Superoxide Anion Radical-Triggered Ca Release from Cardiac Sarcoplasmic Reticulum through Ryanodine Receptor Ca Channel

The ryanodine receptor Ca channel (RyRC) constitutes the Ca-release pathway in sarcoplasmic reticulum (SR) of cardiac muscle. A direct mechanical and a Ca-triggered mechanism (Ca-induced Ca release) have been proposed to explain the in situ activation of Ca release in cardiac muscle. A variety of chemical oxidants have been shown to activate RyRC; however, the role of modification induced by oxygenderived free radicals in pathological states of the muscle remains to be elucidated. It has been hypothesized that oxygenderived free radicals initiate Ca-mediated functional changes in or damage to cardiac muscle by acting on the SR and promoting an increase in Ca release. We confirmed that superoxide anion radical (O2 .) generated from hypoxanthinexanthine oxidase reaction decreases calmodulin content and increases Ca efflux from the heavy fraction of canine cardiac SR vesicles; hypoxanthine-xanthine oxidase also decreases Ca free within the intravesicular space of the SR with no effect on Ca-ATPase activity. Current fluctuations through single Ca-release channels have been monitored after incorporation into planar phospholipid bilayers. We demonstrate that activation of the channel by O2 . is dependent of the presence of calmodulin and identified calmodulin as a functional mediator of O2 .-triggered Ca release through the RyRC. For the first time, we show that O2 . stimulates Ca release from heavy SR vesicles and suggest the importance of accessory proteins such as calmodulin in modulating the effect of O2 .. The decreased calmodulin content induced by oxygen-derived free radicals, especially O2 ., is a likely mechanism of accumulation of cytosolic Ca (due to increased Ca release from SR) after reperfusion of the ischemic heart. Several attempts have been made to integrate the two most popular hypotheses of myocardial stunning and reperfusion injury: (1) accumulation of cytosolic Ca and (2) increase in oxygen-derived free radical production (Kukreja and Hess, 1992; Opie, 1992). According to this unifying hypothesis, oxygen-derived free radicals initiate Ca-mediated functional changes or damage by acting on the SR and promoting Ca entry into the cytosol. In support of this is the finding that oxygen-derived free radicals, generated by the xanthine-xanthine oxidase system, depress SR Ca uptake in canine heart homogenates (Hess et al., 1984) and isolated SR vesicles (Okabe et al., 1983). Because the net Ca uptake in the SR is a result of the activity of CaATPase and of the SR Ca-release channel, an abnormal Ca uptake may be the result of the dysfunction of either or both structures. The site or sites of action for oxygen-derived free radicals damage are unknown, although previous studies on the SR have focused on damage to the Ca pump. Direct effects of oxygen-derived free radicals on SR Ca-release channels may be important in understanding their potential contribution to ischemia/reperfusion injury and developing strategies to protect against such injury. Previously, we found that decreased SR Ca uptake in response to oxygen-derived free radicals is associated with an enhanced Ca loss by the Ca-release process (Okabe et al., 1988, 1991). We now provide evidence that O2 . dramatically alters the gating characteristics of the reconstituted RyRC from the heavy fraction of cardiac SR vesicles due to decreased calmodulin content; this is a novel mechanism for SR Ca release by oxygen-derived free radicals in cardiac muscle. These findings indicate that an elevation in cytosolic Ca due to abnormal Ca handling, This work was supported by Grants 09877356 (E.O.) and 07557119 (E.O.) from Scientific Research Fund of The Ministry of Education, Science, Sports and Culture of Japan and by a grant from the Research Fund of JEOL Ltd., Tokyo, Japan. ABBREVIATIONS: SR, sarcoplasmic reticulum; Cai, intravesicular free Ca ; O2 ., superoxide anion radical; RyRC, ryanodine receptor Carelease channel(s); Po, open probability; MOPS, 3-(N-morpholino)propanesulfonic acid; EGTA, ethylene glycol bis(b-aminoethyl ether)-N,N,N9,N9tetraacetic acid, HEPES, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid; DMPO, 5,5-dimethyl-1-ryrroline-N-oxide; MnO, manganese oxide marker; SOD, superoxide dismutase; DMSO, dimethylsulfoxide; HOz, hydroxyl radical; Ca-ATPase, Ca-stimulated, Mgdependent ATPase; SH, sulfhydryl. 0026-895X/98/030497-07$3.00/0 Copyright © by The American Society for Pharmacology and Experimental Therapeutics All rights of reproduction in any form reserved. MOLECULAR PHARMACOLOGY, 53:497–503 (1998). 497 at A PE T Jornals on Jne 6, 2017 m oharm .aspeurnals.org D ow nladed from such as the increase in Po of RyRC, by the SR may contribute to in vivo reperfusion injury. Materials and Methods Heavy SR vesicles preparation and Ca efflux measurements. Canine cardiac heavy SR was isolated by discontinuous sucrose gradient centrifugation according to a modified method described previously (Valdivia et al., 1991). Briefly, canine ventricular muscles were minced in a food processor and homogenized for 60 sec in buffer containing 0.3 M sucrose, 20 mM MOPS, pH 7.2, and protease inhibitors (1 mg/ml pepstatin and leupeptin, 1 mM iodoacetamide, 0.1 mM phenylmethylsulfonyl fluoride, 0.1 mM benzamidine, and 10 mg/ml aprotinin). The homogenate was centrifuged for 20 min at 11,000 3 g. The supernatant was centrifuged for 60 min at 119,000 3 g. After centrifugation, the supernatant was discarded. The pellet underwent fractionization overnight on a discontinuous sucrose gradient (10%, 31%, 40%, and 50%) in a solution of 400 mM KCl, 20 mM MOPS, pH 6.8, and 100 mM MgCl2, CaCl2, and EGTA in a Beckman Instruments (Columbia, MD) SW27 rotor at 25,000 rpm. The final pellets were resuspended in 0.3 M sucrose and 20 mM MOPS, pH 6.8, in addition to the previously mentioned mixture of protease inhibitors. Protein concentration was determined according to the method of Lowry et al. (1951). The resulting heavy SR vesicles were preincubated overnight on ice in 2 mM CaCl2 (New England Nuclear Research Products, Boston, MA), 150 mM KCl, and 20 mM MOPS, pH 6.8. They then were diluted 20-fold into a Ca-releasing medium containing 150 mM KCl, 20 mM MOPS, pH 6.8, and 1 mM EGTA/Ca buffer to adjust the pCa to 5. Ca efflux was quenched with ice-cold quench solution containing 1 mM LaCl3, 10 mM MgCl2, 150 mM KCl, and 20 mM MOPS, pH 6.8. After filtration through Millipore (Bedford, MA) filters (0.45 mm) and washing of the filters with the quenching solution, the radioactivity retained by the filter was determined by liquid scintillation counting. EGTA washing of heavy SR vesicles. To remove endogenous calmodulin from SR, the vesicle suspension was diluted 1:100 in 20 mM HEPES, pH 7.4, kept 20 min on ice and made hypertonic by the addition of the same volume of 1.2 M KCl, 20 mM HEPES, and 4 mM EGTA, pH 7.4. The hypertonic suspension was centrifuged for 30 min at 150,000 3 g, and the pellet was washed twice with 20 mM HEPES, pH 7.4. The final pellet was resuspended in 0.3 M sucrose and 20 mM MOPS, pH 6.8, containing the mixture of protease inhibitors and used immediately. Confirmation of the depletion of endogenous calmodulin was obtained according to the method of Schulman and Greengard (1978), in which EGTA-washed SR vesicles were phosphorylated with [g-P]ATP (New England Nuclear Research Products) in the presence of 30 mg of washing extract or 0.6 mM calmodulin (from bovine brain; Fluka AG, Buchs, Switzerland) and then subjected to preparative sodium dodecyl sulfate-polyacrylamide slab gel electrophoresis. The washing and boiled extracts of SR stimulated the incorporation of P from 10 mM [g-P]ATP into EGTA-washed SR proteins in the presence of 0.5 mM CaCl2. With the assumption that kinase is activated only by calmodulin, this result demonstrates the presence of calmodulin in the extracts. Moreover, it was found that a hypotonic treatment, followed immediately by a hypertonic wash in the presence of 4 mM EGTA and by several hypotonic washes in the absence of the chelator, resulted in the depletion of calmodulin (Carafoli et al., 1980). ESR analysis. The spin-trapping studies were performed with the desired mixture containing DMPO (Labotec, Tokyo, Japan; 99– 100% pure, gas chromatographic assay by Dojindo Laboratories, Kumamoto, Japan). ESR detection of the spin adduct was carried out at room temperature with a JEOL (model JES-RE3X) X-band spectrometer connected with the JOEL computer system Esprit (Tokyo, Japan). Hyperfine coupling constants were calculated based on the resonance frequency measured with a microwave frequency counter and the resonance field measured with the JEOL field measurement unit model ES-FC5. ESR spectra were recorded at the instrument settings of 0.05-mT (100 kHz modulation amplitude), 10-mT recording range, 2-min recording time, 0.1-sec time constant, 8-mW (9.414GHz microwave power), and 335.6 6 5-mT magnetic field. A quantitative analysis of the spin adducts of O2 . was performed as described previously (Mitsuta et al., 1990). After recording of the ESR spectra, the signal intensity of DMPOOO2 .(OOOH) was normalized as a relative height against the standard signal intensity of MnO. An absolute concentration of DMPOOOOH was determined by a double integration of the ESR spectrum, in which a 1.0 mM concentration of 4-hydroxyl-2,2,6,6-tetramethylpiperidine-N-oxyl solution was used as a primary standard of ESR absorption. Calmodulin content of heavy SR vesicles. Heavy SR vesicle fractions at a protein concentration of 1 mg/ml were incubated for 10 min at 22° or heated for 10 min at 95° in media containing either 20 mM K-piperazinediethanesulfonic acid, pH 7.0, 0.1 M KCl, 100 mM EGTA, and 106 mM Ca (10 mM free Ca) or 0.125 M borate, pH 8.4, 0.075 M NaCl, 0.2% bovine serum albumin, and 1 mM EGTA (,10 M free Ca) in the presence or absence of 20 mM hypoxanthine (Sigma Chemical, St. Louis, MO). Next, vesicles underwent sedimentation for 30 min at 100,000 3 g in a Beckman airfuge. Xanthine oxidase (0.1 unit/ml; activity, 35.8 mM/min; Boehringer-Mannheim Biochemicals, Indianapolis, IN) was added 2.0 min before the sedimentation, and SOD (10 mg/ml; 3000 units/ml; Sigma Chemical) was added 30 sec before the addition of xanthine oxidase. The supernatants of samples not heated at 95° were incubated for 10 min at 95°. The calmodulin content of the supernatant fractions were measured with the use of an I-calmodulin radioimmunoassay kit from New England Nuclear Research Products). Cai. Cai was calculated after passive Ca 21 efflux, Jp, from heavy SR vesicles was measured as described previously (Okabe et al., 1988). Briefly, steady state Ca uptake was measured in the absence of Ca-precipitating anions at 27° by filtration through 0.45-mm Millipore filters of 1.0-ml aliquots from a 10-ml bath containing 100 mM KCl, 20 mM imidazole, pH 7.0, 10 mM NaN3, 100 mM disodium ATP, 2.1 mM MgCl2, 0.1 mCi of Ca/ml, and 4 mM added Ca. Total Ca in the reaction bath was determined by atomic absorption spectrophotometry after wet ashing of the reaction bath including SR. The total Ca associated with the SR was obtained by Millipore filtration and was calculated on the basis of the total Ca in the reaction bath and the Ca in the filtrates of the reaction bath. The uptake reaction was begun by the addition of ATP, Ca, and Mg to an otherwise complete reaction bath. Passive Ca efflux was measured after steady state Ca uptake was reached through quenching of pump-mediated Ca fluxes and observation of the net release in Ca by Millipore filtration. Quenching of the pump-mediated fluxes was produced by the addition of EGTA to a final concentration of 2.5 mM. The initial apparent first-order rate constant, K/v, was obtained by linear regression of the natural logarithm of the Ca uptake determined through Millipore filtration at various times after the addition of EGTA. The initial passive Ca efflux, Jp, was obtained from the product of the first order-rate constant and the initial Ca load. It is assumed in all experiments that total Ca in the reaction bath is distributed among four compartments; these are Cao, the Ca in solution outside the vesicles; Cabo, the Ca 21 bound to the outside of the vesicles; Cabi, the Ca 21 bound to intravesicular binding site; and Cai, Ca 21 free within the intravesicular space. The initial value (Cat 2 Cabo) was determined by extrapolating the first-order efflux curve to the time of the addition of EGTA quench. The initial passive Ca efflux, calculated as Jp 5 K/v (Cat 2 Cabo), is driven by Cai. Cai is not directly measured in the current experimental system. However, the total internal Ca can be calculated as Cai 1 Cabi 5 Cat 2 Cabo, provided Cabo is known. The Jp value from SR vesicles was measured at various loads obtained by actively loading the vesicles in the presence of 0–25 mM EGTA. By plotting each obtained Jp value against the Ca 21 load (Cat 2 Cabo 5 Cai 1 Cabi, the sum of free and bound intravesicular Ca), Cabi can be determined from 498 Kawakami and Okabe at A PE T Jornals on Jne 6, 2017 m oharm .aspeurnals.org D ow nladed from the extrapolated intercept of the line onto the abscissa. Cai was calculated according to Cai 5 Cat 2 Cabo 2 Cabi. Ca-ATPase activity. The ATPase activity was determined from the rate of Pi release from [gP]ATP according to the method of Feher and Briggs (1980). Planar phospholipid bilayer experiments. Single-channel recordings were carried out by incorporating the native or EGTAwashed calmodulin-depleted heavy SR vesicles into planar phospholipid bilayers according to a previous method (Smith et al., 1985). The planar phospholipid bilayers, composed of phosphatidylethanolamine (bovine heart; Avanti Polar Lipids, Birmingham, AL) dispersed in decane at a concentration of 25 mg/ml, were painted across a 200-mm-diameter hole in the styrene copolymer septum between the two experimental chambers containing 5 mM CaCl2, 50 mM choline chloride, and 10 mM HEPES/Tris, pH 7.2. Heavy SR vesicles (10 mg/ml) then were added to the designated cis chamber, and the solution was fortified with choline chloride to produce a 7:1 gradient across the membrane. Vesicle fusion was monitored as steplike conductance increases that resulted in a Cl-specific macroscope current. After fusion, the cis chamber was perfused with 1 mM Ca-EGTA (10 mM free Ca), and 250 mM HEPES/125 mM Tris, pH 7.4, and the trans chamber was perfused with 250 mM glutamic acid, and 10 mM HEPES, adjusted to pH 7.4 with Ca(OH)2, to give a solution with '67 mM free Ca. Channel opening results in a flow of ions across the bilayer, which was amplified by a patch-clamp amplifier (Axopatch; Axon Instruments, Foster, CA), and stored on a videocassette tape recorder through a PCM converter system (RP-880; NF Instruments, Yokohama, Japan) filtered at 1 kHz and digitized at 2 kHz. All experiments were recorded at room temperature (22°). All recordings were made with the cis chamber voltage-clamped at 0 mV relative to ground. Po of channels, and the lifetimes of open and closed events were identified by 50% threshold analysis. Po values were calculated from 3-min records of steady state recordings. Channel openings are presented as upward deflections.