Reversible Inhibition of the Calcium-pumping ATPase in Native Cardiac Sarcoplasmic Reticulum by a Calmodulin-binding Peptide

Calmodulin (CaM) and Ca2+/CaM-dependent protein kinase II (CaM kinase) are tightly associated with cardiac sarcoplasmic reticulum (SR) and are implicated in the regulation of transmembrane Ca2+cycling. In order to assess the importance of membrane-associated CaM in modulating the Ca2+ pump (Ca2+-ATPase) function of SR, the present study investigated the effects of a synthetic, high affinity CaM-binding peptide (CaM BP; amino acid sequence, LKWKKLLKLLKKLLKLG) on the ATP-energized Ca2+uptake, Ca2+-stimulated ATP hydrolysis, and CaM kinase-mediated protein phosphorylation in rabbit cardiac SR vesicles. The results revealed a strong concentration-dependent inhibitory action of CaM BP on Ca2+ uptake and Ca2+-ATPase activities of SR (50% inhibition at ∼2–3 μm CaM BP). The inhibition, which followed the association of CaM BP with its SR target(s), was of rapid onset (manifested within 30 s) and was accompanied by a decrease inV max of Ca2+ uptake, unalteredK 0.5 for Ca2+ activation of Ca2+ transport, and a 10-fold decrease in the apparent affinity of the Ca2+-ATPase for its substrate, ATP. Thus, the mechanism of inhibition involved alterations at the catalytic site but not the Ca2+-binding sites of the Ca2+-ATPase. Endogenous CaM kinase-mediated phosphorylation of Ca2+-ATPase, phospholamban, and ryanodine receptor-Ca2+ release channel was also strongly inhibited by CaM BP. The inhibitory action of CaM BP on SR Ca2+ pump function and protein phosphorylation was fully reversed by exogenous CaM (1–3 μm). A peptide inhibitor of CaM kinase markedly attenuated the ability of CaM to reverse CaM BP-mediated inhibition of Ca2+ transport. These findings suggest a critical role for membrane-bound CaM in controlling the velocity of Ca2+pumping in native cardiac SR. Consistent with its ability to inhibit SR Ca2+ pump function, CaM BP (1–2.5 μm) caused marked depression of contractility and diastolic dysfunction in isolated perfused, spontaneously beating rabbit heart preparations. Full or partial recovery of contractile function occurred gradually following withdrawal of CaM BP from the perfusate, presumably due to slow dissociation of CaM BP from its target sites promoted by endogenous cytosolic CaM.

By regulating cytosolic Ca 2ϩ concentration, the sarcoplasmic reticulum (SR) 1 plays a central role in the contraction-relaxation cycle of heart muscle. Upon excitation of the cardiomyocyte, Ca 2ϩ is released from the SR through Ca 2ϩ -release channels (known as RYR-CRC) to initiate muscle contraction (1)(2)(3)(4)(5). Subsequent muscle relaxation occurs upon sequestration of Ca 2ϩ back into the SR lumen by a Ca 2ϩ -pumping ATPase (Ca 2ϩ -ATPase) present in the SR (1,4,6,7). A well known mechanism for the regulation of the cardiac SR Ca 2ϩ -ATPase involves phosphorylation of another intrinsic SR protein, phospholamban (8 -11). In its unphosphorylated state, phospholamban is thought to interact with the Ca 2ϩ -ATPase exerting an inhibitory effect; phosphorylation of phospholamban by cAMPdependent protein kinase or CaM kinase is thought to disrupt this interaction resulting in stimulation of Ca 2ϩ pump activity (8 -11). In cardiac SR, the RYR-CRC also undergoes phosphorylation by CaM kinase (12)(13)(14), and this may result in stimulation of Ca 2ϩ release from the SR (12,(15)(16)(17).
Recent studies from this laboratory (14, 18 -22) and other laboratories (23)(24)(25)(26) have demonstrated that in cardiac SR, a membrane-associated CaM kinase phosphorylates the Ca 2ϩ -ATPase in addition to RYR-CRC and phospholamban. The phosphorylation occurred at a serine residue and was specific for the cardiac/slow-twitch muscle isoform (SERCA2a) of the Ca 2ϩ -ATPase (18). Site-directed mutagenesis studies by Toyofuku et al. (23) resulted in the identification of Ser 38 as the site in SERCA2a that is phosphorylated by CaM kinase. Studies using native cardiac SR vesicles (14), purified SR Ca 2ϩ -ATPase preparations (14,18), and SERCA2a expressed in HEK-293 cells (23) suggested that Ser 38 phosphorylation of the Ca 2ϩ -ATPase results in activation of the V max of Ca 2ϩ transport. Some studies have, however, questioned the physiological role of Ca 2ϩ -ATPase phosphorylation. Thus, a study by Odermatt et al. (24) showed CaM kinase-mediated phosphorylation of the Ca 2ϩ -ATPase in native rabbit cardiac SR as well as SERCA2a expressed in HEK-293 cells but failed to observe a significant stimulatory effect of phosphorylation on Ca 2ϩ -ATPase function. Another study by Reddy et al. (27) reported failure to observe phosphorylation of the Ca 2ϩ -ATPase in canine cardiac SR or purified Ca 2ϩ -ATPase reconstituted in lipid vesicles. These studies have attributed the stimulatory effect of CaM kinase to the phosphorylation of phospholamban and a consequent increase in Ca 2ϩ affinity of the Ca 2ϩ -ATPase. In native cardiac SR, analysis of the selective effect of Ca 2ϩ -ATPase phosphorylation on Ca 2ϩ -pumping activity of this enzyme is hampered by the concomitant phosphorylation of phospholamban and RYR-CRC by the membrane-bound CaM kinase. Recently, we achieved selective phosphorylation of the Ca 2ϩ -ATPase by the SR-associated CaM kinase by utilizing a phospholamban monoclonal antibody, which inhibits phospholamban phosphorylation, and the RYR-CRC blocking drug, ruthenium red, which was found to inhibit RYR-CRC phosphorylation (22). Under these conditions, Ca 2ϩ -ATPase phosphorylation by endogenous CaM kinase resulted in enhanced V max of Ca 2ϩ transport (22). During the course of these studies we have found that, in addition to the endogenous CaM kinase, SR vesicles isolated from cardiac muscle contains significant amount of calmodulin that is resistant to extraction with high salt (0.6 M KCl). The presence of calmodulin in isolated SR vesicles may mask the true potential of calmodulin-dependent regulation of SR function in in vitro experiments. For example, since both calmodulin and CaM kinase are structured in the SR, introduction of Ca 2ϩ to the assay medium to measure Ca 2ϩ transport would also result in concurrent activation of CaM kinase and other Ca 2ϩ /calmodulin-dependent membrane events. This issue assumes a higher level of complexity given that CaM kinase, once activated, undergoes autophosphorylation and retains activity independently of Ca 2ϩ /calmodulin (28). In the present study, we utilized a previously characterized amphiphilic, high affinity calmodulin-binding peptide (29) to unmask the potential influence of SR-associated calmodulin on cardiac SR Ca 2ϩ -ATPase function. The results presented here demonstrate a strong inhibitory action of CaM BP on the Ca 2ϩ ion-transporting as well as energy-transducing functions of the Ca 2ϩ -ATPase. This inhibition stems from the association of CaM BP with SR membrane target(s) and is readily reversed by calmodulin. These findings imply that a calmodulin-dependent process controls the velocity of Ca 2ϩ pumping in native cardiac SR.

EXPERIMENTAL PROCEDURES
Materials-45 CaCl 2 was purchased from NEN Life Science Products, and [␥-32 P]ATP was from Amersham Pharmacia Biotech. Reagents for electrophoresis were obtained from Bio-Rad. Monoclonal antibody against calmodulin was purchased from Upstate Biotechnology Inc. (Lake Placid, NY). All other chemicals were from Sigma.
Synthesis and Purification of Peptides-A 17-amino acid high affinity calmodulin-binding peptide (designated CaM BP in this report), designed and characterized previously by DeGrado et al. (29), and three fragments of this peptide with overlapping residues were synthesized by the University of Victoria Protein Micro-chemistry Center using a model 430A Applied Biosystems peptide synthesizer. The C termini of the peptides were amidated. All peptides were purified by high performance liquid chromatography, analyzed by mass spectrometry, and sequenced on Applied Biosystems model 473A protein sequencer. The sequences included the following: CaM BP, LKWKKLLKLLKKLLKLG; fragment A, LKWKKLL; fragment B, LLKLLKK; and fragment C, KKLLKLG.
Preparation of SR Vesicles-SR membrane vesicles were prepared from heart ventricles and fast-twitch (adductor magnus) skeletal muscle of New Zealand White rabbits (body weight 2.5-3 kg) as described previously (30). Following isolation, the SR vesicles were suspended in 10 mM Tris maleate (pH 6.8) containing 100 mM KCl and stored at Ϫ80°C after quick-freezing in liquid N 2 . Protein concentration was determined by the method of Lowry et al. (31) using BSA as standard.
Ca 2ϩ Transport and Ca 2ϩ -ATPase Assays-ATP-dependent, oxalatefacilitated Ca 2ϩ uptake by SR was determined using a Millipore filtration technique as described previously (32). The standard incubation medium for Ca 2ϩ uptake (total volume 250 l) contained 50 mM HEPES (pH 7.2), 5 mM MgCl 2 , 5 mM NaN 3 , l20 mM KCl, 0.1 mM EGTA, 5 mM potassium oxalate, 5 mM ATP, 0.l mM 45 CaCl 2 (ϳ8000 cpm/nmol; free Ca 2ϩ , 7.5 M), 25 M ruthenium red, and SR (6 g of protein). In experiments where Ca 2ϩ concentration dependence was studied, the EGTA concentration in the assay medium was held at 0.1 mM, and the amount of total 45 CaCl 2 added was varied in the range 1 to 200 M to yield the desired free Ca 2ϩ . Modifications to the standard incubation medium are specified in the figure legends. Unless indicated otherwise, all assays were carried out at 37°C; the Ca 2ϩ transport reaction was initiated by the addition of SR vesicles after preincubation of the rest of the assay components for 3 min. The initial free Ca 2ϩ concentrations in the assay medium were determined using the computer program of Fabiato (33). The data on Ca 2ϩ concentration dependence on Ca 2ϩ uptake were analyzed by nonlinear regression curve fitting using Sig-maPlot scientific graph program (Jandel Scientific) run on an IBM-PC computer. The data were fit to Equation 1, where v is the measured Ca 2ϩ uptake activity at a given Ca 2ϩ concentration; V max is the maximum activity reached; K 0.5 is the Ca 2ϩ concentration giving half of V max , and n is the equivalent to the Hill coefficient. Ca 2ϩ -ATPase activity of SR membranes was quantified as described previously (34) using an incubation medium identical to that used for Ca 2ϩ uptake except that [␥-32 P]ATP was used instead of nonradioactive ATP and nonradioactive CaCl 2 was used instead of 45 CaCl 2 . Parallel assays were also performed in the absence of Ca 2ϩ (i.e. in the presence of 0.2 mM EGTA with no CaCl 2 added to the assay medium), and the Ca 2ϩ -ATPase activity was defined as the difference in ATP hydrolysis (liberation of 32 P i ) measured in the absence and presence of Ca 2ϩ . The ATPase reaction was initiated by the addition of SR vesicles after preincubation of the rest of the assay components for 3 min at 37°C and was allowed to proceed for 2 min.
Western Immunoblotting-Western blotting analysis of endogenous calmodulin in SR vesicles was performed using a monoclonal antibody specific for calmodulin (35). SR proteins were fractionated on SDSpolyacrylamide (15%, homogenous) mini-gels and then electroblotted to nitrocellulose sheets. The sheets were incubated in 0.2% glutaraldehyde in PBS for 45 min at 24°C and then rinsed in PBS. Nonspecific binding sites were blocked with 2% BSA, 0.1% gelatin in PBS for 60 min at 37°C. Following three 15-min washes with 0.05% Tween 20 in PBS, the sheets were incubated with anti-calmodulin monoclonal antibody (0.5 g/ml in PBS) for 60 min at 37°C and then with alkaline phosphatase-conjugated goat anti-mouse IgG secondary antibody (dilution 1:1000). After five 10-min washes in PBS/Tween, the sheets were rinsed with deionized water, and the immunoreactive peptide band representing calmodulin was visualized following color development using a Bio-Rad assay kit.
Phosphorylation Assay-Endogenous CaM kinase-catalyzed SR protein phosphorylation was measured as described previously (18). The standard incubation medium (total volume 50 l) for phosphorylation by endogenous CaM kinase contained 50 mM HEPES (pH 7.4), 10 mM MgCl 2 , 0.1 mM CaCl 2 , 0.1 mM EGTA, 1 M calmodulin, 0.8 mM [␥-32 P]ATP (specific activity, 300 -400 cpm/pmol), and SR (30 g of protein). The phosphorylation reaction was initiated by the addition of SR after preincubation of the rest of the assay components for 3 min at 37°C. The reaction was terminated after 2 min by the addition of 15 l of SDS sample buffer, and the samples were analyzed in 4 -18% SDSpolyacrylamide gels. The gels were stained with Coomassie Brilliant Blue, dried, and autoradiographed. Quantification of phosphorylation was carried out by liquid scintillation counting after excision of the radioactive bands from the gels (18).
Heart Perfusion and Measurement of Contractile Function-Rabbits were anesthetized with sodium pentobarbital (35 mg/kg, intravenously), and the hearts were excised and immediately cannulated for retrograde aortic perfusion of the coronary arteries with mammalian Ringer solution consisting of 154 mM NaCl, 5 mM KCl, 2.2 mM CaCl 2 , 6 mM NaHCO 3 , and 5.5 mM dextrose. The perfusion buffer was equilibrated with 95% O 2 , 5% CO 2 , which maintained a pH of 7.4; the perfusion temperature was set at 37 Ϯ 0.2°C. The hearts were perfused at a constant flow rate of 25 ml/min using a peristaltic pump. After an initial 15-20 min of perfusion, when the spontaneous beating had stabilized, a latex balloon-tipped cannula filled with degassed H 2 O was inserted into the lumen of the left ventricle for obtaining systolic left ventricular pressure development. The cannula was connected via a pressure transducer (COBE, Bramalea, Canada) to a BioPac System Digital Monitor (model MP100) and a personal computer that allowed on-line monitoring of left ventricular pressure and off-line calculation of developed pressure, rate of pressure development (ϩdP/dt), and rate of relaxation (ϪdP/dt).
Data Presentation-Unless specified otherwise, the experimental values represent the average of at least three independent experiments using separate SR preparations performed in duplicate. The data are presented as mean Ϯ S.E.

RESULTS
Effects of Varying Concentrations of CaM BP and Its Fragments on ATP-dependent Ca 2ϩ Uptake by Cardiac SR in the Absence and Presence of Calmodulin-The ATP-dependent, oxalate-facilitated Ca 2ϩ uptake by SR vesicles is a useful, commonly used parameter to measure the Ca 2ϩ pump (Ca 2ϩ -ATPase) function of SR in vitro (6). The results presented in Fig. 1 demonstrate the effects of varying concentrations of CaM BP and its fragments on ATP-dependent Ca 2ϩ uptake by cardiac SR vesicles measured in the absence and presence of calmodulin in the assay medium. When assays were performed in the absence of calmodulin, CaM BP caused strong, concentration-dependent inhibition of Ca 2ϩ uptake by SR with virtually complete inhibition occurring at Ͻ5 M CaM BP (Fig. 1A). Fragmented molecules of CaM BP (CaM BP fragments A-C) failed to inhibit Ca 2ϩ uptake by SR (Fig. 1A, inset). Addition of low micromolar concentrations of calmodulin to the assay medium prevented the inhibitory action of CaM BP on Ca 2ϩ uptake by SR, in a concentration-dependent manner, and caused appreciable stimulation (ϳ40 -55%) of Ca 2ϩ uptake (Fig. 1, A and B, inset). The results presented in Fig. 1A and the inset in Fig. 1B were obtained under the standard Ca 2ϩ uptake assay conditions with 6 g of SR protein in the assay medium (see under "Experimental Procedures"). In additional experiments, the effect of CaM BP on Ca 2ϩ uptake by SR was determined with varying amounts of SR in the assay medium. The results presented in Fig. 1B show that the concentration dependence curve for CaM BP inhibition of Ca 2ϩ uptake is progressively shifted to the right with increasing concentration of SR in the assay. These findings suggest that the inhibitory action of CaM BP stems from its apparently stoichiometric interaction with one or more targets in the SR, and such interaction is prevented by calmodulin.
Effect of CaM BP on Cardiac SR Ca 2ϩ -ATPase Activity-Since CaM BP inhibited ATP-dependent Ca 2ϩ uptake by cardiac SR, the effect of CaM BP on Ca 2ϩ -ATPase activity (ATP hydrolysis) was investigated. The results presented in Fig. 2 show that, under the assay conditions identical to that used for Ca 2ϩ uptake, CaM BP caused concentration-dependent inhibition of Ca 2ϩ -stimulated ATPase activity. The inhibition of Ca 2ϩ -ATPase activity and Ca 2ϩ uptake by CaM BP occurred at similar concentration range with only a minor difference in K i values for Ca 2ϩ uptake (50% inhibition at ϳ2 M CaM BP) and Ca 2ϩ -ATPase activity (50% inhibition at ϳ2.8 M CaM BP) ( Fig. 2, inset). Thus the observed reduction in Ca 2ϩ uptake is mainly a consequence of a primary inhibition of ATPase activity by CaM BP. Addition of calmodulin (3 M) to the assay medium reversed the inhibitory effect of CaM BP on Ca 2ϩ -ATPase activity (Fig. 2). Fig. 3A shows the time course of ATP-dependent Ca 2ϩ uptake by cardiac SR measured in the absence of CaM BP and in the presence of two selected concentrations of CaM BP (2 and 4 M) with or without calmodulin. The rates of Ca 2ϩ uptake by SR is strongly inhibited by CaM BP; the inhibition was of rapid onset (manifested within 30 s) and the degree of inhibition increased with increasing concentration of CaM BP. Addition of calmodulin (3 M) to the assay medium prevented the inhibitory action of CaM BP.

Effect of CaM BP on the Time Course of Ca 2ϩ Uptake by Cardiac SR-
In the experiments described thus far, the effect of CaM BP was assessed by adding this peptide directly to the Ca 2ϩ uptake assay medium. In order to determine whether the inhibitory action of CaM BP results from its association with SR membrane target(s), in subsequent experiments, the time course of Ca 2ϩ uptake was measured using CaM BP-pretreated and control SR vesicles obtained as follows. Cardiac SR vesicles were incubated with 5 M CaM BP in the absence of calmodulin and in the presence of 3 M calmodulin for 10 min at 24°C. Subsequently, the SR vesicles were recovered by centrifuga- tion, washed to remove free CaM BP and calmodulin, and then the time course of Ca 2ϩ uptake was determined under standard assay conditions in the absence of calmodulin or in the presence of 3 M calmodulin. SR vesicles subjected to the same experimental protocol but without CaM BP in the incubation medium served as control for these experiments. The results from these experiments showed that pretreatment of SR with CaM BP leads to markedly reduced rates of Ca 2ϩ uptake (Fig. 3, B and C); this decline in Ca 2ϩ uptake rates is not observed when CaM BP-pretreatment of SR is performed in the presence of calmodulin (Fig. 3C) or when Ca 2ϩ uptake assays with CaM BPpretreated SR is performed in the presence of calmodulin (Fig.  3B). These findings suggest that the inhibitory action of CaM BP is dependent on its association with the SR membrane and that calmodulin is able to prevent the onset of CaM BP-mediated inhibition as well as reverse pre-existing inhibition induced by CaM BP.
Blockade of Ca 2ϩ Uptake by Addition of CaM BP during the Turnover Cycle of Cardiac SR Ca 2ϩ -ATPase-To investigate the effect of CaM BP on cardiac SR Ca 2ϩ -ATPase during its turnover, CaM BP was added to the Ca 2ϩ uptake assay medium 3 min 15 s after initiating Ca 2ϩ -ATPase turnover. The time course of Ca 2ϩ uptake was monitored prior to and following the addition of CaM BP for several minutes. It was found that addition of CaM BP (3 M) during the turnover cycle of Ca 2ϩ -ATPase resulted in an apparently instantaneous and short-lived release of a small fraction (ϳ25%) of the pre-existing SR Ca 2ϩ load as well as complete cessation of further Ca 2ϩ uptake by SR vesicles (Fig. 4). Addition of calmodulin (3 M) together with CaM BP (3 M) prevented the above effects of CaM BP (Fig. 4). In additional experiments, it was found that the fractional Ca 2ϩ release induced by CaM BP did not exceed 25% of the pre-existing SR Ca 2ϩ load at higher concentrations (up to 5 M) of CaM BP (data not shown). These findings suggest that about 75% of the inhibitory effect of CaM BP on the measured Ca 2ϩ uptake activity of SR stems from inhibition of the SR Ca 2ϩ pump (Ca 2ϩ -ATPase); the remaining 25% of the inhibitory effect may be attributed to CaM BP-induced Ca 2ϩ release. Since calmodulin prevented the effects of CaM BP, it is likely that CaM BP exerts its

FIG. 2. Concentration-dependent inhibitory action of CaM BP on Ca 2؉ -ATPase activity of cardiac SR and correlation between inhibition of Ca 2؉ uptake and ATP hydrolysis.
The Ca 2ϩ -ATPase and Ca 2ϩ uptake reactions were carried out for 2 min in the standard assay medium (see "Experimental Procedures"). The main figure shows the effects of varying concentrations of CaM BP on Ca 2ϩ -ATPase activity measured in the absence of calmodulin (q) and in the presence of 3 M calmodulin (E) in the assay medium; each data point represents mean Ϯ S.E. of four experiments using separate SR preparations. In the experiments shown in the inset, the effects of varying concentrations of CaM BP on Ca 2ϩ uptake and Ca 2ϩ -ATPase activities were determined using the same SR preparation; the results are presented as percent inhibition of Ca 2ϩ uptake or Ca 2ϩ -ATPase activity as a function of CaM BP concentration in the assay medium. In the experiments shown in C, the time course of Ca 2ϩ uptake was measured using control SR (q) and SR pretreated with CaM BP alone (Ⅺ) or CaM BP plus calmodulin (ƒ). CaM BP-pretreated SR was obtained by incubating SR vesicles (250 g of protein) in buffer A (total volume 600 l) containing 50 mM HEPES (pH 7.2), 5 mM MgCl 2 , 5 mM NaN 3 , 120 mM KCl, 0.1 mM EGTA, 0.1 mM CaCl 2 , 5 mM ATP, and 5 M CaM BP, in the absence or in the presence of 3 M calmodulin, for 10 min at 24°C. The SR vesicles were then recovered by centrifugation, washed twice with buffer B (10 mM Tris maleate, 100 mM KCl (pH 6.8)), resuspended in the same buffer, and used for Ca 2ϩ uptake assays. SR vesicles subjected to the same experimental protocol but without CaM BP in buffer A served as the control for these experiments. Each data point in A and B represents mean Ϯ S.E. of three experiments using separate SR preparations. Each data point in C represents the average of duplicate determinations using a single SR preparation. effects by interfering with calmodulin-dependent processes that are normally involved in the control of Ca 2ϩ sequestering and Ca 2ϩ release functions of the SR.
Effect of CaM BP on Ca 2ϩ Uptake by Cardiac SR at Varying Concentrations of Ca 2ϩ and ATP-The results presented in Fig. 5 show the effect of two selected concentrations of CaM BP (1.5 and 3 M) on Ca 2ϩ uptake by cardiac SR at a wide range of Ca 2ϩ concentrations (9 nM to 67 M). CaM BP inhibited Ca 2ϩ uptake at all Ca 2ϩ concentrations tested. At the submaximally effective concentrations of CaM BP used, the inhibitory effect could not be overcome with increasing Ca 2ϩ concentration. On the other hand, addition of calmodulin (3 M) to the assay medium fully reversed the inhibitory effect of CaM BP. The kinetic parameters derived from the data shown in Fig. 5 are summarized in Table I. It can be seen that the inhibitory action of CaM BP is associated with a decrement in V max without appreciable changes in the apparent affinity of the Ca 2ϩ -ATPase for Ca 2ϩ or the Hill coefficient (n H ) for Ca 2ϩ .
As shown in Fig. 6 Reversible Association of CaM BP with Cardiac SR and the Presence of Endogenous Calmodulin in Cardiac SR-As described earlier, pretreatment of cardiac SR with CaM BP (in the absence but not in the presence of calmodulin) resulted in diminished rates of Ca 2ϩ uptake suggesting that the inhibitory action of CaM BP is dependent on its association with the SR membrane (cf. Fig. 3, B and C). In order to visualize the physical association of CaM BP with SR, experiments were performed in which cardiac SR vesicles were pretreated with CaM BP in the absence and presence of calmodulin and then the SR proteins were fractionated by electrophoresis on SDS-polyacrylamide (4 -18% linear gradient) gels. In these experiments, the electrophoresis was terminated when the dye front had reached about 1 cm above the bottom of the gel so that the low molecular weight peptide (CaM BP molecular mass 2062 daltons) could be retained on the gel matrix. The protein profiles in Coomassie Blue-stained gels from these experiments showed association of CaM BP with the SR membrane when pretreatment with CaM BP was carried out in the absence but not in the presence of calmodulin (Fig. 7A). Furthermore, the SRbound CaM BP could be readily dissociated from the membrane when CaM BP-pretreated SR was subsequently incubated with calmodulin (Fig. 7A). These findings clearly demonstrate that CaM BP associates with the SR, and calmodulin prevents and reverses this association.
Western immunoblotting analysis using a monoclonal antibody specific for calmodulin (35) showed considerable amount of endogenous calmodulin in the isolated cardiac SR vesicles (Fig. 7B). Since the procedure used for the isolation of SR vesicles involved extraction of the membranes with high salt (0.6 M KCl, cf. Ref. 30), it appears that the endogenous calmod-  ulin detected is firmly structured in the SR membrane. Also, treatment of SR with CaM BP does not seem to result in appreciable dissociation of calmodulin from the SR (Fig. 7B).

Reversible Inhibition of Endogenous CaM Kinase-mediated Cardiac SR Protein Phosphorylation by CaM BP-Activation
of SR-associated ␦-CaM kinase by calmodulin and consequent phosphorylation of phospholamban, Ca 2ϩ -ATPase, and RYR-CRC are thought to regulate both the Ca 2ϩ uptake and release functions of the SR (8 -17, 22). In view of this, experiments were performed to determine the effects of CaM BP on endogenous CaM kinase-mediated SR protein phosphorylation. The results presented in Fig. 8 demonstrate that CaM BP causes concentration-dependent inhibition of phosphorylation of phospholamban, Ca 2ϩ -ATPase, and RYR-CRC; this inhibition is reversed by increasing the concentration of calmodulin in the phosphorylation assay medium. Thus, CaM kinase, a major calmodulin target in the SR, is inhibited by CaM BP. 2 The inhibitory effects of CaM BP on SR protein phosphorylation and SR Ca 2ϩ pump function are manifested at the same concentration range of CaM BP (e.g. see Fig. 1 and Fig. 8).

Relationship between the Inhibitory Effects of CaM BP on Cardiac SR CaM Kinase and SR Ca 2ϩ
Uptake-In additional experiments, we utilized a synthetic CaM kinase inhibitor peptide (corresponding to amino acid residues 290 -309 of CaM kinase II, cf. Ref. 28) to investigate the potential relationship between the inhibitory effects of CaM BP on cardiac SR CaM kinase and SR Ca 2ϩ uptake. In these experiments, the effects of varying concentrations of CaM kinase inhibitor peptide on Ca 2ϩ uptake by SR was determined in the absence and presence of CaM BP and/or calmodulin in the assay medium. The results are summarized in Fig. 9. CaM kinase inhibitor peptide abolished the stimulatory effect of calmodulin on Ca 2ϩ uptake by SR but did not affect the basal Ca 2ϩ uptake measured in the absence of calmodulin. Interestingly, the ability of calmodulin to reverse the inhibitory effect of CaM BP on Ca 2ϩ uptake by SR was markedly attenuated by the CaM kinase inhibitor peptide. These findings indicate that reversal of the inhibitory effect of CaM BP by calmodulin is dependent, at least in part, on CaM kinase activation.
Effects of CaM BP on Ca 2ϩ -ATPase and Ca 2ϩ Uptake Activities of Fast-twitch Skeletal Muscle SR-It has been demonstrated previously that the Ca 2ϩ -ATPase in cardiac SR, but not fast-twitch skeletal muscle SR, undergoes phosphorylation by endogenous and exogenous CaM kinase (18). Therefore, it was of interest to examine whether CaM BP influenced the Ca 2ϩ -ATPase activity and Ca 2ϩ transport function of fast skeletal muscle SR. As shown in Fig. 10A, under assay conditions identical to those used in the experiments using cardiac SR (cf. Fig.  1A and Fig. 2), CaM BP (0.5-5 M) did not inhibit the Ca 2ϩ -ATPase activity of fast skeletal muscle SR; instead a stimulatory effect was observed. On the other hand, the ATP-dependent Ca 2ϩ uptake activity of fast skeletal muscle SR was inhibited by CaM BP at the same concentration range in which no inhibitory effect on Ca 2ϩ -ATPase activity could be observed (Fig. 10B). Thus, the observed inhibition of Ca 2ϩ uptake re- flects CaM BP-induced activation of Ca 2ϩ release from the SR. This is in direct contrast to the concurrent inhibition of Ca 2ϩ uptake and Ca 2ϩ -ATPase activity by CaM BP observed in cardiac SR (cf. Fig. 1 and Fig. 2). Inclusion of calmodulin (3 M) in the assay medium prevented the effects of CaM BP on Ca 2ϩ -ATPase and Ca 2ϩ uptake activities of fast skeletal muscle SR (Fig. 10, A and B).

Activation of Ca 2ϩ Release upon Addition of CaM BP to Ca 2ϩ -preloaded Fast Skeletal Muscle SR Vesicles-Addition of
CaM BP to the assay medium after initiating Ca 2ϩ -ATPase turnover resulted in rapid and sustained Ca 2ϩ release from fast skeletal muscle SR vesicles (Fig. 11). This response to CaM BP is also different from the short-lived fractional Ca 2ϩ release observed in the case of cardiac SR under identical assay conditions (Fig. 4). Addition of calmodulin (3 M) together with CaM BP (3 M) prevented the Ca 2ϩ release-promoting effect of CaM BP on fast-twitch skeletal muscle SR (Fig. 11).
Effect of CaM BP on Cardic Contractile Function-In view of the strong inhibitory action of CaM BP on Ca 2ϩ uptake by cardiac SR observed in vitro, it was of considerable interest to examine the effect of this peptide on cardiac contractile function. The hydrophobic nature of CaM BP (29) facilitated such investigation using isolated perfused heart preparations. In isolated, spontaneously beating rabbit heart preparations perfused at a constant flow rate, CaM BP (1 and 2.5 M) produced marked concentration-dependent depression of contractile function as evidenced by decrements in developed left ventricular pressure, rates of pressure development and relaxation, as well as pronounced elevation of end diastolic pressure (Fig. 12   FIG. 9. Attenuation of calmodulin- and Table II). These effects were discernible within 2 min after initiating perfusion with CaM BP. The observed depression of contractility and diastolic dysfunction correlate well with the ability of this peptide to inhibit SR Ca 2ϩ pump function. The depression of contractile function induced by a low concentration of CaM BP (1 M) was fully reversible upon reperfusion with normal buffer over a period of 20 -30 min. However, only partial recovery of contractile function was observed upon reperfusion following infusion of a higher concentration (2.5 M) of CaM BP. In these spontaneously beating preparations, the heart rate (beats/min) was not altered significantly during perfusion with CaM BP, but an enhancement in heart rate was observed during reperfusion with normal buffer subsequent to infusion of 2.5 M (but not 1 M) CaM BP ( Fig. 12 and Table II).

DISCUSSION
In this study, we have made the following novel, key observations. (i) At low micromolar concentrations, CaM BP strongly inhibits active Ca 2ϩ sequestration by isolated cardiac SR vesicles. (ii) The inhibition of Ca 2ϩ transport is mainly the consequence of a primary inhibition of the SR Ca 2ϩ -ATPase. (iii) Cardiac SR vesicles contain firmly bound endogenous calmodulin, and exogenously added calmodulin readily reverses the inhibitory action of CaM BP on the energy transduction and Ca 2ϩ ion transport functions of the SR Ca 2ϩ -ATPase. Taken together, these findings suggest a crucial role for SR-associated calmodulin in the regulation of cardiac SR Ca 2ϩ pump function. As discussed below, analysis of the characteristics of Ca 2ϩ -ATPase inhibition by CaM BP, and the reversal of inhibition by calmodulin, has provided insights into the mechanisms of action of CaM BP.
CaM BP inhibits the SR Ca 2ϩ -ATPase rapidly (Fig. 3A), and the inhibition results from the association of CaM BP with SR membrane target(s) (Fig. 3, B and C and Fig. 7). The association of CaM BP with its SR target(s) appears to be stoichiometric (Fig. 1B) although actual stoichiometry could not be determined as CaM BP binding to SR was not quantified in this study. CaM BP is an amphiphilic peptide with high affinity for calmodulin (dissociation constant for binding ϳ0.2 nM, cf. Ref. 29), and several lines of evidence presented here suggest that SR-associated calmodulin is a prime target of CaM BP. Thus, we have found that (i) presence of exogenous calmodulin in the incubation medium prevents association of CaM BP with SR (Fig. 7A) and blocks the inhibition of Ca 2ϩ -ATPase by CaM BP; (ii) exogenous calmodulin is effective in displacing CaM BP previously bound to SR (Fig. 7A) and in reversing pre-existing CaM BP-induced inhibition of Ca 2ϩ -ATPase (Fig. 3B); and (iii) the intact CaM BP molecule capable of high affinity calmodulin binding, but not its truncated fragments that lack the ability to bind calmodulin, inhibits the SR Ca 2ϩ -ATPase (Fig. 1A). From the above evidence, it appears that the effects of CaM BP arise from its ability to interact with endogenous calmodulin in the SR and consequent perturbations in calmodulin-dependent membrane events.
In exploring the mechanistic links downstream of CaM BPcalmodulin interaction, we found that CaM BP strongly inhibited cardiac SR-associated CaM kinase (Fig. 8), a key natural target of calmodulin, implicated in the regulation of both Ca 2ϩ uptake and release functions of the SR through phosphorylation of phospholamban (8 -11), Ca 2ϩ -ATPase (14, 18 -20, 22), and RYR-CRC (12)(13)(14)(15)(16)(17)21). The CaM BP-induced inhibition of CaM kinase was readily reversed by exogenous calmodulin. In addition, we found that a CaM kinase inhibitor peptide markedly attenuated the ability of exogenous calmodulin to reverse CaM BP-induced inhibition of Ca 2ϩ uptake by SR (Fig. 9). Taken together, these findings suggest that the inhibitory effect of CaM BP on cardiac SR Ca 2ϩ pump function is orchestrated, at least in part, through inhibition of SR CaM kinase. It should be noted that unlike CaM BP, the CaM kinase inhibitor peptide did not inhibit the basal Ca 2ϩ uptake activity of SR, although it was effective in blocking the stimulatory effect of exogenous calmodulin on Ca 2ϩ uptake (Fig. 9). The CaM kinase inhibitor peptide used in this study is modeled after the calmodulin-binding region of CaM kinase (amino acid residues 290 -309), so it can be expected to inhibit CaM kinase by scavenging calmodulin added to the assay medium (28). However, this peptide is membrane-impermeant (28) and, hence, cannot access endogenous SR-associated calmodulin, a functionally important pool that may be trapped within the CaM kinase molecule through a process of sequential intersubunit autophosphorylation (36). A hydrophobic molecule like CaM BP, with high affinity for calmodulin, is apparently able to access a functionally important pool of calmodulin structured in the SR membrane matrix and/or trapped within the CaM kinase molecule. It is also worth noting here that KN-62, a widely used CaM kinase inhibitor, with undefined mechanism of action, does not inhibit the endogenous CaM kinase in cardiac SR (37).
Interestingly, fast skeletal muscle SR Ca 2ϩ -ATPase, which is not phosphorylated by CaM kinase (18), is not inhibited by CaM BP (Fig. 10A). On the other hand, the observed CaM BP-induced activation of Ca 2ϩ release from fast skeletal muscle SR vesicles (Fig. 11) and consequent inhibition of SR Ca 2ϩ loading (Fig. 10B) are essentially similar to the effects of calmodulin antagonists such as calmidazolium and compound 48/80 on skeletal muscle SR reported previously by Tuana and MacLennan (38). Also, the effects of CaM BP on skeletal muscle SR Ca 2ϩ uptake, Ca 2ϩ -ATPase activity, and Ca 2ϩ release are readily reversed by calmodulin (Fig. 10, A and B, and Fig. 11). Since the intrinsic functional properties and reaction mechanism of Ca 2ϩ -ATPase are similar in cardiac and skeletal muscle SR (6,39), it is unlikely that the observed divergent effects of CaM BP on cardiac and skeletal muscle SR arise from direct interaction of the peptide with the Ca 2ϩ -ATPase. All of the above observations, on the other hand, suggest that CaM BP interferes with calmodulin-dependent membrane events. The observed stimulatory effect of CaM BP on skeletal muscle SR Ca 2ϩ -ATPase likely results from the collapse of Ca 2ϩ gradient across the SR vesicles (due to activation of Ca 2ϩ release) as build up of high Ca 2ϩ concentration inside the vesicles leads to inhibition of Ca 2ϩ -ATPase (40).
Besides CaM kinase, RYR-CRC is a natural target of calmodulin in the SR, and binding of calmodulin to high (nanomolar) affinity sites on RYR-CRC is known to inhibit Ca 2ϩ release (41, 42). The fractional Ca 2ϩ release from cardiac SR seen upon addition of CaM BP to the assay medium during active Ca 2ϩ transport (Fig. 4) may be a consequence of CaM BP-induced disruption of the normal interaction between calmodulin and RYR-CRC. Alternatively, the observed fractional Ca 2ϩ release may reflect release of ATPase-bound Ca 2ϩ (i.e. Ca 2ϩ ions in transit) to the assay medium owing to CaM BP-mediated structural perturbations in the enzyme molecule (see below).
Analysis of the influence of CaM BP on the kinetic parameters of Ca 2ϩ transport has provided further insights into the mechanism underlying Ca 2ϩ -ATPase inhibition. The inhibitory effect of CaM BP was associated with a decrease in the V max of Ca 2ϩ transport without appreciable changes in the K 0.5 for Ca 2ϩ activation of Ca 2ϩ transport or the Hill coefficient for Ca 2ϩ (Table I). These findings suggest that CaM BP-mediated structural perturbations in the Ca 2ϩ -ATPase does not alter the functional properties of the Ca 2ϩ -binding sites located in the transmembrane region of the ATPase (6, 7). On the other hand, analysis of the effect of CaM BP on ATP concentration dependence revealed that 50% decrease in V max of Ca 2ϩ transport was accompanied by a 10-fold decrease in the apparent affinity of FIG. 12. Effect of CaM BP on cardiac contractile function. Contractile function was assessed in isolated perfused, spontaneously beating rabbit heart as described under "Experimental Procedures." Shown are contractions recorded during perfusion with control buffer, buffer containing 1 M or 2.5 M CaM BP, and following reperfusion with control buffer subsequent to perfusion with CaM BP. The effects of CaM BP on contractile function parameters are summarized in Table II. Results similar to those shown here were obtained in three additional isolated perfused heart preparations studied. LVP, left ventricular pressure. the ATPase for its substrate, ATP (Fig. 6). Since the inhibitory effect of submaximally effective concentration of CaM BP could not be overcome by increasing the concentration of ATP, CaM BP inhibition is non-competitive with respect to ATP. These findings imply that CaM BP-mediated inhibition involved a major alteration at the catalytic site located in the extramembranous region of the Ca 2ϩ -ATPase (6, 7). Our observations on the effects of CaM BP on SR Ca 2ϩ -ATPase and its mechanism of action are unique when compared with those reported for other basic calmodulin-binding peptides such as melittin (derived from bee venom) and C28R2 (derived from the autoinhibitory domain of plasma membrane Ca 2ϩ -ATPase). Melittin has been shown to inhibit skeletal muscle SR Ca 2ϩ -ATPase (43)(44)(45)(46), and C28R2 has been shown to inhibit both cardiac and skeletal muscle SR Ca 2ϩ -ATPase (47). It has been suggested that these peptides inhibit enzyme activity by electrostatically cross-linking Ca 2ϩ -ATPase into large inactive aggregates (43,44,47) or by binding to hydrophilic cytoplasmic domain of the Ca 2ϩ -ATPase without causing enzyme aggregation (45,46). Unlike melittin and C28R2, CaM BP does not inhibit skeletal muscle SR Ca 2ϩ -ATPase. Furthermore, the inhibitory action of melittin on skeletal muscle SR Ca 2ϩ -ATPase is accompanied by a decrease in the enzyme's affinity for Ca 2ϩ and unaltered affinity for ATP (46). In contrast, CaM BP inhibition of cardiac SR Ca 2ϩ -ATPase is associated with a pronounced decrease in the affinity of enzyme for ATP and unaltered affinity for Ca 2ϩ .
The characteristics of Ca 2ϩ -ATPase inhibition by CaM BP also differ from those of other inhibitors of SR Ca 2ϩ -ATPase such as thapsigargin, cyclopiazonic acid, and clotrimazole. Unlike CaM BP, (i) the above drugs produce marked decrease in the Ca 2ϩ -binding affinity of the ATPase (48 -50), and (ii) their inhibitory action on the SR Ca 2ϩ -ATPase is irreversible (48 -50). To our knowledge, the present report is the first to describe inhibition of the SR Ca 2ϩ pump by a synthetic peptide that could be readily reversed by a biologically important, Ca 2ϩ signal-transducing molecule such as calmodulin.
The CaM BP-induced depression of contractility and diastolic dysfunction observed in isolated perfused hearts ( Fig. 12 and Table II) correlate well with the ability of CaM BP to inhibit SR Ca 2ϩ pump function. Interestingly, full or partial recovery of contractile function occurred gradually following withdrawal of CaM BP from the perfusion medium, presumably due to slow dissociation of CaM BP from its target sites promoted by endogenous cytosolic calmodulin.
In conclusion, the findings presented here implicate a critical role for membrane-bound calmodulin in controlling the Ca 2ϩ sequestering activity of the cardiac SR Ca 2ϩ pump. Calmodulin acts on multiple targets in the SR and activates divergent physiological events. It serves to promote Ca 2ϩ sequestration as well as Ca 2ϩ release (9 -11, 12, 15, 41, 42) and protein phosphorylation as well as protein dephosphorylation (9 -11, 18 -22). The topology of calmodulin in the SR and the mechanisms that allow calmodulin to dictate these reciprocal phenomena are currently unknown. Perhaps, calmodulin structured in the SR membrane is segregated functionally as "target-dedicated" molecules, and the interaction of calmodulin with a specific target is governed by its ability to sense changes in cytosolic and/or SR lumenal Ca 2ϩ concentration as well as Ca 2ϩ -regulated conformational status of the target. In this way, membrane-bound calmodulin may serve as a Ca 2ϩ -sensitive molecular switch that controls and coordinates Ca 2ϩ release and Ca 2ϩ sequestration in concert with the contractionrelaxation cycle of the heart.