Excitation-induced force recovery in potassium-inhibited rat soleus muscle

Abstract

1. Excitation markedly stimulates the Na⁺-K⁺ pump in skeletal muscle. The effect of this stimulation on contractility was examined in rat soleus muscles exposed to high extracellular K⁺ concentration ([K⁺]_o).

2. At a [K⁺]_o of 10 mm, tetanic force declined to 58 % of the force in standard buffer with 5.9 mm K⁺. Subsequent direct stimulation of the muscle at 1 min intervals with 30 Hz pulse trains of 2 s duration induced a 97 % recovery of force within 14 min. Force recovery could also be elicited by stimulation via the nerve. In muscles exposed to 12.5 mm K⁺, 30 Hz pulse trains of 2 s duration at 1 min intervals induced a recovery of force from 16 ± 2 to 62 ± 4 % of the initial control force at a [K⁺]_o of 5.9 mm.

3. The recovery of force was associated with a decrease in intracellular Na⁺ and was blocked by ouabain. This indicates that the force recovery was secondary to activation of the Na⁺-K⁺ pump.

4. Excitation stimulates the release of calcitonin gene-related peptide (CGRP) from nerves in the muscle. Since CGRP stimulates the Na⁺-K⁺ pump, this may contribute to the excitation-induced force recovery. Indeed, reducing CGRP content by capsaicin pre-treatment or prior denervation prevented both the excitation-induced force recovery and the drop in intracellular Na⁺.

5. The data suggest that activation of the Na⁺-K⁺ pump in contracting muscles counterbalances the depressing effect of reductions in the chemical gradients for Na⁺ and K⁺ on excitability.

Several studies have shown that in isolated skeletal muscles where contractile performance is depressed by exposure to a high extracellular K⁺ concentration ([K⁺]_o), considerable force recovery can be elicited by acute stimulation of active Na⁺-K⁺ transport with catecholamines, the β₂-agonist salbutamol, insulin or calcitonin gene-related peptide (CGRP) (Tomita, 1975; Clausen & Everts, 1991; Andersen & Clausen, 1993; Clausen et al. 1993; Cairns et al. 1995). This force recovery was shown to be closely correlated to the stimulation of Na⁺-K⁺ pump-mediated K⁺ uptake, and seems to be related to the restoration of membrane potential and the electrochemical gradient for Na⁺ across the sarcolemma (Clausen et al. 1993; Overgaard et al. 1997a). In keeping with these observations, attacks of hyperkalaemic periodic paralysis in human subjects were found to be alleviated by the administration of salbutamol, adrenaline or insulin (Wang & Clausen, 1976), agents that all induce stimulation of active Na⁺-K⁺ transport in skeletal muscle within minutes.

Excitation of isolated muscles leads to a rapid and pronounced acceleration of active Na⁺-K⁺ transport (Hazeyama & Sparks, 1979; Fong et al. 1986; Juel, 1986; Balog & Fitts, 1996; Nielsen & Clausen, 1997). In isolated muscles recovering from excitation, the activity of the Na⁺-K⁺ pump may remain elevated for several minutes leading to a decrease in intracellular Na⁺ by up to 30 % below the level measured in the resting muscles before the onset of stimulation (Nielsen & Clausen, 1997). Likewise, in vivo studies in rats have shown that following electrical stimulation, skeletal muscle fibres undergo hyperpolarization. Since this could be suppressed by ouabain, it was assumed to reflect activation of the electrogenic Na⁺-K⁺ pump (Hicks & McComas, 1989).

These observations prompted the present study, which was carried out in order to determine if excitation-induced stimulation of the Na⁺-K⁺ pump could alleviate the inhibitory effect of high [K⁺]_o on contractility. We found that in isolated rat soleus muscles exposed to high [K⁺]_o (10 or 12.5 mm) substantial or complete force recovery could be elicited by repeated tetanic stimulation. The possible mechanisms of this new phenomenon were explored using ouabain or propranolol, or by modifying the muscle content of CGRP. Part of the present results has been presented in a preliminary form (Clausen & Nielsen, 1996; Clausen et al. 1998).

METHODS

Animals

All handling and use of animals complied with Danish animal welfare regulations. The experiments were performed using fed 4-week-old female or male Wistar rats weighing 60–70 g. The animals had free access to food (Altromin International, Lage, Germany) and water, and were kept in a thermostatically controlled environment (21°C) with a constant light-dark cycle (12–12 h).

Muscle preparation and incubation

Animals were killed by decapitation, and intact soleus muscles dissected out. The wet weight of the muscles ranged from 20 to 25 mg. The standard incubation medium was Krebs-Ringer bicarbonate buffer (pH 7.4 at 30°C) containing (mm): 120.1 NaCl, 25.1 NaHCO₃, 4.7 KCl, 1.2 KH₂PO₄, 1.2 MgSO₄, 1.3 CaCl₂, and 5 D-glucose. When the concentration of K⁺ in the buffer was increased to 10 or 12.5 mm, an equimolar amount of Na⁺ was omitted from the buffer. Before tension recordings, all muscles were equilibrated for at least 30 min in the standard buffer containing 5.9 mm K⁺. All experiments were carried out at 30°C to reduce metabolic requirements and thus ensure sufficient oxygenation of the central muscle fibres. The buffer was continuously gassed with a mixture of 95 % O₂ and 5 % CO₂. This procedure was shown to allow the maintenance of a high intracellular K⁺ : Na⁺ ratio, constant membrane potential (Clausen & Flatman, 1977) and tetanic force for several hours in vitro (Clausen & Everts, 1991).

Calcitonin gene-related peptide (CGRP) is produced in nerve cells in the central nervous system and transported via the axons to nerve endings in skeletal muscles, and in accordance with this denervation has been shown to lead to depletion of muscle CGRP content (Kashihara et al. 1989). To produce such depletion, denervation of the muscles of one leg was performed on seven rats. The muscles of the other leg served as controls. The denervation was performed under light ether anaesthesia at the sciatic level by removing 2 mm of the sciatic nerve. The incisions were sutured, and 7 days later the animals were decapitated and the soleus muscles were prepared for studies of force development or measurement of CGRP. No infections were seen.

Force development

This was recorded as described in detail elsewhere (Clausen & Everts, 1991). Muscles with their tendons intact were mounted isometrically in a thermostated (30°C) chamber containing Krebs-Ringer bicarbonate buffer and were stimulated via two platinum electrodes positioned on each side of the muscle close to the muscle surface. If not otherwise noted supramaximal pulses of 12 V and of 1 ms duration were used for stimulation. Force development was measured with a force displacement transducer (Grass FTO3, MA, USA) calibrated with standard weights. After adjustment to optimal length, active force of single twitches and tetanic contractions (30 Hz for 2 s) were tested. This was followed by a 30 min rest period. During the subsequent experiments contractile performance was assessed by applying 2 s pulse trains of 30 Hz every 10 min, or in some instances at longer intervals. In most experiments the effect of shortening the intervals between these 2 s pulse trains to 1 min was explored. Control experiments, where the force development during 2 s stimulation at 30 Hz was compared with the force development during 5 s stimulation at 30 Hz, showed that a 2 s stimulation allowed full development of force in muscles incubated in the standard buffer containing 5.9 mm K⁺, and development of around 90 % of full force when incubated at a K⁺ concentration of 12.5 mm. A stimulation frequency of 30 Hz was chosen because it is close to the maximal motor-unit discharge rates observed for soleus muscles in vivo (Bigland-Ritchie & Woods, 1984). High resolution measurements showed that at 30 Hz, the maximum fluctuation in force was 0.4 ± 0.3 % (n = 4) of the mean force produced, indicating that the contractions were close to smooth tetani.

In a few experiments the muscles were dissected with approximately 10 mm of the nerve attached. In these experiments the muscles were stimulated via the nerve (pulses of 10 V and of 0.04 ms duration) using a glass suction electrode with a tip diameter of 350 μm that fitted the nerve twig.

Intracellular Na⁺ and K⁺ contents

Intracellular Na⁺ and K⁺ contents were measured as described in detail elsewhere (Everts & Clausen, 1992). In brief, the muscles mounted on the force transducers were quickly transferred to tubes containing ice-cold Na⁺-free Tris-sucrose buffer (pH 7.4) and washed 4 × 15 min during continuous gassing with air in order to remove extracellular Na⁺. The muscles were then blotted on dry filter paper, the tendons resected and after weighing, the muscles were extracted with 0.3 m trichloroacetic acid (TCA) for flame photometry (FLM3, Radiometer, Denmark). Previous studies have shown that some intracellular Na⁺, but no K⁺ is lost during the washout in the cold. From these experiments, a correction factor of 1.46 for the loss of intracellular Na⁺ was calculated and used for correction of the values measured. Control experiments showed that the correction factor is the same for resting and stimulated muscles (Everts & Clausen, 1992).

Determination of CGRP

Intact soleus muscles were frozen in liquid N₂ and kept frozen until extraction and measurement of CGRP. Each muscle was homogenized in 1.0 ml cold (4°C) acidified ethanol (ethanol 16.7 M, HCl 0.22 M) using a Potter-Elvehjelm homogenizer, left overnight at 4°C, neutralized (pH 7) and centrifuged at 4°C and 10 000 g for 30 min. The supernatant was isolated and evaporated to dryness in a centrifugal concentrator (SVC 200H, Savant, Hicksville, NY, USA) and reconstituted in assay buffer.

The extracts were assayed for CGRP using a rat CGRP RIA-kit (RIK-6006, Peninsula Laboratories Inc., Belmont, CA, USA). The antiserum employed cross-reacts less than 0.01 % with rat calcitonin and 23 % with CGRP II. ¹²⁵I-labelled Tyr-0-CGRP (rat) was used as tracer. Bound and free antigen were separated using the double antibody technique. Each muscle extract was assayed in duplicate at two dilutions within the working range of the assay, and the mean of the four measurements was used.

Recovery was tested using a pool of homogenized muscles, divided into aliquots of 1.0 ml. Varying concentrations of synthetic rat CGRP I (Peninsula Laboratories Inc., cat. no. 6006, 1.6 × 10⁻¹¹ m) were added prior to overnight extraction and subsequent centrifugation. The relative recovery of CGRP was subsequently calculated by dividing the measured increase in CGRP immunoreactivity with the expected increase. The recovery varied from 55 to 85 %.

Statistics

All values are given as means ± s.e.m. The statistical significance of any difference was ascertained using Student's two-tailed t test for non-paired observations.

Chemicals

All chemicals were of analytical grade. Rat CGRP was obtained from Peninsula Laboratories Inc., and the peptide content was controlled in our laboratory by amino acid analysis. Capsaicin, ouabain and tetrodotoxin (TTX) were obtained from Sigma Chemicals.

RESULTS

Excitation-induced force recovery in muscles incubated at a [K⁺]_o of 10 mm

Figure 1 shows that increasing the concentration of K⁺ in the incubation medium from 5.9 to 10 mm reduced the tetanic force development to 58 % of the control level within 60 min. When, after 90 min at 10 mm K⁺, the tetanic stimulation was applied every 1 min, the force development started to increase within the first minute and almost complete force recovery (to 97 % of the level measured at 5.9 mm K⁺) was reached within 14 min. In the contralateral controls staying in the standard buffer containing 5.9 mm K⁺ throughout, the force development remained constant and showed no significant change in response to stimulation at 1 min intervals. Other experiments showed that if the tetanic stimulation at 1 min intervals was applied from the onset of the exposure to 10 mm K⁺ the force was only reduced to around 90 % of the control level after 60 min incubation (data not shown).

Figure 1. Effect of direct electrical stimulation on isometric force development in muscles at a [K⁺]_o of 5.9 and 10 mm.

Isometric tetanic contractions were elicited by 30 Hz pulse trains of 2 s duration (1 ms pulses, 12 V). After 30 min incubation in the standard buffer containing 5.9 mm K⁺, [K⁺]_o was either increased to 10 mm (○), or maintained at 5.9 mm (contralateral control muscles, •). Pulse trains were applied as indicated by data points. From time 120 min, the pulse trains were applied at 1 min intervals. Each point represents the mean of observations on five muscles with vertical bars denoting s.e.m. For sake of clarity, s.e.m. is only shown once every 10 min.

In order to determine whether the excitation-induced recovery of force depicted in Fig. 1 was restricted to the use of direct stimulation of the muscle fibres, the experiments were repeated using shorter pulses (12 V, 0.02 ms). As shown in Fig. 2, the reduction of the pulse duration from 1 to 0.02 ms gave a slight decrease in force development and the response to 10 mm K⁺ was somewhat more pronounced than in the previous experiments. When the pulse trains were applied every minute, a force recovery to 73 % of the initial level at 5.9 mm K⁺ obtained with 0.02 ms pulses was achieved within 10 to 15 min. The subsequent addition of tubocurarine (10⁻⁵ m) induced a complete suppression of force within 4 min. This loss of force was immediately alleviated by increasing pulse duration to 1 ms. This demonstrates that when 1 ms pulses were used the muscle fibres were stimulated directly, whereas the stimulation with pulses of 0.02 ms duration was indirect, taking place via the nerve of the muscle. In experiments where contractions were elicited by direct stimulation throughout (12 V, 1 ms), the addition of tubocurarine was without effect on force recovery (data not shown). To test the effect of indirect stimulation further, a group of muscles was stimulated via the nerve using a suction electrode and supramaximal pulses of 10 V and 0.04 ms duration. This set-up gave tetanic forces around 90 % of the force obtained using direct stimulation. In contrast, stimulation after the nerve was expelled from the suction electrode produced no force at all, demonstrating that the suction electrode only excited the muscle via the nerve. In these experiments the exposure to 10 mm K⁺ for 60 min reduced tetanic force to 23 ± 1 %. When the muscles were stimulated every minute with 30 Hz pulse trains an almost 3-fold increase in force (to 61 ± 3 % of the force obtained at 5.9 mm K⁺, n = 4, P < 0.005) was obtained within 19 min.

Figure 2. Effect of indirect electrical stimulation on isometric force development in muscles at a [K⁺]_o of 5.9 and 10 mm.

Experimental conditions as described in the legend to Fig. 1, except that a pulse duration of 0.02 ms was used from time 30 min to time 180 min as indicated. After 60 min incubation in the standard buffer containing 5.9 mm K⁺, [K⁺]_o was either increased to 10 mm (○), or maintained at 5.9 mm (contralateral control muscles, •). Tubocurarine (10⁻⁵ m) was added to the buffers at time 170 min. Each point represents the mean of observations on five muscles with vertical bars denoting s.e.m. For sake of clarity, s.e.m. is only shown once every 10 min.

Control experiments demonstrated that in muscles stimulated with pulses of 1 ms duration in standard buffer containing 5.9 mm K⁺ the addition of 5 × 10⁻⁷ m TTX reduced tetanic force to between 0 and 3 % of control force. When pulse durations below 0.5 ms were used the addition of TTX gave complete suppression of force.

Excitation-induced force recovery in muscles incubated at a [K⁺]_o of 12.5 mm

In order to obtain a more complete suppression of contractile force and a more defined basal level for the study of the excitation-induced force recovery, the experiments described in the following were performed using a somewhat higher [K⁺]_o (12.5 mm). This concentration was also chosen because in previous studies, hormonal stimulation of the Na⁺-K⁺ pump was observed to produce marked force recovery in muscles exposed to 12.5 mm K⁺ (Andersen & Clausen, 1993; Clausen et al. 1993).

As shown in Fig. 3, the incubation of muscles in buffer with 12.5 mm K⁺ produced a progressive inhibition of tetanic force production. When force was tested every 10 min using 30 Hz pulse trains of 2 s duration, a steady-state level at around 13 % of the initial control force at 5.9 mm K⁺ was reached within 70 min. When contractions subsequently were elicited every minute, force development gradually increased, reaching values around 55 % of the initial maximum force within 15 to 20 min. This phenomenon was observed in all experiments where a similar protocol was used. Thus, in fourteen muscles exposed to 12.5 mm K⁺ for 70 min or longer, a reduction in the interval between contractions from 10 to 1 min led to an increase in force from 16 ± 2 to 62 ± 4 % of the initial control force at 5.9 mm K⁺ (P < 0.001). Notably, a similar excitation-induced recovery of force (to 65–70 % of control force at 5.9 mm K⁺) was obtained when the intervals between the 30 Hz pulse trains were decreased from 10 to 2 or 0.5 min. When exposed to buffer with a [K⁺]_o of 15 mm the tetanic force decreased to 12 ± 2 % and applying tetanic pulses every 1 min only led to a recovery of force to 17 ± 3 % of the control force (n = 4).

Figure 3. Effects of CGRP and of shortening the intervals between 30 Hz pulse trains on isometric force development in muscles at a [K⁺]_o of 12.5 mm.

Isometric tetanic contractions were elicited by 30 Hz pulse trains of 2 s duration (1 ms pulses, 12 V) and were applied every 10 min or every 1 min as indicated by data points. After 30 min at a [K⁺]_o of 5.9 mm, [K⁺]_o was increased to 12.5 mm for all muscles as indicated by bar. ○, muscles undergoing two periods with pulse trains every 1 min. After the first period with stimulation every 1 min, the muscles were washed three times in buffer containing 12.5 mm K⁺. CGRP (10⁻⁷ m) was added after 410 min as indicated (n = 7). ▵, muscles undergoing the same experimental protocol but with CGRP added after 70 min of stimulation with pulse trains every 1 min. For sake of clarity only the last 60 min of the experiment are shown (n = 4). The data points represent the means with vertical bars denoting s.e.m. For sake of clarity, s.e.m. is only shown once every 10 min.

In some experiments twitch force was tested before and after the 70 min incubation at a [K⁺]_o of 12.5 mm with tetanic stimulation every 10 min and then again at the time for maximal recovery of tetanic force during subsequent stimulation at 1 min intervals. This showed that the elevation of [K⁺]_o to 12.5 mm reduced twitch force to 16 ± 5 % of the control level measured at a [K⁺]_o of 5.9 mm and that subsequent tetanic stimulation every 1 min recovered twitch force to 45 ± 5 % (n = 4, P < 0.025). In contrast, tetanic stimulation every 1 min was without effect on either twitch or tetanic force in muscles incubated in standard Krebs-Ringer buffer with 5.9 mm K⁺ (data not shown).

The maximum excitation-induced increase in force observed at 12.5 mm K⁺ was only maintained for 5 to 10 min, and when stimulation continued at 1 min intervals, force slowly returned to the level observed using stimulation every 10 min (Fig. 3). This inability to maintain force when stimulated at 1 min intervals could not be related to metabolic exhaustion because addition of CGRP to muscles after 70 min of stimulation at 1 min intervals produced a considerable improvement of force (Fig. 3).

The experiment shown in Fig. 3 also tests the response to two consecutive periods of electrical stimulation at 1 min intervals. After the first period of stimulation the muscles were washed three times in high K⁺ buffer and given a 90 min rest period during which they were stimulated only once every 10 min. In spite of the long resting period, the force during the second period with stimulation every 1 min only increased to 15 % of initial control force at 5.9 mm K⁺. On the other hand, when active Na⁺-K⁺ transport was subsequently stimulated by the addition of CGRP (10⁻⁷ m), a highly significant force recovery was achieved (P < 0.02), reaching values around 40 % of the control level (Fig. 3). The force recovery obtained by the addition of CGRP was of the same magnitude as that previously demonstrated in muscles exposed to 12.5 mm K⁺ (Andersen & Clausen, 1993; Clausen et al. 1993). This indicates that the reduced response to repetition of the stimulation at 1 min intervals was not due to failure of energy supplies. Moreover, in muscles that were returned to buffer with 5.9 mm K⁺ after the first period of stimulation at 1 min intervals, the tetanic force recovered within 10 min to more than 90 % of the control level determined at start of the experiment (data not shown).

The force recovery induced by increasing the frequency of pulse trains to one per minute was completely prevented by the addition of 10⁻³ m ouabain (3/3 observations, data not shown), indicating that it was elicited by activation of the Na⁺-K⁺ pump. This interpretation was further supported by the observation that in muscles incubated at 12.5 mm K⁺ for 70 min, a reduction in the interval between the 30 Hz pulse trains from 10 to 1 min led to a 30 % decrease in the intracellular Na⁺ content within 20 min (from 12.2 ± 0.8 μmol (g wet wt)⁻¹ in muscles stimulated every 10 min to 8.6 ± 0.5 μmol (g wet wt)⁻¹ in muscles stimulated every 1 min, n = 8/8, P < 0.005). In contrast, intracellular K⁺ content showed no change being 85 ± 2 and 84 ± 2 μmol (g wet wt)⁻¹ in muscles stimulated every 10 min and every 1 min, respectively.

Role of β-agonists and CGRP in the excitation-induced force recovery at high [K⁺]_o

We have previously demonstrated that in rat muscles incubated at 5.9 mm K⁺, electrical stimulation per se leads to an increase in the activity of the Na⁺-K⁺ pump by mechanisms that could not readily be related to an increase in [Na⁺]_i or the release of noradrenaline or CGRP from endogenous stores in the muscle preparation (Nielsen & Clausen, 1997). However, since noradrenaline as well as CGRP have been shown to stimulate the Na⁺-K⁺ pump (Clausen & Flatman, 1977; Andersen & Clausen, 1993), leading to force recovery in K⁺-inhibited muscles, a series of experiments was conducted to evaluate a possible contribution from these agents in the excitation-induced recovery of force observed in the present study. The stimulating effect of noradrenaline on the Na⁺-K⁺ pump is mediated via β-adrenoceptors and can therefore be suppressed by propranolol. However, in muscles incubated in buffer containing 12.5 mm K⁺, pre-treatment with propranolol (10⁻⁵ m) for 30 min produced no detectable change in the force recovery elicited by tetanic stimulation at 1 min intervals compared with contralateral control muscles (3/3 muscles, data not shown).

CGRP is known to be released from nerve endings in skeletal muscle in response to electrical stimulation (Uchida et al. 1990; Sakaguchi et al. 1991). Estimated from the decrease in the total CGRP content of the muscle preparation, it can be assumed that the stimulation pattern eliciting force recovery (2 s at 30 Hz at 1 min intervals) also produced a minor (24 %), but significant decrease in the CGRP content of the nerves in muscles incubated at a [K⁺]_o of 12.5 mm (Table 1). It is conceivable, therefore, that part of the excitation-induced force recovery was caused by activation of the Na⁺-K⁺ pump by CGRP released from the nerve endings of the muscle preparation. To test this possibility, capsaicin was used to reduce the CGRP content in the nerve endings in the muscles (Santicioli et al. 1992). Capsaicin was earlier shown to induce stimulation of ⁸⁶Rb uptake and net Na⁺ extrusion in rat soleus muscle (Andersen & Clausen, 1993) and to elicit force recovery in soleus muscles inhibited by high [K⁺]_o, possibly via a release of CGRP from nerve endings in the muscles. As shown in Table 1, capsaicin (5 × 10⁻⁶ m) reduced the CGRP content of muscles stimulated every 10 min at a [K⁺]_o of 12.5 mm by 50 %. As can be seen from Fig. 4, such addition of capsaicin to K⁺-inhibited muscles elicited a force recovery to 45 % of the initial control level (P < 0.005). Following this pre-treatment, the effect on force of reducing the interval between the 30 Hz pulse trains to 1 min was almost completely suppressed. Likewise, pre-treatment with capsaicin abolished the effect of excitation on intracellular Na⁺. Thus, in capsaicin-treated muscles 20 min of stimulation at 1 min intervals only reduced intracellular Na⁺ content by 5 % (from 13.3 ± 1.3 μmol (g wet wt)⁻¹ before stimulation, to 12.7 ± 1.8 μmol (g wet wt)⁻¹ after stimulation, n = 6/6, P > 0.70). Despite these effects of capsaicin, the stimulating effect of subsequent addition of CGRP (10⁻⁷ m) on contractile force was preserved (Fig. 4).

Table 1.

Effects of electrical stimulation and capsaicin on total CGRP content in muscles at a [K⁺]_o of 12.5 mm

Experimental conditions	CGRP (fmol (g wet wt)−1)	P
Control, stimulated every 10 min for 90 min	1870 ± 130 (12)
Stimulated every 10 min for 70 min, and then every 1 min for 20 min	1430 ± 130 (12)	< 0.025
Control, stimulated every 10 min for 140 min	1760 ± 200 (9)
Stimulated every 10 min for 140 min, with capsaicin for the last 70 min	870 ± 140 (6)	< 0.01

The table shows data from two experiments in which muscles were incubated at a [K⁺]_o of 12.5 mm for the intervals indicated. In one experiment a group of muscles was stimulated (2 s at 30 Hz) every 10 min for 70 min and then every 1 min for 20 min whereas the other group of muscles (control) was stimulated at 10 min intervals throughout. In the second experiment both groups of muscles were stimulated at 10 min intervals for 140 min. One of the groups was exposed to capsaicin (5 × 10⁻⁶m) for the last 70min of the incubation whereas the other group served as control. At the end of the experiments the muscles were frozen in liquid N₂ for determination of CGRP. Data are means ±s.e.m. with number of replicates (n) given in parentheses. P indicates significance of difference between treated and corresponding control groups.

Figure 4. Effect of capsaicin on excitation-induced force recovery in muscles at a [K⁺]_o of 12.5 mm.

Experimental conditions as described in legend to Fig. 3. ○, capsaicin (5 × 10⁻⁶ m) added after 100 min and washed away after 170 min as indicated. •, controls not treated with capsaicin. CGRP (10⁻⁷ m) was added to both groups of muscles after 290 min. Each point represents the mean of observations on four muscles with vertical bars denoting s.e.m. For sake of clarity s.e.m. is only shown once every 10 min.

To evaluate the possible role of CGRP in the excitation-induced force recovery further, the CGRP content of soleus muscles was reduced by denervation 7 days prior to the experiment (Kashihara et al. 1989). At the time of the experiment, the CGRP content of the innervated and the denervated contralateral muscles was 1100 ± 20 and 410 ± 40 fmol (g wet wt)⁻¹, respectively (3/3 muscles, P < 0.001). Figure 5 shows that increasing [K⁺]_o from 5.9 to 12.5 mm produced virtually the same relative reduction in force development in the denervated muscles and their contralateral controls. When the interval between stimulation was reduced to 1 min, however, the denervated muscles showed only a very modest force recovery (from 11 ± 4 to 15 ± 5 % of the force measured at 5.9 mm K⁺) compared with that observed in the innervated contralateral control muscles (from 10 ± 4 to 51 ± 9 % of the force measured at 5.9 mm K⁺). These results gave further support to the idea that CGRP released from nerve endings might contribute to the excitation-induced force recovery.

Figure 5. Effect of denervation on excitation-induced force recovery in muscles at a [K⁺]_o of 12.5 mm.

Experimental conditions as described in the legend to Fig. 3. Rats weighing 50–60 g were anaesthetized and denervated at the sciatic level on one side. After 7 days, muscles were prepared for measurement of force development. ○, denervated muscles. •, contralateral controls. Each point represents the mean of observations on four muscles with vertical bars denoting s.e.m. For sake of clarity, s.e.m. is only shown once every 10 min.

DISCUSSION

The present study demonstrates that in muscles where contractile force was depressed by exposure to high [K⁺]_o, repeated tetanic contractions led to substantial recovery of force. The capability of contractile activity to induce recovery of force depended on the level of [K⁺]_o. Thus, at 10 mm K⁺ tetanic stimulation every minute led to almost complete recovery of force whereas the recovery of force at 12.5 mm K⁺ was somewhat less and of shorter duration. This effect of excitation was seen both on twitch and tetanic force and suggests that active muscles may be less sensitive to elevated [K⁺]_o than resting muscles.

In muscles incubated at a [K⁺]_o of 10 mm and stimulated directly with pulses of 1 ms duration, the recovery of force was not affected by blockage of the neuromuscular transmission by tubocurarine. This demonstrates that the phenomenon is not a simple effect of improved excitability of the motor nerve but involved changes within the muscle fibres per se. On the other hand, since the recovery of force was also seen in muscles stimulated via the nerve using pulse durations of 0.02 or 0.04 ms it could not be related to some non-physiological effects of the pulse duration of 1 ms used in direct stimulation. Likewise, since TTX almost completely suppressed the force in muscles stimulated with 1 ms pulses it is unlikely that direct activation of the T-tubular charge sensor by the electrical pulses was involved in the force recovery.

Role of the Na⁺-K⁺ pump in the excitation-induced force recovery at high [K⁺]_o

It is well established that increased [K⁺]_o leads to depolarization with ensuing inactivation of Na⁺ channels and loss of excitability. It has previously been observed that stimulation of the Na⁺-K⁺ pump by hormones such as catecholamines, insulin and CGRP has a protective effect on contractile force in muscles exposed to high [K⁺]_o (Tomita, 1975; Clausen & Everts, 1991; Andersen & Clausen, 1993; Clausen et al. 1993; Cairns et al. 1995) and such an effect has also been demonstrated in muscles depressed by a combination of high [K⁺]_o and low [Na⁺]_o (Overgaard et al. 1997a). Studies of m waves in isolated rat muscles have shown that the effect of Na⁺-K⁺ pump stimulation most probably is related to an increased excitability of the muscle fibres (Overgaard et al. 1997b) caused by hyperpolarization of the plasma membrane and perhaps an increase in the chemical gradient for Na⁺ (Juel, 1988; Clausen & Everts, 1991; Clausen et al. 1993; Cairns et al. 1995; Overgaard et al. 1997a).

Several observations indicate that a similar stimulation of active Na⁺-K⁺ transport was involved in the excitation-induced force recovery demonstrated in the present study. Firstly, compared with hormonal stimulation, excitation of muscles is a very potent stimulus for the Na⁺-K⁺ pump and may increase the rate of active Na⁺-K⁺ transport substantially. This phenomenon has been observed both in vitro and in vivo (Hazeyama & Sparks, 1979; Fong et al. 1986; Juel, 1988; Hicks & McComas, 1989; Balog & Fitts, 1996; Nielsen & Clausen, 1997). Secondly, the recovery of force in the present study was, as in studies where the Na⁺-K⁺ pump was stimulated with hormones, associated with a significant reduction in intracellular Na⁺ content. Since electrical stimulation of muscles per se leads to an increased Na⁺ influx, the reduction in intracellular Na⁺ is indicative of a net increase in the active extrusion of the ion. Thirdly, the excitation-induced recovery of force was fully suppressible by the addition of ouabain which directly indicates that the Na⁺-K⁺ pump was involved.

Studies on muscles exposed to elevated [K⁺]_o have demonstrated an almost sigmoidal relation between membrane potential and tetanic force with a very steep drop in force when the membrane potential is reduced from around −60 to around −55 mV (Cairns et al. 1995, 1997). Due to the steepness of this relationship, a small (2–5 mV) hyperpolarization of the membrane elicited by increased Na⁺-K⁺ pump activity can potentially lead to substantial recovery of force in K⁺-depressed muscles. In this context, it is of importance that excitation of rat muscle in vivo can lead to a hyperpolarization of several millivolts (Hicks & McComas, 1989). Taken together these observations suggest that the excitation-induced recovery of force in K⁺ inhibited muscles takes place via an increase in the activity of the Na⁺-K⁺ pump leading to partial restoration of the membrane potential.

The excitation-induced force recovery at a [K⁺]_o of 12.5 mm could only be maintained for around 15 min, after which a slow decline in force took place (Fig. 3). Since the addition of CGRP to muscles during this decline in force improved force production (Fig. 3), it is unlikely that the inability to maintain force during stimulation at 1 min intervals was related to metabolic exhaustion of the muscle fibres. Likewise, it could not solely be related to an excitation-induced reduction in muscle CGRP content because the recovery of force induced by the addition of CGRP was also only maintained for a limited period of time (Fig. 3). A transient force recovery is also seen when muscles are stimulated with catecholamines at a [K⁺]_o of 12.5 mm (Clausen et al. 1993). One possible explanation is that because the increase in the activity of the Na⁺-K⁺ pump leads to a drop in [Na⁺]_i, the stimulatory effect of excitation or hormones will eventually be dampened. In contrast, at a [K⁺]_o of 10 mm the excitation-induced enhancement of force could be maintained for at least 60 min. This indicates that despite the reduction in [Na⁺]_i, the activity of the Na⁺-K⁺ pump may be sufficient to protect the excitability of the muscle fibres at a [K⁺]_o of 10 mm whereas a higher level of Na⁺-K⁺ pump activity may be necessary to overcome the more extensive depolarization in muscles incubated at a [K⁺]_o of 12.5 mm.

Mechanisms for the excitation-induced activation of the Na⁺-K⁺ pump

Excitation of muscles is associated with an increased influx of Na⁺ and if intracellular Na⁺ accumulates this will constitute a potent stimulus for the Na⁺-K⁺ pump. The stimulation regimen leading to recovery of force in K⁺-inhibited muscles led, however, to a decrease in [Na⁺]_i. Consequently, the activation of the Na⁺-K⁺ pump associated with the electrical stimulation must have been caused by mechanisms other than increased Na⁺ influx. Recently, we demonstrated that a similar activation of the Na⁺-K⁺ pump not caused by increased [Na⁺]_i takes place during electrical stimulation of muscles incubated in standard buffer containing 5.9 mm K⁺ (Nielsen & Clausen, 1997). The mechanism for this excitation-induced activation of the Na⁺-K⁺ pump is at present unknown. Several hormones including CGRP and catecholamines have been shown to activate the Na⁺-K⁺ pump in skeletal muscle. CGRP and catecholamines most probably act via an increase in the intracellular concentration of cyclic AMP (Clausen, 1986; Kobayashi et al. 1987; Andersen & Clausen, 1993). CGRP is known to be released from nerve endings in skeletal muscle in response to electrical stimulation (Uchida et al. 1990; Sakaguchi et al. 1991). Likewise, it could be envisaged that excitation induces the release of catecholamines from endogenous stores in the muscle preparation. It is possible, therefore, that a local release of such hormones contributes to the excitation-induced activation of the Na⁺-K⁺ pump. Indeed, based on experiments on a rat nerve-hemidiaphragm preparation, Uchida et al. (1990) concluded that excitation can produce an increase in the cyclic AMP content of muscles fibres which is caused by a release of CGRP from nerve endings in the muscle.

The excitation-induced recovery of force in K⁺-inhibited muscles was insensitive to pre-incubation of the muscles with propranolol and a role for catecholamines in the activation of the Na⁺-K⁺ pump, therefore, seems unlikely. In contrast, two observations suggest a role for CGRP in the excitation-induced force recovery. Firstly, pre-incubation with capsaicin, which led to a 50 % reduction in the CGRP content of the muscles, almost completely prevented the excitation-induced recovery of force in K⁺-inhibited muscles, whereas the recovery of force induced by the addition of CGRP was unaffected. Secondly, denervation of muscles 7 days prior to the experiment led to a similar reduction in the excitation-induced force recovery which could be secondary to the observed reduction of the CGRP content of the muscles. Together these experiments indicate that the excitation-induced force recovery depended on the CGRP content of the muscles. This notion tallies with the effect of two consecutive periods with electrical stimulation every 1 min (Fig. 3). During the first 70 min period with stimulation at 1 min intervals the force recovered to 65 % of control force and this was associated with a 24 % reduction in the CGRP content of the muscles (Table 1). When a second identical period of stimulation at 1 min intervals was started 90 min later, force only recovered to 15 % of control force. This reduction in recovery may, thus, be related to a diminished availability of CGRP. In contrast, as can be seen from the persistence of the response to CGRP (Fig. 3), it could not be attributed to exhaustion of the muscles. These results suggest that CGRP released from nerve endings within the muscle preparation contributes to the excitation-induced force recovery by activating the Na⁺-K⁺ pump via CGRP receptors on the muscle fibres. This conclusion is supported by the observation that in denervated muscles, where the CGRP content presumably is reduced, direct electrical stimulation fails to induce a decrease in [Na⁺]_i (Nielsen & Clausen, 1997).

The study of Nielsen & Clausen (1997) also demonstrated that in control muscles stimulated via the nerve using 0.02 ms pulses, the inhibition of the motor end plate by tubocurarine completely abolished the activation of the Na⁺-K⁺ pump. Thus, the excitation-induced activation of the Na⁺-K⁺ pump could not be due to excitation of the motor nerve per se. In keeping with this, Sakaguchi et al. (1991) have demonstrated that in the presence of tubocurarine, electrical stimulation of the nerve of rat soleus muscles using pulse intensities 3 times the threshold for the motor nerves produced no release of CGRP from the muscle preparation. In contrast, a significant release of CGRP was obtained with stimulation intensities 50 to 100 times the threshold intensity, which was sufficient to excite most sensory nerves as well. It is, thus, possible that in our experiments with 0.02 ms pulses in the presence of tubocurarine only the motor nerves were excited and, therefore, only a minor or no release of CGRP took place. The activation of the Na⁺-K⁺ pump in muscles stimulated via the nerve but not treated with tubocurarine (Nielsen & Clausen, 1997) could then be explained by a release of CGRP from sensory nerves in response to some activity-induced changes within the muscle, such as the production of force or an excitation-induced rise in interfibre K⁺ concentration. In this context, it has been shown that elevated [K⁺]_o, reduced pH and the presence of lactate at physiological concentrations (5 to 10 mm), which are all factors associated with muscle activity, elicit the release of CGRP from nerve tissue (Uchida et al. 1990; Sakaguchi et al. 1991; Santicioli et al. 1992; Wang & Fiscus, 1997). It is, thus, possible that the excitation-induced activation of the Na⁺-K⁺ pump (Nielsen & Clausen, 1997) is due to CGRP but that the release of CGRP primarily takes place from sensory nerves in response to activity in muscle fibres. In experiments with 0.02 ms pulses, tubocurarine prevents the activation of the muscle. The release of CGRP from sensory nerves, and thus the stimulation of the Na⁺-K⁺ pump, would therefore be missing.

Physiological significance of the excitation-induced activation of the Na⁺-K⁺ pump

It is well documented that exercise is associated with a loss of potassium from the working muscles which may lead to a substantial increase in [K⁺]_o with ensuing depolarization of muscle fibres. Thus, [K⁺]_o values from 10 to 12 mm have been reported in contracting muscles (Gebert, 1972; Hirche et al. 1980). At the same time [Na⁺]_i may increase leading to a reduction in the chemical gradient for Na⁺ (Sreter, 1963; Juel, 1986). Since such changes tend to reduce the excitability of the muscle membrane it has been suggested that reductions in the chemical gradients for Na⁺ and K⁺ contribute to fatigue at least during high intensity exercise (Sjøgaard, 1990). The present study suggests, however, that an increase in active Na⁺-K⁺ transport associated with muscle activity provides substantial protection against the depressing effects of increased [K⁺]_o, possibly by hyperpolarization of the sarcolemma (Hicks & McComas, 1989) and T-tubular membranes. This mechanism may also explain the observation that continued localized activity (e.g. in the muscles of the forearm) can prevent the development of paralysis in the active muscle during otherwise generalized attacks of hyperkalaemic periodic paralysis (McArdle, 1962). Due to the continued activity the Na⁺-K⁺ pump may be stimulated with ensuing protection of the excitability in the contracting muscles. Together these findings suggest that activation of the Na⁺-K⁺ pump in contracting muscles counterbalances the depressing effect of reductions of the chemical gradients for Na⁺ and K⁺ and, thereby, significantly delays the onset of fatigue caused by loss of excitability.