INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

IT HAS BEEN PROPOSED THAT fatigue may be associated with changes in the functional aspects of the contractile apparatus and the sarcoplasmic reticulum (SR) (1, 13, 51-53). However, the mechanism underlying fatigue still remains unclear. Reactive oxygen species (ROS), which would be produced by strenuous exercise (18, 35, 41, 46), may act as a factor in muscle fatigue (1, 13, 51). Exposure of skeletal muscle to xanthine oxidase increases the fatigue rate by generating ROS, whereas antioxidants decrease it (2, 41, 43). A reductant, dithiothreitol, also improves recovery from diaphragm fatigue (12). Fatigue and its recovery, therefore, may be attributable to the redox state of proteins involved in excitation-contraction coupling and contractile proteins.

Glutathione (GSH) and GSH disulfide (GSSG) function as a major redox buffer system within many cells. The ratio of the concentration of GSH to GSSG ([GSH]/[GSSG]) is known to be decreased depending on exercise intensity (22, 45), resulting in a shift of the redox potential toward oxidative states. Hydrogen peroxide (H₂O₂) is a weak oxidant for SR proteins (32). Previous reports indicated that H₂O₂ at millimolar concentrations elicits the release of Ca²⁺ from the SR and increases channel activity of the ryanodine receptor (RyR) in bilayers (6, 14, 29, 31, 32, 39, 47). On the other hand, Ca²⁺ uptake by the SR still remains depressed with a 60-min rest after intermittent exercise (48, 50). These observations may allow us to assume that H₂O₂ oxidizes critical thiols on the RyR and Ca²⁺-ATPase molecules to promote Ca²⁺ release and to inhibit Ca²⁺ uptake. If so, in turn, this would result in a decrease in releasable SR Ca²⁺ content. However, it is questionable whether H₂O₂ at a physiological concentration (below several µM; see Ref. 25) can function as a channel agonist, because a large concentration of H₂O₂ is required for channel activation (2, 29), and low H₂O₂ concentrations (0.1 to ~10 µM) decrease cytoplasmic Ca²⁺ concentration during tetanic stimulation, although not in a dose-dependent manner (3). This issue, therefore, must be reexamined by exposing single fibers to small amounts of H₂O₂ and by exploring RyR channel responsiveness with exposure to near physiological concentrations of H₂O₂.

Reportedly, the RyR molecule is a redox sensor with a well-defined redox potential (15, 54). Our recent finding demonstrates that activation of the RyR1 channel elicited by Ca²⁺ or adenine nucleotides depends on cytoplasmic redox potential (33). If redox potential would alter the responsiveness of the RyR channel to H₂O₂, physiological amounts of H₂O₂ might act as an intense releaser of Ca²⁺ under the resting redox state and then lead to a decrease in releasable Ca²⁺ content in the SR. We, therefore, focused on the effect of H₂O₂ on RyR channel activity under the redox control.

The present study indicates that rabbit RyR1 and frog RyR channel activities were significantly enhanced by exposure to 10 µM H₂O₂ under a specific condition, where intracellular redox potential was defined at the resting state. Exposure of muscle fibers to fatiguing stimulation led to a marked decrease in the releasable Ca²⁺ content after a 60-min recovery from fatigue. This decrease, however, was not significantly enhanced by application of 10 µM H₂O₂. The results suggest that externally applied H₂O₂ may not significantly disturb the Ca²⁺ homeostasis during postfatigue slow recovery in vivo, although it markedly sensitized the RyR channel in vitro.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Single muscle fiber and RyR preparations. The institution's guidelines for the care and use of laboratory animals were followed. Toe muscle (musculus flexor brevis digiti) was dissected in ice-cold Ringer solution (in mM: 115 NaCl, 2.5 KCl, 1.8 CaCl₂, and 3 sodium phosphate buffer, pH 7.0) from bullfrogs (Rana catesbeiana). The solution was gassed with 95% O₂ and 5% CO₂. Single fibers were isolated in Ringer solution and used for tension experiments and histological observation. H₂O₂ was added to the bathing medium 10 min before fatiguing stimulation and was left in the medium throughout the experimental period.

Planar lipid bilayer experiments. The dose-dependent effect of 1 µM to 1 mM H₂O₂ on RyR channel activity under redox potential control was examined by incorporating purified RyR1 channels or frog SR vesicles into lipid bilayers, as previously reported (27, 29). Lipid bilayers consisting of a mixture of L--phosphatidylethanolamine, L--phosphatidyl-L-serine and L--phosphatidylcholine (5:3:2 wt/wt/wt) in n-decane (40 mg/ml) were formed across a hole of ~250 µm in diameter in a polystyrene partition separating cis and trans chambers. The cis (1 ml)/trans (1.5 ml) solutions consisted of (in mM) 500/50 KCl, 20 HEPES-Tris (pH 7.4), and 0.1 CaCl₂ for RyR1 channels and of 250/50 CsCH₃SO₃, 10 CsOH (pH 7.4 by HEPES), and 0.1 CaCl₂ for frog SR vesicles. After the channel incorporation by the occurrence of flickering currents was confirmed, single-channel current was recorded in symmetrical solutions prepared by adding an aliquot of 3 M KCl (pH 7.4 by HEPES-Tris) for RyR1 channels and 2.2 M CsOH (pH 7.4 by HEPES) for SR vesicles to the trans compartment (intraluminal side of the SR). The trans side was held at ground potential, and the cis side was voltage clamped at 40 mV by using 1.5% agar bridges in 3 M KCl and Ag-AgCl electrodes. Then redox potential in cis and trans solutions was defined at 220 and 180 mV, respectively. In such solutions, RyR channel activity was measured on cumulative application of 1 µM to 1 mM H₂O₂ to the cis compartment. Channel activity ascribed to the RyR was confirmed by the responses to 10 µM ryanodine and 5 µM ruthenium red at the end of every experiment. In addition, we designed some experiments to establish whether a redox potential gradient across the membrane is required to enhance RyR1 sensitivity to H₂O₂ or whether just a reduced environment on the cis side is the requirement for channel activation. Redox potential in the solution was calculated from the Nernst equation by using the standard redox potential = 240 mV (17). Redox potentials were generated by different ratios of [GSH]/[GSSG] (mM/mM) of 2:0.469 for 180 mV and of 2:0.0196 for 220 mV. To evaluate the effect of reduction in cis redox potential from 220 to 231 mV, 1.096 mM GSH was further added to cis solution. Experiments were carried out at room temperature (18~22°C).

Isometric contraction measurement and experimental protocol. One tendon of the fiber was clamped to a chamber in Ringer solution, and the other tendon was led to an isometric transducer. After twitch and maximum tetanus tensions (50 Hz for 1 s, 0.05-ms pulse duration) were checked in oxygenated Ringer solution, the fiber was exposed to 0, 10, or 100 µM H₂O₂. Ten minutes later, twitch and tetanus tensions were recorded again to elucidate the effects of H₂O₂ on contraction. Then the fiber was subjected to a fatigue protocol in which contractions were caused for 2 min by electrical stimulation at a frequency of 50 Hz (0.05-ms pulse duration) on a duty cycle of 200 ms every second (SEN-3301, Nihon-Koden, Tokyo). The effect of H₂O₂ on muscle fatigue was estimated by dividing the tension observed just before the end of the fatigue protocol by that found after the first stimulation for fatigue. One, 5, 10, 30, and 60 min after the end of the fatigue protocol, tetanus tension was elicited to elucidate the effects of H₂O₂ on the recovery rate from fatigue. In some experiments, catalase (1,000 U/ml) was added 1 min before the onset of fatigue. As controls, tetanus tensions in 100 µM H₂O₂ were measured in unfatigued fibers. Experiments were done at room temperature.

Estimation of releasable Ca²+ content in the SR. To determine whether application of H₂O₂ to fatigued fibers leads to a decrease in releasable Ca²⁺ in the SR, 25 mM caffeine or 5 µM 4-chloro-m-cresol (4-CmC) was given to 10 or 100 µM H₂O₂-treated fatigued fibers immediately after a 60-min rest. Ca²⁺ content of the SR was estimated from the amplitude of caffeine or 4-CmC-induced contractures. Caffeine and 4-CmC have been used as useful tools to estimate the rapidly releasable Ca²⁺ content in the SR (21, 49).

Histological procedures. Whether a 2-min fatigue protocol after exposure to 100 µM H₂O₂ for 10 min produces gross muscle damage was studied by cytoplasmic fluorescence observation of procion orange and ultrastructural analysis. Procion orange (0.15% wt/vol) was added to Ringer solution to identify sarcolemmal injury of the fiber immediately after the end of the fatigue protocol. After 30 min of staining, the fiber was observed at ×100 magnification and photographed under a fluorescence microscope (Olympus Fluorescent Microscope, BH2-RFCA). To check ultrastructures of subcellular membranes and contractile apparatus, some fibers used for fatigue experiments were rinsed in Ringer solution after 60 min of recovery from fatigue in the presence or absence of H₂O₂. Then the fiber was fixed in 2.5% glutaraldehyde for 10 min at room temperature and postfixed in 1% OsO₄ in 100 mM phosphate buffer, pH 7.0, on ice for 10 min. After dehydration in a series of ethanol and 100% acetone, samples were embedded in epon, as previously described (30). They were sectioned with an LKB Ultratome V with a diamond knife and stained with 2% uranyl acetate solution and then with lead citrate. Sections were observed with a Hitachi HU-11DS electron microscope, and photographs were taken on Kodak electron image film.

Statistical analysis. The results are presented as means ± SE. Statistical analysis was done with one-way ANOVA followed by Fisher's least-significant difference method or Student's t-test. Values of P < 0.05 were regarded as statistically significant.

Chemicals. H₂O₂ (31% stock solution; Mitsubishi Gas Chemical, Tokyo, Japan), caffeine (0.5 M stock solution; Sigma Chemical, St. Louis, MO), GSH (250 mM stock solution; Sigma Chemical) and GSSG (50 mM stock solution; Sigma Chemical) were prepared in ultrapure water (Barnstead, Boston, MA) immediately before application. Catalase (EC 1.11.1.6, 50,000 U/ml stock solution; Sigma Chemical) and ryanodine (1 mM stock solution; Wako Pure Chemical, Osaka, Japan) were dissolved in ultrapure water and ethanol, respectively, and stored at 20°C. Ruthenium red (1 mM stock solution; Sigma Chemical) was dissolved in ultrapure water and stored at 0°C. 4-CmC (1 M stock solution; Wako Pure Chemical) was dissolved in dimethylsulfoxide (Wako Pure Chemical). Procion orange (reactive orange 14, 3% stock solution; Sigma Chemical) was dissolved in Ringer solution and stored in a dark room. Other reagents were of analytical grade.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Effects of H₂O₂ on channel activity of the RyR under redox potential control. Our previous observations demonstrated that redox states of critical sulfhydryls located on the cytoplasmic side of the RyR1 alter responsiveness to channel modulators, such as Ca²⁺, adenine nucleotides, and caffeine (33). RyR1 channels belonged to several distinct populations with different P_o at pCa 4, as previously shown (33). For the present study, we used only RyR1 channels with a P_o of 0.05 at pCa 4 (P_o = 0.232 ± 0.063, n = 6). Frog skeletal muscle expresses two isoforms of the RyR ( and ) (37) with distinct Ca²⁺ dependencies of single-channel activity in bilayers (32). In this experiment, we used the RyR (n = 5) that displayed a bell-shaped curve of P_o against cis Ca²⁺ concentration, similar to that observed in rabbit RyR1 channels. To observe channel activation better, P_o was decreased by increasing cis pCa from 4 to 6-6.3 (P_o = 0.030 ± 0.012 in the RyR1, and P_o = 0.083 ± 0.016 in the frog RyR). Definition of the redox potentials in cis and trans solutions at 220 and 180 mV, respectively, hardly altered the channel activity (0.021 ± 0.006 in the RyR1 and 0.081 ± 0.021 in the frog RyR). As is typically shown in Fig. 1, application of 10 µM H₂O₂ to the cis side of the RyR1 channel under redox control increased P_o 2.5-fold from 0.027 (control) to 0.067. Increase in H₂O₂ to 100 µM markedly enhanced the channel activity to a P_o of 0.219. Numbers of open events increased to 81.3/s in 10 µM H₂O₂ from 15.7/s without H₂O₂, but mean open time remained unchanged (1.21 to 1.13 ms). The open time distribution was best fit by the sum of two exponentials before and after application of 10 µM H₂O₂: _O1 = 0.31 ms (90.2% in relative area) and _O2 = 1.45 ms (9.8%) before H₂O₂, and _O1 = 0.41 ms (92.0%) and _O2 = 1.46 ms (8.0%) after H₂O₂ exposure. Mean closed time decreased to 7.76 from 26.3 ms. The closed time distribution was best fit by the sum of three exponentials: _C1, _C2, and _C3 were altered from 0.94 ms (37.6%), 2.96 ms (33.5%), and 14.45 ms (28.9%) to 0.86 ms (72.7%), 4.63 ms (24.0%), and 30.08 ms (3.3%) after treatment with 10 µM H₂O₂, respectively. Increase in H₂O₂ to 100 µM or 1 mM did not produce further alteration of gating parameters of the open channel, except for the increase in the number of open events, compared with those in 10 µM H₂O₂. As we expected from our laboratory's previous observations (33), reduction of cis redox potentials from 220 to 231 mV, while keeping trans potential at 180 mV, led to a great decrease in P_o due to a decrease in numbers of open events. Similar experiments were repeated with six separate channels, and the results are summarized in Fig. 2. The minimum concentration of H₂O₂ required to increase P_o was between 3 and 10 µM. On the other hand, the redox potential-undefined channels were never stimulated even after exposure to 0.5 mM H₂O₂ (P_o = 0.057 ± 0.020 vs. P_o = 0.045 ± 0.010 in controls; Fig. 2), which is consistent with our laboratory's previous observations (29, 30). Addition of 10 µM H₂O₂ to RyR1 channels, in which only the cis redox potential was set at 220 mV, increased the P_o twofold to 0.041 from 0.02 before H₂O₂ treatment (Table 1). Under this condition, subsequent fixation of the trans potential at 180 mV further activated the channel activity to a P_o of 0.068. Thereafter, reduction of trans potential to 220 mV did not alter the P_o (0.063 ± 0.008). These results indicate that the channel activation by peroxide does not depend on absolute value of redox potential of the trans side, although the channel is stimulated more strongly by both cis and trans potential fixation.

View larger version (62K): [in this window] [in a new window]   Fig. 1.   Effects of hydrogen peroxide (H2O2) on single-channel activity of the rabbit ryanodine receptor (RyR) 1. After the redox potential in cis and trans solution was defined at 220 and 180 mV, respectively, in pCa 6, H2O2 was cumulatively applied to the cis chamber. In presence of 1 mM H2O2, reduction of the redox potential on the cis side to 231 mV produced a marked decrease in open probability (Po). Downward deflection of channel current indicates channel opening. Calibration, 20 pA and 100 ms.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Dose-dependent effects of H2O2 on Po in single RyR1 channel in which redox potential was defined (; n = 6) or not defined (; n = 9) at 220 mV for cis chamber and 180 mV for trans one. Note that channel responsiveness to H2O2 depends on the redox control. Significantly different from corresponding control single-channel activity in channels not treated with H2O2: * P < 0.05 and ** P < 0.01.

Effects of H₂O₂ on muscle fatigue. Exposure of the fiber to 10 µM H₂O₂ for 10 min did not affect the maximum tetanus tension (3.29 ± 0.24 mN of controls to 3.30 ± 0.25 mN, n = 10). Increase in H₂O₂ to 100 µM slightly increased the tension by 6% after a 10-min incubation (3.49 ± 0.36 mN of controls to 3.70 ± 0.25 mN; n = 10). When the fiber not treated with H₂O₂ was subjected to fatiguing stimulation for 2 min, the tetanus tension was greatly decreased by 46%, from 4.46 ± 0.59 mN before fatigue to 2.39 ± 0.43 mN (P < 0.01; n = 8). Addition of 10 or 100 µM H₂O₂ to the external solution did not significantly alter the extent of fatigue, as shown in Fig. 3.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Changes in relative tetanus tension of single muscle fibers subjected to fatiguing stimulation with or without H2O2. , Unfatigued fibers in no H2O2, n = 8; , unfatigued fibers in 100 µM H2O2, n = 5; , fatigued fibers in no H2O2, n = 6; , fatigued fibers in 10 µM H2O2, n = 6; , fatigued fibers in 100 µM H2O2, n = 6. Muscle fiber was stimulated to elicit fatigue for 2 min at a frequency of 50 Hz on a duty cycle of 200 ms every second. Then tetanus tensions (50 Hz for 1 s) were checked to estimate the recovery from fatigue after 1, 5, 10, 30, and 60 min. H2O2 at 10 or 100 µM was given 10 min before application of fatigue protocol. Each tension was calculated as a value relative to the tetanus tension before exposure to H2O2. Note the occurrence of fast and slow recovery after muscle fatigue and inhibition of the slow recovery after application of H2O2. Vertical bars indicate SE. Small bar on the x axis indicates the duration of fatigue stimulation. Significantly different vs. corresponding mean values in unfatigued fibers untreated with H2O2; * P < 0.05 and ** P < 0.01.

Effects of H₂O₂ on postfatigue recovery. Tetanus tension in unfatigued fibers was almost constant throughout the experimental period of 60 min (3.61 ± 0.31 to 3.68 ± 0.43 mN after a 60-min rest; n = 5). The results were summarized in Fig. 3 in relation to the control value for each fiber. A slight but significant decrease in tetanus tensions was observed after further incubation for 10 (P < 0.05) and 30 min (P < 0.01) after a 10-min exposure to 100 µM H₂O₂ compared with tetanus tensions in H₂O₂-untreated fibers.

Caffeine or 4-CmC-induced contracture after 60-min rest in H₂O₂-treated fatigued fibers. It is of interest to know whether a decrease in releasable Ca²⁺ content in the SR is induced by H₂O₂, since results noted above have demonstrated that H₂O₂ at 10 µM significantly activated RyR channels under a defined redox potential and at 100 µM elicited the slowing of relaxation of tetanus after fatigue. Releasable Ca²⁺ content in the SR was estimated from the amplitude of contracture observed by applying 25 mM caffeine or 5 mM 4-CmC immediately after a 60-min rest. The results are summarized in Table 2. Mean amplitude of caffeine contracture in unfatigued fibers not treated with H₂O₂ was 93.5% of control tetanus (group A). Caffeine contracture in unfatigued fibers treated with 10 or 100 µM H₂O₂ (groups C or E) was 7 or 11% smaller (not significant) than that in unfatigued fibers not treated with H₂O₂, respectively. With fatiguing stimulation (group B), caffeine contracture in controls not treated with H₂O₂ was significantly decreased to ~74% of tetanus force (P < 0.05 from group A). The force decrease in 10 or 100 µM H₂O₂-treated, fatigued fibers (groups D and F) was significantly larger (P < 0.05) than in unfatigued fibers to which H₂O₂ had been applied (groups C and E). Fatiguing stimulation in the presence of 100 µM H₂O₂ (group F) caused a significant decrease in caffeine contracture compared with fatigued fibers not treated with H₂O₂ (group B) (P < 0.05). Exposure to 10 µM H₂O₂ decreased caffeine contracture to 0.93 (group C/group A) in unfatigued muscles and to 0.90 (group D/group B) in fatigued muscles. A similar calculation for 100 µM H₂O₂ gave 0.89 (group E/group A) in unfatigued muscles and 0.85 (group F/group B) in fatigued muscles. These differences between unfatigued and fatigued fibers do not seem to be significant. When 4-CmC was used to estimate the Ca²⁺ content in the SR, similar results were observed, although the difference was slightly larger (0.90 vs. 0.85).

Catalase-induced improvement of rate of recovery from muscle fatigue in H₂O₂. As noted, muscle fibers subjected to H₂O₂ failed to show improvement in the slow recovery phase. We used catalase to study whether this effect is produced via the direct effect of externally applied H₂O₂ itself and/or of hydroxyl radicals ( · OH) produced through the Fenton reaction. Catalase (1,000 U/ml) itself did not affect muscle tension during 60-min incubation (2.58 ± 0.26 to 2.53 ± 0.42 mN after a 60-min rest; n = 5) and did not alter the amplitude of fatigue or the fast recovery phase. As shown in Fig. 4, however, muscle contraction recovered to 94% of the control level with catalase after a 60-min rest, comparable to postfatigue recovery observed with muscles not treated with catalase (93%; see Fig. 3). A similar catalase-induced recovery from fatigue was also observed in fibers that had been exposed to 100 µM H₂O₂. These results suggest that catalase ameliorates the effect of extracellularly applied H₂O₂ but does not appear to affect the normal response to fatiguing stimulation.

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Acceleration of the slow recovery from muscle fatigue on exposure to catalase. Catalase (1,000 U/ml) was added 1 min before the onset of fatigue stimulation. , unfatigued fibers in catalase, n = 6; , fatigued fibers in catalase, n = 7; , fatigued fibers treated with catalase 9 min after application of 100 µM H2O2, n = 7. Note that tetanus tensions at 30 and 60 min after fatigue in fibers treated with both catalase and H2O2 do not significantly differ from those in the absence of H2O2. ** Significantly different vs. corresponding mean values in unfatigued fibers treated with catalase (P < 0.01).

Histological observations of muscle membranes and contractile apparatus. Sarcolemmal disruptions after fatiguing stimulation in the presence of 100 µM H₂O₂ were checked by using a fluorescent dye, procion orange. The dye uptake into the myoplasm is a sensitive index of gross sarcolemmal disruption because the intact fiber is impermeable to this dye (7, 16, 38). Figure 5A exhibits no cytoplasmic fluorescent staining in H₂O₂-treated, fatigued fibers. In ultrastructural observations, a longitudinal section of H₂O₂-treated fatigued muscle did not show disalignment of the A-I junction and Z-lines. Focal disruption of the A-band and swelling of the SR, mitochondria, or transverse tubules (often referred to as "exercise-induced muscle damage") (9, 26) were never observed in examined fibers (Fig. 5B). These results indicate no gross muscle damage with fatigue in the presence of H₂O₂ but do not eliminate the possibility that membrane or contractile proteins were damaged to an extent where they could affect the response to fatiguing stimulation.

View larger version (112K): [in this window] [in a new window]   Fig. 5.   Fluorescent (A) and electron (B) micrographs of frog single fibers observed after fatigue stimulation in the presence of 100 µM H2O2. A: light micrograph (top) and fluorescence micrograph (bottom) obtained with a single fiber. Note that there was no cytoplasmic staining with procion orange, but there was staining in connective tissues, which indicated no sarcolemmal disruption after fatigue protocol. B: transversely sectioned ultrastructural micrograph of a fatigued fiber. Sarcoplasmic reticulum (SR), transverse tubule (T), and mitochondria (M) and contractile apparatus show clear structures. A and I, A- and I-bands; G, glycogen granule; Z, Z-line. Vertical bar in A = 100 µm. Horizontal bar in B = 1 µm.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The present study demonstrates that 1) H₂O₂ at 10 µM markedly activated rabbit skeletal muscle RyR1 and frog RyR channels in lipid bilayers when redox potentials were defined at cis/trans = 220 mV/180 mV by using a [GSH]/[GSSG] buffer; 2) when only the cis redox potential was set at 220 mV, RyR1 channels were activated irrespective of the trans potential, but both cis and trans potential fixation stimulated the channel more strongly; 3) releasable Ca²⁺ content in the SR was markedly decreased after a 60-min rest after fatiguing stimulation for 2 min, and its decrease was little affected by exposure to H₂O₂; 4) prolongation of T_1/2 after a 60-min rest was observed in fibers exposed to H₂O₂; and 5) the decrease in tetanus force during the slow recovery process was not potentiated by exposure to H₂O₂. These findings suggest that H₂O₂ may not significantly disturb the intracellular Ca²⁺ homeostasis of skeletal muscle during the process of postfatigue recovery after intense muscle activity despite a marked activation effect of H₂O₂ on RyR channels in vitro.

All of the previous bilayer experiments dealing with effects of H₂O₂ on RyR channel modulation have been performed under undefined conditions of the redox potential in both cis and trans solutions. Under such conditions, high concentrations of H₂O₂ were needed to activate the channel (6, 14, 29-32, 47), consistent with the present study (Fig. 2). This may lead us to assume that H₂O₂ should not contribute to modulation of the RyR channel in vivo. Recently, however, it has been reported that the RyR molecule is a redox sensor and that the initial rate of ryanodine binding to the RyR vesicles depends on redox potentials (15, 54). The intracellular redox state in various cells mainly depends on the [GSH]/[GSSG] ratio. Cytoplasmic total [GSH] and [GSSG] in resting skeletal muscle have been estimated to be ~3 mM and ~50 µM, respectively (22, 45). This means that cytoplasm of resting muscles probably is in a reduced state (approximately 220 mV of redox potential) (2, 42), although the redox environment within the SR may be in an oxidative state (approximately 180 mV of redox potential; see Ref. 15). Our recent report indicates that modulation of RyR 1 channel activity by adenine nucleotide depends on cis, but not on trans, redox potential; the channel was activated by a shift of redox potential from 220 to 180 mV and was inhibited by a shift to 231 mV (33). These previous findings suggest that the redox potential is a primary factor determining the maximum channel activity and may alter the effects of channel modulators. As evidenced here (Figs. 1 and 2), H₂O₂ at 10 µM significantly enhanced RyR1 and frog RyR channel activities when cis redox potential was controlled near resting state. This is the first finding, to our knowledge, of an activating effect of H₂O₂ at near physiological concentration on the RyR channel. Decrease in the cis redox potential from 220 to 231 mV decreased the channel activity fivefold, which was consistent with our laboratory's previous observation (33). Therefore, RyR channel activity depended on the cis redox potential, and the definition of the cis redox potential at the resting state of muscle cell markedly promoted the channel opening after exposure to H₂O₂. At conditions with a definition of either cis or trans redox potential at 231 mV, an increase in cis H₂O₂ concentration to 100 µM did not affect the channel activity (unpublished observations). Addition of 100 µM H₂O₂ to cis chamber in the presence of [GSH]/[GSSG] = 3.096 mM/0.0196 mM (redox potential = 231 mV) was estimated to shift the potential to 207 mV when calculated with an assumption that 100 µM H₂O₂ oxidizes 100 µM GSH to GSSG. If so, this large oxidation should stimulate channel activity. However, it was not true in our experimental system. Therefore, it seems likely that a shift of cis redox potential toward more positive potential, which would be produced by addition of 100 µM H₂O₂, is small and not enough to induce RyR channel activation.

A concentration of 10 µM H₂O₂ may be still a little higher than physiological concentrations that would be produced in response to stimuli, such as strenuous exercise and ischemic reperfusion (less than several µM; Ref. 25). One to 3 µM H₂O₂ failed to stimulate the channel activated at a pCa of 6-6.3, even under the redox potential control (Figs. 1 and 2). RyR channel activity is known to be modulated by various ligands such as Ca²⁺, Mg²⁺, adenine nucleotides, GSH, GSSG, FK506-binding proteins, calmodulin, triadin, and calsequestrin (34), which exist endogenously within muscle cells. In the present study, we defined only [GSH]/[GSSG] and Ca²⁺ concentration and used KCl as an ionic carrier. Other channel modulators were ignored in the present experiment to easily analyze data. This may explain why H₂O₂ at 3 µM failed to activate RyR channels, although a recent report shows that prolonged exposure of mouse skeletal muscle fiber to H₂O₂ at 10 µM, but not at 1 µM, increased resting intracellular free Ca²⁺ concentration (3).

Importantly, we demonstrate that channel activation induced by 10 µM H₂O₂ in the RyR1 under cis redox potential control was further enhanced by subsequent definition of trans redox potential (Table 1). As noted above, RyR channel activity was primarily regulated by cis redox potential. In addition, this observation indicates an important role of the intraluminal GSH buffer system on RyR1 channel modulation, i.e., possible participation of a redox gradient across the SR membrane on channel activity as suggested by Feng et al. (15). However, we do not know the reason why a shift of trans redox potential from 180 to 220 mV did not alter the P_o stimulated by H₂O₂ under a control of cis potential, although we have previously published that RyR1 channel activation induced by Ca²⁺ and adenine nucleotide depended on cis, but not on trans, redox potential (33). Further studies on roles of trans redox potential are required.

Externally applied H₂O₂ crosses the cell membrane (5) to act as an oxidant for SR proteins (32) and has been reported to elicit the release of Ca²⁺ from the SR, as noted in the introduction (see also Ref. 24 for review). When animals are subjected to exercise-induced oxidative stress, intracellular GSH appears to rapidly oxidize to GSSG, resulting in a shift of the redox potential toward less negative values. In addition, the activity of catalase or superoxide dismutase in the gastrocnemius muscle of adult men has been estimated to be 16- or 40-fold less than the respective activities in the liver (19). Therefore, skeletal muscle antioxidant defenses are considered to be poor, as reviewed by Sen (45), so muscles are susceptible to endogenous and/or exogenous ROS. In fact, there are extensive experiments indicating that ROS produced by exercise result in muscle fatigue and damage (11, 40, 41) and that antioxidants or ROS scavengers mitigate muscle fatigue and improve recovery (4, 12, 43, 45). These observations made us expect supplementation of H₂O₂ to induce or enhance muscle damage. Even when the high concentration of 100 µM H₂O₂ was used, however, it failed to enhance the extent of fatigue (Fig. 3) or to produce gross histological damage (Fig. 5). Accidentally, we applied 450 µM H₂O₂ to four separate single fibers for 10 min and then stimulated the fibers repetitively. Fibers contracted spontaneously 22-36 s after the onset of stimulation and showed force with 28.1-32.8% of control tetanus. Such fibers were damaged, as evidenced by cytoplasmic procion orange staining (data not shown). Therefore, an enormous concentration of H₂O₂ has the ability to produce muscle damage. However, H₂O₂ at near physiological concentrations does not seem to be a factor that enhances muscle fatigue or elicits gross membrane damage, although the possibility should be checked that membrane proteins are damaged to an extent at which they could affect the response to fatiguing stimulation.

The mechanism by which H₂O₂ inhibits tension during the postfatigue slow recovery remains unexplained by the present study. Several mechanisms may control the H₂O₂-induced delay: 1) decrease in maximum Ca²⁺-activated force and Ca²⁺ sensitivity of contractile apparatus, 2) increased Ca²⁺ release from the SR, 3) depression of releasable Ca²⁺ content in the SR, and 4) deterioration of myofibrillar proteins. Ca²⁺ sensitivity has been reported to be rather increased after fatigue compared with that of the resting fiber (52). H₂O₂ by itself had little effect on maximum Ca²⁺-activated force and Ca²⁺ sensitivity (8, 10, 25), but Andrade and colleagues (2, 3) reported an increase in myofibrillar Ca²⁺ sensitivity during incubation with H₂O₂. In any case, there is no evidence of decreased sensitivity of contractile proteins to Ca²⁺ after exposure to H₂O₂. On the other hand, 10 µM H₂O₂ activated RyR1 and frog RyR channels in bilayers when the redox potential was defined at the resting state (Figs. 1 and 2). Many investigators indicate that H₂O₂ also enhanced the contraction of unfatigued muscle (3, 20, 32, 42). In addition, Ca²⁺ release from the SR is retained at a decreased level even after 60 min of rest in fibers subjected to fatigue (36). In line with these findings, we observed that rapidly releasable Ca²⁺ in the SR was significantly reduced after a 60-min rest in fatigued fibers. These findings suggest that H₂O₂-induced enhancement of Ca²⁺ release may elicit the decrease in Ca²⁺ content in the SR, and in turn decreased releasable Ca²⁺ content may contribute partially to the delay of postfatigue recovery of force. However, exposure to 10 µM H₂O₂ decreased caffeine contracture to 93% of controls not treated with H₂O₂ in unfatigued fibers and to 90% in fatigued fibers (Table 1). Similar estimation for 100 µM H₂O₂ gave decreases of 89% for unfatigued and 85% for fatigued muscles. This indicates that H₂O₂ may produce a decrease of only several percent of the releasable Ca²⁺ content in the SR. Therefore, it is questionable whether such a little effect of H₂O₂ on the releasable Ca²⁺ content contributes to the H₂O₂-induced decrease in postfatigue tetanus (Fig. 3) as a primary factor. A reduction in SR Ca²⁺-ATPase activity (or decrease in Ca²⁺ sequestration by the SR) has been reported after exhaustive exercise (13). In the present study, a small but significant prolongation of T_1/2 was obtained after a 60-min rest in the presence of H₂O₂. Therefore, the externally applied H₂O₂-induced decrease in tetanus tension during postfatigue recovery may be caused by combined effects of functional alteration of contractile proteins, activation of Ca²⁺ release, and depression of Ca²⁺ uptake by the SR on prolonged exposure to H₂O₂, although H₂O₂ did not histologically produce gross muscle damage (Fig. 5).

H₂O₂ is known to be stable in the absence of transition metals (44), but it is decomposed to produce the highly cytotoxic reactive radical · OH via the Fenton reaction in the presence of transition metals such as iron. The amount of · OH that would be generated in the absence of iron is very small (20). In the present study, catalase improved the slow recovery from fatigue, when H₂O₂ was externally applied. Catalase acts to decompose H₂O₂ to nontoxic substances (H₂O and O₂) but does not produce the · OH radical. Therefore, H₂O₂-induced delay of postfatigue recovery observed here would be caused by the direct effect of H₂O₂ applied to muscles. Although catalase of a large molecular size cannot enter the cell, H₂O₂ entering the cell immediately after extracellular application seems to be slowly decomposed by application of catalase (see DISCUSSION of Ref. 41). Eventually, a decreased intracellular concentration of H₂O₂ would make the RyR channel less open and protect excitation-contraction coupling-related proteins from toxicity of H₂O₂. This may explain why the recovery from fatigue was improved in our experimental conditions.

In conclusion, we found that 10 µM H₂O₂ enhanced channel activity of RyR1 and frog RyRs when redox potentials were defined at values near the resting state and releasable Ca²⁺ amount in the SR and tetanus tension still decreased after a 60-min rest in fatigued fibers, but that H₂O₂ application to muscles did not significantly affect the releasable Ca²⁺ content and the postfatigue recovery of force. These results suggest that H₂O₂ markedly activates the RyR channel in vitro but may not play an important role in the postfatigue recovery process in vivo when externally applied.