The Hydrolytic Cycle of Sarcoplasmic Reticulum Ca2+-ATPase in the ...

THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1987 by The American Society of Biological Chemists, Inc.

Vol. 262, No. 14, Issue of May 15, pp. 6610-6614 1987 Printed in 6.s.~.

The Hydrolytic Cycle of Sarcoplasmic Reticulum Ca2+-ATPase inthe Absence of Calcium* (Received for publication, July 28, 1986)

Paulo C. Carvalho-AlvesS and Helena Maria ScofanoS From the Department of Biochemistry, Institute of Medical Sciences, Federal University of RW de Janeiro, C.P. 68041, RW de Janeiro, RJ 21910, Brazil

The hydrolytic cycle of sarcoplasmic reticulum Ca2+- hypothesis that Ca2+-ATPasedoes not have "basal" activity ATPase in the absence of Ca2+was studied. At pH 6.0, and that purified enzyme would be unable to utilize ATP as 10 O C and in the absence of K+, the enzyme displays a substrate in the absence of Ca2+.The partial reactions of the very low velocity of ATP hydrolysis. Addition of up to catalytic enzyme cycle have been extensively studied only in 15% dimethyl sulfoxide increased this velocity sever- the presence of Ca2+.There are no data in the literature that alfold (from 5-18 nmol of Pi*mgof protein-'oh") and then decreased at higher solvent concentrations. Di- permit a detailed comparison of the reaction mechanisms for methyl sulfoxide increased both enzyme phosphoryla- ATP hydrolysis in the presence and absence of Ca2+. We have previously shown that an exclusively Mg2"detion from ATP and the affinity for this substrate. Maximal levels of 1.0-1.2 nmol of EP*mgof protein" and pendent phosphorylation of purified enzymefrom ATP is detected provided that special reaction conditions are mainapparent KM for ATP of 5 X M were obtained at a concentration of 30% dimethyl sulfoxide. The same tained (pH below 6.0, absence of NaCl or KCl, and presence preparation under optimal conditions (pH 7.5, 10 p~ of Me2SO)' (6). In this paper we study the relationship beCaCl,, 100mM KC1 and no dimethyl sulfoxide at 37 "C) tween the phosphoenzyme formed and thehydrolytic activity displays a velocity of ATP hydrolysis between 8 and in theabsence of Ca2+.We further compare the intermediates 12 X lo6 nmol of Pi*mg of protein"*h" while the formed in our conditions with those formed in the presence phosphoenzyme levels varied between 3.5 and 4.0 nmol of Ca". of m * m gof protein". Enzyme phosphorylation from ATP in the absence of EXPERIMENTALPROCEDURES Ca2+always preceded Pi liberation into the assaymeCa2+-ATPase was purified from sarcoplasmic reticulum vesicles of dia. Two different phosphoenzyme species were formed which were kinetically distinguished by their rabbit white skeletal muscle according to Mac Lennan et al. (1). decomposition rates. The observed steady-state veloc- During this purification the sarcoplasmic reticulum vesicleswere ity of ATP hydrolysis could be accounted for eitherby treated with deoxycholate and then dialyzed to eliminate excess of the decay of the fastcomponent or by the simultaneous detergent which renders a leaky preparation unable to accumulate calcium. The specific ATPase activity of the preparations varied decomposition of both phosphoenzyme species. between 8 and 12 X 106 nmol of Pi.mg protein" h-', and thelevel of The hydrolysis of the phosphoenzyme formed in the phosphoenzyme formed by ( T - ~ ~ P J Avaried T P between 3.5 and 4 nmol absence of Ca2+ was KC1-stimulated and ADP-inde- of EP.mg of protein-' when measured as described by MacLennan pendent. The rate constant of breakdown was equal to et al. (1). Proteinconcentration was measured by the method of that observed for the phosphoenzyme formed in the Lowry et al. (7). ( T - ~ ~ P J Awas T P prepared according to the method presence of Ca2+.It is suggested that therapidly decay- of Glynn and Chapel1 (8) with small modifications (9). The ing phosphoenzyme (and possibly both rapidlyand (T-~~PJATP obtained was neutralized with Tris andstored a t -6 "C. slowly decaying species) are intermediates in the reFor determination of phosphorylation and hydrolytic activity, the action cycle of Mg2+-dependentATP hydrolysis of sar- enzyme was preincubated in the presence of 10 mM EGTA during 2 coplasmic reticulum Ca2+-ATPaseand may represent min in the assay media as described in the figure legends. Previous controls (see Ref. 6) assure that calcium contamination is smaller a bypass of Ca2+activation by dimethyl sulfoxide. than M andthat under our conditions the enzyme cannot be phosphorylated by the contaminant Pi. Reactions were initiated by the addition of (y3'P)ATP or by addition of preincubated enzyme and carried out at 10 "C. Prior to starting thereaction, the pH of the The Ca*+-ATPaseof sarcoplasmic reticulum accounts for medium was adjusted to pH 6.0. The reaction was arrested with 0.25 the major portion of the membrane proteins (1-4). A low rate M perchloric acid containing 4 mM Pi at the indicated times. After of ATP hydrolysis exclusively dependent on M$+ is observed three to four washings, the phosphoenzyme formed was determined in native vesicles and has been named "basal activity." This as described by de Meis and Carvalho (10). 32Piliberation was measactivity is greatly reduced following enzyme purification by ured after stoppingthe reaction with perchloric acid and then diluting different methods (1, 2, 5). This observation has led to the the medium to a concentration of 5-8% MezSO. Aliquots (0.5 ml) were mixed with 0.5 ml of a suspension containing 25% w/v charcoal activated with 0.1 N HCl (11). This mixture was stirred for 15 s and * This work was supported by grants from the Conselho Nacional de Desenvolvimento Cientifico e Tecnol6gico and Convbnio FundacHo centrifuged. Aliquots of the supernatant were counted in a liquid Jose Bonifacio/Financiadora de Estudose Projetos. The costs of scintillation counter. Rate constants for phosphoenzyme decay and percentage of the publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "adver- different components in steady state were determined as already tisement'' in accordance with 18U.S.C. Section 1734 solelyto indicate described (12, 13, 15). this fact. The abbreviations used are: Me2S0, dimethyl sulfoxide; EGTA, 4 Recipients of a fellowship from the Conselho Nacional de Desen[ethylenebis(oxyethylenenitrilo)]tetraaceticacid. volvimento Cientifico e Tecnol6gico.

6610

Hydrolytic Cycle of SR Ca2+-ATPase

6611

each 10% incrementin organic solvent concentration (Fig. 2). M is Temporal Relationship between Phosphoenzyme Formation With 30% Me2S0, an apparent KM for ATP of 5 X observed. At lower solvent concentrations the maximal initial and Pi Release from ATP inthe Absence of CaZ'-It has been proposed that during the reaction cycle, the sarcoplasmic velocity of enzyme phosphorylation is increased (see values reticulum Ca2+-ATPaseexistsin two different enzymatic at lod3 M ATP) and saturation becomes difficult to obtain due to the scattering of the data at millimolar ATP concenforms, *E and E (3). In anaqueous medium, ATP binds with trations. Therefore, it was not possible to obtain a precise high affinity to theenzymatic form E in the presence of Ca2+, determination of apparent KM at lower solvent concentrawhereas Pi binds with low affinity to theenzymatic form * E tions, but they areclearly higher than the value measured at in the absence of Ca2+(14, 16). Binding in both conditions is 30% Me2S0. At 30% Me2S0, the apparent KM for ATP is followed by phosphorylation of an aspartyl residue at the comparable to the apparent KM for ATP in the presence of catalyticsite of the enzyme (17-21). We have previously Ca2+(3, 4). shown that addition of Me2S0 allows the enzyme to be The exclusively M$'-dependent ATP hydrolysis displayed phosphorylated by ATP in the absence of Ca2+(6). M e 8 0 a complex behavior as a function ofMe,SO concentration. inhibits the velocity of phosphoenzyme decay and thus leads The rate of hydrolysis increased approximately 4-fold with an to itsaccumulation (6). The phosphoenzyme so formed is an increase in MezSO from 0 to 15% (Fig. 3). Further additions acylphosphate-type compound, as itexhibits an acid stability of solvent inhibited the M$'-dependent ATPase activity, similar to thatof the intermediate formed during the reaction whereas the phosphoenzyme steady-state levels increased cycle of the sarcoplasmic reticulum Ca2+-ATPase(6, 19, 21). continuously. Comparison of the reaction conditions used for The following experiments were designed to investigate Fig. 3 with the dataof Fig. 2 suggests that theincrease in the whether, from a kinetic point of view, the phosphoenzyme velocity of ATP hydrolysis in the range 0-15% MezSO is due formed in the absence of Ca2+is a true intermediate of an exclusively Me-dependent cycle of ATP hydrolysis or a parallel and nonrelatedphosphorylation at some other siteon - 0.3 - 0.6 the enzyme. In all solvent concentrations tested, enzyme phosphorylation preceded Pi liberation in the reaction medium. Fig. 1 - 0.2-.0.4 shows the data obtained in the presence of 30% Me2S0. The values observed are much smaller than those obtained in the presence of calcium (for a comparison see values at optimal - 0.1 -0.2 conditions described under "Experimental Procedures"). Pi liberation had a lag phase in pre-steady-state conditions, and ( 0 ) (0) a linear release was observed when the EP level attained a w I I I steady state. These results arecompatible with a mechanism -6 -5 -4 -3 of ATP hydrolysis that includes one or more phosphoenzyme log ATP intermediates. FIG.2. Effect of MeaSO on the ATP dependence of enzyme Effects of Me2S0 on the Hydrolytic Cycle in the Absence of phosphorylation in the absence of Ca". Purified Ca2+-ATPase Cu2+-Upon measuring the initial velocity of phosphorylation (0.3 mg/ml) was incubated at 10 'C in media containing 30 mM Tris as a function of ATP concentration at several MeSO concen- maleate buffer, pH 6.0, 5 mM MgC12, 10 mM EGTA, and 10% (v/v) Me2S0 (O), 20% (v/v) Me2S0 (A), or 30% (v/v) Me2S0(0).Reaction trations, it was observed that the concentration of ATP was initiated by adding up to 0.05 ml of ty3'P}ATP toeach milliliter required for half-maximal velocity decreased severalfold for of reaction medium to give the final concentrations indicated in the RESULTS AND DISCUSSION

-

ubcissa, and arrested at 30 s for 10%Me'SO, 1.5 min for 20% Me,SO, and 3 min for 30% Me2S0 in order to ensure that initial velocities of enzyme phosphorylation were measured. EP was determined according to "Experimental Procedures." 1 20

I

0 1 2 3 4 5

IO TIME (minutes)

I

I

15

20

0

K)

20

30

DMSO (% "/V

40

"

)

FIG.1. Time course of M&+-dependent enzyme phospho-

FIG.3. Effect of MeaSOon phosphoenzymelevels and M&+-

rylation and hydrolytic activity. Purified Ca2+-ATPase(0.7 mg/ ml) was preincubated for 2 min at 10 "C in media containing 10 mM EGTA, 30 mM Tris maleate buffer, pH 6.0, 5 mM MgCl,, and 30% (v/v) Me,SO. Reaction was initiated by adding 0.05 ml of 10 m M (T-~'P]ATP toeach milliliter of reaction medium. Pi release from ATP (A) and phosphoenzyme level (0)were determined according to "Experimental Procedures."

dependent hydrolytic activity. Pi release from ATP (A) and phosphoenzyme levels (0)were measured at 10 "C in media containing 60 mM Tris maleate buffer, pH 6.0, 5 mM MgCl,, 10 mM EGTA, 0.5 mM fy3'P)ATP, and Me's0 concentrations as indicated. Reaction was initiated by adding purified Ca2+-ATPase (0.7 mg/ml), previously incubated for 2 min in the same medium in the absence of ATP, and was arrested after 30 min to 1 h.

6612

Hydrolytic Cycle of SR Ca"-ATPase

to the decrease in the apparent KM for ATP, whereas the inhibition observed at higher solvent concentrations with concomitant accumulation of phosphoenzyme is consistent with the secondary effect of MezSO in diminishing the rate of phosphoenzyme breakdown (6) (see Table I). Relationship between theRate of Phosphoenzyme Decay and Hydrolytic Activity in the Absence of CaZ+-The radioactive phosphoenzyme formed in the absence of Ca2+decays monotonically and two components may be distinguished kinetically. Their decay was measured by diluting the phosphorylation media in a mixture containing nonradioactive ATP (Fig. 4). Approximately 70% of the phosphoenzyme decayed rapidly in the presence of 15% Me2S0, whereas only 50% decayed rapidly in the presence of 30% Me2S0. The rates of both components (fast and slow) decreased at the higher solvent concentration. Both phasesof phosphoenzyme decomposition obeyed first-order kinetics, allowing the fraction of the slowly decomposing phosphoenzyme present initially to be determined by extrapolation. Therate of the fast component (dotted line, Fig. 4) was calculated by subtracting the extrapolated values of the slow component from the observed values at the initial times and replotting the resulting values (12, 15). The apparent first-order rate constants governing phosphoenzyme decay at different solvent concentrationsare shown in Table I. Theoretical values of ATP hydrolysis were calculated for three possible cases, supposing that: 1) only the decay of the fast component was responsible for Pi liberation, 2) only the decay of the slow component was responsible for Pi liberation, or 3) Pi liberation was caused by simultaneous decomposition of both components. Comparison with the experimental values of ATP hydrolysis in the steady state (Fig. 3) easily disqualified the second possibility. Therefore, the decomposition of the slow component does not represent the ratelimiting step of a reaction cycle. Cases l and 3, however, are indistinguishable since the contribution of the slow component is relatively small. Both cases are in good agreement with the experimental values and indicate that at least the

010 4

ll

I

I

0 5 10 20

I

40

I

,

60 TIME (minutes )

,

,

,

,

90

FIG. 4. Rate of phosphoenzyme decay. Purified Ca2+-ATPase (5 mg/ml) was incubated for 2 min in a medium containing 60 mM Tris maleate buffer, pH 6.0, 5 mM MgCI2, 10 mM EGTA, and 15% (v/v) Me2S0 (A) or 30% (v/v) Me2S0 (B).Reaction was initiated by to 1.5 ml ofthe reaction medium adding 0.05 ml of 15 mM (y-32P)ATP and carried out during 20 min (A) or 30 min (B) at 10 "C. Dephosphorylation was initiated by diluting the labeled phosphoenzyme with 20 volumes of the same medium used for enzyme phosphorylation except that 0.5 mM nonradioactive ATP was used.EP was determined as described under "Experimental Procedures." EP, refers to the amount of phosphoenzyme at time t of dephosphorylation, EP, is the amount of phosphoenzyme remaining after 180 min, and EP.,, is the maximal amount of phosphoenzyme formed (zero time of dephosphorylation). The dotted line shows the fast component of EP decay, after substraction of the slow component.

fast component (and possibly both) is a competent intermediate of the reaction cycle of ATP hydrolysis in the absence of Ca2+.It is possible that theinhibition of the ATP hydrolysis shown in Fig. 3 at high Me2S0 concentrations is due to the accumulation of slowly decaying component. Comparison of the Intermediate Formed in the Presence and TABLE I Absence of Ca2+-Several phosphoenzyme intermediates are Rate constantsfor EP decay and velocities of ATP hydrolysis in the sequentially formed during the reaction cycle in the presence absence of Ca2+ of Ca2+.These forms are kinetically distinguished by their The constants for phosphoenzyme decay and thepercentage of the sensitivity to ADP and KC1. The initially formed, ADPdifferent components in the steady state were obtained from Fig. 4, A and B. The maximal EP levels observed in these preparationswere sensitive phosphoenzyme transfers its phosphate to ADP in 1.27 nmol of E P . mg of protein" in the presence of 30% Me2S0 and the presence of Ca2+concentrations sufficient to saturate the 0.36 nmol of E P . mg of protein" in the presence of 15%Me2S0. The low-affinity Ca2+-bindingsites (12, 25-28). The ADP-'insentheoretical velocity of Pi liberation attributable to thedecomposition sitive phosphoenzyme is sensitive to KC1 (25-28). of each of the different components was calculated by multiplying In a medium containing 30% MezSO,the rate constants of kobs by the amount of that component present in the steady state decay of the intermediates formed in theabsence of Ca2+were according to u = hb.(EP).The experimental velocities of Pi liberation similar to the rates of decay of the intermediates formed in were taken from Fig. 3. the presence of Ca2+(Fig. 51, despite large differences in levels M e 8 0 (v/v) of phosphoenzyme accumulated (1.1 nmol of EP.mg of pro15% 30% tein" in the absence of Ca2+and approximately 4 nmol of EP fast decay EP .mg of protein" in its presence). Addition of KC1 to the 50% 69% 3' 6 of total dephosphorylation media greatly enhanced the velocity of 1.05 min" 0.225 min" kobs decay of both phosphoenzymes (Fig. 5A). In contrast, the intermediates formed in the presence of MezSO were insenE P slow decay sitive to ADP, regardless of whether Ca2+was present or not kobs min" 0.013 0.056 min" (Compare Fig. 5, A and B ) .This is in accordance with previous Theoretical velocities (nmol Pi observations of de Meis et al. (22) that Me2S0 inhibits the mg" h-'1 transfer of phosphate from the phosphoenzyme to ADP. 8.6 15.6 From EP fast Shigekawa and Akowitz (12) showed that, during the steady 0.4 From EP slow 0.5 state of the Ca2+-activatedhydrolytic cycle in the presence of 16.0 From EP fast + EP9.1slow organic solvents, the KC1-sensitive form ispredominant. Our 10.417.5 Experimental velocity (nmol Pi results are suggestive of a certaindegree of similarity between mg" h-') the KC1-sensitive phosphoenzyme formed in the presence of

Hydrolytic Cycle of SR Ca2+-ATPase

I, 0

6613

1

I

20

40

I

60 KCL (mM)

I

80

100

FIG. 6. Effect of KC1 on phosphoenzyme levels and M9+dependent hydrolytic activity. Purified Ca2+-ATPase (0.7 mg/ml) was preincubated for 2 min in media containing 10 mM EGTA, 30 I I I I mM Tris maleate buffer, pH6.0,5 mM MgCI,, 30%(v/v) Me2S0,and 0 30 60 90 120 the indicated KC1 concentrations. Reaction was initiated by adding TIME(secmds) 0.05 ml of 10 mM (-y-32PJATP to each milliliter of reaction medium FIG. 5. Effect of KC1 and ADP on the decay of the phos- and carried out at 10 "C during 2 h.Pireleasefrom ATP (A) and phoenzyme formed in the presence of Me2S0. In A, enzyme (5 phosphoenzyme levels (0)were determined according "Experimental mg/ml) wasincubated in media containing 60 mM Tris maleate buffer, Procedures." The level of Ep in the absence of KC1 was 1.2 nmol of pH 6.0,lO mM MgCIz,30% MezSOand 10 mM EGTA (A,A) or 5 mM EP/mg of protein, and Pi release from ATP in the absence of KC1 CaClz ( 0 , O ) . Phosphorylation was initiated by adding 0.05 ml of 15 was 11.2 nmol of PJmg of protein/h. mM {-y-32PIATP to 1.5 ml of the reaction medium and carriedout for 30 s a t 10 "C in the presence of Caz+, orfor 30 min in its absence. Dephosphorylation was initiated by diluting the labeled phosphoen- shifted to the micromolar range in the presence of M e 8 0 zyme with 20 volumes ofthe same medium used for enzyme phospho- when Ca2+is absent. rylation except that 0.5 mM nonradioactive ATP was used and the It has been proposed that Pi and furylacryloylphosphate dephosphorylationmedia contained either no additions (0,A) or 100 bind to the enzymatic form *E(22, 30). Fast kinetic experimM KC1 (0, A). In B, enzymewasphosphorylatedin the same ments have suggestedthat ATP can also bind to the enzymatic conditions described forA in the presence of 10 mM EGTA (A)or 5 form *E with low affinity, displacing the distribution of the mM CaC& (0).Dephosphorylation was initiated by diluting the labeled phosphoenzyme with 20 volumes of the same medium used for enzymatic form in favor of formation of E (32). It may be enzyme phosphorylation exceptthat 0.5 mM unlabeled ATP was used that, in the presence of organic solvents and the absence of and 2.66 mM ADP was added. The initial levels of EP formed were Ca2+, ATP binds with high affinity to the enzymaticform *E 1.1nmol of EP/mg of protein in the absence of Ca2+ and4.0 nmol of and isutilized as substrate through this form, bypassing Ca2+ EP/mg of protein in the presence of 5 mM CaCl2. stimulation. That explanation would be in agreement with a previous proposal in the literature that the enzymatic form Ca2+ and that which accumulates in itsabsence. *E has a hydrophobic catalytic site and that the additionof Effect ofKC1 on the Hydrolytic Cycle in the Absence of organic solvents to an aqueous medium inthe absence of Ca2+ Ca'+"At saturating Ca2+ concentrations, KC1 increases the increases the affinity of a hydrophilic substrate for thehydrorate of breakdown of the KCI-sensitive phosphoenzyme and phobic site of the enzyme (22). Additionof Ca2+would expose the rateof ATP hydrolysis without changing the total steady-the catalytic site to the hydrophilic medium (22,23). The and state phosphoenzyme level (12, 25, 26, 29). I n Fig. 5A we Hasselbach (24) demonstrated that, at saturating Ca2+conshowed that, in our conditions, KC1 accelerates the velocity centrations, organic solvents have no effect on the apparent of dephosphorylation of the intermediates formed both in theKMfor ATP. These results can be accounted for by a shift in presence and in the absenceof Ca2+.The effects of KC1 were the equilibrium between the enzymatic forms containinghyin the same order of magnitude under both conditions. In drophobic and hydrophilic catalytic sites since under these contrast, in the absence of Ca2+,KC1 did not accelerate the conditions the enzymaticform E:: is predominant. velocity of ATP hydrolysis and greatly decreased the steadyAn alternative explanationis suggested by the observation state phosphoenzyme level (Fig. 6). The effect of KC1 on the that Me2S0 modifies the iodoacetamide spin label signal of the enzymatic form *E in the same direction that Ca2+does steady-state level of phosphoenzyrne is probably due to the acceleration of the rateof phosphoenzyme breakdown. These (33). It may be that the presenceof solvent stabilizes a form results suggest that, in the absence of Ca2+,KC1 not only that is morelike E than *E with respect to its substrate accelerates thehydrolysis of phosphoenzyme, but also changes affinity. In any case, Fig. 2 shows that the enzymatic form is the rate-limiting step of the reaction cycle t o some step that distinct from the form containing bound Ca2+ and that it binds ATP as well as Pi and furylacryloylphosphate. is involved in phosphoenzyme formation. Our results show that purified Ca2+-ATPase from sarcoAcknowledgment-The technical assistance of Monica Freire is plasmicreticulumdisplays a n exclusivelyM$+-dependent greatly appreciated. hydrolysis of ATP which is enhanced by Me's0 and that, during the reactioncycle, at least one phosphoenzyme interREFERENCES mediate is formed. The enzyme alsohydrolyzes furylacryloyl1. MacLennan,D. H., Seemaan, P., Iles,G. H., andYip, C. C. (1971) phosphate, a pseudo-substrate, in the presence of M$+ alone. J. Biol. Chem. 246,2702-2710 The hydrolysis of this compound is similarly activated by 2. Barrabin, H. W., Scofano, H. M., and Inesi, G. (1984) BiochemMe2S0, and a phosphoenzyme is formed during the cycle (30). istry 23,1542-1548 It is noteworthy that the affinity of substrates as different as 3. de Meis, L. (1981) in The Sarcoplasmic Reticulum: Transport and EnergyTransduction.TransportintheLifeSciences, Vol. 2 Pi (22,31), ATP(Fig. 2), andfurylacryloylphosphate (30) are

6614

Hydrolytic Cycle of SR Ca2+-ATPase

(Bittar, E.,ed) pp. 69-100, Wiley Interscience, New York 4. Ikemoto, N. (1982) Annu. Reu. Physiol. 44,297-317 5. Meissner, G., Conner, G. E., and Fleischer, S. (1973) Biochim. Biophys. Acta 298,246-269 6. Carvalho-Alves, P. C., and Scofano, H. M. (1983) J. Bwl. Chem. 258,3134-3139 7. Lowry, 0.H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193,265-275 8. Glynn, I. M., and Chappel, J. B. (1964) Biochem. J. 90,147-149 9. de Meis, L. (1972) Biochemistry 11,2460-2465 (1974) Biochemistry 13, 10. de Meis, L., and Carvalho, M.G.C. 5032-5038 11. Grubmeyer, C., and Penefsky, H. S. (1981) J. Biol. Chem. 256, 3718-3727 12. Shigekawa, M., and Akowitz, A. A. (1979) J. Biol. Chem. 254, 4726-4730 13. Froehlich, J. P., and Taylor, E. W. (1975) J. Bwl. Chem. 250, 2013-2021 14. de Meis, L.,de Souza Otero, A., Martins, 0. B., Alves, E. W., Inesi, G., and Nakamoto, R. (1982) J. Biol. Chem. 257,49934998 15. Inesi, G., Kurzmack, M., Kosk-Kosicka, D., Lewis, D., Scofano, H., and Guimarles-Motta, H. (1982) 2. Naturforsch. 37c, 685691 16. Carvalho, M. G. C., de Souza, D. G., and de Meis, L. (1976) J. Biol. Chem. 251,3629-3636

17. Makinose, M. (1969) Eur. J. Biochem. 10,74-82 18. Martonosi, A. (1969) J. Biol. Chem. 244,613-620 19. Bastide, F.,Meissner, G., Fleischer, S., and Post, R. L.(1973) J. Biol. Chem. 248,8385-8391 20. Degani, C., and Boyer, P. D. (1973) J. Biol. Chem. 248, 82228226 21. Masuda, H., and de Meis, L.(1973) Biochemistry 12,4581-4585 22. de Meis, L.,Martins, 0. B., and Alves, E. W. (1980) Biochemistry 19,4252-4261 23. de Meis, L., and Inesi, G. (1982) J. Biol. Chem. 257,1289-1294 24. The, R., and Hasselbach, W. (1977) Eur. J. Biochem. 74, 611621 25. Shigekawa, M., and Dougherty, J. P. (1978) J. Biol. Chem. 253, 1458-1464 26. Shigekawa, M., Dougherty, J. P., and Katz, A.M. (1978) J. Biol. Chem. 253,1442-1450 27. Yamada, S., and Ikemoto, N. (1980) J. Biol. Chem. 256, 47264730 28. Dupont, Y. (1980)Eur. J.Biuchem. 109,231-238 29. Duggan, P. F. (1977) J.Biol. Chem. 252,1620-1626 30. Inesi, G., Kurzmack, M., Nakamoto, R., de Meis, L., and Bernhard, S. A. (1980) J. Biol. Chem. 255, 6040-6043 31. Champeil, P., Guillain, F., Venien, C., and Gingold, M. P. (1985) Biochemistry 24,69-81 32. Scofano, H.M., Vieyra, A., and de Meis, L. (1979) J. Biol. Chem. 254,10227-10231 33. Coan, C. (1983) Biochemistry 22,5826-5836