Coincidence of H+ Binding and Ca2+ Dissociation in the Sarcoplasmic Reticulum Ca-ATPase during ATP Hydrolysis*

H+ and Caz+ concentration changes in the reaction medium following MgATP addition at pH 6.0 were determined with the partially purified Ca-ATPase from sarcoplasmic reticulum vesicles in the presence of 25-50 PM CaClz and 6 mM MgClz at 4 OC. Previously, we showed a sequential occurrence of H+ binding and H+ dissociation in the Ca-ATPase during ATP hydrol- ysis and further suggested that the H+ binding takes place inside the vesicles (Yamaguchi, M., and Kana- zawa, T. (1984) J. Biol. Chern. 259,9626-9531). The present results demonstrate that the H+ binding oc- curred coincidently with Ca2+ dissociation from the enzyme upon conversion of the phosphoenzyme (EP) intermediate from the ADP-sensitive form to the ADP-insensitive form in the catalytic cycle of ATP hydrolysis. As KC1 decreased in the medium, the extent of the H+ binding increased almost proportionately with the extent of either the Ca2+ dissociation or the accumula- tion of ADP-insensitive Ep. Both the H+ binding and the Ca2+ dissociation were prevented by a modification of the specific SH group of the enzyme essential for the conversion of ADP-sensitive EP to ADP-insensitive EP. In the late stage of the reaction, H+ dissociation from the enzyme occurred coincidently with Ca2+ bind- ing to the dephosphoenzyme which was formed by EP decomposition. These results are consistent with the possibility that

* This work was supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science, and Culture of Japan. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
$ To whom reprint requests should be addressed. verted to ADP-insensitive EP which is incapable of reacting with added ADP (11-14). Upon the EP conversion, the Ca2+ affinity of the transport sites is greatly reduced and as a result the Ca2+ dissociates inside vesicles (8-10, [15][16][17]. Next, ADPinsensitive EP is hydrolyzed, and this hydrolysis is stimulated by K+ (11). Finally, the transport sites return to the outer surface of vesicles.
Recently, several investigators (18-24) indicated that H+ is ejected from vesicles during the Ca2+ uptake. Thus, an important problem to be solved is whether this H+ ejection occurs through a specific mechanism directly involving the Ca-AT-Pase. In this regard, our previous experiment showed a sequential occurrence of H+ binding and H+ dissociation in the Ca-ATPase during ATP hydrolysis (24). This finding is in favor of the possibility that the Ca-ATPase serves as a H+ carrier in the H+ ejection.
In the present experiment, we provide evidence that the H+ binding and the, H+ dissociation occur coincidently with the Ca2+ dissociation from the enzyme and the subsequent Ca2+ binding to the enzyme in the catalytic cycle of ATP hydrolysis.
The results are consistent with the possibility that the H+ ejection during the Ca2+ uptake takes place through a Ca2+/ H+ exchange directly mediated by the membrane-bound Ca-ATPase.

EXPERIMENTAL PROCEDURES
Preparation of Sarcoplasmic Reticulum Vesicles-Sarcoplasmic reticulum vesicles were prepared from rabbit skeletal muscle as described previously (7). They were suspended in 0.3 M sucrose, 0.1 M KCI, and 1.0 mM MES/Tris (pH 6.0) and stored at -80 "C.
Preparation of Partially Purified Ca-A TPase-Vesicles were treated with deoxycholate according to the method of Meissner and Fleischer (25). The treatment resulted in a partial purification of the Ca-ATPase and a complete loss of the capacity for Ca2+ accumulation. All of the measurements in the present experiment were performed at 4 "C by using this purified Ca-ATPase. In order to prevent an undesirable aggregation (26), the enzyme was sonicated for 15 s at 0 "C with a sonicator (Branson, Sonifier-200) immediately before the start of the measurement.
Treatment of the Ca-ATPase with NEM-The purified enzyme (3 mg/ml) was incubated with 1.5 mM NEM in the presence of 1 mM App(NH)p for 20 min by the method of Kawakita et al. (27).
Measurement of p H Changes in the Reaction Medium-pH changes in the reaction medium were determined with a pH meter (Radiometer, PHM84) or determined by following the change in the fluorescence intensity of FITC-dextran with a spectrofluorometer (Shimadzu, RF-503A) as described previously (24). The reaction was initiated by adding 2.2 pl of 30 mM MgATP (30 mM Tris-ATP containing 36 mM MgC12, adjusted with Tris to pH 6.0) to 2.2 ml of the medium containing the enzyme and other reagents as described under "Results." When the enzyme was absent, no significant pH change occurred upon addition of MgATP (see the controls in Fig.  1). Under the present experimental conditions, at pH 6.0, complete hydrolysis of MgATP with an extremely low concentration of the enzyme did not significantly cause H+ production and H+ absorption.
Measurement of Ca2+ Concentration Changes in the Reaction Medium-Ca2+ concentration changes in the reaction medium were 4896 determined by two different methods. ( a ) They were followed with a calcium electrode (Radiometer, F2112Ca). The reaction was initiated by adding 2.5 pl of 30 mM MgATP (pH 6.0) to 2.5 ml of the medium. (b) They were followed with a dual-wavelength spectrophotometer (Shimadzu, UV-300) by using the metallochromic indicator murexide. A wavelength pair of 540 and 700 nm was chosen. The reaction was initiated by adding 2.2 pl of 30 mM MgATP (pH 6.0) to 2.2 ml of the medium.
Determination of EP-Phosphorylation of the Ca-ATPase was carried out with [T-~~PIATP as described under "Results." The amount of E P was determined essentially in the same way as described previously (28). Rapid kinetic measurements of EP formation were performed with a handmade rapid mixing apparatus (29).
Materials-Na2ATP was purchased from Boehringer Mannheim.
Na2ADP, Lid-App(NH)p, EGTA, and FITC-dextran (average M, = 66,000; 0.02 mol of FITC/mol of glucose residue) were obtained from Sigma. Murexide was from Nakarai Chemicals. Removal of free FITC contaminating FITC-dextran and conversion of Na2ATP and Na2ADP into Tris form were performed as described previously (24).
[ T -~~P ] A T P was prepared according to Post and Sen (30). Protein concentrations were determined by the method of Lowry et al. (31) with bovine serum albumin as a standard.

MgATP-induced Changes of H+ and Ca2+ Concentrations in
the Medium-Changes of H' and Ca2+ concentrations in the reaction medium following addition of 30 ~L M MgATP at a final concentration were determined with hydrogen and calcium electrodes in the presence of 2.5 mg of the enzyme/ml, 25 FM CaC1, added: 5 mM MgCl,, and 100 mM KC1 (Fig. 1A). The pH of the medium before the start of the reaction was 6.0. After the addition of MgATP, an alkalinization of the medium occurred and then the pH returned to the initial level (truce u in Fig. 1A). The data agree with our previous observations (24) that the alkalinization and its disappearance represent H' binding and H+ dissociation in the Ca-ATPase during ATP hydrolysis.
Under similar conditions, the Ca2+ concentration in the medium increased after the MgATP addition and then it returned to the initial level (trace c in Fig. lA). These results show Ca2+ dissociation from the enzyme in the early stage of the reaction and Ca2+ binding to the enzyme in the late stage of the reaction (see "Discussion"). When KC1 was absent (LiCl was added to maintain a constant ionic strength), the H' binding and the Ca2+ dissociation were substantially enhanced ( Fig. 1B). In addition, the H+ dissociation and the Ca2+ binding in the late stage of the reaction were markedly delayed.
It appeared that the H+ binding and the H' dissociation almost coincided with the Ca" dissociation and the Ca2' binding, respectively, either in the presence of KC1 or in the absence of KCl. However, it was difficult to exactly compare the time courses since the responses of the electrodes used were slow. H+ Binding and Accumulation of ADP-insensitive EP in the Initial Phase-The H' binding in the initial phase of the reaction was determined fluorometrically with FITC-dextran in the presence (Fig. 2 A ) and absence (Fig. 2B) of KC1. EP formation was also determined with a rapid mixing apparatus under similar conditions. The H' binding was much slower than EP formation in the presence and absence of KCI. This finding agrees with our previous observations (24). The data showing no detectable H+ dissociation and H' binding upon EP formation are consistent with the recent findings from the calorimetric study by Kodama et al. (32).
Contaminant Ca2+ in the reaction mixture was about 15 KM, as determined by atomic absorption spectrophotometry. Most of the Caz+ was derived from the enzyme. Thus, total Caz+ concentration in the mixture was estimated to be 40 p~. Both the H' binding and the accumulation of ADP-insensitive EP were markedly enhanced when KC1 was absent. The enhancement of the accumulation of ADP-insensitive EP is compatible with the observation (11) that hydrolysis of ADPinsensitive EP is inhibited in the absence of K' . The time course of the H+ binding essentially agreed with that of the accumulation of ADP-insensitive EP either in the presence or absence of KC1.
Comparison of the Time Course between the H+ Binding and the Ca2' Dissociation-In order to compare the time course of the H' binding with that of the Ca2' dissociation with reasonable accuracy, the Ca2+ dissociation was determined spectrophotometrically by using murexide (Fig. 3). The H+ binding was determined with FITC-dextran under the same conditions. The H' binding and the Ca2+ dissociation were markedly enhanced when KC1 was absent. This is consistent with the results obtained by using hydrogen and calcium electrodes (cf. Fig. 1  dissociation increased in parallel with the amount of ADPinsensitive EP as KC1 was replaced by LiCl, while the total amount of EP remained almost constant (Fig. 4)

DISCUSSION
The implications of the results may be discussed conveniently in terms of the minimum scheme (Fig. 5) tentatively proposed for the Ca-ATPase. This scheme is based on present and earlier findings (8-10, 24). E and EP denote the state of the enzyme which has transport sites with a high affinity for Caz+ and a low affinity for H+. In this state, the transport sites face on the external medium of vesicles. *E and * E P indicate the state of the enzyme which has transport sites with a low affinity for Ca2+ and a high affinity for H' . In this state, the transport sites face on the internal medium of vesicles. According to this scheme, EP is ADP-sensitive and *EP is ADP-insensitive. The stoichiometries of Ca2+ and H+ are omitted for simplicity.
The observed alkalinization (Figs. 1-3) agrees with our previous findings (24), which showed a sequential occurrence of H+ binding and H+ dissociation in the Ca-ATPase during ATP hydrolysis and suggested that the H+ binding to the enzyme occurs inside vesicles. The Ca2+ concentration changes ( Figs. 1 and 3) are also consistent with previous findings (15-17, 33) that Ca2' dissociation and Ca" binding in the transport sites occur in Steps 4 and 1, respectively, given in Fig. 5 .
The coincidence of the H' binding and the Ca2+ dissociation ( Figs. 1 and 3) gives evidence that the H' binding occurs in Step 4. This is further supported by the findings (Fig. 2) that the H' binding almost coincided with the accumulation of ADP-insensitive EP, since the Ca" dissociation in Step 4 results from the conversion of ADP-sensitive EP to ADPinsensitive EP (15)(16)(17)33). The enhancement of the H' binding upon a reduction in KC1 concentration corresponded well to the increase in the extent of either the Ca2+ dissociation or the accumulation of ADP-insensitive EP (Fig. 4). This correspondence also adds probability to the H+ binding in Step 4.
The coincidence of the H' dissociation and the Ca" binding in the late stage of the reaction (Fig. 1) suggests that the H+ dissociation occurs between Steps 5 and 1.
Step 1 is a strong candidate for this H+ dissociation, because it was previously shown by Inesi and co-workers that the titration of the transport sites with Ca2+ in the absence of ATP was accompanied by a stoichiometric H' dissociation (20), that the Ca2+ binding to the transport sites was competitively inhibited by H' (34), and that this H+ competition resulted in an inhibition of EP formation (35).
The inhibition of the conversion of ADP-sensitive EP to ADP-insensitive EP by NEM treatment (Table I) is in accord with the results reported by Yasuoka-Yabe et al. (36). The prevention of the H+ binding by NEM treatment in the presence of KC1 is consistent with lack of the accumulation of ADP-insensitive EP and with lack of the Ca2+ dissociation. On the other hand, the significant occurrence of the H-' binding to the NEM-treated enzyme in the absence of KC1 corresponded to the accumulation of ADP-insensitive EP and the Ca2+ dissociation. The markedly delayed H+ dissociation without KC1 in the late stage of the reaction is also consistent with the markedly delayed Ca2+ binding. These findings with the NEM-treated enzyme give additional support to the proposed scheme.
Thus, all of the results discussed hitherto are compatible with the possibility that at least at acidic pH the H+ ejection during the Ca2+ uptake takes place through a Ca2+/H+ exchange directly mediated by the Ca-ATPase. However, at present, it is not clear whether the observed H+ binding plays an essential role in the process of the Ca2+ uptake.
The Ca2+ binding to the NEM-treated enzyme in the presence of KC1 (Table I)  The stoichiometric ratio of the H+ binding to the Ca2+ dissociation was 1.5 mol/mol under the conditions in which almost all of the EP formed was ADP-insensitive (Fig. 4). On the other hand, Chiesi and Inesi (20) indicated the stoichiometric ratio of the H' ejection to the Ca2+ uptake to be 1.0 mol/mol. Previously Inesi et al. (37) found that functional residues of the enzyme with different pK values (pK 7.7 in *EP and pK 5.8 in *E. Pi) are involved in the reaction of Step 5. This finding is supported by the pH effects on EP hydrolysis (Step 5) (35) and Pi/H20 exchange (dynamic reversal of Steps 5 and 6) (38). It seems, therefore, likely that some part of the observed H+ binding and H+ dissociation is ascribed to the protonation and deprotonation of these residues and is possibly unrelated to the H+ ejection.
It is difficult to unequivocally define the sidedness of H+ binding from our previous results (24) showing that the alkalinization was enhanced by making the vesicles leaky since, as reported by Meissner and Young (39) and Meissner (401, sarcoplasmic reticulum membranes are highly permeable to H+. Therefore, the possibility cannot be eliminated that the observed H+ binding and H' dissociation in the present study represent conformational changes of the enzyme being not directly related to the H+ ejection. Thus, it is still possible that the H+ ejection during the Ca2+ uptake with the intact vesicles is mediated by a H+ channel rather than by an obligate exchange process directly coupled to Ca2+ translocation. Our previous observation (cf. Fig. 7 in Ref. 24) showed that the alkalinization of the medium in the absence of KC1 occurred almost coincidently with the accumulation of ADPinsensitive EP. This finding agrees well with the present results. However, in the previous experiment, the alkalinization in the presence of KC1 appeared to be somewhat faster than the accumulation of ADP-insensitive EP. This rapid alkalinization was not found under the conditions used in the present experiment.