Creatine Kinase Is the Main Target of Reactive Oxygen Species in Cardiac Myofibrils
Abstract Reactive oxygen species (ROS) have been reported to alter cardiac myofibrillar function as well as myofibrillar enzymes such as myosin ATPase and creatine kinase (CK). To understand their precise mode and site of action in myofibrils, the effects of the xanthine/xanthine oxidase (X/XO) system or of hydrogen peroxide (H2O2) have been studied in the presence and in the absence of phosphocreatine (PCr) in Triton X-100–treated cardiac fibers. We found that xanthine oxidase (XO), with or without xanthine, induced a decrease in maximal Ca2+-activated tension. We attributed this effect to the high contaminating proteolytic activity in commercial XO preparations, since it could be prevented by a protease inhibitor, phenylmethylsulfonyl fluoride (PMSF), and it could be mimicked by trypsin. In further experiments, XO was pretreated with 1 mmol/L PMSF. Superoxide anion production by the X/XO system, characterized by electron paramagnetic resonance spin-trapping technique, was not altered by PMSF. A slight increase in maximal force was then observed either with X/XO (100 μmol/L per 30 mIU/mL) or H2O2. pMgATP–rigor tension relationships have been established in the presence and in the absence of PCr to separate the effects of ROS on myosin ATPase and myofibrillar-bound CK. In the presence of PCr, pMgATP50, the pMgATP necessary to induce half-maximal rigor tension, was reduced from 5.03±0.17 (n=21) to 4.22±0.22 (n=4) after 25 minutes of incubation in the presence of 30 mIU/mL XO and 100 μmol/L xanthine or to 4.04±0.1 (n=11) after incubation in the presence of 2.5 mmol/L H2O2. The ROS effects were partially prevented or antagonized by 1 mmol/L dithiothreitol. No effect was observed on pMgATP50 when PCr was absent. pCa-tension relationships have been evaluated to assess the effects of ROS on active tension development. Incubations with H2O2 induced an increase in Ca2+ sensitivity and resting tension when MgATP was provided through myofibrillar CK (PCr and MgADP as substrates) but not when MgATP was added directly. These results suggest that myofibrillar CK was inhibited by ROS. Active stiffness and the time constant of tension changes after quick stretches applied to the fibers were dose-dependently increased by H2O2 only in the presence of PCr. In addition, myofibrillar CK but not myosin ATPase enzymatic activity was depressed after incubation with either ROS. These results suggest that ROS mainly alters CK in myofibrils, probably by the oxidation of its essential sulfhydryl groups. Such CK inactivation results in a decrease in the intramyofibrillar ATP-to-ADP ratio. The effects of ROS on cytosolic and bound CKs may take part in the overall process of myocardial stunning after cardiac ischemia and reperfusion.
Oxygen-derived free radicals have been implicated in various types of myocardial injury, especially in ischemia/reperfusion injury (for review, see References 1 and 21 2 ). Electron spin resonance investigations have demonstrated that a brief episode of ischemia and the subsequent reperfusion are associated with free radical generation.3 4 5 In reperfused hearts, it has been shown that both the decrement of oxygen-derived free radicals after the initial burst and the recovery of the stunned myocardium are accelerated by agents that either scavenge oxygen radicals, such as SOD and CAT,6 7 or prevent their generation, such as allopurinol and desferrioxamine.8 Alterations in the recovery of PCr, following the peak of oxygen radical production, were shown to be largely prevented by the administration of SOD.9
Different ROS have been shown to induce negative inotropic effects, contracture, cell damage, an increase in lipid peroxidation, and a decrease in high-energy phosphate contents in the myocardium10 11 12 (for review, see Reference 11 ). Since the contractile apparatus plays the central role in cardiac function, deleterious effects of ROS on myofibrils would be of critical importance. Thin and thick filaments contain SH groups that could be modified by oxidizing agents. The main myofibrillar enzymatic systems, myosin ATPase and CK, have been shown to be susceptible to different ROS. The marked changes in myofibrillar ATPase and CK induced by X/XO and hydrogen peroxide (H2O2) may be mediated by the oxidation of SH groups in the enzyme.13 14 15 Incubation of cardiac myofibrils with diamide (an oxidizing agent), H2O2, X/XO, or hypochlorous ions (HOCl) increases basal ATPase activity and abolishes its Ca2+ dependence.13 16
Skinned fibers have been used to assess the consequences of myofibril exposure to different ROS on the intrinsic mechanical properties of contractile proteins. ROS have been shown to induce quite specific alterations, the mechanisms of which have not been identified. A sharp and irreversible decline in maximal force, without changes in Ca2+ sensitivity, has been described with the X/XO system.17 It was shown that HOCl mainly increased rigor tension and the Ca2+ sensitivity of active tension, whereas H2O2 only marginally affected mechanical properties.18
The aim of the present study was to characterize the functional consequences of the exposition of myofibrillar proteins to ROS. Triton X-100–treated cardiac fiber bundles were used to assess intrinsic mechanical properties of myofibrils in controlled ionic and substrate environment. By studying the action of ROS in the presence or in the absence of PCr, we sought to compare the susceptibility of myofibrillar CK and myosin ATPase. MgATP was provided either directly in the solution or through endogenous ATP production catalyzed by myofibrillar-bound CK in the presence of PCr and MgADP. The results show that the main target of ROS in myofibrils seems to be the essential SH groups of CK. Oxidation of these groups would induce an inhibition of the enzyme and an alteration of the contractile properties as a consequence of a decreased ATP-to-ADP ratio in the myofibrillar compartment.
Materials and Methods
Rats were anesthetized with an intraperitoneal injection of urethane (2 g/kg) according to the recommendations of the institutional Animal Care Committee (INSERM, Paris, France). Muscle fiber bundles (≈200 μm in diameter) were dissected from papillary muscles of the left ventricle of rats in a zero-Ca2+ Krebs’ solution, pH 7.4. Fibers were incubated for 1 hour in a relaxing solution (solution A; see solutions below) containing 1% Triton X-100 to solubilize the membranes and were then transferred to the relaxing solution without detergent. After the skinning procedure, the fiber was mounted in the experimental apparatus. It was adjusted to slack length, stretched by 20%, and subjected to an activation/relaxation cycle. Sarcomere length was adjusted to 2.2 μm using laser diffraction (10 mW He-Ne laser, Spectra Physics Inc) after the first activation cycle. The length and diameter of the muscles were measured using a graticule in the dissecting microscope. Muscles were immersed in 2.5-mL chambers arranged around a disk in a temperature-controlled bath positioned on a magnetic stirrer. Each solution was well stirred at high speed (>1000 rpm). All experiments were performed at 22°C.
One fiber was snared at both ends with hair emerging from stainless steel tubes, between a transducer (model AE 801, SensoNor Microelectroniks) and a vibrator.19 The bandwidth of the transducer and tube was 2 kHz. The permanent magnet and coil came from a standard loudspeaker (Pioneer TS-130A, Pioneer Electric and Research Corp). The coil was glued to a glass tube axis (2 mm in diameter) driven in an axial ball bearing (total moving mass, <1.5 g). A flag with a narrow window was glued on the glass axis between a lamp and a position detector (type S1543, Hamamatsu), allowing measurement of the displacement length. A feedback with the length signal combined with a power amplifier allowed control of muscle length. The system had a rise time of ≈1 millisecond without overshoot. Length and tension changes were driven by a pulse generator (GP1-GE 219, CEA-DAM France) and monitored on an oscilloscope (model OS4020, Gould, Inc). Tension and length tracings were digitized at 20 kHz (12-bit analog-to-digital converter), analyzed on-line using a computer (Deskpro 486, COMPAQ Corp), and stored on a videotape.
Solutions were calculated using the computer program of Fabiato20 and are listed in Table 1⇓. Rigor solutions were obtained by mixing solutions A4 and E′ with PCr or solutions C and G without PCr. Activating solutions were obtained by mixing two solutions of pCa 9 and 4.5 (solutions A and B, respectively). EGTA, PMSF, xanthine, H2O2, CAT, SOD, DMPO, and desferrioxamine were purchased from Sigma Chemical Co. XO (X4376) was obtained from Sigma as lyophilized compound. PCr (Neoton, Schiapparelli Farmaceutica) was a kind gift of Prof E. Strumia. All solutions for ROS generation were prepared immediately before use. Experiments were performed at pH 7.1 and 22°C.
pMgATP–Rigor Tension Relationships
Before and after incubating the fiber in solution A containing either the X/XO system or H2O2 for 25 minutes, pMgATP–rigor tension relationships were established by a stepwise decrease in ATP concentration until maximal rigor tension was obtained. Data were fit using the Hill equation: where T is the relative rigor tension, K is a constant, and nH is the Hill coefficient. pMgATP50 [(−log10K)/nH] was calculated for each experimental condition (between 10% and 90%) using linear regression analysis.
Resting tension was the tension in relaxing solution A at a sarcomere length of 2.2 μm. Tmax of each fiber was obtained at pCa 4.5. The pCa-tension relationships were established under isometric conditions by stepwise increasing [Ca2+] until Tmax was reached. Data were fit using the Hill equation: where T is the relative tension. pCa50 [(−log10K)/nH] was calculated for each experimental condition (between 10% and 90%) using linear regression analysis. Active tension was total tension at pCa 4.5 minus resting tension. Tension was expressed as millinewtons per square millimeter.
Stiffness and Kinetic Measurements
Fiber stiffness was estimated by the quick length change technique as previously described.19 A series of seven stretches and five releases of 0.3% to 3% initial muscle length were applied in relaxing or activating solutions. Only responses to stretch were used for calculations. The spike of tension in phase with the length change characterizes the elastic phase.21 Stiffness was the extreme tension reached during stretch (in millinewtons per square millimeter) divided by the length change (in micrometers). Each value for each fiber is the mean of five to seven determinations after stretches of varying amplitudes in a given experimental condition. A first series of length changes was imposed in the relaxing solution to assess the passive properties of the fiber. Resting stiffness was calculated by linear regression analysis on the responses to stretches. Then a second series of length changes was initiated in the control activating solution. The tension level before the first stretch was taken as the Tmax and was used for normalization. Active stiffness was calculated as the difference between total stiffness and resting stiffness.
The spike of tension occurring after a quick stretch is followed by a rapid drop in tension to a minimum value. The time constant (in milliseconds) of this tension drop was calculated by a least-squares regression analysis.
Enzyme Activity Measurements
CK and myosin ATPase activities were measured according to Hoerter et al22 on skinned isolated cells in conditions similar to those for the mechanical experiments.
Adult rat cardiac myocytes were isolated by collagenase perfusion according to Pucéat et al.23 After isolation, the cardiomyocytes were skinned for 5 minutes in solution A containing 0.5% Triton X-100 and washed twice. Freshly isolated skinned cells were incubated for 25 minutes at 22°C in solution A (≈3 mg protein per milliliter) in the presence or in the absence of either H2O2 or the X/XO system. After incubation, aliquots were used for the assays.
CK activity was assayed using the coupled-enzyme system of G-6-PDH and hexokinase. CK activity was determined by measuring NADPH production at 340 nm (Gilford Spectrophotometer) in solution E′ (Table 1⇑) also containing (mmol/L) ADP 1, glucose 20, and NADP 0.6, along with 2 IU/mL of both hexokinase and G-6-PDH at a pH 7.1 and 22°C. Reaction was initiated by PCr addition.
Myosin ATPase activity was assayed using the coupled-enzyme system of pyruvate kinase and lactate dehydrogenase. ATPase activity was determined by measuring NADH consumption at 340 nm in solution C (Table 1⇑) containing, in addition, 0.5 mmol/L phosphoenolpyruvate, 0.15 mmol/L NADH, and 2 IU/mL of both pyruvate kinase and lactate dehydrogenase at pH 7.1 and 22°C. Sodium azide (5 mmol/L) was added to inhibit possible residual mitochondrial ATPase activity. The Mg-ATPase reaction was started by addition of ATP. Ca,Mg-ATPase activity was measured after the addition of CaCl2 to obtain pCa 4.5. Ca-ATPase activity was the difference between Mg-ATPase and Ca,Mg-ATPase activities. The protein content of skinned isolated cells was determined by the Lowry assay.24
Free Radical Generation
Free radical species generated by the different systems used were characterized by EPR spin-trapping techniques using the spin trap DMPO at a final concentration of 50 mmol/L. DMPO was further purified by double distillation. Care was taken to keep the DMPO solution protected from light-induced degradation.
EPR spectra were recorded in a flat-type quartz cell at room temperature with an X-band Bruker ESP 300E spectrometer with a TM110 cavity using a modulation frequency of 100 kHz, modulation amplitude of 0.5 G, microwave power of 20 mW, time constant of 0.16 second, and a scan rate of 100 G/42 s. Results are expressed in arbitrary units.
Values are expressed as mean±SEM. Student’s t test was used to compare the means. Statistical significance was reached at P≤.05. Linear regression analysis was carried out using the least-squares method.
Effects of H2O2 and X/XO on Maximal Ca2+-Activated Tension
The effects of H2O2 on Tmax were studied after activating the fiber at pCa 4.5. As shown in Fig 1⇓, increasing the concentration of H2O2 from 0.5 to 5 mmol/L induced a slight and irreversible rise in Tmax. After relaxing the preparation in the absence of H2O2, the resting tension failed to return to the control level, so that active tension per se diminished.
The X/XO system is widely used to produce specific ROS, mainly superoxide anion (O2−•) and H2O2. We found that X/XO, in contrast to H2O2, induced an irreversible slow decline in Tmax, followed by a complete recovery of resting tension as previously described.17 This decline was accelerated by increasing activities of XO with a threshold for 2 to 3 mIU/mL. When XO was applied without xanthine, Tmax nevertheless decreased. DTT (1 or 2 mmol/L), a reducing agent, did not prevent or block the decrease. When the fiber was incubated in the presence of 400 IU/mL SOD, which catalyzes the dismutation of superoxide anions into H2O2, together with 100 IU/mL CAT, which reduces H2O2 into H2O, further application of 50 mIU/mL XO alone still induced the same decrease in maximal force, whereas further addition of the substrate (xanthine, 50 μmol/L) did not change the slope of the tension fall until complete disappearance of force (Fig 2A⇓).
Mammalian xanthine dehydrogenase/XO is an interconvertible enzyme, and the conversion from xanthine dehydrogenase to XO occurs either reversibly by oxidation of the SH residues or irreversibly by proteolysis.25 Commercial preparations of XO thus contain a contaminating proteolytic activity.13 In order to see whether a proteolytic activity can mimic the XO effects on Tmax, we applied 10 mIU/mL trypsin to a maximally Ca2+-activated fiber. Trypsin induced a fall in Tmax similar to that induced by XO, which could be inhibited by a specific inhibitor of trypsin-like proteases, leupeptin (10 and then 20 μmol/L, Fig 2B⇑). Leupeptin did not completely inhibit the decrease in Tmax induced by XO. Yet, another protease inhibitor, PMSF, could prevent the decrease. When 1 mmol/L PMSF was applied, tension first abruptly decreased and then stabilized, but further addition of XO (50 mIU/mL) with or without xanthine (50 μmol/L) did not induce further decline; moreover, PMSF could stop the decline induced by XO (50 mIU/mL) alone (Fig 2C⇑). Finally, since PMSF alone induced a sharp and small decrease in Tmax, the stock solution of XO was pretreated with 1 mmol/L PMSF to completely block the protease activity, so that the final concentration of PMSF applied to the fiber never exceeded 10 μmol/L. Application of PMSF-pretreated XO (30 mIU/mL) alone was without effect on Tmax (Fig 2D⇑). Utilization of pretreated XO in the presence of xanthine did not induce any decrease in Tmax, but a small increase was generally observed for high X/XO activity (100 μmol/L per 30 mIU/mL). In the following experiments, pretreated XO was always used.
Free Radical Generation
Since many substances have been described to exert free radical–scavenging effects, we analyzed the production of free radicals by the X/XO system and H2O2 in the solutions used for skinned fiber studies and in the presence or absence of PMSF to ensure that this molecule did not act as a free radical scavenger.
As shown in Fig 3a⇓, the EPR spectrum obtained with the X/XO system in solution A consisted of a prominent 12-line signal arising from the DMPO-superoxide adduct, DMPO-OOH, and a smaller quartet signal originating from the hydroxyl adduct, DMPO-OH. The DMPO-OH signal could be due to the decomposition of DMPO-OOH or to direct trapping of · OH radicals. The addition of ethanol (10% [vol/vol]), which could potentially react with ·OH radicals and give ·CH(OH)CH3 radicals, was not associated with the appearance of the DMPO-CH(OH)CH3 adduct, suggesting that under our experimental conditions, DMPO-OH generation occurred mostly from the breakdown of DMPO-OOH rather than from direct ·OH trapping. PMSF at a final concentration of 10 μmol/L, used in the mechanical studies, did not modify the EPR spectrum and therefore did not show any superoxide-scavenging property (Fig 3b⇓). However, SOD (10 IU/mL) scavenged the superoxide radicals, preventing the formation of DMPO adducts (Fig 3c⇓). The superoxide generation continued for ≈10 minutes and was ≈8.4 arbitrary units per milliliter per minute (≈10 nmol per milliliter per minute) in the presence of 30 mIU/mL XO and 100 μmol/L xanthine. With the lower xanthine concentration (30 μmol/L), the superoxide generation intensity was reduced to ≈1.8 arbitrary units per milliliter per minute but was prolonged until 25 minutes (Fig 3d⇓). In solution A, H2O2 used at a concentration of 2.5 mmol/L in the mechanical studies did not induce the formation of a DMPO adduct, even when the DMPO concentration was increased to 100 mmol/L, suggesting the absence of or very low free radical generation.
Effects of H2O2 or X/XO on Rigor Tension Development
In order to elucidate the mechanisms by which free radicals alter the myofibrillar function, we first studied the effects of ROS on rigor tension development. Rigor tension develops in the absence of Ca2+, when MgATP is depleted close to the myosin ATPase. We have shown previously that when PCr is present, myofibrillar-bound CK can increase the local ATP-to-ADP ratio, and the pMgATP–rigor tension curve is shifted toward a lower MgATP; the amplitude of the shift is thus an estimate of the myofibrillar CK efficacy (for review, see Reference 2626 ). Since the ROS are potentially able to inhibit both myosin ATPase and myofibrillar CK, the effects of H2O2 and of the X/XO system on rigor tension development were assessed in the presence and absence of PCr to dissociate their influence on either myosin ATPase or CK function in myofibrils.
In the following experiments, fibers were incubated with either ROS for 25 minutes in the relaxing solution to compare their long-lasting effects on rigor tension. pMgATP–rigor tension relationships were obtained by bathing the fibers in a set of solutions of decreasing MgATP before and after exposure to the ROS.
Effects of 2.5 mmol/L H2O2 on typical fibers are shown in Fig 4⇓. In the absence of PCr, the threshold for rigor tension development was in the millimolar range, and the MgATP concentration for half-maximal rigor tension was ≈400 μmol/L (Fig 4A⇓). Incubation of the fiber with H2O2 was without effect on rigor tension development.
In the presence of PCr, under control conditions, rigor tension developed at much lower MgATP concentrations, with an MgATP concentration for half-maximal tension of ≈10 μmol/L, as previously described.26 27 28 Fig 4B⇑ shows the behavior of a representative fiber. After the control pMgATP–rigor tension relationship in the presence of PCr had been obtained, the fiber was first incubated for 25 minutes in the presence of 2.5 mmol/L H2O2 and CAT (100 IU/mL). No shift in the pMgATP–rigor tension relationship was observed, showing that CAT completely prevented the effects of H2O2. When the incubation was performed without CAT but in the presence of 2.5 mmol/L H2O2 together with 1 mmol/L DTT (a concentration used to prevent myosin ATPase and CK inhibition by XO, H2O2, or HOCl13 14 15 ), the pMgATP–rigor tension curve was shifted slightly to the right. However, when this fiber was then incubated in the presence of H2O2 alone, the shift was much more pronounced, reaching a value close to control conditions in the absence of PCr (see Table 2⇓ for mean values). Finally, in the presence of 1 mmol/L DTT, the curve returned to the level attained after incubation in the presence of both H2O2 and DTT. These results suggest that not all the effects of H2O2 could be attributable to DTT-sensitive SH group oxidation and that this long protocol did not significantly alter the fiber response. Effects of H2O2 on rigor tension development in the presence of PCr were not prevented by the addition of 100 μmol/L desferrioxamine, a concentration sufficient to inhibit iron-mediated CK inactivation29 (data not shown), suggesting that hydroxyl radicals were not involved. Exposure of the fibers to lower H2O2 concentration (100 μmol/L) for up to 25 minutes had no effect on resting tension or rigor tension with or without PCr (data not shown).
The same kinds of experiments were performed with the X/XO system (Fig 5⇓ and Table 2⇑). Results similar to the H2O2 effects were obtained. X/XO incubation did not alter rigor tension development in the absence of PCr (Fig 5A⇓). In the presence of PCr (Fig 5B⇓), incubation with X/XO in the presence of 400 IU/mL SOD and 100 IU/mL CAT was without effects. X/XO alone induced a rightward shift of the pMgATP–rigor tension relationship, which could be partially reversed by 1 mmol/L DTT. Increasing DTT concentration to 2 mmol/L did not further modify rigor tension (data not shown). When the fibers were incubated with 3 mIU/mL pretreated XO and 30 μmol/L xanthine, no effects on rigor tension were observed.
Since the ROS were effective only when PCr was present, the results suggest that the alterations in rigor tension development are mainly the consequence of an inhibition of myofibrillar CK.
Effects of H2O2 on Active Tension Development
ROS have been shown to increase Ca2+ sensitivity of tension development.18 Ca2+ sensitivity is a complex parameter that depends on many factors, such as regulatory proteins on the thin filament, protein interactions within the thin filament, and the kinetics of crossbridges. In order to identify the possible target and to dissociate between the direct effects of ROS on the thin-filament regulatory proteins and the indirect effects via myosin ATPase or CK alterations, Ca2+ sensitivity of skinned fibers was assessed by stepwise increasing the Ca2+ concentration in the presence of different substrates. One set of solutions (solutions C and D in Table 1⇑) contained MgATP (3.16 mmol/L) as a substrate, whereas in the other set (solutions E and F in Table 1⇑), MgATP was provided exclusively through the rephosphorylation of MgADP by the endogenous CK in the presence of PCr. When MgATP was the only substrate, the pCa-tension relationships obtained before and after incubation of a representative fiber with 2.5 mmol/L H2O2 for 25 minutes were similar (Fig 6A⇓). The averaged pCa50 and Hill coefficient were, respectively, 5.98±0.13 and 1.44±0.08 (n=3) before and 5.96±0.13 and 1.27±0.06 (n=3) after incubation. As already described,28 when MgATP was provided by the CK reaction, the pCa-tension relationship was shifted toward higher Ca2+ concentrations, and the slope became more steep (pCa50, 5.53±0.04 [n=3, P<.05]; nH, 2.63±0.25 [n=3]; P<.05 compared with MgATP alone). After incubation with H2O2, a high increase in resting tension was observed together with an increase in Ca2+ sensitivity (Fig 6B⇓). After normalization, the relationship appeared shifted toward lower Ca2+ concentrations (pCa50, 6.07±0.19 [n=3]; P<.01), and Hill coefficient decreased to 1.03±0.21 (n=3, P<.01). Since changes in Ca2+ sensitivity were observed only when MgATP was provided through the CK reaction, the results strongly suggest that H2O2 mainly acts through an inhibition of the myofibrillar CK and that this inhibition is responsible for Ca2+ sensitivity changes. Similar results were obtained after incubation with XO (30 IU/mL) and xanthine (100 μmol/L) (data not shown).
Effects of H2O2 on Tension Kinetics and Stiffness
We have shown previously that inhibition of myofibrillar CK changes the local ATP-to-ADP ratio and alters the mechanical properties.28 Local deficiency in MgATP or more possibly MgADP accumulation26 27 tends to inhibit crossbridge detachment (or the transition from strong to weak binding states) and thus to slow down the overall crossbridge cycling rate. Since a higher number of crossbridges is attached at a given moment, the Ca2+ sensitivity of tension development and the overall amount of Ca2+ bound to the regulatory proteins are increased.26 30 31 32
When an activated fiber is subjected to quick changes in length, the immediate tension response to the length change is determined by the number of attached crossbridges; the kinetics of the return of tension to control level depends on the crossbridge cycling rate.21 Quick length changes were thus applied to test whether the ROS affect crossbridge kinetics and stiffness. Tension responses to quick length changes were first determined in the relaxing solution to assess passive mechanical properties. Responses in relaxing solutions were then subtracted from responses in control activating solution (pCa 4.5) to determine active mechanical properties. The preparations were then successively bathed in solutions containing increasing concentrations of H2O2 in the presence of ADP and PCr. We observed that fibers studied in solutions that did not contain a protector of SH groups (DTT) were less stable and evidenced a time-dependent increase in the time constant of tension recovery after stretches as well as a decrease in tension and absolute stiffness. The rundown of the preparations was measured in a parallel series of experiments, and the values obtained with H2O2 were corrected accordingly. The results of such an experiment are shown in Fig 7⇓. Results obtained on four different fibers showed that H2O2 increased the time constant of tension recovery in a dose-dependent manner, ie, decreased the cycling rate of the crossbridges. These changes were already significant at 0.5 mmol/L H2O2, whereas stiffness normalized to force started to increase for higher (1 to 2 mmol/L) concentrations. An increase in this parameter indicates a decreased force generated per crossbridge.
The time constant of tension changes was also estimated before and after incubation of the fiber for 25 minutes with 2.5 mmol/L H2O2. In the presence of MgATP alone (solution D), the time constant of tension recovery was not altered by H2O2 incubation (37.0±2.0 versus 36.2±1.5 milliseconds in control, n=4). In the presence of PCr (solution F), the time constant of tension changes was dramatically increased from 19.1±1.6 to 49.5±3.3 milliseconds (n=5, P<.001) after incubation with H2O2.
Effects of H2O2 or X/XO on Myosin ATPase and Myofibrillar CK–Specific Activities
The enzymatic activities of myofibrillar CK and myosin ATPase have been measured on Triton-skinned isolated cardiac cells in conditions similar to those of mechanical experiments with skinned fibers. Isolated cardiac cells were incubated in 1 mL solution A for 25 minutes at 22°C in different conditions. Enzymatic activities were measured for each sample before and after incubation.
Control myofibrillar CK activity was 1.27±0.03 IU/mg protein (n=26). As shown in Fig 8A⇓, CK was inhibited by incubation with 2.5 mmol/L H2O2; this inhibition was prevented by 100 IU/mL CAT. Incubation with 10 μmol/L PMSF and 100 μmol/L xanthine did not modify CK activity. XO (30 mIU/mL) decreased CK activity in a xanthine concentration–dependent manner. This inhibition was prevented by the addition of both SOD (400 IU/mL) and CAT (100 IU/mL) as well as by the SH group reducing agent DTT (1 mmol/L).
Mg-ATPase activity of control cells was 32.8±1.6 mIU/mg protein (n=8) and Ca-ATPase activity was 97.7±4.9 mIU/mg protein (n=8). Myosin ATPase activities either in the absence or in the presence of Ca2+ were unaffected by 25 minutes of incubation with X/XO or H2O2 at 22°C (Fig 8B⇑), thus confirming the conclusions obtained from the mechanical experiments.
It is largely recognized that ROS may affect myocardial function. Numerous possible sites of action have been reported, and among them myofibrillar proteins have been involved. The results of the present study show the following: (1) The decrease in maximal force previously reported in experiments using the X/XO system could be attributed to a high contaminating proteolytic activity of commercial XO preparations. (2) The effects of ROS on rigor tension development, Ca2+ sensitivity, and rate constant of tension recovery were dependent on substrate and were not observed in the absence of PCr. (3) The main effect of ROS was a strong inhibition of myofibrillar CK without alterations of myosin ATPase. (4) The effects of H2O2 and XO were similar. (5) The effects were largely prevented by DTT. These results suggest that the main effect of ROS in myofibrils is the oxidation of the essential SH groups of CK, inducing an inactivation of the enzyme and a decrease in the local ATP-to-ADP ratio.
ROS and Tmax
In the present study, we attributed the fall in Tmax, observed after application of XO on maximally Ca2+-activated fibers, to the contaminating proteolytic activity of commercial XO preparations. This conclusion was drawn after several observations: (1) The fall in maximal Ca2+-activated tension could be observed in the absence of substrate for XO and could not be prevented by high activities of both CAT and SOD. (2) These effects were very similar to those of trypsin, a well-known proteolytic enzyme. (3) They could be antagonized by PMSF, an inhibitor of proteases that did not interfere with the production of superoxide by the X/XO system. In addition, commercial preparations of XO do contain proteolytic activity (eg, see Sigma catalog 1995, p 1052).
Mammalian XO exists originally as a dehydrogenase form and is converted to an oxidase form either irreversibly by proteolysis or reversibly by SH oxidation of the protein molecule.25 The pretreatment of XO with PMSF, which prevented tension decrease, did not interfere with the production of O2−•. We could thus clearly conclude that the decrease in maximal force induced by crude XO and described in skinned cardiac fibers (References 17 and 1817 18 and the present study) appeared to be the consequence of the contaminating proteolytic activity of commercial XO preparations.
Inhibition of CK Is Responsible for the ROS-Induced Alterations of Myofibrillar Function
Incubation with either X/XO (100 μmol/L per 30 mIU/mL) or 2.5 mmol/L H2O2 induced several alterations of myofibrillar function, ie, an increase in resting tension at low ATP, an increase in Ca2+ sensitivity, a decrease in Hill coefficient, and a slight augmentation of maximal force. These effects appeared to be directly related to ROS, since they could be antagonized by CAT and SOD, enzymes catalyzing the degradation of the ROS. We found that these alterations in myofibrillar function were highly dependent on the pattern of energy supply. Withdrawal of PCr induced an increase in Ca2+ sensitivity as well as a slight rise in resting tension.28 However, incubation with the ROS did not produce a further change in Ca2+ sensitivity. By contrast, when ATP was generated through the phosphorylation of ADP by myofibrillar-bound CK and PCr, strong increases in Ca2+ sensitivity and in resting tension were observed. Likewise, the dependence of rigor tension development as a function of MgATP was not altered by either X/XO or H2O2 in the absence of PCr but was greatly altered in its presence. The same held true for the time constant of tension changes. Increases in Ca2+ sensibility and in the time constant of tension changes and the shift of the pMgATP-tension dependence to higher MgATP concentrations are reminiscent of the effects brought about by a PCr drop.28
Ca2+ sensitivity of tension development may be influenced by a number of factors. Among them, the decreased rate of crossbridge cycling due to substrate deficiency may induce an increase in resting tension, an increase in Tmax and stiffness, and a shift toward lower Ca2+ concentrations of the pCa-tension relationship together with a decrease in cooperativity.28 30 This effect can be observed when comparing the pCa-tension relationships obtained with or without PCr (Fig 6⇑ and Reference 2828 ) and can be explained by the efficacy of bound CK to maintain a high local ATP-to-ADP ratio. Under conditions of restricted exchange of adenine nucleotides between the cytosol and the myofibrillar compartment, bound CK can provide MgATP, remove MgADP, and buffer proton production in the immediate environment of myosin ATPase (for review, see Reference 2626 ). Inhibition of the CK function leads to a slowing of the off rate of crossbridges and a prolonged duration of the duty cycle of crossbridges.
These observations give evidence that the myofibrillar function is not altered directly through modifications of the regulatory proteins or myosin ATPase but indirectly through an inhibition of myofibrillar CK–inducing substrate deprivation of myosin ATPase and secondary slowing down of crossbridge kinetics. This was confirmed by the measurements of myofibrillar CK and myosin ATPase activities in our experimental conditions. Although incubation with both H2O2 and X/XO was able to dramatically decrease myofibrillar CK activity, neither myosin Mg-ATPase nor Ca2+-activated ATPase was affected.
Sensitivity of Myosin ATPase and Myofibrillar CK to ROS
In the present experimental conditions, we did not obtain evidence of alterations of myosin ATPase or regulatory proteins with the ROS, whereas myofibrillar CK activity was greatly impaired. This is in contrast with the fact that both enzymes have been shown to be sensitive to ROS13 14 15 and that many SH groups are present in myofibrils. The fact that no detectable alteration in myofilament properties could be observed does not exclude the occurrence of contractile protein alterations without clear functional counterpart. In this respect, the increase in resting tension observed after exposure to ROS in high Ca2+ (Reference 1818 and Fig 1⇑ in the present study) could be explained in such a way. On the other hand, a lower sensitivity of myosin ATPase compared with CK to ROS could be involved. Indeed, Kaneko et al15 reported a complete inhibition of myofibrillar CK by X/XO within a few minutes at 37°C with 0.5 mmol/L xanthine and 7.5 mIU/mL XO, whereas others13 showed that a maximum of 60% inhibition of myosin ATPase required 30 minutes at 30°C with 30 mIU/mL XO and 2 mmol/L xanthine. Indeed, prolonged exposure to ROS might reveal myosin ATPase alterations, and the observations of rigor tension finally exceeding Tmax after prolonged exposure to HOCl18 may be explained by the combined inhibition of both CK and Ca2+-activated myosin ATPase.
CK System in Cardiac Tissues
CK is an important cellular enzyme involved in energy transduction in muscle cells. In addition to the traditional role of CK as a spatiotemporal ATP/ADP buffer, it has been proposed that the location of the CK isoenzymes in cardiac muscle facilitates the transduction of high-energy phosphates throughout the cell and acts to fine-tune the regulation of energy utilization and production.33 34 In myofibrils, CK is an integral part of the M line (for review, see Reference 3535 ) and is also distributed across the entire filament.36 There is a high activity of myofibrillar CK (1 to 2 IU/mg protein). This myofibrillar-bound CK can induce a specific kinetic enhancement of the myofibrillar ATPase37 and rephosphorylate enough MgADP produced by the myosin ATPase reaction to ensure optimal contractile capacities as well as normal Ca2+ sensitivity and crossbridge cycling rate in the absence of MgATP and thus can meet the contractile protein energy requirements.26 28 35 On the other hand, deprivation of PCr, even in the presence of ATP, decreases the rate of cycling of crossbridges, increases the Ca2+ sensitivity, and increases the susceptibility of rigor tension development to a decrease in MgATP. We have recently shown that the inhibition of myofibrillar function following the sharp decrease in PCr seen during ischemia or hypoxia seems to be responsible for rigor tension development, which is the basis of the ischemic contracture.27
The CK isoenzymes contain easily oxidized SH groups in the active site, which are required for the enzymatic activity.38 Inhibition of mitochondrial as well as cytosolic forms of CK by the ROS has been described previously.13 15 39 40 Thus, alteration of myofibrillar CK by the ROS may be one of the mechanisms involved in the paradoxical reduction of contractile activity at reperfusion despite complete and fast recovery of PCr content.41 Indeed, altered myofibrillar CK activity was shown upon reperfusion in open-chest dogs subjected to 15 minutes of left anterior descending coronary artery occlusion.42 The failure to observe altered myofibrillar function and myofibrillar CK in isolated rat hearts after ischemia and reperfusion43 44 could be due either to low ROS generation in isolated heart preparations or to the presence of radical scavengers or reducing agents in these in vitro experiments.
ROS and Myofibrillar Function
In our experiments, the X/XO system and H2O2 produced similar effects on myofibrillar properties and CK. The production of free radicals has been measured in both conditions, in the relaxing solution and in the absence of skinned fibers. EPR spectra confirmed the generation of superoxide radicals by the X/XO system45 but not hydroxyl radicals. H2O2 did not generate any significant hydroxyl radical or superoxide, as previously reported.46 In fact, H2O2 can be reduced to hydroxyl radicals in the presence of iron, which facilitates the Fenton reaction. In the presence of fibers, it could be possible that endogenous iron catalyzes the reaction and that hydroxyl radicals are produced with the two free radical–generating systems. However, since desferrioxamine, an iron chelator, was without effect, it is reasonable to consider that hydroxyl radicals were not importantly involved and that this was merely due to the presence in the solutions of EGTA, another metal chelator. In experiments performed in intact papillary muscles, Schrier and Hess10 reported that the main negative inotropic radical species produced by X/XO was the superoxide anion and that H2O2 itself was not involved. H2O2 toxicity could be due to direct effects,46 but the formation of derived-toxic substances could also be involved.
The exact mode of action of ROS is not well understood. Inhibitory effects of X/XO, HOCl, and H2O2 on myosin ATPase and CK were shown to be prevented by thiol-reducing agents such as DTT,13 14 15 suggesting that oxidation of SH groups was the main effect of ROS. DTT by itself does not decrease the high sensitivity of myofibrils to ATP in the presence of PCr, since pMgATP50 values are identical with or without DTT (values of 5.07 in Reference 27 and 527 5 .03 in the present study). In our hands, DTT did not completely reverse the effects of ROS on rigor tension development and was without effect on HOCl-induced mechanical alteration.18 These observations are not entirely consistent with a simple oxidation of thiol groups and require further analysis.
Relevance to In Vivo Conditions
The high effective concentrations of ROS in the present study require further discussion. In the present study, mechanical modifications were observed only for relatively high ROS concentrations (30 mIU/mL XO or 2.5 mmol/L H2O2), whereas lower concentrations of H2O2 (100 μmol/L) or XO/X [(3 mIU/mL)/(30 μmol/L)] were without noticeable effects. In the study of MacFarlane and Miller,18 prolonged incubations with H2O2 concentrations up to 10 mmol/L only marginally increased resting tension, Tmax, and Ca2+ sensitivity. On the other hand, Banerjee et al41 and Suzuki et al14 observed an inhibition of CK by 100 μmol/L H2O2. Limited intracellular diffusion of these compounds seems to be excluded. H2O2 is a small molecule, and fibers skinned with Triton X-100 are permeable to ions, substrates, and enzymes. Moreover, steady state effects were analyzed after prolonged times of incubation. Similarly, superoxide production with XO should take place both inside and outside the fibers. It is highly possible that some compounds present in the solutions used for skinned fiber experiments may have a protective effect against H2O2 or other ROS action, thus explaining the discrepancies in the dose-response relationship of myofibrillar CK to H2O2. In addition, temperature differences can also be involved. Our experiments were performed at 22°C, a temperature that is known to increase the stability of the preparations and that allows comparisons with other studies. Increasing the temperature to more physiological values will undoubtedly potentiate the ROS production and/or effects. Indeed, Banerjee et al reported a 50% inhibition of CK activity for 30 μmol/L H2O2 after incubation for 15 minutes at 37°C. Similarly, the higher sensitivity of myosin ATPase and CK to ROS observed by Suzuki et al13 and Kaneko et al15 may well be explained by the higher temperature used in these studies. In this respect, an ED50 of 30 μmol/L H2O2 for CK is in the plausible range of physiologically relevant H2O2 concentrations at reperfusion.41 A similar involvement of temperature may also apply to the X/XO system. Changes in a variety of oxidative indexes in isolated perfused rat heart tissue subjected to direct infusion of 10 mIU/mL XO for 10 minutes at 36°C were shown to be similar in magnitude to those induced by as little as 10 minutes of hypoxia.47 Although comparisons are difficult to make, in the conditions of the present study, the intensity of free radical generation with 30 IU/mL XO and 100 μmol/L xanthine was only two to three times higher than that observed after reperfusion of an isolated ischemic heart.4 Moreover, it should be kept in mind that in skinned fiber experiments, because of the presence of EGTA, hydroxyl radicals (the most reactive of the ROS) were not generated. However, Khalid and Ashraf48 recently provided evidence that OH· formation by isolated cells during cycles of anoxia/reoxygenation was not the major ROS inducing cell damage and enzyme leakage. Other in vivo sources of free radicals and ROS may also be involved, such as the very reactive hypochlorous acid that was shown to induce similar effects on skinned fibers at very low concentrations.18 Taken together, these observations suggest that inhibition of myofibrillar CK may occur in vivo under conditions of ischemia/reoxygenation.
ROS Inhibition of CK and Cardiac Pathophysiology
A number of studies indicate a possible role of ROS in the prolonged depression of the mechanical function following a brief period of ischemia, known as myocardial stunning.49 After restoration of the coronary flow, the cellular ATP pool remains depleted for several days, whereas PCr content often quickly returns back to preischemic levels or even higher. During myocardial ischemia, a series of metabolic events promotes the univalent reduction of molecular oxygen and initiates free radical chain reactions. During ischemia or hypoxia and reoxygenation, involvement of free radical generation has been documented. Formation of superoxide radicals and H2O2 has been detected during recirculation.4 41 The proteolytic conversion of xanthine dehydrogenase to XO during Ca2+ overload by the activation of Ca2+-sensitive proteases has been proposed as one of the possible sources of free radicals (see Reference 11 for review). However, even though XO activity is high in rat myocardium, it appears to be absent in human myocardium.50 The involvement of XO in reperfusion injury is an attractive hypothesis, but it is still a matter of debate.25
It is well documented that reperfusion after moderate ischemia is accompanied by a rapid recovery of PCr to supranormal levels (PCr overshoot), whereas contractile activity recovers slowly, suggesting an impairment in PCr utilization.41 51 52 53 During prolonged mild to moderate ischemia in anesthetized pigs, the return of PCr levels to control values after the first minutes of initial decrease, despite ATP concentrations of <50% of control values, suggested a limitation of the cells’ ability to produce ATP by use of PCr.54 The decrease in total CK activity at reperfusion has been correlated with the generation of H2O2.41 Thus, paradoxically, elevated levels of PCr and depressed myocardial function during myocardial stunning were suggested to result from specific inhibition of cytosolic and myofibrillar CK by ROS.40 41 The present results are also consistent with this interpretation. However, apart from the fact that myocardial stunning is of multifactorial origin, this explanation requires that mitochondrial CK not be inhibited by free radicals, an assumption that is difficult to reconcile with electron flow through the respiratory chain being an important source of oxygen radicals.55 In addition, reductions in both cytosolic and mitochondrial forms of CK activity have been reported to occur in hearts subjected to ischemia and reperfusion.40 42 56 Such inhibition may induce serious alterations of cardiac function. Recently, using iodoacetamide to inhibit CK in isolated heart, Hamman et al57 concluded that CK activity was essential for the expression of the full dynamic range of myocardial performance and that myofibrillar CK inhibition could be the cause of altered kinetics of the myofibrillar ATPase and could contribute to a substantial reduction of contractile reserve.
Decreased bound as well as cytosolic CK activities induced by the ROS after ischemia and reperfusion could be of importance for the regulation of energy fluxes and the integration of utilization and production of high-energy phosphate compounds in the myocardium. The present study brings evidence that myofibrillar CK is the first and main target of the ROS in the myofibrillar compartment and that alteration of PCr utilization at the myofibrillar side of the PCr shuttle may participate in the damaging effects of ROS in cardiac cells. This effect may be partly responsible for the poor and slow recovery of mechanical function after ischemia and reperfusion.
Selected Abbreviations and Acronyms
|EPR||=||electron paramagnetic resonance|
|pCa50||=||pCa for half-maximal activation|
|pMgATP50||=||pMgATP for half-maximal activation|
|ROS||=||reactive oxygen species|
This study was supported in part by grants from the Association Française contre les Myopathies, Fondation pour la Recherche Médicale, and Fondation de France. Dr Ventura-Clapier was supported by the Centre National de la Recherche Scientifique. The authors wish to acknowledge F. Lefèbvre for skillful technical assistance, P. Lechêne for engineering assistance, and R. Fischmeister for continuous support.
- Received October 30, 1995.
- Accepted March 5, 1996.
- © 1996 American Heart Association, Inc.
Bolli R, Patel BS, Zhu W, O’Neil PG, Hartley CJ, Charlat ML, Roberts R. Demonstration of free radical generation in ‘stunned’ myocardium of intact dogs with the use of spin trap alphaphenyl N-tert-butyl nitrone. J Clin Invest. 1988;82:476-485.
Zweier JL. Measurement of superoxide-derived free radicals in the reperfused heart. J Biol Chem. 1988;263:1353-1357.
Myers ML, Bolli R, Lekich RF, Hartley CJ, Roberts R. Enhancement of recovery of myocardial function by oxygen radical scavengers after reversible regional ischemia. Circulation. 1985;72:915-921.
Przyklenk K, Kloner RA. Superoxide dismutase plus catalase improve contractile function in the canine model of the ‘stunned’ myocardium. Circ Res. 1986;58:148-156.
Bolli R, Patel BS, Zhu W, O’Neil PG, Charlat ML, Roberts R. The iron chelator desferrioxamine attenuates postischemic ventricular dysfunction. Am J Physiol. 1987;253:H1372-H1380.
Schrier GM, Hess ML. Quantitative identification of superoxide anion as a negative inotropic species. Am J Physiol. 1988;255:H138-H143.
Burton KP, Jones JG, Le TH, Sherry AD, Malloy CR. Effects of oxidant exposure on substrate utilization and high-energy phosphates in isolated rat hearts. Circ Res. 1994;75:97-104.
MacFarlane NG, Miller DJ. Depression of peak force without altering calcium sensitivity by the superoxide anion in chemically skinned cardiac muscle of rat. Circ Res. 1992;70:1217-1224.
Mayoux E, Coutry N, Lechene P, Marotte F, Hoffmann C, Ventura-Clapier R. Effects of acidosis and alkalosis on mechanical properties of hypertrophied rat heart fiber bundles. Am J Physiol. 1994;266:H2051-H2060.
Hoerter JA, Kuznetsov A, Ventura-Clapier R. Functional development of the creatine kinase system in perinatal rabbit heart. Circ Res. 1991;69:665-676.
Pucéat M, Clement O, Lechene P, Pelosin JM, Ventura-Clapier R, Vassort G. Neurohormonal control of calcium sensitivity of myofilaments in rat single heart cells. Circ Res. 1990;67:517-524.
Lowry OH, Rosenbrough NJ, Farr AL, Randall RJ. Protein measurement with the folin phenol reagent. J Biol Chem. 1951;193:265-275.
Nishino T. The conversion of xanthine dehydrogenase to xanthine oxidase and the role of the enzyme in reperfusion injury. J Biochem. 1994;116:1-6.
Ventura-Clapier R, Veksler V, Hoerter JA. Myofibrillar creatine kinase and cardiac contraction. Mol Cell Biochem. 1994;133/134:125-144.
Ventura-Clapier R, Veksler V. Myocardial ischemic contracture: metabolites affect rigor tension development and stiffness. Circ Res. 1994;74:920-929.
Ventura-Clapier R, Mekhfi H, Vassort G. Role of creatine kinase in force development in chemically skinned rat cardiac muscle. J Gen Physiol. 1987;89:815-837.
Brandt PW, Cox RN, Kawai M, Robinson T. Regulation of tension in skinned muscle fibers: effect of crossbridge kinetics on apparent Ca2+ sensitivity. J Gen Physiol. 1982;79:997-1016.
Güth K, Potter JD. Effect of rigor and cycling crossbridges on the structure of troponin C and on the Ca2+ affinity of the Ca2+ specific regulatory sites in skinned rabbit psoas fibers. J Biol Chem. 1987;262:13627-13635.
Saks VA, Khuchua ZA, Vasilyeva EV, Belikova OY, Kuznetsov AV. Metabolic compartmentation and substrate channeling in muscle cells: role of coupled creatine kinases in in vivo regulation of cellular respiration: a synthesis. Mol Cell Biochem. 1994;133:155-192.
Wallimann T, Wyss M, Brdiczka D, Nicolay K, Eppenberger HM. Intracellular compartmentation, structure and function of creatine kinase isoenzymes in tissues with high and fluctuating energy demands: the phosphocreatine circuit for cellular energy homeostasis. Biochem J. 1992;281:21-40.
Wallimann T, Eppenberger HM. Localization and function of M-line-bound creatine kinase M-band model and creatine phosphate shuttle. In: Shay JW, ed. Cell and Muscle Motility. New York, NY: Plenum Publishing Corp; 1985;6:239-285.
Krause SM, Jacobus WE. Specific enhancement of the cardiac myofibrillar ATPase by bound creatine kinase. J Biol Chem. 1992;267:2480-2486.
Watts DC. Creatine kinase (adenosine 5′-triphosphate-creatine phosphotransferase). In: Boyer PD, ed. The Enzyme. New York, NY: Academic Press; 1973;8:383-455.
McCord JM, Turrens JF. Mitochondrial injury by ischemia and reperfusion. Curr Top Bioenerg. 1994;17:173-195.
Banerjee A, Grosso MA, Brown JM, Rogers KB, Whitman GJR. Oxygen metabolite effects on creatine kinase and cardiac energetics after reperfusion. Am J Physiol. 1991;261:H590-H597.
Greenfield RA, Swain JL. Disruption of myofibrillar energy use: dual mechanisms that may contribute to postischemic dysfunction in stunned myocardium. Circ Res. 1987;60:283-289.
Krause SM. Effect of global myocardial stunning on Ca2+-sensitive myofibrillar ATPase activity and creatine kinase kinetics. Am J Physiol. 1990;259:H813-H819.
Kuppusamy P, Zweier JL. Characterization of free radical generation by xanthine oxidase. J Biol Chem. 1989;264:9880-9884.
Josephson RA, Silverman HS, Lakatta EG, Stern MD, Zweier JL. Study of the mechanisms of hydrogen peroxide and hydroxyl free radical-induced cellular injury and calcium overload in cardiac myocytes. J Biol Chem. 1991;266:2354-2361.
Park Y, Kanekal S, Kehrer JP. Oxidative changes in hypoxic rat heart tissue. Am J Physiol. 1991;260:H1395-H1405.
Khalid MA, Ashraf M. Direct detection of endogenous hydroxyl radical production in cultured adult cardiomyocytes during anoxia and reoxygenation: is the hydroxyl radical really the most damaging radical species? Circ Res. 1993;72:725-736.
Braunwald E, Kloner RA. The stunned myocardium: prolonged, postischemic ventricular dysfunction. Circulation. 1982;66:1146-1149.
Eddy LJ, Stewart JR, Jones HP, Engerson TD, McCord JM, Downey JM. Free radical-producing enzyme, xanthine oxidase, is undetectable in human hearts. Am J Physiol. 1987;253:H709-H711.
Neubauer S, Hamman BR, Perry SB, Bittl JA, Ingwall JS. Velocity of the creatine kinase reaction decreases in post ischemic myocardium: a 31P-NMR magnetization transfer study of the isolated ferret heart. Circ Res. 1988;63:1-15.
Schaefer S, Carr LJ, Kreutzer U, Jue T. Myocardial adaptation during acute hibernation: mechanisms of phosphocreatine recovery. Cardiovasc Res. 1993;27:2044-2051.
Pantely GA, Malone SA, Rhen WS, Anselone CG, Arai A, Bristow J, Bristow JD. Regeneration of myocardial phosphocreatine in pigs despite continued moderate ischemia. Circ Res. 1990;67:1481-1493.
Ambrosio G, Zweier JL, Duilio C, Kuppusamy P, Santoro G, Elia PP, Tritto I, Cirillo P, Condorelli M, Chiariello M, Flaherty JT. Evidence that mitochondrial respiration is a source of potentially toxic oxygen free radicals in intact rabbit heart subjected to ischemia and reflow. J Biol Chem. 1993;268:18532-18541.
Bittl JA, Weisfeld ML, Jacobus WE. Creatine kinase of heart mitochondria: the progressive loss of enzyme activity during in vivo ischemia and its correlation to depressed myocardial function. J Biol Chem. 1985;260:208-214.
Hamman BL, Bittl JA, Jacobus WE, Allen PD, Spencer RS, Tian R, Ingwall JS. Inhibition of the creatine kinase reaction decreases the contractile reserve of isolated rat hearts. Am J Physiol. 1995;269:H1030-H1036.