM KCI, 20 mM PIPES-K', pH 7.2,5 m M MgCI,, 5 p~ CaM, 50 p~ Arsenazo 111, and 2.5-3.0 pg of proteoliposomal proteidml. Additional CaCI, (5 p~ increments) was added after obtaining the absorption base line in order to standardize the absorption changes. Then the reaction was started by the addition of 0.2 m M ATP. Ejection of H' was measured by monitoring the fluorescence intensity of pyranine entrapped within the lumen of proteoliposomes, using A,,, 460 nm and A,, 510 nm. Standardization of the lumenal pH changes was obtained by adding small amounts of 0.2 N solution to reaction media containing proteoliposomes, valinomycin (100 nm), and FCCP (100 n~), and measuring in parallel fluorescence (lumenal pyranine) and pH (electrode) changes in the equilibration medium. The stoichiometry of lumenal H' producing fluorescence changes was estimated on the basis of the lumenal volume of the proteoliposomes (Levy et al., 1992). The reaction mixture for pH change measurement was the same as that for Ca2+ uptake measurement, except for the omission of Arsenazo 111.
The internal volume of the proteoliposomes was determined by reconstituting the proteoliposomes in the presence of a membrane impermeable radioactive tracer ("C-inulin carboxyl), washing out the tracer outside the vesicles by column chromatography (Sephadex G-150), and measuring the radioactive tracer trapped in the lumen of the vesicles. The volume of 417 pl/mg protein obtained by this method matched closely the volume (370 pVmg protein) calculated on the basis of the vesicle's dimensions determined by electron microscopy.
The development of transmembrane electrical potential was monitored by measuring the differential absorption of oxonol VI in a dual wavelength (603 uersus 625 nm) spectrophotometer. The reaction mixture was the same as that for Ca2+ uptake, except for the presence of 1 All functional measurements were carried out a t 30 "C, unless otherwise specified in the Figures. The volume of the reaction mixture in each measurement was 1 ml and the reaction was started by adding of 0.2 mM ATP after obtaining a steady base line.

RESULTS
Characterization of the Purified ATPase-The extent of ATPase purification obtained in our experiments is shown in Fig. 1. In the whole membrane electrophoretic pattern the ATPase is hardly visible, while the purified preparation reveals one strong band migrating with an apparent molecular mass of 130 kDa. From the functional standpoint, the purified enzyme displays approximately a 150-fold increase in ATPase activity, as compared to the erythrocyte ghost (Table I).
ATP-dependent Ca2+ Pansport-Reconstitution of proteoliposomes with purified ATPase and exogenous lipids yields vesicles which are able to sustain a high Ca2+ uptake activity, as shown in Fig. 2. This activity is stimulated by CaM and the H+ ionophore FCCP, and is inhibited by vanadate. Addition of the Ca2+ ionophore A23187 causes sudden release of the Ca2+ taken up by the vesicles, demonstrating that the ATP-dependent uptake does in fact produce a transmembrane Ca2+ gradient which collapses if the membrane is rendered passively permeable to Ca2+. We estimate that lumenal Ca2+ concentrations as high as 20 rn were reached as a consequence of transport activity.
As for the coupling stoichiometry of Ca2+ uptake and ATPase activity, we find that under conditions of optimal stimulation (FCCP and calmodulin) the enzyme operates with a ratio of 1. The stoichiometric ratio of 1 is demonstrated better at 12 "C than at 30 "C, since at the higher temperature the rate of Ca2+ transport begins to decline after a few minutes, thereby yielding stoichiometric ratios lower than 1 (Fig. 3).
ATP-dependent R Ejection-In addition to ATP-dependent Ca2+ uptake, H+ ejection from the lumen of the vesicles can be demonstrated by monitoring the fluorescence intensity of the pH indicator pyranine which is trapped in the lumen of the proteoliposomes during their reconstitution. Lumenal measurements, as compared to measurements in the outside medium, are much more sensitive to H+ ejection, and are not affected by H+ production by the hydrolytic ATPase reaction.
In analogy to Ca2+ uptake, H+ ejection is also stimulated by  calmodulin. If we perform our measurements at 12 "C, H+ ejection and Ca2+ uptake proceed initially with a stoichiometric ratio of approximately 1, which then declines after 3-4 min (Fig. 4A). On the other hand, if we measure the two activities at 30 "C, the Ca2+/H+ stoichiometric ratio is lower than 1 (Fig.  4B). Since we found in parallel experiments that the passive permeability of the proteoliposomes to H+ is quite low in the absence of ATP, the variable H+/Ca2+ countertransport ratio suggests a temperature-dependent slippage of H' during the operation of the pump. Under comparable conditions, slippage of the pump appears to effect H+ more than Ca2+ (compare Figs.

and 4).
Notwithstanding the variable Ca2+/H+ ratio, a maximal (net) H+ ejection of approximately 2 pmoVmg protein was obtained a t steady state (Fig. 43). Based on a lumenal volume of 400 pVmg protein, this corresponds to a 5 mM drop of lumenal concentration, including free H+ and H' released by buffers. The actual pH rise is 0.35 units, as determined by measurements with a pH electrode during the standardization procedures (see "Materials and Methods").
It is noteworthy that H' ejection is not affected significantly by valinomycin in the presence of K+, but is totally reversed by FCCP (Fig. 5). Therefore, H' ejection is not driven by any electrical gradient, but is a primary event linked to ATP-dependent Ca2+ uptake in the form of countertransport.
Establishment of Electrical Potential-It can be shown by the use of the indicator oxonol VI that ATP-dependent Ca2+ uptake generates an electrical potential across the proteoliposomal membrane, reaching steady state levels of 13-15 mV (Fig. 6). Establishment of the electrical potential is favored by the addition of FCCP. In this case the steady state potential increase up to 30-40 mV (Fig. 6), likely due to collapse of the H' gradient and removal of its contribution to charge neutralization, and also due to stimulation of Ca2+ transport activity (Fig. 2). In the absence of FCCP and CaM, the initial rates of electrical potential formation are 1.8 2 0.1 and 5.1 2 0.3 mV min" at 12 "C and 30 "C, respectively (Fig. 7). In the presence of FCCP and CaM the initial rate rises to 22.1 2 1.6 mV min" a t 30 "C (Figs. 6 and 7).
Regulation of Dunsport and ATPase Activities-It is shown in Fig. 8 that a marked stimulation of Ca2+ uptake, ATPase activity, and electrical potential development is obtained by the addition of CaM. FCCP was added in these experiments to prevent inhibition by lumenal alkalinization, thereby allowing occurrence of the CaM effect without interference. Comparable stimulation of Ca2+ uptake, ATPase activity, H+ ejection, and charge transfer demonstrates that these four parameters are subject to the same regulation and, therefore, linked by the same mechanism. It is noteworthy that stimulation by CaM, which is quite evident with native erythrocyte membrane (Schatzmann, 1993;Gopinath and Vincenzi, 1977;Penniston, 1977, 1987), is much more modest in the purified ATPase preparation (Kosk-Kosicka and Bzdega, 1988). On the other hand, we find that following its reconstitution in proteoliposomes, the enzyme regains its property of being greatly (5-fold) stimulated by CaM (Table I).  30 "C ( B ) . The reaction mixture for Ca2+ uptake was the same as that described for Fig. 2, in the presence of 5 pg of C W m l and 4 p~ FCCP. The reaction mixture for ATP hydrolysis was same as that used for Ca2+ uptake, but for the omission of Arsenazo 111 and the presence of 25 units of pyruvate kinase, 25 units of lactic dehydrogenase, 2 m M phosohoenolpyruvate, 150 M NADH. The reaction was started by the addition of 2 mM ATP.
In the presence of CaM, collapse of the electrical gradient by valinomycin does not affect the rate of Ca2+ uptake (Fig. 9A). On the contrary, collapse of the H' gradient by FCCP causes a pronounced stimulation of Ca2+ uptake (Fig. 9B). The stimulation is greater if FCCP is added during the steady state when the H+ gradient is maximal (Fig. 9B), as compared to the beginning of the reaction when the H+ gradient is minimal (Table I).
These findings indicate that while a rise of lumenal pH inhibits Ca2+ uptake (see also Fig. 2), the electrical potential developed under our condition produces only a minor inhibition if any. This latter conclusion is further substantiated by experiments in which 65, 130, or 260 mM KC1 was added to proteoliposomes formed in the presence of 65 mM KC1 in the lumenal medium. It is shown in Fig. 10 that in the absence of valinomycin the KC1 additions had no effect. In the presence of valinomycin, addition of 130 or 260 RIM KC1 is expected to generate a 18or 36-mV diffusion potential, which is in the range of the potential produced by the pump. We found only a minor inhibition of the rate of transport by the 36 mV diffusion potential (curve f i n Fig. lo), while the 18-mV potential had no significant effect.
It is noteworthy that the ATPase activity is subject to a regulatory pattern which is consistent with its linkage to Ca2+ transport. In addition to the effect of CaM (Table I), the consistent pattern of regulation includes ionophores and inhibitors. It is shown in Fig. 11 that collapse of the electrical potential by valinomycin does not change the ATPase rate, while FIG. 7. Initial rates of electrical potential development at different temperatures. The reaction mixture was the same as that in Fig. 6, without valinomycin and FCCP. The measurement was carried out at 30 "C ( a ) and 12 "C ( b ) , respectively. collapse of the H' gradient by FCCP increases the ATPase activity. A further increase is produced by collapsing the Ca2+ gradient with A23187, thereby relieving the inhibition by high lumenal Ca2+ and promoting uncoupled ATPase turnover. Fi-nally, it is shown in Fig. ll that the PM ATPase is inhibited by vanadate which inhibits also Ca2+ uptake (Fig. 2).

DISCUSSION
As previously observed with the sarcoplasmic reticulum Ca2+-ATPase (Levy et al., 1990;, reconstitution of proteoliposomes with low protein to lipid ratios has the advantage of providing a large lumenal volume per enzyme unit. This allows transport of large amounts of Ca2+ through repeated enzyme cycles over a relatively long time of experimental observation. Considering the 1:150 (w:w) protein to lipid ratio used in our experiments with the PM Ca2+-ATPase, we estimate that no more than 2 or 3 molecules of enzyme were reconstituted in each proteoliposome. This estimate was confirmed by electron microscopic observations on frozen and fractured proteoliposomes.
With respect to the transport properties of the proteoliposomes, we obtained our best results when we used a 10% phosphatidylcholine and phosphatidic acid (1O:l) mixture for the liposomes, and C,,E, as the detergent for the reconstitution procedure. The transport rates of 1.00 2 0.12 pmol of Ca2+/mg of protein min-l observed in our experiments with reconstituted PM ATPase are in the same range as those obtained by  with reconstituted SR ATPase, and by Niggli et al. (1981) with PM ATPase.
The Ca2+/ATP stoichiometric ratio of 1 observed in our experiments is consistent with that reported by Niggli et al. (1981) for the PM ATPase, but is lower than the ratio of 2 sustained by the SR Ca2+-ATPase under optimal conditions. In fact, the Ca2+lATP ratio tends to decrease after a few minutes of reaction, possibly due to "slippage" of the pump when the vesicular Ca2+ rises, as also observed with the SR ATPase (Inesi and de Meis, 1989). It is difficult to demonstrate unambiguously whether a ratio lower than 2 is due to detergent exposure during purification, or is rather related to intrinsic features of the PM ATPase which may differ from those of the SR ATPase.
It should be pointed out that sequence alignment of the PM Ca2+-ATPase and of a number of SR Ca2+-ATPase isoforms (SERCAATPases) reveals extensive homology within the group of SERCA isoforms, and within the group of PM isoforms. However, very little homology is found between the SERCA (Mac-Lennan et al., 1985) and the PM ATPases (Verma et al., 1988), except in the region neighboring the phosphorylation site. In particular, the 6 residues involved in Ca2+ binding (Clarke et al., 1989) within the SERCA ATPase transmembrane region are only partially retained by the PM ATPase. Therefore, there is no reason to expect necessarily an identical stoichiometric behavior from the SERCA and PM Ca2+ ATPases.
An advantage of the system reconstituted with purified enzyme is the elimination of channel proteins that allow rapid electrolyte leak in the native membrane and prevent detection of H' countertransport and net charge transfer. In the reconstituted proteoliposomes we found the stoichiometric ratio of H+/Ca2+ countertransport to be 1 under optimal conditions, i.e. early time of reaction at low temperature (Fig. 4A). The ratio becomes lower following a few minutes of reaction with ATP, especially at higher temperatures (Fig. 4B). This suggests slippage of H+ (and to a lesser extent of Ca") through the pump. It is important to realize, however, that in all cases a rise of lumenal pH occurs to the extent (0.35 units) of producing inhibition of the Ca2+ pump. The inhibition is promptly relieved when the H ' gradient is collapsed by the addition of FCCP ( Figs. 2 and 11).
Even under conditions permitting the highest (i.e. 1) ratio of H+ countertransport, net transfer of positive charge into the lumen of the vesicles is sustained by the Ca2+ pump. Accord-

electrogenic properties of SR and PM ATPases reconstituted in proteoliposomes
Initial rates of Ca2+ transport and voltage development were measured at 12 "C as for Fig. 8, under comparable conditions for PM and SR ATPase proteoliposomal preparations. Transport related voltage development (dV/dt) was calculated according to: dVldt = (dQ/dt) (l/C,), where dQ/dt is the rate of charge transfer per unit membrane area, and C,, the membrane capacitance. The observed Ca2+ transport rates (mol/mg protein x s) were first multiplied by the Avogadro number, and related to the membrane area/mg protein in the particular proteolipo-soma1 preparation. Assuming a 1:l H' and Ca2+ countertransport (therefore one net charge per transport cycle), the resulting value was divided by the elementary particle (e, = 1.6 x lo-' coulomb) and divided by C,, (1 pF ern-?. Note that the calculation is influenced by the membrane area per protein unit, which is larger for the PM than for the SR ATPase proteoliposomes used in these experiments. ingly, development of an electrical potential can be demonstrated during the early time of reaction following the addition of ATP, reaching steady state levels of 13-15 mV. Formation of the electrical gradient is favored (up to 30-40 mV) by collapse of the H+ gradient with FCCP. The effect of FCCP is due to additional positive charge associated with H+ flux into the lumen of the vesicles, and also due to reversal of Ca2+ pump inhibition by high lumenal pH. On the other hand, collapse of the electrical potential by the addition of valinomycin does not produce significant stimulation of Ca2+ transport. Furthermore, in agreement with Niggli et al. (1982) we find that imposition of a diffusion potential of 38 mV produces very little inhibition of the rate of Ca2+ transport. Our experiments suggest that the electrogenic step has only a minor rate-limiting influence on the turnover of the pump, in the presence of the voltage range developed under our conditions. Countertransport and electrogenic features of the PM ATPase are also observed with the SR ATPase reconstituted in a very similar proteoliposomal system . The reconstituted PM ATPase, however, appears to be electrogenically less efficient than the SR ATPase. In fact, the initial rate of voltage development by the SR ATPase corresponds quite closely to the rate of charge transfer calculated from the initial rates of cation transport (Table 11). On the contrary, the initial rate of voltage development by the PM ATPase, under comparable conditions (10 "C, FCCP and calmodulin), is lower than the rate of charge transfer calculated from the initial rate of cation transport. This suggests significant slippage of compensating electrolytes, and may explain previous difficulties in detecting the electrical potential generated by the PM ATPase (Niggli et al., 1982) when less favorable detergents may have been used.
Our experience suggests that various detergents have different influences on the catalytic and transport functions of various enzymes, as well as on the apparent electrogenic properties of the reconstituted systems. For instance, in the reconstitution of the SR ATPase we obtained the best results by solubilizing separately the native SR vesicles with C,,E, and the liposomes with octylglucoside, and then mixing the two solutions together before removal of the detergents. On the other hand, when we used CI2E8 for solubilization of both SR vesicles and liposomes, the resulting proteoliposomes displayed the same transport properties, but unfavorable electrogenic properties. As for the PM ATPase reconstitution, we found that the use of octylglucoside during the reconstitution procedure caused impairment of catalytic and transport activity. On the other hand, the use of Triton or cholate permitted retention of good enzyme activity, but interfered with the development of electrical potential. For this reason we chose C,,E,, resulting in production of proteoliposomes displaying very good transport activity, but less than ideal electrogenic properties. As for the composition of the reaction media for the functional assays, we found that the use of C1as the main anion allows higher PM ATPase transport rates than SO:-. This is in contrast with our experiments with the SR ATPase, in which we found that minus SO:allows higher activity than C1- .
With regard to regulation of the PM ATPase, it is noteworthy that the CaM effect is largely lost in the purified enzyme, but is regained following reconstitution of the proteoliposomal system ( Table I and Fig. 2). This apparent loss of the CaM effect has been attributed to catalytic enhancement by enzyme oligomerization following purification (Kosk-Kosicka and Bzdega, 1988). Our finding that the CaM regulatory mechanism is regained following removal of detergent and reconstitution indicates that dilution of the enzyme in the proteoliposomal membrane promotes its native state and conformation.
In conclusion, our experiments indicate that the reconstituted erythrocyte PM Ca2+-ATPase sustains a (1:l) H+ and Ca2+ countertransport, which is electrogenic due to the excess charge associated with Ca2+ transport. Rise of the lumenal pH by H+ ejection inhibits the pump turnover, while formation of electrical potential up to 30-40 mV produces very little inhibition. Demonstration of electrogenicity places a limit ( i e . less than 2) to the H+/Ca2+ stoichiometric ratio which is very difficult to determine accurately by direct measurements.