The calcium binding sites involved in the regulation of the purified adenosine triphosphatase of the sarcoplasmic reticulum.

Abstract The role of the interaction of Ca2+ with the purified Ca2+-dependent ATPase in the regulation of enzyme activity has been investigated. It appears that the sensitivity to Ca2+ of the sarcoplasmic reticulum, resulting in activation and inhibition of ATPase activity, is intrinsic to the ATPase moiety of the membrane. Three types of Ca2+ binding sites have been found in equilibrium dialysis studies. In the absence of ATP there is approximately one of each per 105 daltons; the binding constants are 4 x 106 m-1 (α site), 4 x 104 (β site), and 1 x 103 (γ site). Addition of 1.5 mm ATP slightly increases the affinity of all sites and reduces the apparent capacity of the α and β sites. Study of ATP hydrolysis in parallel with calcium binding has shown that Ca2+ binding at the α site activates it and binding to the γ site inhibits it, while binding of Ca2+ to the β site appears not to be involved in the enzymatic regulation.


SUMMARY
The role of the interaction of Ca2+ with the purified Ca2+dependent ATPase in the regulation of enzyme activity has been investigated.
It appears that the sensitivity to Ca2+ of the sarcoplasmic reticulum, resulting in activation and inhibition of ATPase activity, is intrinsic to the ATPase moiety of the membrane.
Three types of Ca*+ binding sites have been found in equilibrium dialysis studies.
In the absence of ATP there is approximately one of each per 10" daltons; the binding constants are 4 x lo6 M-' ((Y site), 4 X lo4 (0 site), and 1 x lo3 (y site).
Addition of 1.5 m&r ATP slightly increases the affinity of all sites and reduces the apparent capacity of the a! and /I sites.
Study of ATP hydrolysis in parallel with calcium binding has shown that Ca2+ binding at the a site activates it and binding to the y site inhibits it, while binding of Ca*+ to the 0 site appears not to be involved in the enzymatic regulation.
The sarcoplasmic reticulum plays its role in the regulation of muscle function by accumulating Ca*+ and reducing the concentration of Ca*+ in the cytoplasm below 1O-7 M (l-6) thereby producing relaxation.
The lowering of the concentration of Ca*+ is achieved by the transport of Ca*+ across the membrane through a process tightly coupled to ATP hydrolysis catalyzed by an enzyme in the membrane (1,3). At [Ca'+] of lOWa -1OV M, the transport of Ca*f accompanied by increased hydrolysis of ATP and the formation of a phosphorylated protein intermediate takes place (1,5,(7)(8)(9)(10)(11).
The increase of [C&l in the vesicles of fragmented sarcoplasmic reticulum leads to a decrease in ATPase activity and a leveling off of Ca2+ uptake (5,12,13). This paper describes results of a simultaneous study of binding of Ca*f to, and ATP hydrolysis by, the isolated ATPase enzyme of the sarcoplasmic reticulum.
As will be shown, one can distinguish among various Ca*+ binding sites of the ATPase enzyme and their roles in the regulation of the enzyme activity.

EXPERIMENTAL PROCEDURE
Fragmented sarcoplasmic reticulum was prepared from rabbit skeletal muscle as described previously (14 ATPase fraction, which sedimented at 36,500 x g (15) after solubilization of the fragmented sarcoplasmic reticulum with Triton X-100 followed by chromatography on Sepharose 4B, was used for this study.
The conventional dialysis method for Ca*+ binding was modified in order to shorten the time required for establishing equilibrium to about 60 min, thereby maintaining the enzyme activity intact and permitting a simultaneous assay of ATPase.
A sample (0.4 ml) of the purified ATPase enzyme solution (3.0 mg of protein per ml) was placed in narrow dialysis tubing (0.22.inch diameter, inflated, Fisher Scientific, Pittsburgh, Pa.) into which a round-ended glass rod, 112 mm in length and 6 mm in diameter, was subsequently inserted. This procedure led to a thin enzyme layer (0.12 mm average thickness).
Prior to equilibrium dialysis, contaminating Simultaneous assays of Caz+ binding and ATPase activity were carried out by dialysis at 0" versus Medium A or B containing 1.5 mM ATP which had been treated with Dowex 50 to eliminate contaminating Ca2+ (16). Solutions containing 0.1 M KC1 and 5 mM MgC12, an optimal ionic milieu for ATPase assay, were contaminated with 2.5 x 1O-6 M Ca2f as determined by atomic absorption spectrometry.
This was taken into account in the calculation of Ca*+ concentration with the use of a computer program (17). Constants used for the calculation are listed in the legend to Fig. 1. For the Ca*f binding assay, 0.2-ml portions, taken from the dialysis medium at 0 time and at 3 hours and from the sample inside the dialysis tubing at 3 hours, were dried on filter paper strips and subjected to scintillation counting (21). For ATPase assay, portions were taken from the dialysis medium at 1, 2, and 3 hours and Pi was determined according to Fiske and SubbaRow (22). The time course of Pi liberation measured outside the dialysis tubing was linear between 1 and 3 hours of dialysis.
It is reasonable to assume that the rate of diffusion across the dialysis membrane of Pi is proportional to the Pi concentration inside the bag. Since the back-diffusion is negligible because of the large dialysis volume the linear increase in Pi outside the bag indicates that (a) the concentration of Pi inside the bag has reached a steady state, and (b) the rate of appearance of Pi on the outside equals the rate of Pi liberation from ATP inside the bag. Throughout the period of measurement the ATP concentration remained essentially constant because of the large dialysis volume.

Solubilization
of the fragmented sarcoplasmic reticulum with Triton X-100 leads to a considerable increase in ATPase activity (15) while the dependence of ATPase activity on the Ca2+ concentration, both activation and inactivation, remains unchanged (13). Even with the purified ATPase enzyme, the activity of which is about 3 times that of the ATPase of unfractionated sarcoplasmic reticulum activated by Triton X-100 (15), the Ca2+ dependence is identical with that of intact fragmented sarcoplasmic reticulum.
A similar Ca*+ dependence has been reported for the ATPase enzyme purified from fragmented sarcoplasmic reticulum solubilized with deoxycholate (23,24). These results indicate that the mechanism by which ATPase is regulated as a function of Ca2f concentration is intrinsic to the ATPase enzyme moiety of the membrane, and does not depend on the interaction of the enzyme with other membrane components. Fig. 1 depicts the results of Ca*+ binding studies in the absence and presence of ATP.
The binding data can be fitted by a least squares method with the use of a computer assuming that the binding of Ca*f to the ATPase enzyme takes place to three noninteracting classes of sites, (Y, /3, and y in Table I. Each class has a binding capacity of approximately 1 mole per lo5  25,26). In the presence of 1.5 mM ATP the affinity of all sites to CaZf is increased by about the same factor (1.9 times for Q: site and 1.6 times for /3 and y sites) while the capacity for Ca2+ binding decreased in the case of a! and /3 sites. Fig. 2 shows the ATPase activity determined simultaneously with the calcium binding study; calculated plots of Ca2f binding to each site are also included in the figure.
Binding of Ca2+ to the cy, 0, and y sites is well correlated with the three phases of the Ca?+ dependence of ATPase: binding of Ca*+ to the (Y site parallels the activation of ATPase; there is little change in ATPase when Ca2+ binding to the fi site takes place; and parallel with the binding of Ca 2+ to the y site inhibition occurs.

DISCUSSION
The results presented here have an important bearing on the mechanism of the regulation of calcium transport in the sarcoplasmic reticulum.
The association constant of the a! site in the presence of ATP agrees well with the reported K, values for ATPase and phosphorylation of the fragmented sarcoplasmic reticulum (8, 10, 12, 27). Thus, it seems that the (Y site is the one involved in the activation of the ATPase enzyme, and presumably in Ca2f transport across the-membrane. The second type of site (0 site) has little or no involvement in the regulation of ATPase.
Since the Ca2+ dependence of the phosphorylated intermediate formation shows a plateau (lo), as does ATP hydrolysis, in the concentration range of CaZf of 1OF -10F4 M, the /3 site seems not to be involved in the regulat.ion of phosphorylation either.
Binding of Ca'+ to the third, y, site appears to be involved in the inhibition of ATPase. It is known that if the Ca2+ concentration within the vesicles increases to about 1 mM as a result of ATP-dependent transport (28-30), ATPase and Ca2+ uptake are inhibited (12,13). At that concentration, binding of Ca *+ to the y site could take place, which in turn would produce inhibition of ATPase and, in the intact system, presumably of the Cay+ transport.
It is worth noting that in the range of Ca2+ concentrations where binding takes place to the y site there is little dependence on Ca2+ of the formation of the phosphorylated intermediate (31). This would indicate that binding of Ca*+ to the y site primarily affects the breakdown or transformation of the E -P complex. A Cat+ binding site which has an association constant of 5 x lo5 -1.5 x lo6 M-I has been reported in the fragmented sarcoplasmic reticulum (32, 33) and in the "calcium pump" protein (34, 35).
A puzzling feature of the present results is the apparent reduction in the presence of ATP of the binding capacity of the QI and /3 sites and a higher value for the y site, particularly since the values are nonintegral in terms of the estimated molecular weight of 100,000. A tentative interpretation suggests that in the presence of ATP a dimer of subunits, each of which has a molecular weight of 100,000, is involved, in which only one of the two cy sites or of the two /3 sites would be able to bind Ca2+. Changes in conformation, probably caused by subunit interaction, are also suggested by the changes in the Ca2+ binding constants brought about by ATP.