In several pathological and experimental conditions, the passive membrane conductance of muscle fibers (Gm) and their excitability are inversely related. Despite this capacity of Gm to determine muscle excitability, its regulation in active muscle fibers is largely unexplored. In this issue, our previous study (Pedersen et al. 2009. J. Gen. Physiol. doi:10.1085/jgp.200910291) established a technique with which biphasic regulation of Gm in action potential (AP)-firing fast-twitch fibers of rat extensor digitorum longus muscles was identified and characterized with temporal resolution of seconds. This showed that AP firing initially reduced Gm via ClC-1 channel inhibition but after ∼1,800 APs, Gm rose substantially, causing AP excitation failure. This late increase of Gm reflected activation of ClC-1 and KATP channels. The present study has explored regulation of Gm in AP-firing slow-twitch fibers of soleus muscle and compared it to Gm dynamics in fast-twitch fibers. It further explored aspects of the cellular signaling that conveyed regulation of Gm in AP-firing fibers. Thus, in both fiber types, AP firing first triggered protein kinase C (PKC)-dependent ClC-1 channel inhibition that reduced Gm by ∼50%. Experiments with dantrolene showed that AP-triggered SR Ca2+ release activated this PKC-mediated ClC-1 channel inhibition that was associated with reduced rheobase current and improved function of depolarized muscles, indicating that the reduced Gm enhanced muscle fiber excitability. In fast-twitch fibers, the late rise in Gm was accelerated by glucose-free conditions, whereas it was postponed when intermittent resting periods were introduced during AP firing. Remarkably, elevation of Gm was never encountered in AP-firing slow-twitch fibers, even after 15,000 APs. These observations implicate metabolic depression in the elevation of Gm in AP-firing fast-twitch fibers. It is concluded that regulation of Gm is a general phenomenon in AP-firing muscle, and that differences in Gm regulation may contribute to the different phenotypes of fast- and slow-twitch muscle.
In our companion paper (see Pedersen et al. in this issue), we presented a new technique that allows determination of Gm in action potential (AP)-firing muscle fibers with a temporal resolution of seconds. Using this approach, we identified biphasic regulation of Gm in AP-firing fast-twitch extensor digitorum longus (EDL) muscle fibers. In the first phase, Phase 1, the onset of AP firing led to a reduction in Gm that was caused by reduced resting membrane conductance for Cl− (GCl) via underlying inhibition of ClC-1 channels. The second phase, Phase 2, was initiated after prolonged AP firing and was characterized by a dramatic rise in Gm that reflected greatly elevated resting membrane conductance for K+ (GK) in combination with elevated GCl. We further showed that the ion channels underlying the elevated GK and GCl during Phase 2 were KATP and ClC-1 channels, respectively. Intriguingly, the elevations in GK and GCl occurred synchronously during Phase 2, indicating that the KATP and ClC-1 channels responded to similar cellular signaling.
In this study, we further explore Gm dynamics in AP-firing muscle fibers by comparing Gm between fast- and slow-twitch muscle fibers. This comparative approach proved useful for two reasons. First, it allowed us to assess whether regulation of Gm is a general phenomenon in skeletal muscle or only restricted to fast-twitch muscle. Thus, it was possible to evaluate whether differences in Gm dynamics in active muscle fibers could contribute to phenotypic differences between fast- and slow-twitch muscle fibers. Second, because Gm dynamics in fast-twitch muscle fibers predominantly reflected underlying regulation of ClC-1 and KATP channels and these channels are expressed in both fiber types, the comparative approach could reveal how ion channel regulation depends on the cellular environment. In particular, because slow-twitch fibers have much larger oxidative capacity than fast-twitch fibers (Jackman and Willis, 1996; Mogensen and Sahlin, 2005), the comparative approach allowed the same ion channels to be studied under different settings of cellular maintenance of metabolic status during repeated AP firing.
MATERIALS AND METHODS
Animal handling and muscle preparation
The aim of this study was to explore Gm dynamics during muscle activity in fast- and slow-twitch muscle fibers. Consequently, experiments were performed using either rat EDL muscle to represent typical fast-twitch fibers or rat soleus muscles that almost exclusively contain type I slow-twitch fibers (Edström et al., 1982). Experiments for Gm determinations used muscle from 12–14-wk-old female Wistar rats (∼230 g), whereas for contraction experiments, muscles from male or female 4–5-wk-old animals (∼70 g) were used. The handling and killing of animals followed Danish animal welfare regulations. The standard Krebs-Ringer bicarbonate solution and the solutions containing reduced or no Cl− were prepared as described in our companion paper (Pedersen et al., 2009). Solutions with 10 rather than 4 mM K+ were made by substitution of KCl for NaCl. All experiments were conducted at 30°C. GF109203X, Gö6993, N-benzyl-p-toluene sulphonamide (BTS), dantrolene, phorbol 12,13-dibutyrate (Pbdu), and blebbistatin were dissolved in DMSO. The maximal concentration of DMSO in experimental solutions was 0.15%, which did not affect resting conditions in muscles.
Determination of the passive membrane conductance in AP-firing fibers
To determine Gm in AP-firing fibers, the technique and methods of calculating GCl and GK developed in our previous study (Pedersen et al., 2009) were used. In brief, direct measurements of the input conductance in between successive AP trains were used to extract Gm in AP-firing fibers. This required that the passive membrane conductance was assessed before AP firing (GmS) in a subset of muscle fibers under the appropriate experimental conditions. Values of Gms under the various experimental conditions are shown in Table I. The protocol for current injections, data sampling, filter settings, electrode resistance, and electrode solutions (2 M K-citrate) were all identical to those in our previous study.
Blebbistatin was used to block contractile activity in soleus muscles
To measure the input conductance in between successive AP trains without encountering problems of microelectrode breakage due to contractile activity, the myosin II of the contractile proteins was inhibited using specific inhibitors. In most experiments with fast-twitch fibers, myosin II was inhibited using 50 µM BTS, which only reduces the energy consummation of active EDL muscles by ∼20% (Zhang et al., 2006). However, because BTS is specific for myosin II of fast-twitch fibers (Cheung et al., 2002; Macdonald et al., 2005), an alternate myosin II inhibitor, blebbistatin (25 µM) (Straight et al., 2003; Limouze et al., 2004), which inhibits contractions in both fast- and slow-twitch fibers (Fig. S1), was used in experiments with soleus muscles. To evaluate whether blebbistatin was an appropriate experimental tool for the present study, it was important to ensure that it did not markedly affect the excitability of the muscle fibers. Thus, a series of experiments was conducted to investigate for effects of blebbistatin on the excitability and force in both EDL and soleus muscles. In brief, these control experiments showed that blebbistatin was able to block the contractile force in whole EDL and soleus muscle (Fig. S1). It did not interfere with the resting membrane potential or GmS (Table S1). There was no change in the ionic contents of Na+ or K+ with the drug and it did not interfere with the Na+/K+-pump activity (Table S1). Blebbistain did not affect the AP peaks, nor did it in any dramatic way affect the shape of the APs (Table S1). The only consistent effect of blebbistatin was an increase in the rheobase current; i.e., the current required to be injected to elicit an AP (Table S1). To further evaluate blebbistatin as an experimental tool, a series of experiments was conducted to compare Gm dynamics in AP-firing EDL fibers that had been treated with blebbistatin with Gm dynamics in BTS-treated fibers (Fig. 1). In experiments with blebbistatin-treated fibers, a biphasic development of Gm was observed that appeared indistinguishable from observations in BTS-treated fibers. Collectively, these experiments showed that blebbistatin could be used as an experimental tool to block contractile activity during AP firing in both fast- and slow-twitch muscle fibers.
Intracellular pH (pHi) measurements
Measurements of pHi were performed using an epifluorescence system as described previously (de Paoli et al., 2007). In brief, BTS-treated EDL muscles were loaded with 20 µM 2′,7′-bis(2-carboxyethyl)-5(6)-carboxyfluorescein (BCECF-AM) for ∼30 min and then washed. Whole muscles were stimulated to fire APs using field stimulation with protocols similar to those used in the single-electrode experiments. The emission ratio of BCECF-loaded muscle was calibrated to pHi by the K+-nigericin technique (Thomas et al., 1979).
Contractions in whole muscle
Force production and M-waves were assessed in intact EDL and soleus muscles using an experimental setup that has been extensively described previously (Nielsen et al., 2001; Macdonald et al., 2005; Pedersen et al., 2005). Tetanic contractions were elicited via field stimulation at 30 V/cm for 2 (soleus) or 0.5 s (EDL) using 60-Hz trains of 0.02-ms pulses every 10 (soleus) or 20 min (EDL).
Quantification of Gm dynamics and statistics
In our companion paper (Pedersen et al., 2009), the biphasic dynamics of Gm in AP-firing fast-twitch fibers was well described by Eqs. 1 and 2. This approach for quantification was also used in the present study.
Thus, Phase 1 was described by a single-exponential function containing information on the amplitude of the reduction in Gm during Phase 1, A1, and how fast this reduction occurred, τ1. Phase 2 was described by a sigmoidal function that quantified the magnitude of the rise in Gm during Phase 2, A2, when Gm rose, τ2, and how fast Gm rose, β, during Phase 2. The dynamics of Gm in soleus fibers was reminiscent of Phase 1 in EDL fibers, and it was well quantified by Eq. 1. Statistical analysis of data was performed as described in Pedersen et al. (2009).
Chemicals and isotope
All chemicals were of analytical grade. BTS and blebbistatin were from Toronto Research Chemicals Inc., and GF109203X (bisindolylmaleimide I), Gö6976, Pbdu, and dantrolene were from Sigma-Aldrich. BCECF-AM was from Invitrogen. 86Rb+ (1,590 GBq mmol−1) was from PerkinElmer.
Online supplemental material
The online supplemental material contains the outcome from experiments where the effects of blebbistatin on the excitability of EDL and soleus muscles were evaluated. It is available at http://www.jgp.org/cgi/content/full/jgp.200910291/DC1.
Comparison of Gm dynamics in AP-firing fast- and slow-twitch fibers
In the initial series of experiments, Gm dynamics were determined in AP-firing muscle fibers from fast-twitch EDL or slow-twitch soleus muscles. Fig. 1 (B and C) shows representative recordings of the membrane potential from an EDL fiber (Fig. 1 B) and from a soleus fiber (Fig. 1 C) during experiments in which the fibers were activated to fire 3.5-s AP trains at 15 Hz repeatedly every 7 s. In the EDL fiber, the onset of AP firing was associated with an increased ΔV, and when AP firing was continued beyond ∼1,800 APs, ΔV suddenly dropped to less than half of its value before AP firing. These observations in the EDL fiber reflect the biphasic dynamics of Gm in AP-firing EDL fibers that we presented in the companion paper (Pedersen et al., 2009). In the soleus fiber, the onset of AP firing was similarly associated with increased ΔV but, in marked contrast to the observations in the EDL fiber, the drop in ΔV never appeared, even after >15,000 APs. Such experimental observations of ΔV in AP-firing fibers were next converted to Gm using Eq. 3 from Pedersen et al. (2009). Fig. 1 D shows the average Gm from EDL fibers that were stimulated to fire AP trains of either 15 or 30 Hz, and Fig. 1 E shows the dynamics of Gm in AP-firing soleus fibers that were stimulated to fire AP trains of 6, 15, 30, or 60 Hz. Fig. 1 (D and E) shows that the onset of AP firing was associated with substantial reductions in Gm in both muscle fiber types and at all frequencies. When comparing these initial reductions of Gm between fast- and slow-twitch fibers, the reductions appeared to be of similar magnitudes, but they developed substantially faster in EDL fibers than in soleus fibers. Indeed, when the reduction in Gm at the onset of AP firing was fitted to Eq. 1, the magnitudes of the Gm reductions, A1, reached 45–70% and 43–55% of Gm at the start of the experiments in soleus and EDL fibers, respectively, whereas the time required for the development of the reductions, τ1, was substantially longer in the slow-twitch fibers (60 ± 9 and 76 ± 9 s at 15 and 30 Hz, respectively; Table II) than in the fast-twitch fibers (32 ± 4 and 17 ± 1 s at 15 and 30 Hz, respectively; Table II). However, the most striking difference between fast- and slow-twitch muscle fibers was the complete absence of a rise in Gm with prolonged AP firing in slow-twitch fibers. Thus, comparing Fig. 1 (D and E) also shows that in clear contrast to the massive rise in Gm after prolonged AP firing in EDL fibers (Phase 2), Gm remained low throughout all experiments and at all frequencies in the soleus fibers, even if the AP firing was continued for much longer.
Note also that the observations in the EDL fibers presented in Fig. 1 D were obtained using blebbistatin as a myosin II inhibitor. These very closely replicated the observations in our previous study, where BTS rather than blebbistatin was used to inhibit contractile activity during AP firing. Thus, none of the parameters used to quantify Phase 1 (A1 and τ1) and Phase 2 (A2, τ2, and β) were different between EDL fibers treated with the two types of myosin II inhibitors (Table II).
Ion channels involved in Gm dynamics in AP-firing slow-twitch fibers
To explore which ion channels that were involved in the reduction in Gm in AP-firing soleus fibers, GK and GCl were determined using the approach described in our previous study (Pedersen et al., 2009). This required experimental observation of Gm dynamics at two extracellular Cl− concentrations (Eq. 4 in Pedersen et al., 2009). Thus, Fig. 2 A shows the dynamics of Gm at 50 and 127 mM Cl− with an AP-firing frequency of 30 Hz. Two major effects of reduced Cl− were apparent. First, the resting membrane conductance before AP firing was reduced substantially (Table I). Second, AP firing was associated with an initial increase in Gm, followed by a slight reduction in Gm, whereas at 127 mM Cl−, Gm became reduced from the onset of AP firing. This indicated that the reduction of Gm in AP-firing soleus fibers could largely be ascribed to reduced GCl. This notion was confirmed by calculations of GK and GCl (Fig. 2 B), which demonstrated that the reduced Gm in AP-firing soleus fibers reflected the net outcome of a pronounced reduction in GCl and a slight increase in GK. Similar changes in GK and GCl were observed in soleus fibers stimulated with AP trains of 6 and 60 Hz (not depicted). Because ClC-1 channels are the major Cl− channel in muscle (Koch et al., 1992; Lueck et al., 2007) this shows that, similar to the situation in EDL fibers (Pedersen et al., 2009), the onset of AP firing in soleus fibers leads to ClC-1 channel inhibition and a minor activation of K+ channels.
Cellular mechanism underlying ClC-1 inhibition at the onset of AP firing
The above experiments showed that inhibition of ClC-1 channels at the onset of AP firing was common to both fiber types. It has previously been demonstrated experimentally that the resting membrane conductance for Cl− can be reduced in muscle fibers either by acidosis (Hutter and Warner, 1967; Pedersen et al., 2004, 2005) or by activation of PKC (Tricarico et al., 1991; Rosenbohm et al., 1999; Pierno et al., 2007; Dutka et al., 2008).
First, to explore for a role of PKC in the ClC-1 channel inhibition at the onset of AP firing, experiments were conducted in which the muscles were pretreated with a PKC inhibitor, GF109203X (Toullec et al., 1991). Under such conditions, the reduction in Gm at the onset of AP firing was largely abolished in both fiber types. Thus, Fig. 3 A shows observations from an EDL muscle fiber pretreated with the inhibitor during the first, the 22nd, and the 70th AP train. The enlarged recordings of ΔV below the AP trains show that the increase in ΔV that was usually observed at the onset of AP firing was completely abolished by the inhibitor, whereas the drop in ΔV during prolonged AP firing was still present with PKC inhibition. Fig. 3 B shows that the increase in ΔV at the onset of AP firing could also be abolished by the PKC inhibitor in a soleus fiber. Fig. 3 (C and D) presents the average Gm with the inhibitor in 11 EDL fibers (Fig. 3 C) and in 5 soleus fibers (Fig. 3 D). In the soleus fibers, the PKC inhibitor completely abolished the reduction in Gm at the onset of AP firing, whereas in the EDL fibers, a minor reduction of Gm that reached 81 ± 4% of the Gm at the start of the experiments in EDL could still be detected with the inhibitor. This reduction in EDL fibers was, however, much less than the reduction of Gm in control fibers, where it became reduced to 45 ± 4% of its value at the start of the experiment. Furthermore, the reduction in Gm in the EDL fibers with the PKC inhibitor appeared to occur later than the reduction in control fibers and, accordingly, Gm in the fibers treated with the PKC inhibitor could not be quantified by fitting to Eq. 1. In contrast, to the effect of PKC inhibition on the reduction of Gm at the onset of AP firing, the PKC-inhibited EDL fibers developed a pronounced increase in Gm with prolonged AP firing that was very similar to the increased Gm during Phase 2 in control fibers. In an additional series of experiments with four EDL fibers, which had been pretreated with an alternate PKC inhibitor, Gö6976, Gm was only reduced to 79 ± 11% of the Gm at the start of the experiment during Phase 1, whereas a pronounced increase of Gm was also observed during Phase 2 in these fibers.
In our previous study (Pedersen et al., 2009), the ClC-1 channel inhibition during Phase 1 in EDL fibers was demonstrated to be independent of whether the APs were triggered using inserted electrodes or whether the APs were triggered via the more physiological approach of motor nerve stimulation. To ensure that PKC activation was also activated during such motor nerve stimulation, whole muscles were stimulated to fire 686 APs, after which the two electrodes were rapidly inserted (<1 min) to monitor the recovery of Gin after AP firing. Although muscle contractions were clearly visible during the motor nerve stimulation, Fig. 5 E shows that after the AP firing, Gin was still high and did not display any change throughout 7 min of recovery in the fibers treated with the PKC inhibitor. This contrasted the rise in Gin that was observed after identical motor nerve stimulation in muscles without the inhibitor (Fig. 3 E).
Second, despite the very clear involvement of PKC in the fast reduction of Gm at the onset of AP firing (Fig. 3), it was possible that acidosis played a role for activating PKC or was involved in the delayed reduction of Gm during Phase 1 in the EDL fibers treated with the PKC inhibitor. Thus, to evaluate the role of acidosis for the reduction in Gm at the onset of AP firing, recordings of pHi in whole, field-stimulated BTS-treated EDL muscle was compared with the observed Gm dynamics during Phase 1. Fig. 4 A shows that the development of pHi had similar magnitudes when muscles were stimulated using AP trains of 6, 15, or 30 Hz. Also, the changes in pHi appeared to depend on the number of APs fired rather than on the duration of the experiment (Fig. 4 A). Hence, during the first 500 APs, pHi either rose a little or remained constant, and then it began to drop. The lowest pHi occurred after ∼1,500 APs. In Fig. 4 (B–D), the dynamics of pHi and Gm have been plotted against time for observations at 6 (Fig. 4 B), 15 (Fig. 4 C), and 30 Hz (Fig. 4 D). Comparing the dynamics of pHi and Gm shows that although pHi did not change markedly during the first 500 APs, this was actually the period when the largest decline in Gm occurred. This strongly indicated that the PKC-mediated ClC-1 channel inhibition at the onset of AP firing was not triggered by acidification. However, in the EDL fibers exposed to PKC inhibition, a minor reduction in Gm during the early stages of AP firing was still observed (Fig. 3 C). When the dynamics in pHi during AP firing were compared with the dynamics of Gm in the AP-firing fibers treated with the PKC inhibitor, a close temporal correlation was observed (Fig. 3 E). Thus, although the majority of the reduction in Gm at the onset of AP firing was caused by PKC-mediated ClC-1 channel inhibition, the close temporal correlation between the dynamics of pHi and Gm in PKC-inhibited EDL fibers indicates that acidification may have imposed a minor inhibitory effect on the ClC-1 channels.
The three groups of PKC isoforms that have thus far been discovered are distinguished by the cellular signals that convey their activation (Steinberg, 2008). Supposedly, the PKC inhibitors GF109203X and Gö6976 have the highest affinity for conventional PKC isoforms (Martiny-Baron et al., 1993), which require the presence of diacylglycerol as well as a rise in the free cytosolic Ca2+ for activation (Steinberg, 2008). To test whether the release of Ca2+ from the SR could be involved in activating the PKC-mediated ClC-1 channel inhibition, at the onset of AP firing a series of experiments was conducted in EDL fibers in which the Ca2+ release channels of the SR had been partially blocked by 25 µM dantrolene. In these fibers, the biphasic Gm dynamics that were usually observed in AP-firing EDL fibers were still observed (Fig. 4 F). However, with dantrolene, the reduction in Gm during Phase 1 was slower and less pronounced. Accordingly, the fitting of the observation to Eqs. 1 and 2 showed that τ1 of Phase 1, which reflects how fast the PKC-mediated ClC-1 channel inhibition proceeded, was markedly increased with dantrolene, whereas A1, which reflects the magnitude of the reduction of Gm, was not significantly affected (Table III). Hence, the primary effect of dantrolene was to slow the ClC-1 channel inhibition during Phase 1 and not to abolish it as occurred with the PKC inhibition.
Cellular mechanism underlying the activation of ClC-1 and KATP channels during Phase 2
An intriguing difference in the Gm dynamics between EDL and soleus fibers was the complete absence of Phase 2 in soleus fibers (Fig. 1). Our previous study (Pedersen et al., 2009) proposed that Phase 2 arose when the fibers reached a critical level of metabolic depression. This hypothesis is supported by the absence of Phase 2 in soleus fibers (Fig. 1 E) because these fibers have a higher content of mitochondria and a larger oxidative capacity than fast-twitch fibers (Jackman and Willis, 1996; Mogensen and Sahlin, 2005). The absence of elevated Gm with prolonged AP firing does not exclude that Gm can rise in these fibers because in experiments with carbonyl cyanide m-chlorophenylhydrazone, an uncoupler of mitochondria, the input conductance of resting soleus fibers increased immensely (not depicted). Still, to further evaluate a role of the metabolic state in the etiology of Phase 2, two more series of experiments were conducted in EDL fibers.
First, experiments were performed in which 6 min of rest were introduced after every 20th train of AP. The average observations from three such experiments are presented in Fig. 5. These observations show that by introducing resting periods of 6 min, Gm completely recovered after every set of 20 AP trains, and upon resumed AP firing, Phase 1 reappeared. Moreover, when compared with experiments without the intermittent resting periods, the appearance of Phase 2 became substantially delayed.
Second, a series of experiment was conducted with EDL fibers in glucose-free extracellular solution. Fig. 6 A shows that at the onset of AP firing, Gm was reduced to a similar extent in the fibers in glucose-free and control conditions. Indeed, quantification showed that the magnitude of the reduction during Phase 1, A1, was similar in the two situations, and that the time required for this reduction, τ1, was also similar in the control fibers and in the fibers under glucose-free conditions (Table III). However, with prolonged AP firing, the fibers under glucose-free conditions entered Phase 2 before the control fibers, and Gm became substantially higher during Phase 2 in the fibers under glucose-free conditions than in the control fibers. Thus, quantification of Phase 2 with Eq. 2 showed that under glucose-free conditions, Phase 2 arose faster (τ2 was 2,417 ± 127 APs and 2,127 ± 163 APs in control and in glucose-free conditions, respectively; P < 0.05) and reached a higher level than the in control fibers (A2 was 5,426 ± 494 µS/cm2 and 11,056 ± 1,143 µS/cm2 in control and in glucose-free conditions, respectively; P < 0.05). Additionally, after 1 min of recovery from firing 4,998 APs, Gm only recovered partly in the absence of glucose, whereas a more complete recovery of Gm was observed in control conditions (Fig. 6 B).
Effect of Gm dynamics for the excitability of muscle
The physiological importance of the PKC-mediated ClC-1 channel inhibition for muscle excitability was next considered using two different experimental approaches.
First, the effect of PKC inhibition on muscle excitability was evaluated. Fig. 7 (A and B) demonstrates that if AP firing was ceased during the initial reduction in Gm, it lasted ∼5 min for full recovery of Gm to be reached in both fiber types under control conditions. In contrast, neither EDL fibers nor soleus fibers showed any change in Gm during AP firing or during the subsequent recovery period if the fibers were pretreated with the PKC inhibitor. This slow recovery from the PKC inhibition of ClC-1 channels during Phase 1 in the control fibers was exploited to evaluate the role of ClC-1 channel inhibition for the fiber excitability by comparing the rheobase currents before and 1 min after AP firing in control fibers and in fibers with the PKC inhibitor. Fig. 7 C shows experimental recordings from a representative EDL fiber under control conditions, whereas Fig. 7 D shows recordings from a representative EDL fiber exposed to the PKC inhibitor. These recordings illustrate that when the rheobase current was assessed 1 min after AP firing, a marked reduction, indicative of increased excitability, was observed under control conditions. In contrast, in the fiber with the PKC inhibitor, the rheobase was substantially elevated after AP firing. Fig. 7 E shows the average rheobase currents before and after AP firing under control conditions and with PKC inhibition in both EDL and soleus fibers. In both fiber types, the rheobase was reduced after the short duration of AP firing under control conditions, whereas increased rheobase current was observed if PKC was inhibited. Note that similar changes in the resting membrane potential were observed after AP firing in the absence and presence of the PKC inhibitor in the two muscle fiber types (Fig. 7 F).
Second, the effect of pharmacological PKC activation for muscle fiber excitability was assessed in muscles at elevated extracellular K+, which repeatedly has been used to experimentally compromise muscle excitability and contractile force (Nielsen et al., 2001; Pedersen et al., 2005). Fig. 8 shows representative recordings of M-waves (Fig. 8 A) and tetanic force (Fig. 8 B) from an intact soleus muscle when stimulated every 10 min with a 2-s long 60-Hz train. When extracellular K+ was elevated from 4 to 10 mM, a pronounced decline in both M-waves (Fig. 8 A) and force (Fig. 8 B) took place. After 1 h at elevated extracellular K+, PKC was activated by the addition of 5 µM of PKC activator, Pbdu, which resulted in a pronounced recovery of both the excitability (M-waves) and force. In four similar experiments, the addition of Pbdu to muscles at 10 mM K+ caused the force to increase from 43 ± 5 to 76 ± 12% of the control force at 4 mM K+. Similar results were obtained in two EDL muscles at 13 mM K+ (not depicted).
This study demonstrates that in active, AP-firing skeletal muscle, the ion channels that determine Gm undergo substantial regulation in both fast- and slow-twitch muscle fibers. We show that in both fiber types, ClC-1 channels are rapidly inhibited at the onset of AP firing via a PKC-dependent pathway that becomes activated via AP-mediated SR Ca2+ release. Such ClC-1 channel inhibition increases the excitability of the fibers. We further show that the synchronous openings of ClC-1 and KATP channels with prolonged AP firing in EDL fast-twitch fibers do not occur in slow-twitch soleus fibers. We propose that a significant reduction in the metabolic state of the fibers may induce such ion channel activations in fast-twitch fibers and thereby support recent suggestions that ClC-1 channels are sensitive to the metabolic state of muscle fibers (Bennetts et al., 2005; Tseng et al., 2007; Zhang et al., 2008).
Increase in Gm after prolonged AP firing only occurs in fast-twitch fibers
In slow-twitch fibers, the onset of AP firing was associated with a reduction in Gm that, despite developing at a slower rate, was reminiscent of the ClC-1 channel inhibition that caused a reduction in Gm during Phase 1 in the fast-twitch fibers (Pedersen et al., 2009). Further support for this notion was provided by the observation that in both fiber types, the reduction in Gm at the onset of AP firing primarily reflected a pronounced reduction in GCl. Because the ClC-1 channel is generally considered to be the main Cl− channel in muscle (Koch et al., 1992), and because GCl was reduced by as much as 70–80% at the onset of AP firing, reduced GCl must have also reflected ClC-1 channel inhibition in soleus fibers. Collectively, these observations show that the onset of muscle activity triggers a pronounced reduction in GCl that reflects underlying inhibition of ClC-1 channels in both fast- and slow-twitch muscle fibers.
In marked contrast to the similar Gm dynamics in the two fiber types during early stages of muscle activity, the two fiber types displayed strikingly different Gm dynamics after prolonged AP firing. Thus, in contrast to the development of a very large Gm after firing more than ∼1,800 APs in EDL fibers, Gm in slow-twitch fibers remained low even after firing as many as 15,000 APs. If Phase 2 reflected a reduced metabolic state of the EDL fibers, as it will be argued below, it can be speculated that Phase 2 did not appear in slow-twitch fibers because these fibers have a larger oxidative capacity and hence would have a better preservation of their energetic state during repeated AP firing. Indeed, when the cellular metabolic state is evaluated after intensive contractions in muscles containing both fast- and slow-twitch fibers, the reduction in ATP and elevation in IMP are much more pronounced in the fast-twitch fibers as compared with slow-twitch fibers (Esbjörnsson-Liljedahl et al., 1999).
SR Ca2+ release triggers PKC-mediated ClC-1 channel inhibition at the onset of AP firing
Previous studies have demonstrated that ClC-1 channels can be inhibited by acidification and via intracellular pathways involving PKC (Hutter and Warner, 1967; Tricarico et al., 1991; Rosenbohm et al., 1999; Pedersen et al., 2004, 2005; Pierno et al., 2007; Dutka et al., 2008). Hence, to explore for a role of pHi in ClC-1 channel inhibition at the onset of AP firing, the dynamics of pHi in whole EDL muscle were measured during field stimulation and compared with the corresponding Gm dynamics. This showed that the majority of the ClC-1 channel inhibition occurred while the fibers were in an alkaline state, presumably caused by breakdown of creatine phosphate. Hence, the rapid ClC-1 channel inhibition at the onset of AP firing was not caused by acidification. In contrast, in both muscle fibers, the initial inhibition of ClC-1 channels during AP firing was largely abolished by PKC inhibition. The ClC-1 channel inhibition could also be partly prevented by blockage of SR Ca2+ channels with dantrolene. This indicates that Ca2+ release from SR during AP firing was important for activation of PKC. Such interactions between increased cytosolic Ca2+ and PKC for ClC-1 channel inhibition implicate the conventional PKC isoform in the inhibition, and this is supported by a recent study, which suggests that statin treatment of rats causes a chronic elevation in cytosolic Ca2+ that leads to PKC-mediated ClC-1 channel inhibition and hyperexcitability (Pierno et al., 2009). Because the reduction in Gm at the early stages of AP firing was well described by a single-exponential function in both muscles, it is likely that there was one rate-limiting step in the ClC-1 channel inhibition. Because dantrolene slowed down the process of ClC-1 inhibition at the onset of AP firing, it can be speculated that the rise in cytosolic Ca2+ represents this rate-limiting step in PKC-mediated ClC-1 inhibition in active muscle. Our observations demonstrate that the modulation of ClC-1 channels by statins (Pierno et al., 2009) that act via an elevated cytosolic Ca2+ reflects a physiological mechanism for inhibition of ClC-1 channels that occurs in active muscles. Thus, as summarized in Fig. 9, our findings demonstrate that during muscle activity, SR Ca2+ release triggered by AP firing activates PKC, which then inhibits the ClC-1 channels causing Gm to drop.
In both fiber types, full recovery from ClC-1 channel inhibition was observed after ∼5 min. It is interesting that this was substantially longer than the time required for recovery of Gm after Phase 2 in the EDL fibers. This indicated that in fast-twitch fibers, the PKC-mediated inhibition of ClC-1 channels during Phase 1 disappeared once Phase 2 was initiated. The role of PKC for ClC-1 channel inhibition also implies that some proteins, possibly the ClC-1 channels, became phosphorylated. Thus, as it will be argued below, if Phase 2 reflected a substantial reduction in the metabolic state of the fibers, it is possible that dephosphorylation becomes predominant rather than PKC-mediated phosphorylation, when the ATP levels have declined substantially.
Further evidence of reduced metabolic state being the trigger for Phase 2
In our companion paper (Pedersen et al., 2009), it was demonstrated that the elevated Gm during Phase 2 in fast-twitch fibers reflected the opening of ClC-1 and KATP channels. In that study, it was hypothesized that the opening of these channels during Phase 2 reflected a substantial reduction in the metabolic state of the fibers. Three lines of evidence in favor of this hypothesis have been shown in the present study. First, the absence of Phase 2 in slow-twitch fibers supports the hypothesis that Phase 2 occurred in fast-twitch fibers when their metabolic state reached a critical level of depression. This support comes from the observations that slow-twitch fibers show less reduction in their cellular energetic state during prolonged contractile activity when compared with fast-twitch fibers (Esbjörnsson-Liljedahl et al., 1999). Second, a role of the metabolic state of the fibers for the induction of Phase 2 was implied by the early appearance and larger magnitude of Phase 2 in EDL fibers that were incubated under glucose-free conditions. These fibers were likely to have a moderately reduced metabolic state even before the AP firing and consequently would reach the level of metabolic depression required for the induction of Phase 2 faster than control fibers. Third, when resting periods of 6 min were introduced after every 20th train of AP, the appearance of Phase 2 was substantially postponed. This suggests that by introducing resting periods, the fibers were recovering their metabolic state in the resting periods sufficiently to delay the onset of Phase 2. Collectively, Phase 2 was accelerated if the muscles were exposed to either glucose-free conditions or prestimulated with field stimulation (Fig. S3 in Pedersen et al., 2009), it was delayed when AP trains had the lowest frequency (6 Hz) and when intermittent resting periods were introduced during the AP firing, and, finally, it was completely absent when explored in soleus fibers that have a larger oxidative capacity and show a better maintenance of the energetic state during activity (Esbjörnsson-Liljedahl et al., 1999). These findings can all be explained by the metabolic state of the fibers being a key determinant for the induction of Phase 2.
Functional significance of Gm dynamics for muscle function
Generally, the resting membrane conductance of muscle fibers and their excitability appear to be inversely related. Accordingly, the PKC-mediated ClC-1 channel inhibition at the onset of AP firing was associated with enhanced muscle fiber excitability, as indicated by a reduction in rheobase current after short-duration AP firing. Note that this effect could not be related to changes in the resting membrane potential because this was not affected by the PKC inhibitor before or after AP firing. Also, in experiments where the muscle excitability was initially depressed by exposing whole muscles to elevated extracellular K+, PKC activation with a phorbol ester markedly recovered the excitability. Using a similar experimental approach of elevated extracellular K+, we have previously shown that ClC-1 channel inhibition either via acidification or directly by the addition of 9-AC can recover muscle excitability in both fast- and slow-twitch muscles (Nielsen et al., 2001; Pedersen et al., 2004, 2005). Thus, the addition of the phorbol ester, which is known to reduce the resting GCl in rat muscle (Tricarico et al., 1991), is likely to have mediated the recovery of excitability and force in K+ depressed via inhibition of ClC-1 channels, although other mechanisms may contribute.
Collectively, these observations suggest a scenario in which at the onset of muscle activity, AP-mediated SR Ca2+ release leads to an increased PKC activity, which in turn inhibits the ClC-1 channels and thereby enhances the excitability of the working muscle (Fig. 9). Conversely, the openings of ClC-1 and KATP channels during Phase 2, which was exclusive for fast-twitch fibers, resulted in increased Gm and dropout of APs (Fig. 9). It can be speculated that such a mechanism may contribute to the larger fatigability of fast-twitch fibers, and could serve as a protective mechanism that, by hindering prolonged muscle activation, limits metabolic depletion and excessive fiber damage during exercise. This notion is supported by observations of severe muscle damage in KATP channel knockout mice after treadmill running (Thabet et al., 2005; Cifelli et al., 2007).
In combination with our previous study (Pedersen et al., 2009), our findings demonstrate for the first time that acute regulation of Gm is a general phenomenon in active muscle. Our studies also show that this Gm regulation results from changes in the functions of ClC-1 and KATP channels and, thereby, reveal new physiological aspects on ClC-1 and KATP channels in skeletal muscles. In so doing, our observations complement several previous studies that have focused on specific pathways for ion channel regulation. Our approach allows for the determination of whether and when Gm is regulated in AP-firing fibers and which cellular signaling pathways that are involved. Because the observed changes in Gm had significant effects on muscle fiber excitability, our observations indicate that the excitability of skeletal muscles is far more dynamically regulated during muscle activity than previously expected.
Furthermore, the regulation of Gm differs significantly between fast- and slow-twitch fibers, with fast-twitch fibers being much more prone to a regulated reduction in excitability via massively elevated Gm. This fits with slow-twitch fibers often being involved in low intensity but prolonged contractile activity in the intact organism, whereas fast-twitch fibers usually are involved in intensive contractions of short duration. As such, the observed regulation of Gm may contribute to these well-established differences in the phenotypes of fast- and slow-twitch fibers.
We thank T.L. Andersen, V. Uhre, and M. Stürup-Johansen for technical assistance. Dr. J.A. Flatman is acknowledged for helpful discussions of the experiments and for assistance in writing the manuscript. Professor Christian Ålkjær is acknowledged for technical assistance in experiments performed to determine pHi.
This work was supported by The Danish Research Medical Council (to T.H. Pedersen and O.B. Nielsen) and the Faculty of Health Science, University of Århus (to T.H. Pedersen and F. de Paoli).
Christopher Miller served as editor.
extensor digitorum longus