We examined 1) whether the effects of lowered trans-sarcolemmal Na+ gradient on force differed between nonfatigued fast- and slow-twitch muscles of mice and 2) whether effects on action potentials could explain the decrease of force. The Na+ gradient was reduced by lowering the extracellular [Na+] ([Na+]o). The peak force-[Na+]o relationships for the twitch and tetanus were the same in nonfatigued extensor digitorum longus and soleus muscles: force was maintained over a large range of [Na+]o and then decreased abruptly over a much smaller range. However, fatigue was significantly exacerbated at a lowered [Na+]o that had little effect in nonfatigued soleus muscle. This finding suggests that substantial differences exist in the Na+ effect on force between nonfatigued and fatigued muscle. The reduced contractility in nonfatigued muscles at lowered [Na+]o was largely due to 1) an increased number of inexcitable fibers and threshold for action potentials, 2) a reduction of action potential amplitude, and 3) a reduced capacity to generate action potentials throughout trains.
- sodium gradient
- muscle contraction
- action potential train
- extensor digitorum longus
during repetitive activation of skeletal muscle, a net influx of sodium ions (Na+) occurs down their electrochemical gradient during each action potential (13, 25, 29). The plasma [Na+] is usually unchanged or undergoes only a small increase during exercise (24-26). The intracellular [Na+] ([Na+]i), on the other hand, increases during high-intensity exercise or intense electrical stimulation because the Na+-K+ pump is unable to extrude all of the Na+ that has entered the muscle fiber (13, 25). During fatiguing exercise in humans, [Na+]i increases from 6-13 to 11-24 mM (24, 26), and during fatiguing electrical stimulation, [Na+]i of mammalian muscles increases from 8-14 to 19-58 mM (14, 17, 19, 27). Moreover, it has been postulated that the Na+ that leaves the lumen of the transverse (t) tubules and moves into the myoplasm is not readily replaced by diffusion from the interstitial space, leading to a depletion of t-tubular Na+ (4-6, 12, 16, 22). Overall, the increase in intracellular and decrease in t-tubular [Na+] results in a rundown of the transsarcolemmal Na+ gradient, which has been suggested to cause fatigue, i.e., a reduced muscle force production. This is known as the Na+ hypothesis for fatigue.
Several lines of evidence support the Na+ hypothesis for fatigue. Action potential failure in the t tubules is regarded as an important mechanism for fatigue during prolonged tetani (6, 12, 16, 30). Indirect evidence suggests that reduced extracellular [Na+] ([Na+]o) impairs action potential conduction in t-tubular membranes (4, 12, 20). Experimentally reducing [Na+]o lowers force in nonfatigued amphibian (4, 5, 20) and mammalian skeletal muscle (6, 8, 12, 16, 22). Lowered [Na+]o exacerbates fatigue during continuous tetanic stimulation in mammalian muscle (6, 16).
One major argument against Na+ as an important factor in fatigue is that nonphysiologically large reductions in [Na+]o are required to reduce force in nonfatigued muscle (5, 30). Here this issue is addressed from the perspective of the effect of lowered [Na+]o/[Na+]i ratio on force production. An inward Na+ current generates the upstroke of the muscle action potential (10, 25, 29), and the driving force for the Na+ current is the difference between the membrane potential (Em) and the Na+ equilibrium potential (ENa), where ENa = RT/F·loge ([Na+]o/[Na+]i).
Considering that the gas constant (R) and the Faraday constant (F) are fixed values, then at a given temperature (T) the determinant of ENa is the ratio [Na+]o/[Na+]i and not merely [Na+]o. Hence, the [Na+]o/[Na+]i ratio is likely to be important in fatigue. Therefore, when examining the influence of [Na+]o on force in nonfatigued muscle, values consistent with the [Na+]o/[Na+]i ratio measured during fatigue must be used. For example, when the [Na+]i is doubled during fatigue (17, 26), an equal reduction of the [Na+]o/[Na+]i in nonfatigued muscle would require a reduction of [Na+]o by one half.
Fast-twitch muscle has lower fatigue resistance than slow-twitch muscle. Furthermore, [Na+]i is reported to increase relatively more during repeated contractions in fast- than in slow-twitch muscle (13, 19, 27). Thus the lower fatigue resistance of fast-twitch muscle may be due to higher [Na+]i. However, the effects of a lowered transsarcolemmal Na+ gradient on force have been quantified only in slow-twitch mammalian muscle (6, 8, 16, 22), and only one study determined the peak tetanic force-[Na+]o relationship (22). The possibility that lowered [Na+]o has greater effects in fast- than in slow-twitch muscle thus remains. Therefore, a principal aim was to quantify and determine whether the effects of lowered Na+ gradient on force differ between fast- and slow-twitch muscles. To do this, we used the fast-twitch mouse extensor digitorum longus (EDL), which is composed primarily of type IIB (68%) and IIX (20%) fibers (23), and the slow-twitch mouse soleus, which is composed of type I (40%) and IIA (60%) fibers (31).
It is also of considerable interest to establish the mechanisms by which lowered [Na+]o impairs the contractility of nonfatigued muscle. It is well established that lowered [Na+]o reduces the amplitude of the action potential (1, 4, 10, 21). However, it is not fully understood whether this Na+ effect influences force. It has often been suggested that any decrease in action potential amplitude reduces the amount of Ca2+ released by the sarcoplasmic reticulum. However, using tetrodotoxin, Yensen et al. (32) recently showed that the peak twitch force of mouse EDL does not decrease until the overshoot becomes <10 mV. It remains to be determined whether the same relationship can be observed when the Na+ gradient is reduced. Furthermore, to understand fully how lowered Na+ gradient affects tetanic contractions, it is important to extend the study to trains of action potentials. Although this has been done in amphibian muscle (4), similar experiments have not been performed in mammalian muscle. Therefore, another major aim was to investigate whether effects of lowered Na+ gradient on both single action potentials and trains of action potentials could explain the reduction of force in nonfatigued muscle.
MATERIALS AND METHODS
The Animal Ethics Committee of The University of Auckland and the Animal Care Committee of the University of Ottawa approved the protocol and animal use for this project. Experiments were performed on adult mice (Swiss CD-1) of either sex (20-30 g body wt). Animals were either killed by cervical dislocation or anesthetized with sodium pentobarbital (Somnotol; intraperitoneal injection 0.8 mg/10 g body wt). Intact soleus and EDL muscles were dissected in the control solution, which was continuously gassed with carbogen (95% O2-5% CO2) at room temperature (21-23°C).
Solutions and Chemicals
The control solution contained 147 mM Na+, which is close to the [Na+] in the extracellular fluid of mammals (24-26). The solution was composed of (in mM) 122.2 NaCl, 25.1 NaHCO3, 2.8 KCl, 1.2 KH2PO4, 1.2 MgSO4, 1.3 CaCl2, and 5 d-glucose and was equilibrated with carbogen (pH 7.4). Low-[Na+] (30-120 mM) solutions were prepared by replacing NaCl with equimolar choline chloride (Sigma, Oakville, Canada) or N-methyl-d-glucamine (NMDG) (Sigma); when necessary, pH was adjusted to 7.4 with either hydrochloric or sulfuric acid. Choline and NMDG are well-known substitutes for Na+ (4, 5, 10, 12, 22). The [Na+] in the solutions was checked by flame photometry. The experimental temperature for all measurements was 25°C.
Contractions were elicited by electrical stimulation, using either platinum plates or wire electrodes when action potentials were measured. The platinum plates (2 × 5 cm), hereafter referred to as plate electrodes, were positioned on either side of the muscle, 0.77 cm apart. Bipolar stimulation pulses were computer generated and delivered to the plate electrodes via a purpose-built power amplifier (MOSFET). Twitch contractions were elicited with supramaximal pulses (0.1 ms, 28 V/cm), unless specified otherwise. Tetanic contractions were elicited with a train of the same pulses at 125 Hz for 2 s for soleus and at 200 Hz for 0.5 s for EDL. The platinum wires, hereafter referred to as wire electrodes (1.0 mm in diameter, 5 mm long) transversed the muscle in a plane above and below the muscle. For wire electrodes, supramaximal stimulation pulses (0.3 ms, 10 V) were generated with a Grass S88 stimulator and Grass SIU5 isolation unit (Grass, West Warwick, RI). Addition of 30 μM curare, a neuromuscular transmission blocker, had no effect on peak tetanic force in EDL and soleus muscles at 147 mM Na+ when either plate or wire electrodes were used, i.e., all muscle fibers were activated directly rather than by nerve terminals.
Experiments were initially carried out by using the plate electrodes and later using the wire electrodes to prevent artifacts with the measurements of action potentials. The dimension and position of plate electrodes ensured that muscles were located in a uniform electrical field, which presumably triggers an action potential simultaneously at all sites along the length of the surface membrane. Wire electrodes presumably trigger an action potential over a small area of membrane that then propagates along the entire surface membrane. Some effects of lowered [Na+]o differed significantly with plate and wire electrode stimulation, indicating an effect on action potential propagation along the surface membrane.
When plate electrodes were used, isometric force alone was measured. Muscles were suspended vertically by their tendons in a muscle chamber containing 100 ml of solution, which was oxygenated with carbogen bubbled from the bottom of the chamber. The temperature was set at 25°C by immersing the chamber in a temperature-controlled water bath. The force transducer consisted of a horizontal magnesium alloy bar on which were bonded, to both the upper and lower surfaces, semiconductor strain gauges (KSP-2-E3, Kyowa, Japan). Twitch and tetanic contractions were continuously monitored on a chart recorder (Gould model 2400, Valley View, Ohio) and recorded (at 400 Hz) on a computer using an analog-to-digital (A/D) system consisting of an input/output board (NB-MOI-16), a 32-bit data acquisition board (NB-DMA-2800), and custom-written Labview software (National Instruments, Austin, TX).
When wire electrodes were used, force and membrane potentials were measured from the same muscles. Muscles were suspended horizontally in a 2-ml chamber and attached to a Cambridge force transducer (model 300; Boston, MA). A one-way flow-through system at 15 ml/min permitted oxygenated solution, preheated to 25°C, to enter the muscle chamber. Twitch and tetanic contractions were continuously monitored on a chart recorder, and selected contractions were recorded on a computer using a Keithley Metrabyte A-D board (model DAS-50; Keithley, Edmonton, Canada). The sampling rate was 20 kHz for the twitch and 2 kHz for the tetanus.
For all experiments, peak twitch and tetanic forces were calculated as the difference between the maximum force during a contraction and the baseline (which was constant throughout all experiments). All muscles were stretched to the length that gave peak tetanic force, which on average was 141 ± 3 mN (n = 45) for soleus and 203 ± 11 mN (n = 10) for EDL (plate electrodes). Similar forces were obtained in soleus with wire electrodes. Fade, defined as the decline of force within a tetanic contraction, was calculated as the force after 500 ms for EDL or 2 s for soleus and was expressed as a percentage of the peak tetanic force.
Effect of lowered [Na+]o. In nonfatigued muscle, lowered [Cl-]o has only a small transient effect on resting membrane potential (11) or peak tetanic force (9). Hence, for the initial recording of tetanic contractions (see Fig. 2A), either choline chloride or NMDG were used as a substitute for Na+. However, the recovery of peak tetanic force of both EDL and soleus muscles upon return from lowered [Na+]o to the control solution was faster with NMDG (5-10 min) than with choline chloride (30-40 min). Furthermore, in EDL muscle, force recovery was usually incomplete when choline chloride was used (data not shown). Hence, as reported for amphibian muscle (5), NMDG appears to be a better Na+ substitute than choline chloride for mammalian muscle. Thus NMDG was the Na+ substitute used for all remaining experiments, except for those with fatigued muscle.
Contractions were recorded in the control solution while tetani were evoked every 5 min until a steady-state force was achieved. Muscles were then exposed to either one or up to six lowered [Na+]o at random before being exposed again to control solution to test whether the intervention was reversible. The control force was obtained by correcting for “force rundown” by interpolating between the peak force at 147 mM Na+ immediately before, and with maximum recovery after, exposure to lowered [Na+]o. Data from experiments with a rapid (>0.2% min-1) rundown of peak tetanic force after exposure to lowered [Na+]o were eliminated.
Fatigue test. For these experiments, paired muscles were used from the same animal such that one muscle was tested at 147 mM Na+ and the contralateral muscle at 80 mM Na+, because peak tetanic force did not fully recover after fatigue at 80 mM Na+ as it did at 147 mM Na+ (data not shown). At 80 mM Na+, sodium chloride was replaced only by choline chloride because NMDG lacks Cl- and reduced extracellular [Cl-] ([Cl-]o) influences fatigue (6, 7, 9, 25). Fatigue was induced using intermittent tetanic stimulation at 125 Hz: 100 tetani of 500 ms in duration were evoked at a rate of one per second.
Action Potential Measurements
All action potential measurements were carried out by using wire electrodes only because plate electrodes generate large stimulus artifacts.
Single action potentials. Em were measured by using conventional glass microelectrodes under the same conditions as for force measurements. Microelectrodes (tip potentials <5 mV, tip resistances 7-15 MΩ) and a reference electrode (tip resistance of 1 MΩ) were filled with 3 M KCl. Em were recorded simultaneously on a chart recorder and a computer. Single action potentials were elicited with a single 10-V, 0.3-ms pulse. They were recorded using a WPI electrometer (model M-707; Sarasota, FL) and digitized at a sampling rate of 200 kHz. Action potentials were recorded first at 147 mM Na+, then at one lowered [Na+] (30, 40, 60, or 100 mM) when force had reached a steady-state, and again when force had recovered at 147 mM Na+. Action potentials were analyzed if all three of the following criteria were met: 1) the potential measured on the chart recorder showed a sharp drop upon microelectrode penetration, 2) the resting Em in the control solution was more negative than -65 mV, and 3) the stimulus artifact did not interfere with the upstroke of the action potential.
The resting Em was measured from the baseline of the action potential. Overshoot was defined as the difference in potential between the peak of the action potential and 0 mV. Action potential amplitude was defined as the difference between resting Em and the peak potential attained. Action potential width was defined as the duration at one-half of the amplitude. The first derivative of each action potential (dV/dt) was obtained by calculating the slope over every 10 data points. Maximum rates of depolarization were measured as the peak of the first derivative during the depolarization phase, whereas the maximum rate of repolarization was measured as the minimum value attained during the repolarization phase. Excitability was calculated as the percentage of the total number of fibers penetrated that generated an action potential.
Trains of action potentials. To reduce movement artifacts, trains of action potentials (50 and 125 Hz) were measured after the muscle had been stretched to 1.3-1.5 times its optimal length. After this stretch, one stimulating electrode was located about half way along its length, with the other stimulating electrode at the end of the muscle. Action potential trains were digitized at 10 kHz. This sampling frequency was high enough to determine precisely the overshoot but too low to determine the other parameters described for single action potentials. Considering that stretching may have damaged some fibers, a train of action potentials was analyzed only if the values of the Em prior to, and the overshoot of, the first action potential in the train were within the range of values obtained from the measurements of single action potentials in unstretched muscle.
Data are given as means ± SE for the number (n) of muscles (force) or fibers/muscles (action potential). Statistical analyses involved split plot analyses of variance (ANOVA). Split plot designs were used because force and action potential measurements at different times (fatigue) or [Na+]o were made using the same muscle. ANOVA were calculated using the General Linear Model procedures of the SAS statistical software (SAS Institute, Cary, NC). When a main effect or an interaction was significant, the least-squares difference was used to locate any significant difference (28). Only those effects significant at P < 0.05 are reported.
Influence of Lowered [Na+]o on Force
Nonfatigued muscle. Figure 1 shows representative force records whereby lowered [Na+]o caused a reduction of peak tetanic force in nonfatigued EDL and soleus muscles (plate electrodes). Lowered [Na+]o also increased the extent of fade (decline of force within a tetanus) but only in EDL: the force at 500 ms of stimulation was 85 ± 1% (n = 6) of the peak force at 147 mM Na+, but this was markedly reduced to 38 ± 6% (n = 6) at 40 mM Na+. The peak tetanic force-[Na+]o relationships for soleus and EDL were not significantly different (Fig. 2A). For example, a reduction of [Na+]o from 147 to 80 mM induced only a moderate suppression of force to 88% in soleus and to 90% in EDL. When [Na+]o was further reduced, the force fell abruptly over a narrow range of [Na+]o. The influence of lowered [Na+]o on twitch contractions was also studied (Fig. 2B). The peak twitch force-[Na+]o relationships for soleus and EDL show that force was well maintained between 147 and 60 mM Na+. At [Na+]o below 60 mM, peak twitch force fell sharply and was similar in both muscle types, albeit a single significant difference appeared at 40 mM Na+. Considering that in most cases there was no difference in the Na+ effect on peak force between EDL and soleus, all remaining experiments were performed using only soleus muscles.
The Na+ effect on force was also determined when muscles were activated by wire electrode stimulation (see materials and methods). The effects of lowered [Na+]o on peak twitch force (from 147 to 30 mM Na+) and on peak tetanic force (from 147 to 60 mM Na+) were the same in soleus, whether plate or wire electrodes were used (data not shown). However, several notable differences appeared for tetanic contractions at the lowest [Na+]o, being most apparent at 40 mM Na+ (see Fig. 3). 1) Peak tetanic force (125 Hz) fell to 19% of the control with wire electrodes, in contrast to the reduction to 45% observed with plate electrodes. 2) With wire electrodes, a reduction of stimulation frequency from 125 to 50 Hz partially restored peak force to 53% of the control. Interestingly, this effect was not observed with plate electrodes, and the peak force (50 Hz) was the same with wire and plate electrodes. 3) Although fade was not observed in soleus with plate electrodes at 125 Hz (Fig. 1A), it appeared when wire electrodes were used: at 2 s of stimulation, the force was 39 ± 9% (n = 9) of the peak tetanic force (Fig. 3, inset). Moreover, fade was markedly reduced at 50 Hz to 78 ± 8% (n = 8) of the peak force.
Fatigued Muscle. To further test the importance of Na+ during fatigue, we examined the effects of 80 mM Na+ on peak tetanic force of soleus muscle during fatigue induced by intermittent tetanic stimulation (plate electrodes). This value of lowered [Na+]o was used because it causes only a moderate decline of peak tetanic force in nonfatigued muscle (Fig. 2A). The decrease of force was greater at lowered [Na+]o over the entire stimulation period (Fig. 4). For example, the decline of peak tetanic force at 80 mM Na+ was double that at 147 mM Na+ over the first 30 s of stimulation, whereas at 100 s of stimulation, the final forces at 80 and 147 mM Na+ were 19 and 34% of the initial force, respectively.
Mechanisms of Force Depression at Lowered [Na+]o
Contractile tests. Stimulation pulses of 0.1-ms duration depolarize the sarcolemma enough to reach action potential threshold. Pulses of 1.0-ms duration delivered with plate electrodes, on the other hand, generate longer depolarizations, which are additive to those of action potentials and give rise to a greater myoplasmic [Ca2+] ([Ca2+]i) (8). If action potential impairment at lowered [Na+]o reduces force, then increasing pulse duration from 0.1 to 1.0 ms should allow some force recovery. We tested this notion and obtained the following results. At 147 mM Na+, the peak tetanic force evoked with 1.0-ms pulses (125 Hz, 2 s) was 103 ± 1% (n = 15) of that evoked with 0.1-ms pulses. At 80 mM Na+ (n = 9), the peak tetanic force evoked with 0.1-ms pulses was reduced to 89 ± 3% but was fully restored to 101 ± 1% with 1.0-ms pulses. At 40 mM Na+, the peak tetanic force evoked with 0.1-ms pulses fell to 35 ± 9% compared with 74 ± 4% (n = 8) with 1.0-ms pulses. We infer that an impairment involving action potentials is a mechanism for the force depression at lowered [Na+]o.
One mechanism for action potential impairment is an increased threshold for action potential generation. To test this hypothesis, we determined the effect of lowered [Na+]o on the peak twitch force-stimulation strength relationship. Indeed, in muscles exposed to 40 mM Na+, a large shift toward greater stimulation strengths occurred that reversed upon return to the control solution (Fig. 5). The stimulation strength necessary to elicit 50% of maximum twitch force increased from 3.0 ± 0.2 V/cm at 147 mM Na+ to 12.5 ± 1.5 V/cm (n = 5) at 40 mM Na+ (plate electrodes). Similar shifts also occurred when wire electrodes were used, being progressive from 147 to 60, 40, and 30 mM Na+ (data not shown).
Influence of lowered [Na+]o on single action potentials. Given the obligatory role of Na+ in generating action potentials, we examined the consequences of lowered [Na+]o on single action potentials. When [Na+]o was lowered from 147 to 100, 60, 40, or 30 mM, there was no change to the resting Em from its value of -78 mV at 147 mM Na+ (Table 1). At 60, 100, and 147 mM Na+, all fibers were excitable; i.e., they generated an action potential upon stimulation when the standard stimulation pulse was used (10 V, 0.3 ms), which was supramaximal for the twitch. However, 19% of the fibers were inexcitable at 40 mM Na+, a value that increased to 40% at 30 mM Na+. The resting Em of these inexcitable fibers was -75 ± 5 mV, which was similar to those generating action potentials (Table 1).
The effects of lowered [Na+]o on action potential parameters are shown in Table 1 and Fig. 6. The action potential became progressively smaller, with the overshoot falling from 32 mV at 147 mM Na+ to 7 mV at 60 mM Na+ (Fig. 6). No overshoot was observed at either 40 or 30 mM Na+ because the average peaks of the action potentials were -9 and -20 mV, respectively. Similarly, the action potential amplitude fell progressively (Table 1). Action potentials also became progressively broader at lowered [Na+]o (Fig. 6A) and were associated with a decline in the maximum rates of depolarization and repolarization (Table 1). The effects of lowered [Na+]o were fully reversible for all parameters, as shown for the overshoot in Fig. 6B.
Influence of lowered [Na+]o on trains of action potentials. Because most muscular activity and tetanic contractions are evoked by a train rather than a single action potential, we examined the influence of lowered [Na+]o on trains of action potentials. To reduce movement artifacts, we recorded trains of action potentials in the second or third layer of fibers from the surface of soleus muscles after stretch. At 147 mM Na+, the resting Em measured before the first action potential of the train of all stretched fibers used was -77 ± 2 mV, and the overshoot of the first action potential was 29 ± 2 mV (n = 22 fibers/5 muscles). These values are similar to those reported for unstretched fibers (Table 1 and Fig. 6). At 40 mM Na+, the resting Em was -72 ± 1 mV, and the Em at the peak of the action potential was -18 ± 3 mV (n = 9/3) in stretched fibers, which was in the lower range of values obtained for unstretched fibers.
Trains of action potentials recorded at 147 mM Na+ showed decreases in both Em (between action potentials) and overshoot during the train (Fig. 7, A-C). At 50 Hz, the overshoot fell continuously from 27 ± 2 to 16 ± 2 mV while the Em fell from -73 ± 3 to -65 ± 2 mV during the 2 s of stimulation (n = 6/4) (see Fig. 7A). Trains evoked at 125 Hz (n = 16/5) displayed two distinct patterns of action potentials. The first pattern (6 of 16 fibers) was similar to that at 50 Hz showing continuous changes, except that the decreases in Em and overshoot during the train were both larger (Fig. 7B). For this pattern, the mean overshoot fell from 30 ± 3 to -6 ± 3 mV while Em fell from -76 ± 3 to -60 ± 8 mV.
The second pattern at 125 Hz (10 of 16 fibers; Fig. 7C) showed an initial more rapid and much larger decline in Em and overshoot, followed by a sudden and partial restoration of both, hereafter referred to as a “rebound.” On average, the first action potential rebound started at 1,075 ± 133 ms into the 2-s train. The rebound was not due to skipping (intermittent failure to generate action potentials) because all stimuli gave rise to an action potential (see Fig. 7C, inset). Instead, each rebound was preceded by a sudden decrease of Em to -53 ± 2 mV and the action potential peak to -27 ± 4 mV, followed by a partial recovery of Em to -61 ± 2 mV and the action potential peak to -7 ± 3 mV. After each partial recovery, Em and overshoot again decreased rapidly to give rise to the next rebound.
Figure 7, D-F, shows 2-s trains of action potentials evoked at 40 mM Na+. This value of [Na+]o was chosen because it was associated with a marked reduction of peak tetanic force (Figs. 2A and 3). Trains of stimuli delivered at 50 Hz generated action potentials in response to each stimulus (Fig. 7D) in 7 of 10 fibers (3 muscles). In such cases, the sequence of 100 action potentials was associated with only a small decline in overshoot and no decrease of Em. In the other three fibers, only the first stimulus triggered an action potential.
In contrast, trains of action potentials evoked at 125 Hz showed a marked change in pattern (n = 9/3) as shown in Fig. 7, E and F. This involved skipping of action potentials (often several at a time), leading to complete failure of generation. This complete failure often occurred within the first 500 ms (6 fibers) (see Fig. 7E). In fact, only two fibers, with considerable skipping, generated action potentials throughout the 2 s of stimulation. When skipping occurred, the stimulus often arrived during the repolarization phase of the previous very prolonged action potential. This scenario accounted for 73% of the skipping observed in the example of Fig. 7F. When complete excitation failure occurred during a train of stimuli at 125 Hz (Fig. 7E), it was not due to damaged fibers because subsequent trains evoked at 50 Hz in the same fibers produced action potentials in response to every stimulus (4 fibers). Moreover, trains of action potentials (125 Hz) were evoked at both 10 and 18 V in the same fiber on three occasions during exposure to 40 mM Na+. In each case, the effect of increasing the stimulation strength was to increase the number of action potentials in the train, but there was still complete failure later in the train.
The principal new findings of the present study are that 1) the peak force-[Na+]o relationships do not differ between soleus and EDL muscles (Fig. 2); 2) different tetanic contractile responses occur at lower [Na+]o when stimulating with wire or plate electrodes (Fig. 3); 3) lower [Na+]o causes some fibers to become inexcitable; 4) trains of action potentials (2 s) recorded at 147 mM Na+ in soleus fibers reveal a complete loss of overshoot and the “rebound effect” (Fig. 7C); 5) 40 mM Na+ causes a failure to generate action potentials during high-frequency trains (Fig. 7, E and F); and 6) lower [Na+]o exacerbates fatigue induced with repeated tetani (Fig. 4).
Effects of Lowered [Na+]o on Force
A main aim of our study was to determine whether effects of lowered [Na+]o on force differed between fast-twitch and slow-twitch muscles. The peak force-[Na+]o relationships did not differ between soleus and EDL (Fig. 2): force was well maintained despite large reductions of [Na+]o from 147 to 80 mM for the tetanus, or to 60 mM for the twitch, beyond which force fell abruptly. Stability of peak force over a large range of [Na+]o has also been reported for rat soleus (22) and frog sartorius (5). Thus, regardless of muscle type or species, a large safety margin exists for peak force when [Na+]o is reduced, followed by an abrupt force loss over a narrow range of [Na+]o.
Fade, or the inability to maintain force during a short tetanus, is a feature of fast-twitch but not slow-twitch muscles in normal solutions (Fig. 1). With plate electrodes, fade was exacerbated at lowered [Na+]o in EDL (Fig. 1). Fade occurred in soleus, but only when wire electrodes were used and at the lowest [Na+]o (Fig. 3, inset). Thus, contrary to the situation for peak force, the Na+ effect on fade is greater in fast-twitch than in slow-twitch muscle.
Effects of Lowered [Na+]o on Action Potentials
Single action potentials. Lowered [Na+]o causes an increased threshold for action potentials as indicated by a large shift of the peak twitch force-stimulation strength relationship toward higher voltages (Fig. 5) or as assessed from action potential recordings in rat soleus and EDL at 45 mM Na+ (1). A second effect seen only at [Na+]o ≤ 40 mM is a complete loss of sarcolemmal excitability in some fibers despite supramaximal stimulation. This occurrence of inexcitable fibers (and failure to generate action potentials during a train, Fig. 7, E and F) is unlikely to involve impaired neuromuscular transmission, because recordings were from surface fibers located close to the wire stimulating electrode. Another effect seen at all lowered [Na+]o is a reduction of action potential overshoot and amplitude for the excitable fibers (Fig. 6 and Table 1). According to the Nernst equation, the decrease in overshoot should be 59 mV per 10-fold decrease in [Na+]o. However, the decrease was 75 mV per decade (correlation coefficient, r = 0.99), which is comparable to the value previously reported for rat soleus (1). The depression of the maximum rate of depolarization at lowered [Na+]o was as much as threefold greater than the decrease in amplitude. Thus lowered [Na+]o suppresses the overshoot not only by lowering ENa but also by slowing the depolarization.
Trains of action potentials. 147 mm na+. Trains of action potentials, recorded in mouse soleus fibers, revealed two distinct patterns. One pattern involved a continuous depolarization of the sarcolemma and a decrease in overshoot throughout the 2-s train (Fig. 7, A and B). This depolarization is presumably in response to a progressively reduced K+ gradient due to net K+ efflux with each action potential. The diminished overshoot is likely to be a consequence of a decreased Na+ gradient and increased Na+ channel inactivation as the sarcolemma depolarizes. The greater decrease in these parameters at 125 Hz than at 50 Hz is consistent with greater Na+ and K+ fluxes and less time between action potentials for the Na+-K+ pump to reestablish the Na+ and K+ gradients at 125 Hz.
The second pattern (seen only at 125 Hz) consisted of an initial continuous depolarization and decline in overshoot followed by an “action potential rebound,” a pattern that repeated itself cyclically throughout the remainder of the train (Fig. 7C). Each rebound occurred when a very low amplitude action potential was generated as Em approached -53 mV. The partial recovery of the action potential peak during the rebound was clearly linked to repolarization and not to a skipping of action potentials, as reported for frog muscle (4). Perhaps the Na+ and K+ fluxes during the low-amplitude action potential were small enough to allow for partial recovery of Na+ and K+ gradients by the Na+-K+ pump.
40 mm na+. Most fibers generated a small action potential in response to every stimulus during the 2-s train at 50 Hz (Fig. 7D), as seen in frog muscle (4). In contrast, at 125 Hz every fiber failed to respond to some stimuli, i.e., demonstrated skipping (Fig. 7, E and F), and in most cases there was a complete failure to generate action potentials within 500 ms (Fig. 7E). These failures cannot be attributed to a decrease of Em between action potentials during a train because this did not occur at 40 mM Na+ (as it did at 147 mM Na+). This lack of depolarization is possibly because the small action potentials caused diminished activation of delayed rectifier K+ channels and, hence, diminished K+ efflux. An increased action potential duration (Table 1) is likely to cause skipping, because 73% of those stimuli that failed to trigger an action potential occurred during the repolarization phase of the preceding action potential in Fig. 7F, i.e., when the fibers were still in their refractory period. An increased threshold during the train is also likely to cause a failure to generate action potentials, because increasing the stimulation strength from 10 to 18 V increased the number of action potentials generated by trains of stimuli.
Mechanisms for Force Depression at Lowered [Na+]o
Another major aim of this study was to determine whether the Na+ effects on action potentials can explain the force depression at lowered [Na+]o or whether other mechanisms were involved.
Role of the sarcoplasmic reticulum. The release of Ca2+ from the sarcoplasmic reticulum (SR) may be triggered by an action potential and/or involve an effect of Na+ to modulate the SR (2). Whereas 0.1-ms pulses (plate electrodes) initiate action potentials, 1.0-ms pulses additionally accentuate the depolarization generated by action potentials resulting in greater Ca2+ release (7) (Renaud J-M and Chin ER, unpublished observations). Increasing the pulse duration from 0.1 to 1.0 ms resulted in a full recovery of the previously diminished peak tetanic force at 80 mM Na+, suggesting an impairment via action potentials. The incomplete, although large, recovery of force observed at 40 mM Na+ indicates that an impairment of Na+-induced modulation of Ca2+ release from the SR cannot be eliminated.
Role of excitability. All fibers were excitable at values of [Na+]o from 147 to 60 mM, but some fibers became inexcitable at lower [Na+]o (19 and 40% at 40 and 30 mM Na+, respectively). We calculate that this accounts for 44 and 65% of the twitch depression at these [Na+]o, respectively. A similar contribution is calculated for the tetanus at 50 Hz. The occurrence of inexcitable fibers is thus a major factor in the reduction of peak force.
Role of action potential overshoot. The peak twitch force-overshoot relationship shows that in soleus, force decreased by only 10% with a 30-mV reduction in overshoot (Fig. 8). Using tetrodotoxin, Yensen et al. (32) also reported a very small decrease in EDL twitch force for overshoots between 30 and 5 mV. They also showed that prolongation of the action potential, similar to that reported in Table 1, does not compensate for the decrease in overshoot to maintain force. Thus mouse EDL and soleus have considerable safety margins whereby peak twitch force is well maintained for large reductions of overshoot.
A striking observation in the present study was that during tetanic stimulation (125 Hz) at 147 mM Na+, the peak of the action potential could fall to between 0 and -30 mV (Fig. 7C), and this occurred without any associated fade. A possible explanation for this finding is that enough Ca2+ is released at the beginning of the tetanus so that the remaining action potentials are large enough to maintain the [Ca2+]i above the level necessary to generate maximum force. Another possible explanation is that a reduction in action potential peak to -20 mV reduces peak twitch force (and perhaps Ca2+ release) by <20% (Fig. 8). At 40 mM Na+, a long train of action potentials was generated at 50 Hz, where the overshoot had completely disappeared. We cannot state conclusively that these small action potentials contribute to the reduced peak force (on the basis of the above observations at 147 mM Na+). However, we speculate that at 40 mM Na+, the small amplitude action potentials from the start of the train may result in a diminished rate of Ca2+ release, leading to a reduced peak tetanic force.
Role of failure to generate action potentials during a train. At 40 mM Na+ (wire electrodes), a skipping of, or complete failure to generate, action potentials in the surface membrane during a 125-Hz train (Fig. 7, E and F) is likely to cause fade and contribute to the diminution of peak tetanic force (Fig. 3). Two arguments support this conclusion. First, reducing the frequency to 50 Hz largely restored action potential generation during a train (Fig. 7D), increased peak tetanic force, and diminished fade (Fig. 3). Second, the decline of peak tetanic force at 125 Hz was greater, and fade appeared, when wire electrodes were used in contrast to plate electrodes (Figs. 1 and 3). Because wire electrodes generate action potentials in a small area of surface membrane, whereas plate electrodes generate action potentials along the entire length of surface membrane, we infer that the different tetanic responses at 40 mM Na+ reflect a failure of action potentials to propagate along the surface membrane when wire electrodes are used. On the basis of the differences in peak force, we calculate that at least 32% of the peak force decline with wire stimulation was due to propagation failure.
The above failure to generate action potentials in the surface membrane during tetanic stimulation also means that there will be no action potentials traveling down the t-tubular membranes to trigger Ca2+ release. Indeed, the idea of impaired t-tubular membrane conduction of action potentials is well supported by the observations that lowered [Na+]o causes the development of wavy myofibrils (2) and lower [Ca2+]i (12) in the center of fibers than at the periphery during tetanic stimulation.
Implications for Fatigue
Does a reduced transsarcolemmal Na+ gradient cause fatigue? Changes in Na+ gradient during fatigue are mainly due to increases of [Na+]i, because plasma [Na+] is relatively constant (24-26). The maximum increases of [Na+]i during intense exercise in humans are 50-100% (24-26). In mammalian muscle, intermittent tetanic stimulation can elevate [Na+]i by 100% (13, 14, 17), with the largest increases of [Na+]i, three- to fourfold, occurring in fast-twitch muscle (19, 27). The driving force for the Na+ current during action potentials is the difference between Em and ENa, the latter being determined by the transsarcolemmal Na+ ratio ([Na+]o/[Na+]i; see Introduction). Thus decreasing [Na+]o from 147 to 75 mM mimics the change in Na+ ratio that occurs when [Na+]i increases by 100%. Such a decrease of [Na+]o causes only a small decrease (<15%) of peak tetanic force (Fig. 2A), so a doubling of [Na+]i is unlikely to contribute much to fatigue. Only in fast-twitch muscle, and when a three- to fourfold decrease in the Na+ ratio occurs, is Na+ per se likely to contribute to fatigue, because a corresponding decrease of [Na+]o from 147 to 40-50 mM Na+ causes a large depression of force (Fig. 2A).
However, although 80 mM Na+ had little effect in nonfatigued muscles, it exacerbated fatigue induced by intermittent tetanic stimulation (Fig. 4). Similar effects of lowered [Na+]o on fatigue have been seen in amphibian muscle (15) and during prolonged continuous tetanic stimulation of mammalian muscle (6, 16). Clearly, these results suggest that a small reduction of the Na+ gradient can be rate limiting during fatigue. One possibility for the discrepancy between nonfatigued and fatigued muscle is because changes of ttubular [Na+] have not been considered. Bezanilla et al. (4) estimated a 30 mM depletion of Na+ in the t tubules during a single 500-ms tetanic contraction at 60 Hz. Other investigators have also suggested that Na+ depletion occurs in the t tubules because of diffusion limitations (6, 12, 15, 16). Consequently, limiting consideration to changes in [Na+]i is likely to underestimate the true Na+ gradient across t-tubular membranes during fatigue.
Another possibility is that the Na+ effect in fatigue is amplified by other changes that occur during fatigue. Indeed, the depression of peak tetanic force in nonfatigued muscle with small reductions of [Na+]o becomes greater when acting synergistically with raised [K+]o (5, 22) or lowered [Ca2+]o (8); i.e., Na+ may have a greater effect in fatigued than in nonfatigued muscle because of the simultaneous decrease in K+ gradient (17, 19, 26) and, possibly, lower t-tubular [Ca2+] (8, 25). Lowered Na+ gradient may also affect the activity of the Na+/Ca2+ (3, 25) or Na+/H+ exchanger (25), resulting in changes in intracellular [Ca2+] and pH, both of which may influence fatigue.
Do reductions of action potential overshoot cause fatigue? A striking observation from the present study is that the overshoot could fall by almost 30 mV without any depression of peak twitch force (Fig. 8) (32). Because the decreases in overshoot during fatigue are <30 mV (18), our data suggest that changes in overshoot (or action potential peak) are unlikely to cause fatigue. On the other hand, if, during fatiguing stimulation, an intermittent failure to generate action potentials (Fig. 7, E and F) or complete excitation failure (Fig. 7E) occurs, then such action potential changes may indeed cause fatigue because they are associated with force loss.
In summary, we have found that the effect of lowered [Na+]o on peak force is the same in nonfatigued fast-twitch EDL and slow-twitch soleus muscles of the mouse. A large safety margin is observed in which large reductions of [Na+]o (up to 60 mM) or overshoot (up to 30 mV) have little effect on peak twitch and tetanic forces. Larger decreases of [Na+]o (>90 mM) abruptly reduce peak forces, the mechanism involving primarily a complete loss of sarcolemmal excitability in some fibers and a lower capacity to generate action potentials in others. Although Na+ appears as a minor factor in the etiology of muscle fatigue from its effect in nonfatigued muscle, we have provided evidence that Na+ is a limiting factor in fatigue elicited by a high frequency of stimulation because substantial differences exist in the Na+ effect between nonfatigued and fatigued muscle.
Several studies have looked at the separate roles of Na+ and K+ in the etiology of muscle fatigue. However, few studies have looked at the combined effects of Na+ and K+, and even fewer have systematically combined other ions such as Ca2+, Cl-, and H+. Such interactive studies could be important because 1) at least for Na+ and K+, the effects are synergistic and not simply additive (5, 22), and 2) at least for K+, the effects are complex because an increase of [K+]o sufficient to abolish action potential overshoot can actually potentiate, rather than suppress, peak twitch force (32). Experimental temperature and stimulation frequency must also be considered. For example, K+-induced twitch potentiation and Na+-K+ pump activity are greater at 37°C (i.e., at physiological temperature) than at lower temperatures (32). Greater pump activity and low stimulation frequencies should also give rise to different changes in ion concentration gradients, modifying how ions affect force during fatigue.
This work was supported by grants from the Health Research Council of New Zealand (to D. Loiselle), Lottery Grants Board of New Zealand (to S. Cairns), and National Science and Engineering Research Council (to J.-M. Renaud).
We gratefully acknowledge Dr. John Slack for helpful discussion.
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 © 2003 the American Physiological Society