Transient State Kinetic Studies of Ca2+-dependent ATPase and Calcium Transport by Cardiac Sarcoplasmic Reticulum

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 255, No. 5, Issue of March 10. pp. 1985-1992, 1980 Printed in U.S.A. Transient State Kinetic Studies of Ca2+-dependent ATPase and Calcium Transport by Cardiac Sarcoplasmic Reticulum EFFECT OF CYCLIC AMP-DEPENDENT PROTEIN KINASE-CATALYZED PHOSPHORYLATION OF PHOSPHOLAMBAN* (Received for publication, July 20, 1979) Michihiko Tada,+ Makoto Yamada, F umio Ohmori, Tsunehiko Kuzuya, Makoto Inui, and Hiroshi Abe From the Division of Cardiology, First Department of Medicine, Osaka University School of Medicine, Fukushima-Ku, Osaka 553, Japan Phosphorylation of the 22,000-dalton protein phospholamban of cardiac microsomes (sarcoplasmic reticulum) by CAMP-dependent protein kinase was shown to alter profoundly the intermediary steps of microsomal Ca +-dependent ATPase. To define the control of ATPase by its putative regulator phospholamban, we examined transient kinetics of the formation of the phosphorylated intermediate EP of ATPase and calcium binding by the membrane through a rapid quenching device, after microsomes were incubated with and without CAMP-dependent protein kinase. When micro- somes were suspended with ethylene glycol bis( 8-aminoethyl ether)njv tetraacetic acid (EGTA) prior to the assay (Ca +-free microsomes), the initial rates of EP formation and calcium binding were markedly enhanced in phosphorylated microsomes, whereas those incubated with calcium/egta buffer (Ca2+-bound microsomes) exhibited less pronounced enhancement. In Ca +-free microsomes, these effects were evident within a wide range of ionized Ca2+ between 0.1 and 20 pm. Under these conditions, the initial rates and maximal amounts of EP formation were enhanced significantly by microsomal phosphorylation above 10 p~ ATP, while those were reduced at lower ATP (1 to 6 p~). Latter effects were attributed to the predominance of a protein kinase-induced increase in EP decomposition at lower ATP ranges, where the EP formation is ratedetermining. These results indicate that CAMP-dependent phosphorylation of phospholamban produces marked increases in the rate of a step associated with calcium binding to the ATPase enzyme and the rate at which EP is subsequently formed, resulting in an increased rate at which calcium is translocated across the microsomal membrane. Together with our previous report (Tada, M., Ohmori, F., Yamada, M., and Abe, H. (1979) J. Biol. Chern. 264,319-326) that phospholamban phosphorylation induces a marked increase in the rate of EP decomposition, the present observations suggest that phospholamban, through modulation of elemen- * This work was supported by research grants from the Ministry of Education, Science and Culture of Japan and by a Grant-in-Aid from the Muscular Dystrophy Association of America. A preliminary report of this paper was presented at the XIth International Congress of Biochemistry, July, 1979, Toronto, Canada (I), and at the Second Assembly of the Japanese Section of the International Society for Heart Research, November, 1979, Osaka, Japan (2). The costaof publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 4 TO whom correspondence concerning this paper should be addressed. 1985 tary steps of ATPase, could function as a regulator controlling the active calcium transport by cardiac sarcoplasmic reticulum. Our previous report (3) indicated that CAMP-dependent protein kinase (EC 2.7.1.37, ATP:protein phosphotransferase)-catalyzed phosphorylation of a 22,000-dalton protein phospholamban of cardiac sarcoplasmic reticulum results in increased Ca2+-dependent ATPase activity (EC 3.6.1.3, ATP phosphohydrolase) of cardiac sarcoplasmic reticulum, through an enhancement of the rate of decomposition of the phosphorylated intermediate EP of the ATPase. These observations suggest that the phosphorylated and unphosphorylated states of phospholamban (4) can determine the extent at which the elementary step(@ of Ca2+-dependent ATPase is stimulated, and are thus in support for the view that phospholamban may function as a regulatory factor controlling the active calcium transport by cardiac sarcoplasmic reticulum (5-7). This proposal was consistent with a number of findings that the rates of calcium transport and Ca2+-dependent ATP hydrolysis can be augmented when CAMP-dependent protein kinase cata- lyzes phosphorylation of a microsomal protein of 20,000 to 22,000 daltons (4, 8-16). Resultant increases in the rate and amount of calcium accumulation into sarcoplasmic reticulum may account for the relaxing and inotropic responses of the myocardium when exposed to agents that increase intracellular CAMP levels (5, 6, 17, 18). Active calcium transport by sarcoplasmic reticulum from skeletal and cardiac muscles is considered to be mediated by an ATPase enzyme of about 100,OOO daltons, which serves as an energy transducer and a translocator of calcium ions, undergoing rapid and complex series of elementary steps in enzyme states during the translocation of calcium across the membrane (19). Our previous study demonstrated that CAMPdependent phosphorylation of phospholamban increases the rate at which the intermediate (EP) of the ATPase is decomposed (3). However, in view of the extreme rapidity at which EP is formed (20-23), it is absolutely necessary to examine by employing a high performance rapid mixing device whether the rate of EP formation, which takes place within tens of milliseconds (19), is enhanced by CAMP-dependent phosphorylation of phospholamban. It is also necessary to determine whether such alterations in enzymatic events could accom- The abbreviations used are: protein kinase, adenosine 3 :5 monophosphate-dependent protein kinase; EGTA, ethylene glycol bis(8- aminoethyl ether)n,n tetraacetic acid; EP, the phosphorylated intermediate of Ca2 -dependent ATPase of cardiac sarcoplasmic reticulum; pca, - log [ionized Ca2+].

1986 Protein Kinase Effects on Transient Kinetics of Ca2+-ATPase pany a change in the rate of translocation of calcium ions. We therefore investigated in the present paper the transient state kinetic properties of Ca2+-dependent ATPase and calcium transport by cardiac sarcoplasmic reticulum, whose phospholamban is at unphosphorylated and phosphorylated states, through a chemical quench flow apparatus, originally designed by Froehlich et at. (24) to meet measurements of extremely rapid reactions. EXPERIMENTAL PROCEDURES the ATPase enzyme to react consecutively with up to two substrates, or ligands, or both, prior to quenching the reactions with acid or chelating agents. In most experiments, the enzyme syringe contained cardiac microsomes and various concentrations of EGTA (200 p~ to 4.64 mm), the substrate syringe contained [Y-~'P]ATP(2 to 100 PM; 10 to 100 pci/pmol) and 200 p~ CaC12, and the trichloroacetic acid syringe contained 20% trichloroacetic acid with 2 m ATP and 0.5 mm KHzPO4. The standard vehicle solution for the enzyme and substrate syringes contained 100 mm KC], 3 mm MgC12, 5 m~ NaN3,.and 20 mm Tridmaleate (ph 6.8), unless otherwise stated. The temperature was maintained at 20 & 0.5'C by a constant temperature circulator. In some experiments, the enzyme syringe contained cardiac microsomes and calcium/egta buffers which gave various concentrations of ionized Ca'+, and the substrate syringe contained [y- 3ZP]ATP and calcium/egta buffer, whose concentration was identical with that in the enzyme syringe. For the experiments to determine phosphorylated enzyme (EP), a constant volume (1.8 to 2.0 ml) of the assay solution was added to a solution containing 1.26 mg of Materials-Cardiac microsomes, which consist largely of fragmented sarcoplasmic reticulum, were prepared from dog heart ventricle by the procedures of Harigaya and Schwartz (25) with slight modifications (3, 11). Microsomes were stored on ice and used within several hours after preparation. CAMP-dependent protein kinase was prepared as described previously (3). Protein concentrations were measured by the method of Lowry et al. (26). [y-3zp]atp(20 to 35 bovine serum albumin, and the amount of "P incorporated was mci/pmol) and "CaC12 (-3 mci/pmol) were purchased from New England Nuclear or the former was synthesized by the method of Glynn and Chappell (27). Orth~[~~PJphosphate (carrier-free) was obtained from Japan Radioisotope Association, Tokyo. ATP, ADP, CAMP, and calf thymus histone (type 11-A) were obtained from Sigma; Dowex 1-X8 (100 to 200 mesh) from Dow Chemical; DEAEcellulose (DE52) from Whatman; EGTA from Dojin Chemicals, Todetermined by washing with perchloric acid, as described previously (3). The amount of 3zPi liberated was determined by the procedures of Martin and Doty (29), slightly modified as described previously (3). Henceforth, cardiac microsomes with EGTA in the enzyme syringe will be designated as Ca2+-free microsomes (enzyme), and those with calcium/egta buffer will be designated as Ca'+-bound microsomes (enzyme). kyo; MiUipore filters (type HA, 0.45 pm) from Millipore Cow. Cal- Calcium-binding Studies-"Ca2' binding by cardiac microsomes cium/egta buffers contained, in final concentrations, 100 p~ CaC12 and various concentrations of EGTA, ionized Ca2+ concentrations at was examined under conditions essentially similar to those for the ATPase assay with replacing unlabeled ATP for [y-"'piatp, accordph 6.8 being calculated with the association constant of4.5 X IO5 ing to the method of Sumida et al. (23, 30). Briefly, the substrate M" for the calcium. EGTA complex (28). For preparing reagents and solutions for rapid quenching studies, glass-distilled deionized water was filtered through Millipore filters to remove fine particles and threads which could otherwise occlude the thin channels in Berger ball mixer. All other chemicals were of analytical grade. Microsomal Phosphorylation Catalyzed by CAMP-dependent Protein Kinase-Microsomal protein was phosphorylated at 25 C for 2 to 5 min as described previously (3) in the presence of y3'p-labeled (Procedure I) or unlabeled (Procedure 11) ATP. The former was performed to monitor the extent of phospholamban phosphorylation syringe contained 20 p~ ATP and 200 h~ 45CaC12 (3 to 4 pci/pmol), the enzyme syringe contained cardiac microsomes and various concentrations of EGTA, and the EGTA syringe contained 6 mm EGTA in experiments with Ca*+-free microsomes. The vehicle solution was the same as that used in the ATPase assay. In some experiments with Ca2+-bound microsomes, conditions essentially similar to those for the ATPase assay were used (replacing unlabeled ATP and labeled Ca2+ for labeled ATP and unlabeled Ca", respectively). Solutions from the enzyme and substrate syringes were mixed together at 20 C in the fvst mixer, and after various times (5 to 200 ms) were quenched in each microsomal preparation, and the latter was performed as with a solution containing 6 mm EGTA in the EGTA syringe. Impreincubation prior to assay for kinetics of ATPase and calcium mediately after quenching, 2 mlof the assay solution was passed transport. In the latter, cardiac microsomes (4.4 to 8.0 mg of protein/ ml) were incubated with or without 2.2 to 4.0 mg/d of CAMPdependent protein kinase in 1 mm ATP, 1 p~ CAMP, 30 mm KC1, 2 mmmgc12, and 20mM histidine/hcl (ph 6.8) in a total volume of 3 to 5 ml. After 2 to 5 min of incubation at 25"C, 4 volumes of ice-cold 20 mm Tris/maleate (ph 6.8) was added and the mixture was applied to a column of Dowex ]-X8 (1.3 X 2 cm) pre-equilibrated with the through a Millipore filter, and the precipitate with retained 46Ca was washed with 1 mm EGTA in 50 mm Tris/maleate (ph 6.8) and 100 mm KC1 at 0-4'C. After drying the Millipore filter was liquefied with acetone and counted for radioactivity. Sodium Dodecyl Sulfate-Polyacrylamide Gel Electrophoresis- Gel electrophoresis of microsomal protein was performed according to the procedures of Weber and Osborn (31), with slight modifications same buffer at 0-4 C. Microsomes preincubated without or with (4), on 0.1% sodium dodecyl sulfate and 8% polyacrylamide gels. protein kinase, which are henceforth designated as control (unphosphorylated) and phosphorylated microsomes, respectively, were RESULTS eluted from the column at a flow rate of 1 d/min at 0-4'C, and were subjected to assay for E"'P formation, 32P1 liberation, and 45Ca2+ Effect of CAMP-dependent Protein Kinase on Presteady binding, as described below. In some experiments, microsomes prein- State Time Courses of Ca2+-dependent EP Formation and cubated with 2 mm EDTA and protein kinase served as a control. Calcium Binding-A brief incubation of cardiac microsomes When examined by incubating cardiac microsomes and [Y-~'P]ATP with CAMP-dependent protein kinase resulted in marked enby Procedure I, the incubation with protein kinase resulted in signif- hancement of the presteady state time courses of EP formaicant phosphorylation of phospholamban (0.6 to 0.8 pmol of phosphate tion and calcium binding. These effects are depicted in Figs. incorporated/g of microsomal protein), whereas the incubation under identical conditions without protein kinase or with EDTA and protein 1 and 2, in which control and phosphorylated microsomes kinase gave virtually no phosphorylation (t0.02 pmol of phosphate/ were allowed to react with ATP and calcium after pretreatg). ment with EGTA (Ca2+-free microsomes) and calcium/egta Transient State Kinetic Analysis of Ca'+-dependent ATPase- buffer (Ca2+-bound microsomes), and the amounts of EP Cardiac microsomes were phosphorylated by CAMP-dependent pro- formed and calcium bound were determined with time intertein kinase and unlabeled ATP and were passed through the anion vals between 5 and 200 ms. As shown in Fig. 1, the initial rate exchange resin according to Procedure I1 (3), as described above, in of EP formation in Ca"-free microsomes was greatly enorder to remove unreacted ATP and products ADP and P,, which would otherwise interfere with kinetic properties of ATPase (19). To hanced by CAMP-dependent phosphorylation, whereas that the resultant eluate (IO to 20 ml) was added 8 ml of a solution that in Ca2+-bound microsomes, which occurred much more rapgave concentrations of 3 mm MgCL, 100 mm KCl, 5 mm NaN3, and 20 idly than the former, exhibited less evident increase after mm Tris/maleate (ph 6.8). The final concentrations of these in the CAMP-dependent phosphorylation. When the time at which ATPase assay medium were maintained by the inclusion of a vehicle half of the maximal amount of EP was attained (tl/z) was solution in the substrate syringe (see below). R * rid mixing experidetermined, CAMP-dependent phosphorylation produced n(h ments were carried out employing a slightly modified version of a chemical quench flow apparatus of Froehlich et al. (24), which was alteration in Ca2+-bound microsomes (t1/2 = 12 to 13 msh, devised for us by Commonwealth Technology, Inc., Alexandria, VA. while in Ca2+-free microsomes the t1/2 value of phosphorylated This device, equipped with four syringes and three mixers, allowed microsomes (21 ms) was much smaller than that of control

Protein Kinase Effects on Transient Kinetics of Ca2+-ATPase 1987 (43 ms). Fig. 2 shows time courses of calcium binding in the same microsomal preparation under identical conditions. In agreement with the data on EP formation (Fig. l), CAMPdependent phosphorylation produced significant increases in the initial rates of calcium binding in both Ca2+-free and Ca2+- symbols) and with (closed symbols) CAMP-dependent protein kinase (3.8 mg/ml) under the standard conditions. After dilution and treatbound microsomes. However, in the latter microsomes the ment with the anion exchange resin, they were added to the standard calcium initially bound by the enzyme (at time zero) before vehicle solutions containing EGTA (0, 0) or 4"CaC12/EGTA buffer ATP had been added contributed to increase the total amount (A, A). Calcium binding was subsequently started by adding 10 p~ ATP and 0.1 mm 45CaC12 of calcium binding in both control and phosphorylated micro- (0, 0) which gave 14 PM ionized 4sCa'+ by combination with EGTA in the enzyme syringe, or by adding 4"CaC12/ somes. The tl/2 values in both cases exhibited marked shorten- EGTA buffer with 14 p~ ionized 4sCa2' (A, A), through rapid quenching by CAMP-dependent phosphorylation. They were 21 ms ing apparatus as described under "Experimental Procedures." The for phosphorylated microsomes and 38 ms for control in Ca2+- amount of 45Ca2' bound was determined as described previously. free microsomes; corresponding values in Ca2'-bound microsomes were 12 and 19 ms. TABLE I It is of significance to point out that the apparent stoichi- Effect of CAMP-dependentphosphorylation on the stoichiometry ometry between the amount of EP formed and calcium bound between the initial rates of EP formation and calcium binding by during the initial period was maintained at about 1:2, in agreement with the kinetic analysis of this system (IS), as long as the both determinations were performed simultaneously in the same preparation. Thus, when the data in Figs. 1 and 2 are compared, the initial rates of EP formation and calcium binding in both Ca"-free and Ca2+-bound microsomes showed approximately 1:2 stoichiometry, and it was well maintained after augmentation by CAMP-dependent phosphorylation (Table I). In view of the relative complexity of transient time courses in Ca2+-bound microsomes, in which EP was formed rapidly and was decomposed precipitously and calcium binding was obscured by the calcium initially bound by microsomes (Figs. 1 and 2), the following studies to analyze the initial rates of EP formation and calcium binding were largely performed by employing Caz+-free microsomes, whose time courses followed the fvst order kinetics. Also, in analyzing the initial rates of EP formation and calcium binding, they were assessed by determining directly (micromoles of phosphate and calcium per g of protein in 10 ms) by taking the initial few points during the rise of the time courses, rather than taking the tl/l values, because the latter did not represent the true value due to a significant increase in the levels of EP formed and calcium bound in phosphorylated microsomes (Figs. 1 and 2). It should also be added that a brief incubation of microsomes with different concentrations of EGTA at ph 6.8, conditions under which microsomes were kept in the enzyme syringe in studies with Ca2'-free microsomes, did not appreciably alter the ATPase and calcium-binding activities of the microsomes for con- REACTION TIME (msec) siderable period of time (up to 2 h). Effect of CAMP-dependent Phosphorylation on Ca2+ De- FIG. 1. Effect of treatment with CAMP-dependent protein kinase on the initial rate of Ca2+-dependent formation of phos- pendence Profile of EP Formation-When the initial rate of phorylated intermediate EP of ATPase of CaZ+-free and Ca2+- EP formation and the maximal amount of EP formed were bound cardiac microsomes. Cardiac microsomes (7.8 mg/ml) were examined as function of pca in Ca2+-free enzyme (Fig. 3), the preincubated under standard conditions with (closed symbols) and without (open symbols) 3.9 mg/ml of CAMP-dependent protein kinase at 25 C for 2 min in a total volume of 3.0 ml. After application of the diluted reaction mixture to a column of Dowex 1-X8 at 0-4"C, the eluted microsomes were added to the standard vehicle solution containing EGTA (0,O) which gave 14 p~ ionized Ca" by combination with 0.1 mm CaCL in the substrate syringe, or calcium/egta buffer with 14 p~ Ca2+ (A, A). The ATPase reaction was subsequently started by the addition of 10 PM [y3'p]atp and 0.1 mm CaCL (0, 0) or calcium/egta buffer (A, A), respectively, through the rapid chemical quench flow apparatus, as described under "Experimental Procedures." Microsomal protein concentration in the fmal ATPase assay medium was 0.41 mg/ml. The amount of EP formed was determined as described previously. Cardiac microsomes, preincubated with [y3'p]atp and protein kinase under conditions identical with the preincubation prior to the ATPase assay, exhibited phosphoester phosphorylation of 0.63 pmol of phosphate/g of microsomal 0 i' J protein, whereas those preincubated without protein kinase gave 0.02 0 50 IO0 I50 200 REACTION TIME (msec) pmol of phosphate/g of microsomal protein. FIG. 2. Effect of treatment with CAMP-dependent protein kinase on the initial rate of calcium binding by Ca2+-free and Ca2+-bound cardiac microsomes. The cardiac microsomal preparation was the same as that used in the experiment shown in Fig. 1. Cardiac microsomes (7.6 mg/ml) were preincubated without (open Ca2+-free and Ca2+-bound cardiac microsomes From the initial phase of time course of EP formation and calcium binding shown in Figs. 1 and 2, respectively, the initial velocities of EP formation and calcium binding were determined, and they were compared in terms of the ratio Ca/EP. Initial velocity Microsomes Ca"-free microsomes CJ+-bound microsomes EP for- Ca Ca/ EP for- Ca Ca/ mation binding EP" mation binding EP" pmol/lo ms.gprotein pmol/io ms.gprotein Control 0.011 0.024 2.18 0.098 0.192 1.96 Phosphoryl- 0.029 0.060 2.07 0.153 0.304 1.99 ated Calculated by dividing the rate of calcium binding by the rate of EP formation.

Protein Kinase Effects on Transient Kinetics of Ca2+-ATPase depicted in Fig. 5, in which the maximal amount of EP formed (Fig. 5A) and the initial rate of EP formation (Fig. 5B) determined from data in Fig. 4 were plotted against ATP concentrations. Fig. 5A shows that the amount of maximal EP formation was lower at lower ATP concentrations (1 to 5-0 20 g ""015 Ca".free Enzyme I4,,M Ca2' "_ 1 ATP concentrattons FIG. 3. Effect of CAMP-dependent phosphorylation on the Ca2+ dependence profiles EP formation in Ca2+-free enzyme. Cardiac microsomes (7.8 mg/ml) were incubated under standard conditions without (open symbols) or with (closed symbols) CAMPdependent protein kinase at 25 C for 3 min in a total volume of 4.0 ml, and were passed through the anion exchange resin after dilution with buffer solution. The resultant eluates were added to the standard vehicle solutions containing different amounts of EGTA. The ATPase reactions were started by the additions of 10 p~ [y3'p]atp and 0.1 mm CaCL through the rapid quenching apparatus, and the amount of EP formed was determined by the standard procedures. Microsomal protein concentration in the final calcium-binding assay medium was 0.43 mg/ml. The initial rates of EP formation (0,O) were estimated by determining the amounts of EP formed per 10 ms, taking the initial slopes formed by initial few points during the upstrokes of the time courses (micromoles of phosphate/g of microsomal protein- 10 ms). Maximal amounts of EP formation (A, A) were those formed at 200 rns after starting the reaction, where the steady state of reactions were almost attained. The inset represents Ca2+ dependence profdes of the half-time (til2) of EP formation taken from the same experimental data. Cardiac microsomes, preincubated with [y-32]atp and with or without CAMP-dependent protein kinase, gave phosphoester phosphorylation of 0.61 and 0.015 pmol of phosphate/g of microsomal protein, respectively. I 0 w 8 2 010 initial rate of EP formation was increased markedly by phosphorylation of phospholamban at concentrations of CaZ+ ranging from 0.2 to 14 p ~ While. the effect of phospholamban phosphorylation on the initial rate is still evident when Ca2' concentrations were 0.2 p~ or lower, as was partially evidenced from the plots of t1/2 against pca (inset, Fig. 3), the levels of EP at the steady state were almost unaltered by phosphorylation at Ca2+ concentrations below 1 p ~ At. higher Ca2+ concentrations of 20 p~ or higher the protein kinase effects on the rate and amount of EP formation were evident. In Ca2+bound microsomes, the stimulatory effects of phospholamban phosphorylation were also seen (Fig. l), although the accurate 0" I 1 analyses were precluded because the reaction was more rapid 0 5 10 IS 20 25 30 [ATPI Cum) and the effect of phosphorylation was less evident. Effect of CAMP-dependent Phosphorylation on EP For- FIG. 5. Effect of CAMP-dependent phosphorylation on the ATP dependence profiles of the maximal amount and initial mation ut Different ATP Concentrations-Fig. 4 typically rate of EP formation. The maximal amounts of EP formation in represents time courses of EP formation by control and phos- control (0) and phosphorylated (0) cardiac microsomes were those phorylated microsomes (Caz+-free enzyme) in the presence of formed at 200 ms after starting the reaction, where the steady state 14 PM Ca2+ and various concentrations of ATP. At higher of the reaction was almost attained, in the experiments shown in Fig. ATP concentrations above 10 p ~ CAMP-dependent, phospho- 4. The inset represents the ATP dependence profiles of the apparent rylation of phospholamban resulted in the stimulation of EP initial velocity (u,) of EP formation by control (0) and phosphorylated (0) microsomes, which was determined by taking the initial slope formation, whereas at lower concentrations of ATP below 5 (micromoles of phosphate/g of microsomal protein- 10 ms), formed p~ phosphorylation of phospholamban produced inhibition of by initial few points during the upstroke of the time course shown in EP formation. These intriguing phenomena are more clearly Fig. 4. a W b!i Oo5 0 0 i/ so 1 0 0 I50 200 REACTION TIME Imsd FIG. 4. Time courses of EP formation by Ca*+-fiee enzyme of control and phosphorylated cardiac microsomes. Cardiac microsomes (8.0 mg/ml) were preincubated under standard conditions without (open symbols) and with (closed symbols) CAMP-dependent protein kinase at 25OC for 3 min in a total volume of 6.0 ml, and were passed through the anion exchange resin after dilution with buffer solution. The resultant eluates were added to the standard vehicle solutions containing 0.2 m~ EGTA (0.1 mm in the final assay medium, giving 14 p~ ionized Ca" by combination with 0.1 m~ CaClz from the substrate syringe). The ATPase reaction was started by the addition of 0.1 mm CaClz and different concentrations of [Y-~'P]ATP, through the rapid quenching apparatus. Microsomal protein concentration in the final ATPase assay medium was 0.42 mg/ml. Cardiac microsomes, preincubated with [y3'p]atp and with or without CAMP-dependent protein kinase, exhibited phosphoester phosphorylation of 0.64 and 0.02 pnol of phosphate/g of microsomal protein, respectively. I

S,E.c, Protein Kinase Effects on Transient Kinetics of Ca2+-ATPase 1989 p ~ and ) higher at higher ATP concentrations (E10 pm) in phosphorylated microsomes than in control. Fig. 5B also shows that the initial rate of EP formation by phosphorylated microsomes was markedly higher at higher ATP concentrations above 10 F, whereas at lower ATP concentrations the rate was lower in phosphorylated microsomes, compared with control (cf Fig. 6). In the latter the observed initial velocity of EP formation is considered to represent merely an apparent rate since the initial rate of Pi liberation exhibited a significant increase in the presence of lower ATP concentrations (1 to 5 p ~ ) where, EP formation was inhibited, when microsomes were phosphorylated by CAMP-dependent protein kinase (Fig. 6 also CJ Figs. 4 and 5). The apparent inhibition of EP formation in phosphorylated microsomes is probably related to the increased rate of EP decomposition (3), which becomes more apparent at lower ATP concentrations where the rate of EP formation is lower, becoming comparable to that of EP decomposition (see "Discussion"). Effect of CAMP-dependent Phosphorylation on Calcium Binding by Cardiac Microsomes-Fig. 7 shows the Ca2+ dependence profile of the initial rate of 45Ca binding control by and phosphorylated microsomes in the assay system in which Ca2+-free microsomes were reacted with 10 p~ ATP and various concentrations of 45Ca2+. The initial rates and maximal amounts of calcium binding were dependent upon Ca2+ concentrations between 0.2 and 14 PM, and exhibited marked increases after CAMP-dependent phosphorylation of phospholamban. The effects were evident even when Ca2+ concentrations were 0.1 p~ or lower. In accord with findings in the EP REACTION TIME (sed FIG.6. Effect of CAMP-dependent phosphorylation on the rate of Pi liberation and Ep formation in Caa+-free microsomes. Cardiac microsomes (7.8 mg/ml) were preincubated under standard conditions without (open symbols) and with (closed symbols) CAMPdependent protein kinase at 25 C for 5 min in a total volume of 6.0 ml, and were passed through the anion exchange resin after dilution with buffer solution. The resultant eluates were added to the standard vehicle solution under conditions identical with those in Fig. 4. The ATPase reaction was started by the addition of 0.1 mm CaCL with 1 ~ L M (0, 0, A, A), and 2 ~ I[yX2P]ATP U (0, M, V, V), and the amounts ofpi liberated (0, 0, 0, H) and EP formed (A, A, V, V) were determined by the standard procedures. Microsomal protein concentration in the tinal ATPase assay medium was 0.40 mg/ml. The inset represents the ATPase dependence profiles of the initial rate (u) of Pi liberation by control (0) and phosphorylated (0) microsomes. Cardiac microsomes, preincubated with [Y-~*P]ATP and with and without CAMP-dependent protein kinase, exhibited phosphoester phosphorylation of 0.63 and 0.01 pmol of phosphate/g of microsomal protein, respectively. - z 1 / 0-O 6 5 PCa FIG. 7. Effect of CAMP-dependent phosphorylation on Caz+ dependence profiles of calcium binding by Caz+-free cardiac microsomes. Cardiac microsomes (8.0 mg/ml) were preincubated under standard conditions without (open symbols) and with (closed symbols) CAMP-dependent protein kinase (4.0 mg/ml) at 25 C for 2 min in a total volume of 6.0 ml, and were passed through the anion exchange resin after dilution with buffer solution. The resultant eluates were added to the standard vehicle solutions containing different concentrations of EGTA. The calcium-binding reactions were started by the additions of 10 p~ ATP and 0.1 m~ 45CaC12 through the rapid quenching apparatus, and the amounts of45ca bound were determined as described under "Experimental Proce- dures.'' The initial rate of calcium binding (0,O) was determined by taking the initial slope (micromoles of "Ca/g of microsomal protein. 10 ms), formed by initial few points during the upstroke of the calcium binding curve, and the maximal amounts of calcium binding (A, A) were those at 200 ms, where the steady states of the reactions were attained. Microsomal protein concentration in the final calcium-binding assay medium was 0.56 mg/ml. Cardiac microsomes, preincubated with [y-3zp]atp and with and without CAMP-dependent protein kinase, exhibited phosphoester phosphorylation of 0.62 and 0.02 pmol of phosphate/g of microsomal protein, respectively. TABLE I1 Effect of CAMP-dependentphosphorylation on the half-time of calcium binding by cardiac microsomes From the initial phase of time course of calcium-binding studies performed under conditions identical with Fig. 7, the half-time (t1/2) to attain the maximal amount of calcium binding was estimated at five different Ca2+ concentrations. Microsomes Half-time (tlt2) of maximal calcium binding at Ca" concentration of Mean ~ 0.2UM 0.5UM 2UM 10UM 14 UM rns rns Control 38 37 36 40 37.6 f 1.4 Phosphorylated 20 22 18 17 18.8 17 -t 1.9 Averages (mean f S.E.) of determinations at five different Ca2' concentrations are shown since the tl,a values are apparently independent of Ca2' concentrations. assay, the t1/2 values of calcium binding was greatly shortened by CAMP-dependent phosphorylation of phospholamban (Table 11), the t ~ values, ~ being almost independent ofca'+ concentrations within a wide range between 0.2 and 14 p ~.

1990 Protein Kinase Effects on Transient Kinetics of Ca2+-ATPase DISCUSSION The present study was aimed at examining the effects of CAMP-dependent protein kinase-catalyzed phosphorylation of phospholamban on the transient state kinetic properties of Ca'+-dependent ATPase and calcium transport by cardiac sarcoplasmic reticulum, since the phosphorylated state of phospholamban was previously shown to be related to a marked increase in the overall rates of ATP hydrolysis and calcium uptake (4, 8-16) through an enhancement of the elementary step(s) of Ca2+-dependent ATPasenzyme of cardiac sarcoplasmic reticulum (3, 6, 7). In view of the extreme rapidity at which the ATPase enzyme reacts with ATP and calcium (19-23), a chemical quench flow device (24) was used to follow the transient time courses of EP formation, Pi liberation, and calcium binding by cardiac sarcoplasmic reticulum with time intervals as short as 5 ms. It was found that CAMP-dependent phosphorylation of phospholamban results in marked increases in both the initial rates and maximal extents of EP formation and calcium binding. The effects were more evident when cardiac microsomes pretreated with EGTA to remove ea2+ bound to the enzyme were allowed to react with ea2+ and ATP. Thus, under such conditions the initial velocities of EP formation and calcium binding were stimulated more than 2-fold within the physiological range of ea2+ (0.1 to 20 p ~ and ) at ATP concentrations above 10 p ~, a major portion of the reaction intermediate (19), which is when CAMP-dependent protein kinase catalyzed phosphoryl- supposed to react with ADP to form ATP, was indicated by ation of phospholamban. It was also found that the stoichio- the experiments in which the sensitivity of EP to ADP was metric coupling between EP formed and calcium transported examined virtually all of the EP in control and phosphorylis well maintained at 1:2 under these conditions, in accord ated microsomes was immediately decomposed when ADP with the thermodynamic analysis of this system (19, 32). was added to EP simultaneously with or shortly after addition Observed alterations in the enzymatic activity produced by of EGTA.~ phospholamban phosphorylation are probably brought out by increases in both the rate at which calcium ions and ATP bind to the ATPase enzyme and the rate at which EP is subsequently formed in these membranes, i.e. either or all of Steps i, ii, and iii in the following equation: ATP" ADP" Q) (i), (iiy, E + 2Ca"d E.Can" 7- E%b 7 ADP" ATP" (1) (iv) ErCa2' 7 E +- Pi" + 2Ca' where i and o indicate the inside and outside of the membrane vesicles, respectively; E represents the ATPase enzyme. At the outer surface of the membrane, 2 mol of Ca" and 1 mol of ATP bind 1 mol of E to form the Michaelis complexe3;. This is followed immediately by the formation of the phosphorylated intermediate E y (referred to as EP throughout the present report) when calcium is translocated from outside to inside the membrane. Calcium is subsequently released from the enzyme to the interior of the vesicle, with the simultaneous decomposition of the intermediate into E and Pi". Among these intermediary steps, Step iv is considered to be rate-determining in the presence of saturating concentrations of ATP and Ca2+, whereas Step ii is not the ratedetermining step unless the ATP concentration is extremely lowered (19). In the present study the overall rate of EP formation (Steps i, ii, and iii) was determined by measuring the initial rate of EP formation during the transient time course, after microsomes were incubated with appropriate concentrations of calcium and ATP, while the apparent initial rate of calcium binding (Step i) was assessed to some extent by determining the rates of calcium binding and EP formation Mg after Ca2+-free and Ca2+-bound microsomes were incubated with calcium and ATP. The apparent initial rate of calcium binding, coupled with EP formation, determined from incubation of Ca2+-free microsomes with calcium and ATP exhibited a marked increase after CAMP-dependent phosphorylation of phospholamban (Figs. 1 and 2). Such an increase was evident within a wide range of Ca2+ concentrations and as low as 0.1 PM Ca2+ in experiments on both calcium binding (Fig. 7) and EP formation (Fig. 3). The apparent half times (t1,2) of calcium binding (Table 11) and EP formation (inset, Fig. 3), which were virtually independent of Ca2+ concentration, also exhibited marked shortening when phospholamban was phosphorylated, although the latter assessment does not represent the true initial rates (see above). These results indicate that the apparent initial rate of calcium binding by the ATPase enzyme (mainly Step i in Equation 1) is markedly enhanced by CAMPdependent phosphorylation of phospholamban.' It was also indicated from these studies and those on the ATP concentration dependence (Figs. 4 and 5) that the rate at which ATP binds to the enzyme (Step ii), or that at which EP is formed from the Michaelis complex (Step iii), or both, was greatly enhanced by phospholamban phosphorylation. The possibility that the observed EP formation represents that of E-P, The present studies on the effect of CAMP-dependent phosphorylation were mainly carried out employing EGTA-treated microsomes (ea"'-free enzymes) in which the effects of protein kinase on the initial rates were more evident and more readily detectable. However, in view of the observations made by Sumida et al. (23) that in Ca2+-bound microsomes the initial amount of transported calcium (at time zero) represents the calcium that was bound to the vesicles before ATP had been added and that the additional calcium would be transported at the rate essentially compatible with that of the Caz+-free microsomes, the protein kinase-induced alterations in enzymatic parameters similar to what were found in Ca2+-free microsomes described above would also occur in Ca2+-bound microsomes as it turns over during the translocation of ea2'. Indeed, we found an increase in the rates and amounts of EP formation (Fig. 1) and calcium binding (Fig. 2) in Ca"-bound microsomes after phosphorylation of phospholamban, although the effects were obscured by the rapidity of the initial rate in the former and the significant amount of calcium initially bound by the enzyme in the latter. We assumed in the previous discussion that the formation of E. Cas" occurs at one step (Step i). However, Sumida et al. (23) showed that Ca2+-dependent ATPase from skeletal and cardiac muscle microsomes could exhibit two forms of the enzyme states, E and E', the former having high affinity for Ca2+ and the latter having extremely low affinity for Can+; the rate of conversion of E to E is much slower than that of E. Ca2" formation as follows: (i') E'- E -C 2Ca"+ (i) E-CaZ" (2) Therefore, the rate-limiting step would be that at which E' is converted to E (Step i'), rather than that at which calcium binds to E (Step i). Thus, it is feasible to assume that one of the malor steps that is altered by protein kinase-induced phosphorylation is Step i'. Taking these into consideration, we refer this step as to the apparent initial step of calcium binding. M. Tada, M. Yamada, M. Inui, and H. Abe, unpublished observations.

Protein Kinase Effects on The present demonstration that the initial rate and maximal amount of EP formation in phosphorylated microsomes were lower at lower ATP concentration ranges (Figs. 4 and 5) is compatible with our previous observations (3) that the rate of EP decomposition (Step iv in Equation 1) is greatly enhanced by CAMP-dependent phosphorylation of phospholamban. Such enhancement is presumably more effective in reducing the apparent velocity and maximal amount of EP formation at lower ATP concentration ranges (Figs. 4 and 5), where the EP formation (Step ii, or iii, or both) is slower and comparable to the EP decomposition (Step iv), whereas the initial velocity and maximal amount of EP formation were greatly augmented in phosphorylated microsomes at higher ATP concentration ranges (Figs. 4 and 5), where the EP formation occurs much faster than the EP decomposition. In fact, CAMP-dependent phosphorylation of phospholamban resulted in the signifcant increases in the initial rates ofpi liberation at lower ATP concentration ranges (Fig. 6), while the rate and amount of EP formation remained lower at these concentrations of ATP, suggesting that the observed reduction in EP formation at lower ATP concentrations in phosphorylated microsomes can be attributed to the predominance of a protein kinase-induced increase in the EP decomposition at lower ATP concentration ranges which becomes manifest due to relatively reduced rate of EP formation. The present observations that the initial rates of EP formation and calcium binding by cardiac microsomes are markedly enhanced by protein kinase-catalyzed phosphorylation with the maintenance of the stoichiometric coupling of 2 mol of calcium transported/mol of EP formed, together with our previous report (3) that the rate of EP decomposition is also enhanced by phosphorylation, are compatible with the accumulated body of evidence that the overall rates of production of Pi" (3, 9, 16) and ADP" (3) and accumulation of Ca' (4, 8-15) are greatly enhanced by protein kinase-catalyzed phosphorylation of a microsomal protein of 20,000 to 22,000 daltons (phospholamban), and thus suggest that the ATPase enzyme itself, but not the efficiency of coupling, is directly affected by protein kinase-catalyzed phosphorylation of phospholamban. The increase in the presteady state rates of calcium binding, coupled with the elementary steps of ATPase, produced by phospholamban phosphorylation, may largely account for the previous findings that the stimulatory effects of protein kinase on the overall rates of ATP hydrolysis and calcium uptake is more evident at lower Ca2+ concentration ranges (3, 9, 33). Thus, it was indicated that the probable sites of action of phospholamban may be those steps associated with calcium and ATP binding to the enzyme and a step at which the reaction intermediate EP is subsequently formed (present study); the rate at which EP is decomposed is also enhanced (3), which can be a predominant factor in affecting the transient kinetics at lower ATP concentration ranges. Such effects produced by protein kinase can increase the apparent affinity of the Ca*'-dependent ATPase system for Ca", resulting in the marked stimulation of the turnover rates of Ca'+-dependent ATPase and active calcium transport by cardiac sarcoplasmic reticulum. These observations thus lent support for the hypothetical view that phospholamban can serve as a regulator controlling the active calcium transport by cardiac sarcoplasmic reticulum (3-7). It is of interest to consider the molecular mechanism by which the ATPase enzyme of cardiac microsomes could be controlled by CAMP-dependent phosphorylation of phospholamban. Although the preliminary results are suggestive of the possibility that these effects are induced by the direct molectllar interactions between the ATPase and phospholamban (7), those could also be caused by an indirect effect such Transient Kinetics CaZ+-ATPase of 1991 as that involving the lipid-protein interactions within the membrane. If the direct protein-protein interactions are functional, phospholamban which is presumably located at the outer surface of the membrane (4, 7) could exhibit an interaction with the ATPase in a manner in which the rate of turnover of the ATPase enzyme is greatly altered as it recognizes and translocates calcium, depending upon phosphorylated and unphosphorylated states of phospholamban. It would still be premature to speculate further the relationship between the enzyme states of ATPase and the states of phospholamban, since no detailed information on the structurefunction relationship between the ATPase enzyme and phospholamban is available at the present time. It would also be intriguing to consider the possible involvement of other potent factors like calcium-binding regulatory protein (34) in these systems. Possible functional relationship between these key components in manifestation of the observed control mechanism should be clarified better by purifying these and reconstituting the functioning membrane vesicles from such components and defined phospholipids. Acknowledgments-We are greatly indebted to Professor Yuji Tonomura and Dr. Michihiro Sumida for their valuable comments and suggestions during the course of this study. We also thank Dr. Jeffery P. Froehlich for his discussion during preparation of the manuscript and for his invaluable support for constructing the rapid chemical quench flow apparatus. REFERENCES 1. Tada, M., Yamada, M., Ohmori, F., Kuzuya, T., and Abe, H. (1979) Proceedings of the XIth International Congress of Biochemistry, July 8-13, 1979, Toronto, Canada 2. Yamada, M., Tada, M., Ohmori, F., Kuzuya, T., Inui, M., and Abe, H. (1979) J. Mol. Cell. Cardiol. 11 (suppl) 34 3. Tada, M., Ohmori, F., Yamada, M., and Abe, H. (1979) J. Biol. Chem. 254,319-326 4. Tada, M., Kirchberger, M. A., and Katz, A. M. (1975) J. Biol. Chem. 250,2640-2647 5. Katz, A. M., Tada, M., and Kirchberger, M. A. (1975) Adu. Cyclrc Nucleotide Res. 5,453-472 6. Tada, M., Ohmori, F., Kinoshita, N., and Abe, H. (1978) Adu. Cyclic Nucleotide Res. 9,355-369 7. Tada, M., Yamada, M., Ohmori, F., Kuzuya, T., and Abe, H. (1979) in Cation Flux Across Biomembranes (Mukohata, Y., and Packer, L., eds) pp. 179-190, Academic Press, London 8. LaRaia, P. J., and Morkin, E. (1974) Circ. Res. 35, 298-305 9. Tada, M., Kirchberger, M. A., Repke, D. I., and Katz, A. M. (1974) J. Biol. Chem. 249,6174-6180 10. Tada, M., Ohmori, F., Nimura, Y., and Abe, H. (1977) J. Biochem. (Tokyo) 82,885-892 11. Kirchberger, M. A., Tada, M., and Katz, A.M. (1974) J. Biol. Chem. 249,6166-6173 12. Kirchberger, M. A., and Tada, M. 11976) J. Bwl. Chem. 251,725-729 13. Kirchberger, M. A., and Chu, G. (1976) Biochim. Biophys. Acta 419,559-562 14. Schwartz, A., Entman, M. L., Kaniike, K., Lane, L. K., van Winkle, W. B., and Bornet, E. P. (1976) Biochim. Biophys. Acta 426, 57-72 15. Will, H., Blanck, J., Smettan, G., and Wollenberger, A. (1976) Biochim. Biophys. Acta 449,295-303 16. Wray, H. L., and Gray, R. R. (1977) Biochim. Biophys. Acta 461, 441-459 17. Tsien, R. W. (1977) Adu. Cyclic Nucleotide Res. 8, 363-420 18. Fabiato, A., and Fabiato, F. (1975) Nature 253, 556-558 19. Tada, M., Yamamoto, T., and Tonomura, Y. (1978) Physiol. Reu. 58. 1-79 20. Froehlich, J. P., and Taylor, E. W. (1975) J. Biol. Chem. 250, 2013-2021 21. Froehlich, J. P., and Taylor, E. W. (1976) J. Biol. Chem. 251, 2307-2315 22. Takisawa, H., and Tonomura, Y. (1978) J. Biochem. (Tokyo) 83, 1275-1284

1992 Protein Kinase Effects Transient on Kinetics CaZ+-ATPase of 23. Sumida, M., Wang, T., Mandel, F., Froehlich, J. P., and Schwartz, A. (1978) J. Biol. Chem. 253,8772-8777 24. Froehlich, J. P., Sullivan, J. V., and Berger, R. L. (1976) Anal. Biochem. 73,331-341 25. Harigaya, S., and Schwartz, A. (1969) Circ. Res. 25, 781-794 26. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193,265-275 27. Glynn, I. M., and Chappell, J. B. (1964) Biochem. J. 90, 147-149 28. Schwarzenbach, G. (1960) Die Komplexometrische Titration, F. Enke, Stuttgart 29. Martin, J. B., and Doty, D. M. C. (1949) Anal. Chem. 21,965-967 30. Sumida, M., and Tonomura, Y. (1974) J. Biochem. (Tokyo) 76, 283-297 31. Weber, K., and Osborn, M. (1969) J. Biol. Chem. 244,4406-4412 32. MacLennan, D. H., and Holland, P. C. (1975) Annu. Reu. Biophys. Bioeng. 4,377-404 33. Hicks, M. J., Shigekawa, M., and Katz, A. M. (1979) Circ. Res. 44,384-391 34. Katz, S., and Remtulla, M. A. (1978) Biochem. Biophys. Res. Commun. 83, 1373-1379