Ca2’-ATPase Membrane Crystals in Sarcoplasmic Reticulum THE EFFECT OF TRYPSIN DIGESTION*

Vanadate induces the formation of two-dimensional crystalline arrays of Ca2+-ATPase molecules in sarco- plasmic reticulum. The Ca2+-ATPase membrane crystals are evenly distributed among the terminal cister- nae and longitudinal tubules of sarcoplasmic reticulum, but very few crystals were observed in the T tubules. Tryptic cleavage of the Ca2+ transport ATPase into two major fragments (A and B) did not interfere with the vanadate-induced formation of membrane crystals. The ability of Ca2+-ATPase to crystallize was lost after further cleavage of the A fragment into the AI and Az subfragments that is known to be accompanied by loss of Ca2+ uptake. Vanadate (0.1-5 mM) inhibited the secondary cleavage of Ca2+-ATPase by trypsin sug- gesting that the susceptibility of the tryptic cleavage sites is influenced either by the conformation of the enzyme or by the formation of ATPase crystals. negative of E, conformation transport equilib-rium

Vanadate induces the formation of two-dimensional crystalline arrays of Ca2+-ATPase molecules in sarcoplasmic reticulum. The Ca2+-ATPase membrane crystals are evenly distributed among the terminal cisternae and longitudinal tubules of sarcoplasmic reticulum, but very few crystals were observed in the T tubules.
Tryptic cleavage of the Ca2+ transport ATPase into two major fragments (A and B) did not interfere with the vanadate-induced formation of membrane crystals. The ability of Ca2+-ATPase to crystallize was lost after further cleavage of the A fragment into the AI and Az subfragments that is known to be accompanied by loss of Ca2+ uptake. Vanadate (0.1-5 mM) inhibited the secondary cleavage of Ca2+-ATPase by trypsin suggesting that the susceptibility of the tryptic cleavage sites is influenced either by the conformation of the enzyme or by the formation of ATPase crystals.
Treatment of sarcoplasmic reticulum vesicles with vanadate induces the formation of two-dimensional crystalline arrays of the Ca2+ transport ATPase, which can be seen by negative staining with uranyl acetate (1) or by freeze-fracture (2). The effect of vanadate is attributed to the stabilization of the E, conformation of the Ca'+ transport ATPase. These observations clearly indicate that Ca2+-ATPase molecules possess specific sites for interactions and suggest a dynamic equilibrium between ATPase monomers and oligomers in the membrane.
Here we report data on the formation of membrane crystals in subfractions of muscle microsomes enriched in vesicles orginating from the terminal cisternae and the longitudinal elements of sarcoplasmic reticulum; Ca2+-ATPase crystals were only rarely found in the T tubule fraction. Observations are also presented on the effect of limited proteolysis upon the formation and stability of Ca2+-ATP membrane crystals and on the effect of vanadate upon the susceptibility of Ca2+-ATPase to cleavage by trypsin.

MATERIALS AND METHODS
Sarcoplasmic reticulum vesicles prepared from rabbit skeletal muscle were separated hy sucrose gradient centrifugation into fractions * This work was supported by Research Grant AM 26545 from the National Institutes of Health and by a grant from the Muscular Dystrophy Association. 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.
$ On leave from the Institute of Biochemistry, School of Medicine, University of Szeged, Hungary. enriched in T tubules, terminal cisternae, or the longitudinal tubules of the sarcotubular system ( 3 ) .
For crystallization of the Ca*+-ATPase the membrane preparations (1 mg of protein/ml) were incubated a t 2 "C in 0.1 M KC1, 10 mM imidazole, pH 7.4, 5 mM MgCL, 0.5 mM EGTA,' and 5 mM vanadate for times ranging from a few minutes to several days. The vesicle suspensions were placed on carbon-coated parlodion films, stained with freshly prepared 1% uranyl acetate (pH 4.3), and viewed with a Siemens Elmiskop I electron microscope a t 60 kV accelerating voltage. Magnification was calibrated using catalase crystals negatively stained with uranyl acetate (4).

RESULTS A N D DISCUSSION
The formation of two-dimensional membrane crystals of Ca2+ transport ATPase was observed in sarcoplasmic reticulum or purified Ca'+ transport ATPase preparations treated with 5 mM vanadate (1, 2). The crystal lattice covers the surface of a major portion (40-60%) of the vesicles present in rabbit skeletal muscle microsomes (Fig. 1). Vesicles with extensive crystallization frequently assume a cylindrical shape with an average diameter of about 600-700 A, but crystals are also present on spherical membrane profiles. The lattice lines are oriented diagonally across the surface of the cylindershaped vesicles (Fig. 2), and frequently the rows of negatively stained surface particles run in pairs, separated from neighboring pairs by wider bands of negative stain. The unit cell dimensions are consistent with ATPase dimers.
Interestingly even under optimum conditions for crystallization about one-fourth to one-third of the vesicles did not develop ATPase crystals. Therefore, the question arises whether the Ca2+-ATPase molecules in vesicles derived from different regions of the sarcotubular system show differences in their propensity for the formation of membrane crystals.

Ca"-ATPase Membrane Crystals in Subfractions of Muscle
Microsomes-The sarcoplasmic reticulum of skeletal muscle is divided into two morphologically distinct regions, the terminal cisternae and the longitudinal tubules (5). The Ca2+ transport ATPase is the major protein component of both types of membranes, and vesicles derived from the two regions show similar Ca" transport activity. However, the terminal cisternae contain large amounts of calsequestrin, and there are differences between the two membranes with respect to other proteins as well.
The transverse or T tubules are invaginations of the surface membrane. The Ca2+ transport ATPase of T tubules (3, 6) is immunologically (7-9) and kinetically (6) distinct from the Ca2+ transport ATPase of sarcoplasmic reticulum.
In crude sarcoplasmic reticulum preparations the vesicles derived from the three types of membranes are mixed in varying proportions but can be separated by sucrose gradient Ca"-ATPase M e m b r a n e Crystals in Sarcoplasmic Reticulum E ' I C ;~. 1-5. Electron micrographs of sarcoplasmic reticulum and T tubule preparations. Rabbit sarcoplasmic reticulum preparations were separated into subfractions as described by Lau ~t 01. ( 3 ) . exposed to 5 mM Na,VO, for 24 h, and stained with 1% uranyl acetate, as described under "Materials and Methods." 1, sarcoplasmic reticulum preparation, which contains vesicles derived from various regions of the sarcotubular system; X 105,000.

centrifugation (3).
Vanadate (5 mM) induced massive crystallization of the Ca2+ transport ATPase in the two sarcoplasmic reticulum fractions. In preparations enriched in terminal cisternae the crystallization occurred most frequently on spherical membrane profiles (Fig.  3), while in the putative longitudinal tubule fraction the crystals were usually seen on the surface of elongated cylinders (Fig. 4).
T h e T tubule vesicles contained only few poorly developed crystalline regions, and even these may be attributable to a slight admixture of sarcoplasmic reticulum membranes (Fig.  5 ) .
As the amount of the 100,000-Da protein, which is tentatively identified as the Ca" transport ATPase, does not differ widely between the various membrane fractions, the absence of significant Ca"-ATPase crystal formation in the T tubule vesicles suggests the following possible explanations.
1. T h e Ca')+ transport ATPase of T tubules, although similar in apparent molecular weight, may be structurally different from the sarcoplasmic reticulum enzyme. This possibility is supported by the immunological (7-9) and kinetic (6) differences between the Ca" transport ATPases of the T tubule and the sarcoplasmic reticulum membranes.
2. Structural constraints in the T tubule vesicles may prevent the formation of Ca"-ATPase crystals even if the AT-Pase has the propensity for crystallization. Such constraints may arise from the presence of proteins unique to T tubules, such as the 80,000-and 30,000-Da components (3), that may interact with the Ca" transport enzyme. Alternatively the absence of crystallization may be related to the small diameter of the spherical vesicles derived from T tubules. Even in sarcoplasmic reticulum tubules covered with extensive arrays of Ca"-ATPase crystals, the lattice is usually absent on the two hemispherical ends of the tubules, as if the sharp curvature of the membrane would be incompatible with the formation of a regular crystal lattice. The crystals seen in spherical terminal cisterna membranes are usually in larger vesicles that may provide a more accommodating surface contour.
A choice between these and other alternatives requires purification and reconstitution of the T tubule Ca"-ATPase.
Tile Effect of Trypsin Treatment on the Formation of Membrane Crystals-As shown in Scheme 1 (10) the Ca" transport ATPase of sarcoplasmic reticulum is first cleaved by tr-ypsin at a highly susceptible site (TI) into two major fragments (A and B) (Fig. 6A). This cleavage occurs without inhibition of ATPase activity or Ca2+ transport (11). Fragment A has an approximate molecular weight of 57,000 and contains the active site aspartyl group that is phosphorylated by ATP (12). The apparent molecular weight of fragment B is about 52,000; its function is unknown. Further cleavage of fragment A at site T, yields subfragments A, and A, (Fig. 6A), accompanied by inhibition of Ca?+ transport (ll), but significant ATPase activity is still retained (not shown).
Cleavage of the Ca"-ATPase at the TI site inhibited only slightly the crystallization of the enzyme induced by subsequent addition of vanadate (Fig. 7C) and had no detectable effect on the previously formed crystal lattice in the presence of 5 mM vanadate (Fig. 7, D, E, and F). Therefore, even after cleavage into two subunits the Ca"-ATPase retains sufficient structural integrity to transport calcium and to develop crystalline arrays. Further cleavage of the A into A, and A, fragments (Fig. 6A) prevented the crystallization of the Ca"-ATPase (Fig. 7, H and I) together with inhibition of Ca" transport but with only moderate change in ATP hydrolysis.
Vanadate (5 mM) slightly decreases the rate of cleavage of the Ca"-ATPase at the TI site (Fig. 6B) but completely inhibits the cleavage of the A fragment into the A, and A, subfragments at the T? site (Fig. 6R). As a result the A and B fragments are present even after 480-min exposure to trypsin (Fig. 6R), and the Ca"-ATPase crystals are also fully preserved (Fig. 7, D, E, and F). Vanadate (5 mM) does not inhibit the activity of trypsin with Azocoll as substrate (13) while the hydrolysis of phosphorylase (94,000-Da band) is actually activated (Fig. 6R).
Vanadate is an analog of inorganic phosphate that inhibits the ATPase and Ca2+ transport activities of sarcoplasmic reticulum by stabilizing the E, enzyme conformation (14, 15).
The inhibition of the tryptic cleavage of Ca"-ATPase at the T, site by vanadate suggests that conversion of the enzyme from the El into the E, conformation decreases the sensitivity of the T, site to trypsin. An analogous difference between the proteolytic cleavage pattern of the El and E2 forms of the (Na,K)-ATPase was observed earlier by Jorgensen (16). The protection of Ca"-ATPase against trypsin was observed even a t a vanadate concentration as low a s 0.1 mM. Vanadate increases the rate of hydrolysis of phosphorylase by trypsin, H . prior to trypsin digestion the preparation was exposed to 5 nlM Na:,V04 in 0.1 M KCI, 10 mM imidazole, pH 7.4, 0..5 mM EGTA. and 5 mM MgCI, at 2 "C for 36 h. T h e samples were soluhilized in a solution of IOr; sodium dodecyl sulfate, 20 mM Tris-HCI pH 8.0, 2 7 13-mercaptoet hanol. 2 0 5 glycerol, 0.1% bromphenol blue, and applied to sodium dodecyl sulfate-polyacrylamide gradient gels (6-18c;) for electrophoresis. The Coomassie blue stained gels were analyzed with an 1,KH lrltrascan laser densitometer coupled with a Hewlett-Packard integrator plotter (X390A). KII, kilodalton.
suggesting that in this case the conformation favored by vanadate is more sensitive to proteolysis. I t is reasonable to assume that vanadate alters the conformation of phosphorylase by forming a stable complex at the phosphate binding site of the enzyme. The 63,000-Da Ca2+ binding protein, which is a usual component of sarcoplasmic reticulum (17), was not digested by tr-ypsin under either condition. Sucrose (1 M ) partially protects the Ca"-ATPase against inhibition by 5 mM vanadate and under similar conditions prevents the vanadate-induced crystallization of the enzyme. The protection by sucrose is not attributable to contaminating calcium as it is observed at free Ca" concentrations less than lo-' M; sucrose obtained from four different manufacturers gave similar results. Sucrose (1 M ) also slowed the rate of The experiment was carried out as described in Fig. 6, A and H, except that 1 M sucrose was added to the digestion medium. A, no vanadate; H, 5 mM Na:$V04 was added 36 h (2 "C) prior to trypsin digestion to induce crystallization of the Ca"-ATPase. KD. kilodalton. hydrolysis at the T, site of the Ca2+-ATPase and provided significant protection of the phosphorylase against tryptic cleavage (Fig. 8). The vanadate inhibition of the trypt.ic cleavage at the T2 site remained essentially complete even in the presence of 1 M sucrose (Fig. 8).
The mechanism of the effect of sucrose is unknown, but it is likely to involve changes in the conformation of the enzyme either by interaction of sucrose with the hydrophilic portion of the ATPase molecule or by reducing the range of motions available to the enzyme due to increase in medium viscosity.