We investigated the hypothesis that muscles of different oxidative potential would display differences in sarcoplasmic reticulum (SR) Ca2+ handling responses to repetitive contractile activity and recovery. Repetitive activity was induced in two muscles of high oxidative potential, namely, soleus (SOL) and red gastrocnemius (RG), and in white gastrocnemius (WG), a muscle of low oxidative potential, by stimulation in adult male rats. Measurements of SR properties, performed in crude homogenates, were made on control and stimulated muscles at the start of recovery (R0) and at 25 min of recovery (R25). Maximal Ca2+-ATPase activity (Vmax, μmol·g protein−1·min−1) at R0 was lower in stimulated SOL (105 ± 9 vs. 135 ± 7) and RG (269 ± 22 vs. 317 ± 26) and higher (P < 0.05) in WG (795 ± 32 vs. 708 ± 34). At R25, Vmax remained lower (P < 0.05) in SOL and RG but recovered in WG. Ca2+ uptake, measured at 2,000 nM, was depressed (P < 0.05) in SOL and RG by 34 and 13%, respectively, in stimulated muscles at R0 and remained depressed (P < 0.05) at R25. In contrast, Ca2+ uptake was elevated (P < 0.05) in stimulated WG at R0 by 9% and remained elevated (P < 0.05) at R25. Ca2+ release, unaltered in SOL and RG at both R0 and R25, was increased (P < 0.05) in stimulated WG at both R0 and R25. We conclude that SR Ca2+-handling responses to repetitive contractile activity and recovery are related to the oxidative potential of muscle.
- skeletal muscles
- contractile activity
- calcium handling
the generation of specific force levels by the muscle cell is intimately dependent on effectively translating the excitation signal into a precisely regulated increase in the cytosolic free Ca2+ concentration ([Ca2+]f). In skeletal muscle, the regulation of [Ca2+]f is primarily dependent on the kinetics of Ca2+ release and Ca2+ uptake by the sarcoplasmic reticulum (SR) (3). The rate at which Ca2+ release occurs is dependent on the open state of the calcium release channel (CRC), also known as the ryanodine receptor (RyR), whereas Ca2+ uptake is dependent on recruitment of Ca2+-ATPase, the enzyme responsible for sequestering [Ca2+]f from the cytosol into the lumen of the SR (3). The factors involved in the acute regulation of both the CRC and the Ca2+-ATPase to achieve a given [Ca2+]f transient, and consequently an expected mechanical event, remain largely undefined (31).
It has become clear that the SR Ca2+ handling can display a variety of behaviors depending on the characteristics of the task. The particular Ca2+ handling response also appears to depend on the fiber type characteristics of the muscles involved. Repetitive contractile activity, as an example, results in reductions in Ca2+ uptake (5, 18, 54) and Ca2+ release (16, 26) in soleus and in red gastrocnemius/red vastus lateralis muscles of the rat. The decreases that are observed in these properties appear to be due to structural alterations in the Ca2+-ATPase (30) and CRC (16), respectively. In contrast, no reductions in SR Ca2+ cycling properties have been observed in white gastrocnemius/white vastus lateralis muscles following repetitive stimulation (5, 16, 18, 40). Because the red vastus lateralis and gastrocnemius muscles are composed primarily of fast-twitch (type II) fibers and the soleus of slow-twitch (type I) fibers (11), it would appear that the “red” character of the fiber, which associates with properties supporting oxidative phosphorylation (39), may be important in the SR Ca2+ cycling responses observed in response to repetitive stimulation. In addition, it is also possible, given that prolonged treadmill running was used to induce repetitive muscle activity, that the SR Ca2+ handling responses observed between muscles also may reflect different patterns of recruitment during the activity (2). Studies using high-intensity treadmill exercise have reported decreases (35) in maximal Ca2+-ATPase activity and Ca2+ uptake immediately following exercise in rat white gastrocnemius/vastus lateralis. In contrast, short-term intense activity induced by electrical stimulation caused an increase in maximal Ca2+-ATPase activity but not Ca2+ uptake in white gastrocnemius (12).
Reports also have been published describing an overshoot in Ca2+-ATPase activity (17, 40) and Ca2+ uptake (40) during recovery following exercise. To date, these observations have been restricted to muscles of predominantly fast, white composition. In muscles that display a reduction in Ca2+ cycling behavior with exercise, which, as emphasized, are highly oxidative in character, it is unclear what occurs during the recovery period following exercise. However, given the observation that normalization of Ca2+-ATPase activity following exercise-induced structural damage to the enzyme appears to take several days (33), it might be expected that Ca2+-ATPase activity would remain depressed in the late recovery period following exercise.
Differential responses in SR Ca2+ cycling behavior between muscles in response to repetitive activity might also be expected to associate with changes in mechanical behavior, because disturbances in Ca2+ handling have been mechanistically linked to fatigue (1). If such is the case, low-frequency fatigue, observed at low frequencies of stimulation and known to persist for hours after exercise, should be restricted to muscles showing disturbances in Ca2+ cycling behavior and, in particular, depressions in Ca2+ release (52). It also would be expected that prolongations in relaxation rate would accompany reductions in Ca2+ uptake (14). Reductions in both Ca2+ release and Ca2+ uptake have been shown to accompany the decrease in the [Ca2+]f transient observed in isolated single fibers at fatigue induced by repetitive stimulation (1).
The primary purpose of this study was to investigate the effects of repetitive contractile activity and contractile activity with subsequent recovery on the regulation of the SR Ca2+ cycling properties in rat muscles of different fiber type composition and oxidative potential. We hypothesized that in response to repetitive stimulation, the changes in SR Ca2+ handling properties would be directly related to the oxidative potential of the muscle and not to the major fiber type composition as dictated by the myosin heavy chain composition and contractile speed. Moreover, fatigue in the soleus, both with exercise and recovery, would be directly related to the reduction in Ca2+ cycling properties. As a subsidiary problem, we also examined fatigue manifestations during exercise and recovery in the whole gastrocnemius muscles. Because this muscle contains, on average, the same oxidative potential as the soleus but has a different fiber type composition (11), we were only able to investigate the hypothesis that fatigue, both with exercise and recovery, could be more pronounced in a muscle of predominantly type II fiber composition compared with a muscle of predominantly type I composition.
Adult male Sprague-Dawley rats were randomly divided into different groups to determine the effects of contractile activity plus recovery on homogenate SR Ca2+ cycling properties and “in situ” mechanical properties. The animals were fed water and laboratory chow ad libitum and were housed in an environmentally controlled room on a reverse light-dark cycle until the time of the study. All experiments were conducted at approximately the same time, between 8:00 AM and 3:00 PM, to limit diurnal variations in muscle glycogen (7). The experimental protocols were approved by the University of Waterloo Animal Ethics Committee.
To investigate the effects of contractile activity and contractile activity plus recovery on soleus (SOL), red gastrocnemius (RG), and white gastrocnemius (WG) muscles, we applied electrical stimulation directly to the sciatic nerve of one limb, whereas the opposing limb acted as the control. Adult male Sprague-Dawley rats were randomly divided into two different groups: one group, weighing 416 ± 3.0 g (mean ± SE), was used for studying crude homogenate (HOM) Ca2+ cycling properties (n = 20); a second group, weighing 421 ± 2.0 g, was used to determine the mechanical properties of isolated muscles stimulated in situ (n = 16). The first group of rats, used for the SR measurements, was randomly assigned to one of two subgroups (n = 10 per subgroup): one group received stimulation, whereas the other group received the same stimulation plus 25 min of recovery. The second group of rats also was divided into two subgroups: one subgroup was used for assessment of SOL contractile characteristics, and the second subgroup was used for assessment of the whole gastrocnemius (GAS). In general the conditions investigated were patterned on earlier work for which investigators in our group reported both increases in Ca2+-ATPase activity with short-term exercise (22) and an overshoot in Ca2+-ATPase activity in recovery (41). The actual specifics of the stimulation protocol were refined by pilot work. Using electrical stimulation, we were able to sample the tissue much more rapidly after the repetitive contractions.
The stimulated muscles received 150 ms/s trains at supramaximal voltage (determined at optimal length, Lo) at 60 Hz for 5 s, followed by 5 s of relaxation, applied for a duration of 5 min. Immediately after the stimulation and at 25 min of recovery, the muscles were removed and separated into the SOL, RG, and WG tissues, and a portion of each muscle was rapidly frozen in liquid N2. We labeled the recovery times, namely, immediately after the exercise and after recovery, as R0 and R25, respectively. The RG and WG tissue were separated visually by color, and samples were always collected from similar areas; RG was taken at the proximal end of the muscle belly just superficial to the plantaris (which was previously removed and discarded), and WG was taken from the superficial muscle belly just lateral to the saphenous artery. The RG and WG tissue would appear to represent the proximal and distal compartments, respectively, that have been characterized by de Ruiter et al. (8). For the muscles and regions selected, oxidative potential, estimated by maximal citrate synthase (CS) activity was 5.03 ± 0.33, 7.50 ± 0.34, and 2.72 ± 0.26 μmol·g protein−1·min−1 for the SOL, RG, and WG, respectively. Maximal CS activity, measured as previously described (20), was different (P < 0.05) among each group of muscles.
The tissue was diluted 1:10 (wt/vol) in ice-cold homogenizing buffer (pH 7.5) containing (in mM) 250 sucrose, 5 HEPES, 0.2 phenylmethylsulfonyl fluoride (PMSF), and 0.2% sodium azide (NaN3) and was mechanically homogenized on ice using a Polytron homogenizer (PT 3100) at 16,500 rpm for two 30-s bursts separated by a 30-s break. Approximately 1,000 μl of HOM sample were frozen in liquid N2 and stored at −80°C for later analysis of Ca2+ cycling properties.
The additional subgroups of animals also underwent the same stimulation protocol; however, in this group mechanical data were collected both before and during recovery following the 5 min of stimulation. Mechanical characteristics were collected in separate experiments for isolated SOL and whole GAS. Surgery was performed similar to that for tissue sampling, except that this time, instead of leaving the whole plantar flexors intact, either the SOL or the GAS was isolated. These isolations were accomplished by separating the SOL at the proximal end of the Achilles tendon, just distal to the muscle belly. The entire SOL was isolated from the GAS and the plantaris. If the SOL was the muscle of interest, the other plantar flexors were cut from the Achilles tendon; if, on the other hand, the GAS was the muscle of interest, the SOL tendon was cut. In addition, the plantaris was isolated from the rest of the GAS and severed close to the Achilles tendon.
Stimulation was applied directly to the sciatic nerve with a Grass S48 stimulator. Before the 5-min stimulation (rest), after the 5-min stimulation (R0), and during recovery at 5 (R5), 10 (R10), 15 (R15), 20 (R20), and 25 min (R25), mechanical properties were measured for both control and stimulated limbs. Tetanies were produced at Lo and at various frequencies of stimulation (10, 20, 30, 50, and 100 Hz), using a 0.2-ms pulse at a supramaximal voltage, and were utilized to determine the maximal rates of relaxation (−dP/dtmax) and force development (+dP/dtmax) and the peak force (Pt). Invoked twitches were also used to characterize these properties, as well. During the determination of Lo, muscle twitches were separated by 30 s to prevent fatigue and potentiation. All mechanical parameters were expressed relative to the initial preexercise resting value, which was set at 100% (rest). Because of the length of this article, only the twitch properties have been included. Force data were collected online by using a 640 A signal interface (Aurora Scientific) connected to a National Instruments 16-bit analog-to-digital card and were analyzed using the Dynamic Muscle Control and Data Acquisition and Dynamic Muscle Analysis software (Aurora Scientific). The general procedures used and properties assessed can be found in earlier reports from our laboratory (6). An independent shunt calibration was performed daily for each force transducer.
Animal Surgical Preparation
The same surgical preparation was used for all experimental groups and was similar to that reported by Tupling et al. (47). Rats were initially anesthetized using intraperitoneal injections of pentobarbital sodium (6 mg/100g body wt). Throughout the experiments, the animals were given booster injections as required to maintain anesthesia. For HOM experiments, hindlimbs were shaved, connective tissue and overlying fascia were removed, and the distal section of the Achilles tendon was dissected, cut, and tied with surgical silk to an S-hook that later was attached to a Grass linear variable differential transducer. A vertical incision was made through the gluteal muscles, and the sciatic nerves were isolated from surrounding connective tissue. Rats were then placed on their stomachs in a neutral position and positioned on a heated holding platform. Both legs were prepared identically, and random assignment was used to determine the S leg. The knee joint was stabilized at 90° flexion by inserting a steel needle through the capsule of the knee (hip also under 90° flexion with no rotation), the previously secured S-hook was attached to the transducer, and electrodes (attached to a Grass model FT03 stimulation apparatus) were placed around the sciatic nerve. During surgical preparation and throughout the experimental protocol, muscle tissue was covered by gauze and repeatedly moistened with warm saline (9% NaCl) to prevent dehydration. Muscle surface temperature was monitored with a surface thermometer and maintained between 28 and 32°C.
To investigate the changes in SR Ca2+ handling properties with our experimental protocol, we measured Ca2+-ATPase, Ca2+ uptake, and Ca2+ release in crude homogenates.
SR Ca2+-ATPase activity.
Measurements of Ca2+-induced SR Ca2+-ATPase activity were performed on crude homogenates prepared from SOL, RG, and WG tissue. Ca2+-ATPase enzyme kinetics were assessed by measuring the Ca2+-ATPase activity at progressively increasing [Ca2+]f. Ca2+-ATPase kinetic parameters of interest included the maximal Ca2+-ATPase activity (Vmax), the Hill coefficient (nH), an indicator of the cooperative Ca2+ binding affinity, and the [Ca2+]f needed to elicit 50% Vmax (Ca50). Ca2+-induced Ca2+-ATPase activity was measured using a Beckman DU900 spectrophotometer according to the regenerating assay developed by Simonides and van Hardeveld (42) with minor modifications (49). Samples were run with and without the Ca2+ ionophore A-23187 to provide information regarding the effects of exercise and recovery on SR membrane integrity. The reaction buffer contained (in mM) 20 HEPES, 200 KCl, 15 MgCl2, 1 EGTA, 10 NaN3, 5 ATP, and 10 phosphoenolpyruvate (PEP), pH 7.0. Immediately before the reaction with Ca2+ additions was initiated, 18 U/ml lactate dehydrogenase and 18 U/ml pyruvate kinase enzyme were added, as well as 0.3 mM NADH, 2.08 μl of Ca2+ ionophore A-23187, and the tissue (HOM). Assays were performed at 37°C and 340 nm. After a 2-min recording of baseline activity, the reaction was initiated by adding 2.0 μl of 100 mM CaCl2 and the absorption was measured for ∼2 min. At this time, and every 2 min thereafter until Vmax was reached, 0.5 μl of CaCl2 was added and, again, absorption was measured. Once Vmax was obtained, 40 μM (1.0 μl of 40 mM stock) cyclopiazonic acid, a specific inhibitor of the Ca2+-ATPase, was added to determine basal activity. The [Ca2+]f corresponding to the additions of 100 mM CaCl2 were measured separately using dual-wavelength spectrofluorometry and the Ca2+ fluorescent dye indo-1. Details of the procedure for calculating [Ca2+]f have been previously published by our group (50).
The Ca2+-dependent Ca2+-ATPase activity was used to calculate nH and Ca50. These properties were determined by nonlinear regression with computer software (GraphPad Software) using the following sigmoidal dose-response equation: (1) where Ybot is the value at the bottom of the curve, Ytop is the value at the top of the plateau, and log Ca50 is the logarithm of the Ca50 (the pCa value that gives a response halfway between Ybot and Ytop). For calculation of nH, only the linear portion corresponding to 10–90% of Vmax was used.
SR Ca2+ uptake.
Oxalate-supported Ca2+ uptake was measured using the Ca2+ dye indo-1 at 37°C according to the methods outlined by O'Brien et al. (36) with minor modifications by our group (47). Fluorescence measurements were collected on a dual-emission wavelength spectrofluorometer (Ratiomaster system; Photon Technology International). The excitation wavelength was set to 355 nm, and 405 and 485 nm correspond to the emission wavelengths for bound (F) and free (G) indo-I, respectively. The measurement of [Ca2+]f with this system is based on the difference between the maximal emission wavelengths between the F and G indo-I. The reaction buffer contained (in mM) 200 KC1, 20 HEPES, 10 NaN3, 0.005 μM N,N,N′,N′-tetrakis(2-pyridylmethyl)ethylenediamine, 5 oxalate, 15 MgC12, and 10 PEP, pH 7.0. The ratio of F/G (R) was used to calculate [Ca2+]f and decreases with Ca2+ uptake. Using Felix software (Photon Technology International), we calculated [Ca2+]f from the following equation: (2) where Kd represents the equilibrium constant for the interaction between Ca2+ and indo-1. A Kd value of 250 was used (23). Rmin is the minimum value of R, Rmax is the maximum value of R, and Gmax is the maximum value of G (corresponding to the addition of 1 mM CaCl2).
Photon counts were simultaneously collected for each wavelength. Before each analytical session, the background fluorescence was determined in the absence of indo-1 and implemented before each assay was started.
Before data collection was initiated, the assay buffer was heated to 37°C and 1.5 μM indo-1, 2.0 μl of 10 mM CaCl2, and 150 μl of SOL HOM (100 μl of RG HOM and 50 μl of WG HOM) were added to a cuvette containing 2.0 ml of buffer. Once these additions were made, data collection was started. After a brief period, 40 μl of 250 mM ATP were added to initiate Ca2+ uptake. Initial [Ca2+]f, defined as the [Ca2+]f obtained just after addition of ATP and before any noticeable Ca2+ uptake by the SR, was ∼3.0 μM.
The curve generated from Eq. 2, [Ca2+]f vs. time, was smoothed over 21 points using the Savitsky-Golay algorithm. Linear regressions were performed on values of [Ca2+]f at 500, 1,000, 1,500, and 2,000 (range ±100) nM. Differentiating the linear fit curve allowed the determination of Ca2+ uptake rates.
Apparent coupling ratios, an indication of the number of Ca2+ ions transported per ATP hydrolyzed, was determined using the ratio of Ca2+ uptake, assessed at 2,000 nM, to Ca2+-ATPase activity (Vmax). We have used the term “apparent coupling ratio” because the two properties were not determined under identical conditions.
SR Ca2+ release.
Ca2+ release was measured in a similar way to Ca2+ uptake, by using indo-1 with a dual-wave emission spectrofluorometer (Ratiomaster system; Photon Technology International) described in detail previously (47). After active loading of the SR, when [Ca2+]f was calculated to be <100 nM, Ca2+ release was induced with the addition of 5 mM (10 μl of 1 M stock) 4-chloro-m-cresol (4-CMC). Tupling and Green (48) have shown that 4-CMC acts directly on the RyR to induce Ca2+ release, unlike silver nitrate (AgNO3), which also induces Ca2+ release from the Ca2+-ATPase. Ca2+ release was determined on the same homogenate following the Ca2+ uptake measurements. After Ca2+ release, the Ca2+ uptake calibration was applied, and the generated curve was smoothed over 21 points by using the Savitsky-Golay algorithm, yielding a curve of [Ca2+]f vs. time. The generated Ca2+ release curve yields two visually distinct phases, a fast initial phase 1 and a slower secondary phase 2. Linear regressions were performed on Ca2+ release vs. time corresponding to each phase and were differentiated to determine the maximal rate of Ca2+ release.
Muscle glycogen and metabolites.
Muscle metabolites and glycogen were analyzed fluorometrically after extraction from freeze-dried tissue according to previously published procedures (21, 24). The metabolites assessed included ATP, phosphocreatine (PCr), and lactate. Glycogen was determined as glucose in a separate piece of tissue after alcohol precipitation and hydrolysis (29).
For Ca2+ uptake measurements, a three-way ANOVA was used to investigate the effects of stimulation (control vs. stimulated), time (stimulation vs. recovery), and [Ca2+]f (2,000 vs. 1,500 vs. 1,000 vs. 500 nM). For Ca2+ release measurements, a three-way ANOVA was used to investigate the effects of stimulation (control vs. stimulated), time (stimulation vs. recovery), and the biphasic release response (phase 1 vs. phase 2). For all other SR measurements and the muscle metabolites, a two-way ANOVA was used to investigate the effects of stimulation (control vs. stimulated) and time (R0 vs. R25). A two-way ANOVA was used for all mechanical measurements to investigate the effects of stimulation (control vs. stimulated) and time (rest, R0, R5, R10, R15, R20, and R25). The significance level was set at 0.05 (P < 0.05), and when appropriate, a Newman-Keuls post hoc test was used to compare specific means. Where only main effects were observed, the conditions affected are indicated.
Muscle Metabolites and Glycogen
Only in the RG and WG muscles did the stimulation protocol result in a reduction in ATP and PCr concentrations (Table 1). For these two regions of the GAS, reductions ranging between 20 and 22% and between 24 and 45% occurred for ATP and PCr, respectively. By 25 min of recovery, no differences existed between control and stimulated muscles for either of the high-energy compounds, regardless of region. For the SOL, no differences were observed between control and stimulated muscles for both ATP and PCr at either R0 or R25. Lactate concentration increased in all three tissues immediately after the stimulation, with the greatest increase occurring in the WG. At R25, lactate was not different between control and stimulated muscles for each tissue. Muscle glycogen was also reduced with stimulation regardless of tissue type. As indicated by the differences between control and stimulated muscles at R25, glycogen levels remain depressed in each tissue during recovery.
sr Ca2+ Cycling Properties
The kinetic properties of the Ca2+-ATPase activity included Vmax, nH, and Ca50 (Fig. 1 and Table 2). The SOL and RG muscles displayed similar responses to the contraction and recovery protocol, namely, a main effect of stimulation. In the case of the SOL, Vmax was reduced by 22% immediately following the stimulation. After 25 min of inactivity, no recovery in Vmax occurred. For the RG, Vmax was reduced by 15% at R0, which persisted throughout the recovery period. In contrast, in WG, Vmax was increased by 12% at R0 and returned to control levels at R25. The other kinetic properties examined, namely, nH and Ca50, were unaffected by the stimulation and recovery protocol in both SOL and RG. For WG, although no change was observed in Ca50, a main effect of stimulation was observed such that nH was less in stimulated than in control muscle at both R0 and R25. The basal ATPase activity was unaltered by the experimental protocol regardless of the muscle studied.
The ionophore ratio, a measure of the ratio of Vmax measured with and without the Ca2+ ionophore A-23187, was reduced with stimulation, but only in the SOL and RG (Fig. 2). The stimulation-induced reduction in this ratio was a main effect that also persisted throughout the recovery period. Repetitive activity was without effect in altering the ionophore ratio in WG at either R0 or R25.
Ca2+ uptake was assessed at four different [Ca2+]f levels. All three muscles showed a main effect of [Ca2+]f level (2,000 > 1,500 > 1,000 > 500 nM). Because all Ca2+ concentrations showed generally the same effect, we have only presented the data for a [Ca2+]f value of 2,000 nM. As with Vmax, repetitive activity depressed Ca2+ uptake in SOL and RG at R0 by 34 and 13%, respectively (Fig. 3). The depression in Ca2+ uptake persisted at R25. For WG, Ca2+ uptake increased by 9% at R0 and remained significantly elevated at R25. For all muscles, only main effects of the stimulation were observed. The apparent coupling ratio, defined as the ratio between Ca2+ uptake and maximal Ca2+-ATPase activity, was not affected by either stimulation or recovery in any of the muscles examined.
The rate of Ca2+ release was assessed for two distinct phases following active Ca2+ loading (Fig. 4). In contrast to the decrease in Ca2+ sequestering properties observed in SOL and RG, no differences at either R0 or R25 were noted for either phase 1 or phase 2 Ca2+ release. For WG, the repetitive contractile activity resulted in increases in phase 1 and phase 2 at R0 by 21 and 30%, respectively. The increase in Ca2+ release for these phases was not affected by the time of recovery.
The 5 min of intermittent contractile activity resulted in pronounced alterations in the properties of the twitch in GAS both immediately after the stimulation and during the 25-min period of recovery (Fig. 5). Compared with rest, peak tension (Pt) was reduced by ∼36% immediately after the contractile activity. No recovery in Pt was observed during the recovery period. The property used to assess the kinetics of force development during the supramaximal twitch, namely, +dP/dtmax, was also affected by the stimulation protocol. For +dP/dtmax, the depression that was observed at R0 persisted throughout the 25 min of recovery. Relaxation characteristics following the single twitch were measured as −dP/dtmax. At R0, −dP/dtmax was reduced by ∼19% compared with rest. At R5, −dP/dtmax was reduced by an additional 17%. No further change in −dP/dtmax was observed throughout the subsequent recovery periods. No changes were observed in the control limb between rest and each of the recovery periods for any of the twitch properties examined.
Mechanical function in the SOL muscle was also disturbed by the repetitive contractions. For the twitch properties, Pt was reduced by ∼46% at R0 (Fig. 6). No recovery in Pt occurred during the recovery period. A similar effect was observed for +dP/dtmax, namely, a pronounced reduction in the rate of force development at R0 that persisted throughout R25. For −dP/dtmax, persistently lower values were observed for stimulated compared with control muscle throughout recovery. However, these differences were not significant.
As hypothesized, we found that the effects of the repetitive contraction and recovery protocol on SR Ca2+ handling properties depended on the oxidative potential of the muscle. The SOL and RG muscles, which are primarily slow twitch and fast twitch, respectively, but of high oxidative potential (8, 11), displayed similar responses. These responses consisted of a depression in Vmax observed both after the contractile activity and after the 25 min of recovery. In contrast, the WG, a muscle of predominant fast-twitch composition but low in oxidative potential (11), responded with an increase in Vmax that returned to normal during recovery. In general, the changes in Ca2+ uptake mirrored the changes in Vmax, namely, decreases in SOL and RG and increases in WG during the exercise and recovery phases of the protocol. We also found that differences observed in Ca2+ release appeared to relate to the oxidative potential of the muscle. Repetitive mechanical activity resulted in increases in both phase 1 and phase 2 Ca2+ release in WG that persisted during the recovery period. In contrast, no differences between the stimulated and control muscles in Ca2+ release, in either phase 1 or phase 2, were observed for SOL and RG either after exercise or after recovery. From the perspective of the SR and [Ca2+]f regulation, it appears that muscles of high oxidative potential, whether of fast-twitch or slow-twitch composition, experience decreases in Ca2+ uptake with no change in Ca2+ release during a short period of tetanic contraction, whereas muscles of low oxidative potential respond with increases in both Ca2+ release and Ca2+ uptake. Interestingly, at least with the stimulation protocol employed, these effects generally persist for at least 25 min after the activity. Because these measurements were performed in vitro under supposedly optimal activating conditions (42), the changes probably reflect second messenger regulatory influences and/or structural modifications to the Ca2+-ATPase enzyme (3, 32, 46) and/or the RyR (15, 19).
A second objective of the study was to investigate the hypothesis that the depression in mechanical function observed both during the contractile activity and during recovery in SOL would be related to the changes in the SR Ca2+ cycling properties. We employed the twitch to investigate the time-dependent changes of repetitive activity on the mechanical properties of SOL. On the basis of the responses we observed, our hypothesis must be partially rejected, namely, that the mechanical changes observed after exercise and during recovery are associated with the changes in SR Ca2+ cycling behavior. In the SOL, a muscle of high oxidative potential, Pt was depressed in stimulated muscle at R0 and remained depressed at R25. In SOL, although depressions were noted in Ca2+ uptake, no changes occurred in Ca2+ release, regardless of the time of measurement.
Of particular importance are the mechanisms underlying the different responses observed in the SR Ca2+ handling properties with our stimulation protocol and how the oxidative potential of the fiber is implicated. The depression in Vmax with contractile activity was expected on the basis of earlier studies that found similar reductions in both the SOL (54) and RG (5), using a variety of exercise protocols. The reductions in Vmax have been attributed to structural alterations in the enzymes, specifically in the region near the adenine nucleotide binding site (30, 33). This effect appears to be mediated by oxidation and nitrosylation as a result of accumulation of reactive oxygen species (ROS) (28). As in previous studies (5, 54), the reduction in Vmax occurred in isolation, because two other kinetic properties of the enzyme that were examined, namely, nH and Ca50, indicated that the Ca2+ sensitivity had not changed with the tetanic contractions. Because Ca2+ uptake responded in a similar fashion to Vmax in SOL and WG, the apparent coupling ratio was unaltered. This would suggest that the energy cost of the transport of a given amount of Ca2+ ions into the SR was unaltered. This result is similar to that previously observed with both SOL (54) and RG (5) following contractile activity. Somewhat surprising, however, was the reduction in the ionophore ratio observed in the SOL and RG with the repetitive contractions, an effect that persisted throughout recovery. Because the ionophore ratio is based on the ratio of Vmax measured with and without the Ca2+ ionophore A-23187, reductions in this ratio suggest that increases in SR membrane permeability to Ca2+ have occurred as a result of the tetanic contractions. The increased Ca2+ permeability could reduce the accumulation of Ca2+ in the SR and prevents back inhibition of the enzyme (31). Exercise-induced decreases in the ionophore ratio should result in increased leak of Ca2+ back into the cytosol during Ca2+ sequestration, providing an erroneously reduced Ca2+ uptake to what actually occurred (27). It is possible, given the apparent increased SR membrane permeability, that the decrease in Ca2+ uptake that we observed in SOL and RG with the contractile activity in this study could have been even more pronounced if a Ca2+ leak through the membrane was present. Previous studies using different exercise protocols have not reported any changes in the ionophore ratio in SOL (54), red vastus/RG (5), or diaphragm (34).
The limitations of our assay notwithstanding (see appendix), our findings do suggest a causal association between changes in Vmax and Ca2+ uptake that was observed with our stimulation and recovery protocol. It is clear from our results that the dominant mechanism acting to reduce Ca2+ uptake with repetitive activity is a reduced Vmax.
Mechanisms underlying the increase in Vmax that were observed in WG with exercise present a unique challenge given the novelty of these results. Others (17), including investigators in our laboratory (41), have reported an overshoot in Vmax in muscles of predominant fast-twitch fiber composition and low oxidative potential, but only during the recovery period following prolonged exercise in rats. No changes in Vmax were observed immediately after running (41). We have suggested, on the basis of our earlier observations using a NCD-4 probe, that the overshoot could be mediated by increased activation secondary to increased availability of the Ca2+ binding site, possibly induced by changes in the oligomeric status of the Ca2+-ATPase (40). At present, there are no known acute regulatory mechanisms that could result in increasing Vmax of the enzyme, short of decreasing sarcolipin content or increasingly actual Ca2+-ATPase content (46). Fast-twitch muscles such as the RG and WG express the sarco(endo)plasmic reticulum Ca2+-ATPase (SERCA)1a isoform almost exclusively (3, 53). Sarcolipin, a small membrane protein, is known to associate with SERCA1a isoform and decrease its Vmax (43, 46). Because we did not find increases in SERCA1a isoform content in previous studies using a more sustained exercise protocol (49), it is unlikely that the 5 min of contractile activity used in this study elevated SERCA1a content. Similarly, increased degradation of sarcolipin appears unlikely as well, given the brevity of our exercise intervention. It is possible that dissociation of SERCA1a with sarcolipin could occur, but there are no known regulatory mechanisms to cause this.
As with the SOL and RG, we found that the changes in Ca2+ uptake generally track the changes in Vmax, because no changes in the apparent coupling ratio were found. The ionophore ratio in WG also was unchanged with our experimental protocol, suggesting that no changes in the membrane permeability for Ca2+ occurred. However, for the WG, we did find evidence that the Ca2+ cooperativity of the enzyme had changed, because nH was reduced after exercise and after recovery. The reduction in nH was not accompanied by any changes in Ca50, the [Ca2+]f value required for 50% activation of Vmax. Because nH was defined as the slope of the relationship between Ca2+-ATPase activity and [Ca2+]f between 10 and 90% of Vmax, it appears that exercise reduced Ca2+ cooperativity during the steep activation phase. As a consequence, in this range, a higher [Ca2+]f value would be needed to elicit a given Ca2+-ATPase activity and, consequently, Ca2+ uptake. At present, the mechanism for decreases in nH appears unclear, particularly given that Vmax was increased. In previous studies reporting increases in Vmax during recovery from prolonged exercise, Ca2+ sensitivity was not modified (40).
Reductions in Ca2+ release assessed in vitro with exercise have commonly been reported in human (13), rat (16, 37), and frog (51) skeletal muscle. When assessed in vitro, reductions in Ca2+ release have been associated with reductions in 3H-labeled RyR binding, implying structural alterations to the RyR (15). These findings would appear to not apply to our study, because we found either no change (SOL and RG) or an increase (WG) in Ca2+ release with exercise. Because reductions in Ca2+ release with exercise have been reported in both rat SOL (54) and extensor digitorum longus (37), muscles of high oxidative potential but differing fast- and slow-twitch fiber composition (11), the most logical explanation for the different results would appear to be the differences in exercise protocols. Our exercise protocol was intense but intermittent in nature, interspersed with rest periods, whereas other exercise protocols were either submaximal and performed for sustained periods (16) or intense and continuous (37).
The increase in Ca2+ release in WG with our exercise protocol has not been previously reported, regardless of the fiber type composition of the skeletal muscle. The increases that we observed were reflected in both phase 1 and phase 2, the fast and slow kinetic phases, respectively. Previous studies measured what is essentially the slower phase of Ca2+ release (16, 37). We elected to report on two phases of release on the basis of our previous work (48), which demonstrated that two distinct phases occur as demonstrated by a sharp break in the relationship of Ca2+ release vs. time. At present, the functional significance of these phases remains unclear. What is clear from our current study is that the overshoot was reflected in both phase 1 and phase 2 of Ca2+ release.
The overshoot in Ca2+ release could be explained by increases in RyR receptor protein or increases in the open state of the RyR channels. It is highly improbable that increases in RyR protein resulted, given the observation that RyR protein has not been found to change in studies utilizing much more extended exercise protocols. Increases in the open status of the RyR leading to increases in Ca2+ release have been reported to occur in muscle secondary to phosphorylation of calmodulin kinase II, a regulatory RyR protein (45). Whether this occurred in our study and whether detection of this effect is possible in an in vitro assay needs to be determined.
The repetitive stimulation protocol that we employed resulted in changes in force and the maximal rates of force development and relaxation in both SOL and GAS muscles. To characterize these effects, we used the properties of twitch. In both the SOL and GAS, twitch force was reduced at R0, and no recovery in force was detected in either muscle during the 25-min period of recovery.
An inviting possibility to explain the fatigue observed with our repetitive stimulation protocol is loss of SR Ca2+ regulation. Reductions in [Ca2+]f with repeated tetani have been commonly associated with loss of force in single fibers of mouse muscle (1). The reduction in [Ca2+]f has been attributed primarily to a reduction in Ca2+ release (1). In this study, we found no changes in Ca2+ release at R0 in SOL and RG and an increase in WG. These findings would suggest that if fatigue in our study occurred as a consequence of reduced Ca2+ release, it did not occur as a direct modification of the Ca2+ release channel. This possibility has been suggested previously on the basis of the ability of caffeine, a known Ca2+ release agent, to reverse the depression in Ca2+ release and [Ca2+]f (1). The impaired Ca2+ release has been attributed to other mechanisms such as impaired Ca2+ loading into the SR or defective t-tubule (44) SR coupling. The sustained depression in force that we observed particularly at the lower frequencies of stimulation is a characteristic of low-frequency fatigue that occurs secondarily to disturbances in Ca2+ release and reduced [Ca2+]f (4). Even though Ca2+ uptake was depressed in SOL at R25, it is difficult to implicate Ca2+ loading as a problem given the 25-min period of low-frequency fatigue.
The twitch rates of both force development and relaxation were also affected by the repetitive contractions. In the case of +dP/dtmax, a different effect was observed between SOL and GAS. For GAS, +dP/dtmax was reduced at R0 and showed no recovery by R25. For SOL, a reduced +dP/dtmax rate was observed that was not specific to a time point. These findings are inconsistent with our in vitro measurements of Ca2+ release, because no changes in Ca2+ release, both phase 1 and phase 2, were observed in SOL. The twitch relaxation rate also showed a dissociation with our Ca2+ uptake measurements. In SOL, in which we observed a reduction in Ca2+ uptake at R0 and R25, no significant effect on −dP/dtmax was generally observed. Unfortunately, because the whole GAS was used for the mechanical studies, we were not able to investigate the role of SR Ca2+ handling. The major portion of the GAS is dominated by type II fibers of moderate oxidative potential (11). Fibers of high oxidative potential (RG) and fibers of low oxidative potential (WG) represent a relatively small region of the muscle, comprising ∼6 and 8% of the fiber mass, respectively. The calculated oxidative potential of the GAS is estimated to approximate that of the SOL. In this study, we did not isolate different branches of the sciatic nerve and investigate the mechanical responses of different regions of the GAS as has been done previously (8). These two different compartments investigated have been shown to display large differences in oxidative potential and to result in different mechanical and metabolic responses (9, 10, 38). The regions that we sampled and labeled as RG and WG would appear to be generally comparable to the two compartments studied by de Ruiter et al. (8).
Some qualification appears necessary in our study regarding the fiber type composition of the regions of the GAS that were sampled. Delp and Duan (11) in their exhaustive characterization of rat muscle reported an ∼4.5-fold difference between the oxidative potential of a pure red and pure white portion of the GAS. In our study, the difference was ∼2.8-fold, suggesting some contamination with the most dominant intermediate or mixed region of the GAS. Because the pure red portion contains some 50% type I fibers, it is probable that our RG contained a small percentage of type I fibers. Consequently, the possibility remains that the difference in the SR Ca2+ cycling response observed between tissues that have been ascribed to differences in oxidative potential could also be influenced by fiber type composition. Collectively, our results suggest that processes in addition to modifications in SR Ca2+ handling are involved in explaining the fatigue and weakness observed with repetitive stimulation.
With our experimental design, we examined the SR Ca2+ cycling response in response to repetitive muscle contraction and recovery under optimal in vitro conditions. In contrast, the mechanical changes were measured in vivo. As such, the mechanical responses are subject to changes in the intracellular environment mediated by the contractile activity. Some of these changes involve the cellular energy status and metabolic by-product accumulation. We have assessed some of these responses in this study. Our stimulation protocol resulted in only small ATP reductions in energy potential, as assessed by ATP and PCr, and only in the RG and WG. Lactate was elevated, but only modestly, in the oxidative SOL and RG, with the greatest increase in WG as expected (25). All of these properties had returned to normal at R25, the end of recovery.
In summary, by employing a brief period of repetitive stimulation followed by 25 min of recovery, we have been able to demonstrate fundamental differences in SR Ca2+-handling responses measured in vitro between muscles of different oxidative potential. Moreover, we also have demonstrated that the changes observed in the Ca2+-handling properties do not generally associate with the mechanical changes that occurred as assessed by the force, rate of force development, and rate of relaxation observed with the twitch response. Collectively, these results imply that fatigue mechanisms are mediated by events other than structural changes to the Ca2+-ATPase enzyme and the RyR channel. As a consequence, the physiological significance of the changes in the SR Ca2+ handling properties that we have observed remains unknown. An intriguing question remains, how does the oxidative potential of the muscle impact on the Ca2+-handling properties during contractile activity?
Methodological Issues in sr Measurements
There are limitations in our assay that could also affect the changes in Ca2+ uptake that were observed. It is possible that Ca2+ could have leaked out of the SR into the cytosol via either the RyR or the Ca2+-ATPase (27, 31). In our assay, we did not include agents that would inhibit Ca2+-ATPase activity, such as thapsigargin or high concentrations of RyR, to examine these potential effects. We have investigated these possibilities in other studies, however, using other exercise models, and have found no additional effect of the exercise (unpublished observation). Others also have reported no effect (16). An unavoidable limitation in our assays for Ca2+ uptake was the need to use oxalate as the medium to precipitate Ca2+ in SR. It is possible that the rate at which the Ca2+-oxalate precipitate is formed could limit the kinetics of Ca2+ uptake. In addition, it should be noted that Ca2+-ATPase activity was measured with the Ca2+ ionophore A-23187 to minimize inhibition in enzyme kinetics (27). For these reasons, as well as others, we have called the ratio of Ca2+ uptake to Ca2+-ATPase the apparent coupling ratio, to reflect the fact that the ratio has not been determined under identical measurement conditions for both properties. In our hands this was not possible, given the extremely slow rate of Ca2+ loading into SR without oxalate and the low Vmax observed when oxalate is used (unpublished observation).
As with the measurement of Ca2+ uptake, limitations in our assay used to measure Ca2+ release may be involved in our findings. The measurement of Ca2+ release followed the measurement of Ca2+ uptake, using the same assay conditions but with the releasing agent 4-CMC. This assay procedure has been criticized because oxalate is used to precipitate the Ca2+, and the dissociation of Ca2+ from oxalate may be the limiting factor in Ca2+ release and not the conductance properties of the RyR channels. However, we feel that this is highly unlikely because we have shown a graded Ca2+ release with different concentrations of 4-CMC (unpublished observation).
Alternatively, differences in the Ca2+ loading might also influence the Ca2+ release. Because the repetitive contractions also resulted in increases in Ca2+ uptake in WG, it is possible that the amount of Ca2+ loading into the SR was greater and, consequently, Ca2+ release increased. However, this possibility also appears remote, because the Ca2+ uptake assay was continued until no further loading resulted, implying that the SR was fully loaded with Ca2+.
We appreciate the financial support provided by the Natural Sciences and Engineering Research Council (to H. J. Green) for this research.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2006 the American Physiological Society