Analysis of exercise-induced Na + K + exchange in rat skeletal muscle in vivo

Transcription

1 Exp Physiol pp Experimental Physiology Research Paper Analysis of exercise-induced Na + K + exchange in rat skeletal muscle in vivo K.T.Murphy,O.B.NielsenandT.Clausen Institute of Physiology and Biophysics, University of Aarhus, Århus, Denmark We aimed to quantify the Na + K + exchange occurring during exercise in rat skeletal muscle in vivo. IntracellularNa + and K + content, Na + permeability ( 22 Na + influx), Na + K + pump activity (ouabain-sensitive 86 Rb + uptake) and Na + K + pump α 2 subunit content ([ 3 H]ouabain binding) were measured. Six-week-old rats rested (control animals) or performed intermittent running for 1 6 min and were then killed or were killed at 15 or 9 min following 6 min exercise. In the soleus muscle, intracellular Na + was 8% higher than in control rats after 6 min exercise, was still elevated (38%) after 15 min rest and returned to control levels after 9 min rest. Intracellular K + showed corresponding decreases after 15 6 min exercise, returning to control levels 9 min postexercise. Exercise induced little change in Na + and K + in the extensor digitorum longus muscle (EDL). In soleus, the exercise-induced rise in Na + and reduction in K + were augmented by pretreatment with ouabain or by reducing the content of muscular Na + K + pumps by prior K + depletion of the animals. Fifteen minutes after 6 min exercise, ouabain-sensitive 86 Rb + uptake in the soleus was increased by 3% but was unchanged in EDL, andtherewasnoeffectofexerciseon[ 3 H]ouabain binding measured in vitro or in vivo in either muscle. In conclusion, in the soleus, in vivo exercise induces a rise in intracellular Na +, which reflects the excitation-induced increase in Na + influx and leads to augmented Na + K + pump activitywithoutapparentchangeinna + K + pump capacity. (Received 26 February 28; accepted after revision 2 June 28; first published online 27 June 28) Corresponding author K. T. Murphy: Department of Physiology, The University of Melbourne, Melbourne 31, Victoria, Australia. ktmurphy@unimelb.edu.au In skeletal muscle, each action potential is elicited by a Na + influx and K + efflux. These passive fluxes are counteracted by the Na + K + -ATPase (Na + K + pump), which actively transports Na + back out of and K + back into the cell, allowing maintenance of the transmembrane Na + and K + gradients. However, during periods of frequent action potentials, the accelerated passive ion fluxes exceed the activity of the Na + K + pumps, resulting in a rundown of the sarcolemmal Na + and K + concentration gradients and loss of membrane excitability (Sejersted & Sjøgaard, 2; Clausen, 23). Precise quantification of the muscular Na + and K + fluxes occurring during exercise and also in the recovery period following exercise has proved problematic. Early studies in humans calculated intracellular Na + and K + concentrations ([Na + ] i,[k + ] i ) based on measurement of femoral arterial venous [Na + ]and[k + ]differencesand correction for extracellular water content as measured using [ 3 H]inulin (Sjøgaard et al. 1985). Using this technique, it was calculated that 6 min of intense knee extensor exercise elevated [Na + ] i from 13 to 23 mm and reduced [K + ] i from 162 to 129 mm (Sjøgaard et al. 1985). After 3 min recovery, [Na + ] i and [K + ] i had returned to pre-exercise levels. In more recent studies, microdialysis probes have been inserted into human gastrocnemius or vastus lateralis muscle for measurements of exercise-induced changes in interstitial [Na + ]and[k + ], and these studies have found increases in interstitial [K + ]from 4 to12mmand reductions in interstitial [Na + ]from 14 to 13 mm (Nordsborg et al. 23; Street et al. 25). Although the exercise-induced increase in interstital [K + ] returned to pre-exercise levels by 15 min recovery, interstitial [Na + ] remained below pre-exercise levels by 25 min recovery (Street et al. 25). However, the microdialysis technique produces variable results (Nordsborg et al. 23), is dependent on extracellular volume, which is difficult to measure (Sejersted & Sjøgaard, 2), and may also be associated DOI: /expphysiol

2 125 K. T. Murphy and others Exp Physiol pp with muscle damage, which causes increases in interstitial [K + ] (Sejersted & Sjøgaard, 2). Furthermore, during recovery from in vitro electrical stimulation of muscles, there is a decrease in intracellular Na + content below control levels (Juel, 1986; Everts & Clausen, 1994; Nielsen & Clausen, 1997). Since the Na + K + pump is regulated by systemic factors (Clausen, 23), which are absent in in vitro electrical stimulation experiments, and since the diffusion of ions, metabolites and hormones is impaired in vitro compared with in vivo, itispossiblethatthe response of intracellular Na + content following in vivo exercise may be different to that following in vitro electrical stimulation. Moreover, in vivo, temperature, perfusion and the availability of numerous metabolites and hormones provide conditions that are not adequately known and therefore not simulated in in vitro experiments. However, owing to the limitations of previously used techniques, this has not been adequately investigated. Therefore, there is a clear need for the development of a novel technique that more precisely quantifies the changes in intracellular Na + and K + occurring during exercise and in the recovery period following exercise in skeletal muscle in vivo. Little is also known regarding the effects of in vivo exercise on active Na + K + transport. When measured by the rates of ouabain-sensitive 86 Rb + uptake and 22 Na + efflux, Na + K + pump activity has been shown to be substantially elevated following in vitro electrical stimulation in rat muscle (Everts & Clausen, 1994; Nielsen & Clausen, 1997). However, whether in vivo exercise also increases Na + K + pumpactivitywhenmeasured by ouabain-sensitive 86 Rb + uptake is unknown. Such information would greatly improve our understanding of the regulation of Na + and K + in muscle following exercise. The primary aim of this study was therefore to develop a technique that more precisely quantifies the changes in intracellular Na + and K + content occurring during and following in vivo exercise in rat skeletal muscle and to identify the mechanism responsible for any exerciseinduced changes in these parameters. The secondary aim was to investigate whether in vivo exercise increases Na + K + pump activity when measured by ouabain-sensitive 86 Rb + uptake and to determine whether any increase was due to an elevation in the content of functional Na + K + pumps. The third aim was to determine whether reduced Na + K + pump capacity would increase the exerciseinduced changes in intracellular Na + and K + contents. Methods Animals Male or female Wistar rats were selected for the study at 4 weeks of age and randomly separated into control or running groups (Run). Animals were housed in a box of the following dimensions: length, 39 cm; width, 23 cm; and depth, 15 cm; were fed ad libitum and maintained in a temperature-controlled environment (21 C) with constant day length (12 h). All animal experiments were conducted in accordance with the recommendations of the European Convention for the Protection of Vertebrate Animals used for Experimentation and after permission from the Danish Animal Experiments Inspectorate. The method of killing was approved by the University Animal Welfare Officer. Treadmill running The intermittent running trial was carried out when animals were 6 weeks old and weighed 149 ± 19g (mean± S.D.) for the control and 146 ± 15 g for the Run group (n = 5 and 64, respectively, P =.22). One week before the intermittent running trial, both control and Run rats were accustomed to treadmill running by performing two familiarization tests on a rodent treadmill (Economical Exercise Treadmill Model Exer-4, Columbus Instruments, Columbus, OH, USA), with each test consisting of running for 3 min at 12 m min 1 on a +1% gradient. Tests were separated by 2 days. Three days after the final familiarization test, run animals completed an intermittent running trial, consisting of twelve 5 min running intervals, performed at 12 m min 1 andona +1% gradient, with each interval separated by 1 min rest (total running time, 6 min; total running distance, 72 m). Based on the treadmill speed and gradient and the strain of the rats, it was estimated that animals were exercising at an intensity corresponding to 6 7% of their peak oxygen uptake (Bedford et al. 1979; Armstrong et al. 1983). Animals were killed by cervical dislocation, followed by decapitation, either immediately or at 15 or 9 min following the final running interval. Additional experiments were carried out to investigate the effects of 1 or 3 min of running. In these experiments, run animals completed only two or six 5 min running intervals, respectively, and were killed immediately following the final running interval. For all experiments, control animals that did not perform the intermittent running trial were killed on the same day as run animals. A scheme of the running protocols used for investigation of running effects on intracellular Na + and K + contents is shown in Fig. 1. Preparation of muscles Immediately after the animals were killed, intact soleus (predominantly slow-twitch fibres) and extensor digitorum longus (EDL) muscles (predominantly fasttwitch fibres) were dissected out. For experiments measuring intracellular Na + and K + contents, 86 Rb + uptake and 22 Na + influx, intact muscles were mounted

3 Exp Physiol pp Running exercise and Na + K + exchange in rat muscle 1251 at resting length at 3 C in standard Krebs Ringer bicarbonate buffer (KR; ph 7.4), containing the following (mm): NaCl, 25.1 NaHCO 3, 2.8 KCl, 1.2 KH 2 PO 4, 1.2 MgSO 4, 1.3CaCl 2 and 5. D-glucose, which was bubbled continuously with a mixture of 95% O 2 and 5% CO 2. For experiments measuring in vitro [ 3 H]ouabain binding site content, tendons were removed and the muscle was rapidly frozen in liquid N 2 andstoredat 5 C until later analysis. The time elapsed from the cessation of exercise until the muscle was transferred to the tubes containing KR (for incubation) or ice-cold Tris sucrose buffer (for determination of intracellular Na + and K + )or was frozen for later determination of in vitro [ 3 H]ouabain binding site content was 387 ± 64 s (n = 32 muscles). Measurement of intracellular Na + and K + contents Muscles were mounted at resting length; some were then incubated at 3 C for 2 min in standard KR without or with 1 mm bumetanide, to inhibit Na + Cl cotransport (Dørup & Clausen, 1996). For measurement of intracellular ions, the muscles were transferred to icecold Na + -free Tris sucrose buffer and underwent four 15 min washouts to remove extracellular Na +. Following washout, muscles were blotted, tendons cut off, muscle wet weight was determined, and the muscles were soaked overnight in.3 M trichloroacetic acid (TCA) to give complete extraction of ions from the tissue (Clausen et al. 1993). Muscle wet weight (determined after washout) was not significantly altered with exercise in either the soleus (control, 56.4 ± 6.8 mg (mean ± S.D.); 1 min Run, 58.2 ± 12. mg; 3 min Run, 52.2 ± 9. mg; 6 min Run, 52.7 ± 5.3 mg; 6 min Run + 15 min Rest, 58. ± 7.8 mg; 6 min Run + 9 min Rest, 52.8 ± 3.8 mg, P =.21, n = 6 38 muscles) or the EDL (control, 53.5 ± 9.1 mg; 1 min Run, 53.3 ± 6.2 mg; 3 min Run, 48.1 ± 6.8 mg; 6 min Run, 47.6 ± 5.2 mg; 6 min Run + 15 min Rest, 54.7 ± 9.4 mg; 6 min Run + 9 min Rest, 5. ± 4.2 mg, P =.14, n = 6 38 muscles). The Na + and K + contents in the TCA extract were measured by flame photometry (FLM3, Radiometer, Copenhagen, Denmark) with lithium as internal standard. Previous experiments have shown that part of the intracellular Na + is lost during the four 15 min washouts at C (Everts & Clausen, 1992). Control experiments, in which the Na + content was determined in muscles that had undergone washout in ice-cold Na + - free Tris sucrose buffer for 3 18 min, showed that the loss of intracellular Na + during the standard 6 min of washout could be corrected for by applying correction factors of 1.22 and 1.33 in soleus and EDL, respectively. These correction values were somewhat lower than those used for muscles from 4-week-old animals, of 1.46 for the soleus (Everts & Clausen, 1992) and 1.59 for the EDL (Murphy et al. 26), indicating slower efflux of intracellular Na + in the larger muscles from 6-week-old animals. In contrast, the loss of K + during the washout in ice-cold Tris sucrose buffer is minimal (Everts & Clausen, 1992). Effect of ouabain injection on intracellular Na + and K + contents To investigate whether reducing Na + K + pump activity altered the exercise-induced changes in intracellular Na + and K + content, we used ouabain. Six-weekold animals were injected intraperitoneally with saline containing ouabain (2 nmol (g body weight) 1 ) and rested for 3 min. At 2 nmol (g body weight) 1, we have previously shown that, after 6 min, 55 and 44% of Na + K + pumps in rat soleus and EDL, respectively, have bound ouabain and are therefore blocked (Clausen et al. 1982). The Run animals then performed 6 min of intermittent treadmill running as described previously, and control animals rested for a matched duration. Immediately following the final running interval, animals were killed, the soleus and EDL muscles dissected and measured for intracellular Na + and K + content as described above. Effect of K + depletion on intracellular Na + and K + contents Potassium depletion was shown to induce downregulation of the content of Na + K + pumps in rat skeletal muscle (Nørgaard et al. 1981). To investigate whether K + depletion altered the exercise-induced changes in intracellular Na + and K + content, 6-week-old animals were fed Altromin no. 521 K + -depleted fodder (Altromin GmbH & Co., Lage, Germany) for a period of 1 days. The 1 min Run 3 min Run 6 min Run 6 min Run + 15 min Rest 6 min Run + 9 min Rest Time (min) Run Rest Muscle dissection Figure 1. Schematic diagram of running protocol for investigation of the effects of exercise on intracellular Na + and K + contents Animals ran for 1, 3 or 6 min at a treadmill speed of 12 m min 1 (filled bars), followed by various lengths of resting periods (dashed lines). Arrows denote time at which animals were killed and muscles dissected.

4 1252 K. T. Murphy and others Exp Physiol pp K + content of the K + -depleted fodder was 2 mmol kg 1. Potassium-supplemented animals were fed the same fodder soaked with KCl solution to give a final K + content of 26 mmol kg 1. All animals were given distilled water, and metal grids were positioned in the bottom of the cages to prevent the animals from having access to urine and faeces. Run animals from each group then performed 6 min of intermittent treadmill running as described previously, and control animals rested for a matched duration. Immediately following the final running interval, animals were killed, the soleus and EDL muscles dissected and measured for intracellular Na + and K + content as described above. Measurement of 22 Na + influx The effect of exercise on the resting influx of Na + was determined by measuring the rate of 22 Na + uptake into muscle fibres during 2 min incubation in KR containing 22 Na + (2 μci ml 1 ), as previously described (Clausen & Kohn, 1977). After incubation, the muscles underwent four 15 min washouts in ice-cold Na + -free Tris sucrose buffer to remove extracellular 22 Na +. Following washout, muscles were blotted, tendons cut off, muscle wet weight determined, and the activity of the 22 Na + retained in the muscles was determined by γ-counting. After correction for the loss of intracellular 22 Na + during the washout, the Na + influx was calculated from the specific activity of 22 Na + in the incubation buffer (Clausen & Kohn, 1977). Measurement of 86 Rb + uptake rate and Na + K + pump activity Previously, 86 Rb + hasbeenshowntobeareliabletracerfor determination of Na + K + pump-mediated K + transport (Clausen et al. 1987). Specifically, the ouabain-sensitive component of 86 Rb + uptake is used as a measure of Na + K + pump activity. Thus, whereas 22 Na + influx is a measure of the permeability of the cells to Na +,ouabainsensitive 86 Rb + uptake measures the active uptake of K + via the Na + K + pumps. To investigate the effects of exercise on ouabain-sensitive 86 Rb + uptake, muscles were equilibrated for 1 min in KR without or with 1mM ouabain and were then incubated for 1 min in KR containing 86 Rb + (.2 μci ml 1 ) without or with 1mM ouabain. All muscles then underwent four 15 min washouts in ice-cold Na + -free Tris sucrose buffer to remove extracellular 86 Rb +. Following washout, muscles were blotted, tendons cut off, muscle wet weight was determined, and then the muscles were soaked in.3 M TCA. Muscles were then taken for counting of 86 Rb + activity by Cerenkov radiation in aβ-counter. The amount of 86 Rb + activity retained after the washout was calculated and expressed as the relative uptake of the 86 Rb + activity from the incubation medium by the muscle. The K + uptake (expressed as nmol (g wet wt) 1 min 1 )wasthen calculated by converting the relative uptake of 86 Rb + to K + using the concentration of K + in the incubation medium (for details, see Buchanan et al. 22). The study of Buchanan et al. (22) showed that the loss of intracellular 86 Rb + during the washout in ice-cold Na + -free buffer is only around 5% and therefore no correction was made for 86 Rb + uptake values. Measurement of in vitro [ 3 H]ouabain binding site content Muscle content of the Na + K + pump α 2 isoform was determined using the vanadate-facilitated [ 3 H]ouabain binding method, in which muscles are sectioned before incubation (Nørgaard et al. 1983). Previously, it has been reported that the values for [ 3 H]ouabain binding obtained using this method are not significantly different from those obtained in intact muscles (Nørgaard et al. 1983; McKenna et al. 23). Combined with the finding that [ 3 H]ouabain only binds to the outer membranes of intact muscle fibres (Clausen & Hansen, 1974), this shows that the vanadate-facilitated [ 3 H]ouabain binding method provides a reliable measure of the [ 3 H]ouabain binding sites in the outer membranes of the muscles. We can, however, still not exclude the existence of a pool of Na + K + pumps located in intracellular vesicles in such a way that [ 3 H]ouabain cannot gain access to the ouabain binding sites. Nonetheless, only 3% of Na + K + pump α 2 isoforms are reported to be located in intracellular membranes of rat skeletal muscles (Zheng et al. 28). For the assay, approximately 2 mg of muscle was divided into four samples (each 5 mg), and washed twice for 1 min at 37 C in Tris vanadate buffer (25 mm sucrose, 1 mm Tris-HCl, 3 mm MgSO 4 and 1 mm NaVO 4 ; ph ). Muscle samples were then incubated for 12 min at 37 C in the same buffer with the addition of [ 3 H]ouabain (1 6 M, 2μCi ml 1 ) and were then washed for four times for 3 min in ice-cold vanadate buffer to remove any unbound [ 3 H]ouabain. Following washout, samples were blotted, the four pieces from each muscle sample were combined, muscle wet weight was determined,andthemusclesweredriedovernightat6 C. The following day, muscle dry weight was determined and the samples were soaked overnight in 1 ml.3 M TCA containing.1mm ouabain as carrier. Previous studies have shown that drying of muscles does not affect the extraction and recovery of [ 3 H]ouabain counts during the soaking in TCA (McKenna et al. 23). After overnight soaking, 5 μl of the TCA extract was removed for counting with 2.5 ml of scintillation cocktail (Opti-Fluor, Packard, Perkin Elmer, Boston, MA, USA) The content of [ 3 H]ouabain binding sites was calculated on the basis of both the sample wet and dry weights and the specific

5 Exp Physiol pp Running exercise and Na + K + exchange in rat muscle 1253 activity of the incubation medium. The final [ 3 H]ouabain binding site content was then calculated by subtracting the non-specific [ 3 H]ouabain uptake measured using vanadate buffer containing an excess of unlabelled ouabain (Nørgaard et al. 1983) and multiplying by a correction factor of 1.33 to allow for impurity of the [ 3 H]ouabain, loss of specifically bound [ 3 H]ouabain during washout and incomplete saturation. The [ 3 H]ouabain binding site content intra-assay coefficient of variation was 1.3 and 1.47% for the soleus and EDL, respectively (n = 68). Recently, it was argued that because of possible trafficking of Na + K + pumps between the outer membrane and a possible intracellular pool of Na + K + pumps, prolonged incubation with ouabain at physiological temperatures will cause binding to all pools of Na + K + pumps in the muscle fibres (Benziane & Chibalin, 28). Therefore, it has been proposed that incubation at a lower temperature (i.e. 18 C) should be used to prevent Na + K + pump trafficking. However, since the binding rate of ouabain to the Na + K + pumps depends on their activity, it is necessary to increase the incubation time dramatically if the temperature is lowered below 3 C. Control experiments performed at 18 Cshowed that in the soleus, 5 h of incubation was necessary to obtain [ 3 H]ouabain steady-state binding. Thus, while incubation at lower temperatures may slow down the rate of potential trafficking of membrane proteins, at the same time it greatly increases the incubation time, which per se favours possible trafficking or cell damage. Moreover, in muscles that had been equilibrated with [ 3 H]ouabain so as to reach almost complete saturation, a subsequent washout of [ 3 H] activity measured at C showed a linear decrease after 9 min and no increase in the washout rate when the muscles were cut transversely into three segments (Clausen & Hansen, 1974). If [ 3 H]ouabain had accumulated in the cytoplasm, this experiment would have shown a larger leakage of [ 3 H] activity. Measurement of in vivo [ 3 H]ouabain binding site content In an attempt to address the criticism raised against long incubation periods at physiological temperatures causing Na + K + pump internalization, as discussed above, we performed additional experiments to investigate the effects of exercise on in vivo [ 3 H]ouabain binding, using a procedure detailed previously (Clausen et al. 1982). Briefly, control animals or those that had performed 6 min intermittent treadmill running were injected intravenously with saline (154 mm NaCl) containing [ 3 H]ouabain (45 μci ml 1 ) and unlabelled ouabain (12.5 nmol ouabain (g body weight) 1 ). After 15 min, animals were killed, samples of venous blood collected from the vessels, and intact soleus and EDL muscles were dissected while continuously washed with ice-cold saline. Muscles were then transferred into ice-cold K + -free KR and washed four times for 3 min to remove unbound ouabain from the extracellular phase. Following washout, muscles were blotted on ice-cold filter paper, tendons removed and muscle wet weight was determined. Muscles were soaked overnight in 1 ml.3 M TCA and.1 m M ouabain, and the content of [ 3 H]ouabain binding sites was determined as described above. Previous studies have shown that when injected intraperitoneally, [ 3 H]ouabain binding reaches a maximum within 15 min in rat soleus and EDL (Clausen et al. 1982). Furthermore, similar results were found for [ 3 H]ouabain binding in rat soleus and EDL when measured in vitro compared with in vivo (Clausen et al. 1982). Chemicals and isotopes All chemicals were of analytical grade. Ouabain and bumetanide were purchased from Sigma Chemicals (St Louis, MO, USA). [ 3 H]Ouabain (2 45 μci ml 1 buffer or saline), 86 Rb + (.2 μci ml buffer 1 )and 22 Na + (2 μci ml buffer 1 ) were from Amersham International (Aylesbury, UK). Statistical analysis All data are presented as means with S.D. The statistical differences between two groups were analysed using Student s unpaired t test. The statistical difference between three or more groups was analysed using a oneway ANOVA. Differences were located with Student Newman Keuls post hoc test. Significance was accepted at P <.5. Results Effect of exercise on intracellular Na + and K + contents In the soleus muscle, intracellular Na + content was elevated by 18 (P <.1), 5 (P <.1) and 8% (P <.1) compared with control values immediately following 1, 3 and 6 min of exercise, respectively (Fig. 2A). At 15 min following the cessation of 6 min of exercise, intracellular Na + content remained 38% higher than control values (P <.1) and returned to levels not significantly different from control values at 9 min following the cessation of exercise (P =.37, Fig. 2A). Intracellular K + content was 6 (P <.1) and 11% (P <.1) lower than control values immediately following 3 and 6 min of exercise, respectively, and remained 4% (P <.1) lower than control values at 15 min following the cessation of 6 min of exercise (Fig. 2B). When expressed in μmol (g wet weight 1 ), thedecreaseinintracellulark + content was generally equivalent to the increase in intracellular Na + content.

6 1254 K. T. Murphy and others Exp Physiol pp Intracellular K + contents after 1 min of exercise (P =.18) and 9 min following the cessation of 6 min of exercise (P =.64) were not significantly different from control levels (Fig. 2B). In the EDL muscle, 1, Intracellular Na + content Intracellular K + content Control * 1 min Run * * 3 min Run * *# * 6 min Run *# 6 min Run + 15 min Rest # # 6 min Run + 9 min Rest Figure 2. Effect of exercise on intracellular Na + (A) andk + contents (B) in rat soleus Animals did not run (control) or performed 1, 3 or 6 min of intermittent treadmill running and were killed either immediately (1 min Run, 3 min Run, 6 min Run) or at 15 or 9 min following the cessation of exercise (6 min Run + 15 min Rest, 6 min Run + 9 min Rest). Dissected soleus muscles were mounted at resting length and either incubated for 2 min in standard KR and then washed out four times for 15 min in ice-cold Na + -free Tris sucrose buffer [control (n = 23), 6 min Run + 15 min Rest and 6 min Run + 9 min Rest] or immediately washed out [control (n = 13), 1 min Run, 3 min Run and 6 min Run]. Following washout, muscles were blotted, tendons removed, and the muscles weighed and taken for flame photometric analysis of Na + and K + content. Values for Na + content were multiplied by 1.22 to correct for the loss of intracellular Na + during the washout. Data are means with S.D.; n = 6 38 muscles. P <.1 versus control, P <.1 versus 1 min Run, P <.5 versus 3 min Run, #P <.1 versus 6 min Run, P <.1 versus 6 min Run + 15 min Rest. 3 or 6 min of exercise did not induce any significant change in intracellular Na + content (all P >.5, Fig. 3A). However, at 15 min following the cessation of 6 min of exercise, intracellular Na + content was 19% (P <.1) lower than control values and returned to levels not significantly different from control values at 9 min following exercise (P =.9, Fig. 3A). Intracellular K + content was 7% (P <.1) higher than control values immediately following 1 min of exercise, but was not Intracellular Na + content Intracellular K + content Control * 1 min Run 3 min Run 6 min Run *# 6 min Run + 15 min Rest 6 min Run + 9 min Rest Figure 3. Effect of exercise on intracellular Na + (A) andk + contents (B) in rat EDL Animals did not run (control) or performed 1, 3 or 6 min of intermittent treadmill running and were killed either immediately (1 min Run, 3 min Run, 6 min Run) or at 15 or 9 min following the cessation of exercise (6 min Run + 15 min Rest, 6 min Run + 9 min Rest). Muscles were treated as described in the legend to Fig. 2. Values for Na + content were multiplied by 1.33 to correct for the loss of intracellular Na + during the washout. Data are means with S.D.; n = 6 38 muscles. P <.1 versus control, P <.5 versus 1 min Run, #P <.1 versus 6 min Run.

7 Exp Physiol pp Running exercise and Na + K + exchange in rat muscle 1255 significantly different from control levels immediately following 3 or 6 min of exercise, or at 15 or 9 min following the cessation of 6 min of exercise (Fig. 3B). Effect of bumetanide Incubation with bumetanide (2 min, 1 mm), an inhibitor of Na + Cl cotransport (Dørup & Clausen, 1996), had no effect on intracellular Na + content in either rat soleus (in μmol (g wet weight) 1 : control, 12.1± 1.6 versus bumetanide, 11. ±.4, P =.12; 6 min Run + 15 min Rest, 16.7 ± 2.2 versus 6 min Run + 15 min Rest + bumetanide, 16.1 ± 2.1, P =.63, all n = 6 muscles) or EDL (in μmol (g wet weight) 1 : control, 9.6 ±.6 versus bumetanide, 9.1 ± 1.2, P =.35; 6 min Run+ 15 min Rest, 9.2 ±.5 versus 6 min Run + 15 min Rest + bumetanide, 9.1 ± 1.1, P =.84, all n = 6 muscles). Effect of exercise on 22 Na + influx Sixty minutes of exercise did not significantly alter resting 22 Na + influx measured over a 2 min period during subsequent incubation in either the soleus (P =.3, Fig. 4A)ortheEDL(P =.16, Fig. 4B). Effect of exercise on 86 Rb + uptake In the soleus muscle, total 86 Rb + uptake was elevated by 21% (P <.1) at 15 min following the cessation of 6 min of exercise, and returned to control levels at 9 min following exercise (P =.27, Fig. 5A). Ouabain-sensitive 86 Rb + uptake was 3% (P <.1) higher than control values at 15 min following the cessation of 6 min of exercise, and returned to control levels at 9 min following exercise (P =.32, Fig. 5B). In contrast, 6 min of exercise followed by a 15 or 9 min rest did not significantly alter ouabain-insensitive 86 Rb + uptake (P >.21, Fig. 5C). In the EDL, there was no significant effect of 6 min of exercise followed by a 15 or 9 min rest on the total (P =.47, Fig. 6A), ouabain-sensitive (P =.11, Fig. 6B) or ouabain-insensitive 86 Rb + uptake (P =.23, Fig. 6C). Effect of ouabain on intracellular Na + and K + contents In the soleus muscle, intraperitoneal injection of ouabain had no effect on either intracellular Na + (P =.9) or K + content (P =.4) in resting animals (Fig. 7). However, ouabain injection increased the exerciseinduced accumulation of intracellular Na + by 9 μmol (g wet weight) 1 at the end of exercise, compared with control values (P =.1, Fig. 7A). Ouabain also augmented the reduction in intracellular K + content with exercise by 8 μmol (g wet weight) 1,comparedwith control values (P <.1, Fig. 7B). In the EDL, there was still no effect of exercise on intracellular ions, but ouabain tended to increase intracellular Na +, the increase being significant only in the exercised animals (P =.2, Fig. 8). Effects of K + depletion on intracellular Na + and K + contents In the soleus muscle, 1 days of consuming a K + -depleted diet increased intracellular Na + content by 24 μmol (g wet weight) 1 (P <.1) and reduced intracellular K + Na + influx (nmol.g -1.min -1 ) Na + influx (nmol.g -1.min -1 ) Control 6 min Run Figure 4. Effect of exercise on Na + influx in rat soleus (A) and EDL (B) Animals did not run (control) or performed 6 min of intermittent treadmill running and were killed immediately following the cessation of exercise (6 min Run). Dissected muscles were mounted for isometric contractions and incubated for 2 min in standard KR containing 22 Na + (2 μci ml 1 ). Following incubation, muscles were washed four times for 15 min in ice-cold Na + -free Tris sucrose buffer, blotted, tendons removed, and the muscles weighed and taken for counting. Data are means with S.D.; n = 4 muscles. No significant differences were found.

8 1256 K. T. Murphy and others Exp Physiol pp content by 3 μmol (g wet weight) 1 (P <.1) in resting animals (Table 1). Potassium depletion increased the exercise-induced accumulation of intracellular Na + by 7 μmol (g wet weight) 1 at the end of exercise compared with K + -supplemented animals (P <.1, Table 1). In the EDL, K + depletion increased intracellular Na + content by 27 μmol (g wet weight) 1 (P <.1) and reduced intracellular K + content by 31 μmol (g wet weight) 1 (P <.1) in resting animals (Table 1). However, similar to K + -supplemented animals, EDL muscles from animals fed a K + -depleted diet for 1 days showed no significant increase in intracellular Na + content (P =.26) or decrease in intracellular K + content (P =.61) with exercise (Table 1). Effect of exercise on in vitro [ 3 H]ouabain binding There was no significant effect of 1, 3 or 6 min of exercise on in vitro [ 3 H]ouabain binding site content expressed per gram wet or dry weight in rat soleus and EDL muscle (Table 2). To detect any possible exerciseinduced increase in Na + K + pump α 2 content evident in the recovery period following exercise, [ 3 H]ouabain binding site content was also measured at 15 and 9 min following the cessation of 6 min of exercise. However, [ 3 H]ouabain binding site content was not significantly different from control values in both the soleus and EDL muscle (Table 2). We also examined whether K + depletion altered the effect of 6 min of exercise ( min recovery) on in vitro [ 3 H]ouabain binding site content. At rest, [ 3 H]ouabain binding site content was 46 and 52% lower in the soleus and EDL muscles, respectively, of K + -depleted animals compared with K + -supplemented animals (both P <.1). However, in both the K + -supplemented and the K + -depleted animals, there was no significant effect of 6 min of exercise on [ 3 H]ouabain binding site content in the soleus (in pmol (g wet wt) 1 :K + -supplemented control, 612 ± 65; 6 min Run, 579 ± 37; K + -depleted control, 328 ± 52; 6 min Run, 276 ± 57, P =.7, all n = 4 6 muscles) and EDL (in pmol (g wet wt) 1 :K + - supplemented control, 752 ± 47; 6 min Run, 746 ± 32; K + -depleted control, 51 ± 86; 6 min Run, 44 ± 42, P =.27, all n = 4 6 muscles). Figure 5. Effect of exercise on total (A), ouabain-sensitive (B) and ouabain-insensitive 86 Rb + uptake (C) in rat soleus Animals did not run (control) or performed 6 min of intermittent treadmill running and were killed at 15 or 9 min following the cessation of exercise (6 min Run + 15 min Rest, 6 min Run + 9 min Rest). Dissected soleus muscles were mounted at resting length, equilibrated for 1 min in standard KR without or with 1 mm ouabain and then incubated for 1 min in standard KR containing 86 Rb + (.2 μci ml 1 ) without or with 1 mm ouabain. Following incubation, Effect of exercise on in vivo [ 3 H]ouabain binding When [ 3 H]ouabain was injected intravenously immediately following the cessation of the 6 min muscles were washed four times for 15 min in ice-cold Na + -free Tris sucrose buffer, blotted, tendons removed, and the muscles weighed and taken for counting. Ouabain-sensitive 86 Rb + uptake was calculated as the difference between total and ouabain-insensitive 86 Rb + uptake. Data are means with S.D.; n = 8 26 muscles. P <.1 versus control, P <.5 versus 6 min Run + 15 min Rest.

9 Exp Physiol pp Running exercise and Na + K + exchange in rat muscle 1257 Total 86 Rb + uptake (nmol.(g wet wt) -1.min -1 ) Ouabain-sensitive 86 Rb + uptake (nmol.(g wet wt) -1.min -1 ) running exercise and animals were then allowed to rest for 15 min, there was no effect of exercise on the amount of [ 3 H]ouabain taken up and retained by either the soleus (control, 526 ± 56 pmol (g wet weight) 1 ; 6 min Run, 554 ± 68 pmol (g wet weight) 1, P =.3, both n = 1 12 muscles) or EDL muscle (control, 669 ± 82 pmol (g wet weight) 1 ; 6 min Run, 67 ± 16 pmol (g wet weight) 1, P =.97, both n = 1 12 muscles). Measurements of [ 3 H] activity in plasma prepared from venous blood showed no significant difference between resting (131 ± 69 c.p.m. ml 1 )and Intracellular Na + content * * Ouabain-insensitive 86 Rb + uptake (nmol.(g wet wt) -1.min -1 ) Control 6 min Run + 15 min Rest 6 min Run + 9 min Rest Figure 6. Effect of exercise on total (A), ouabain-sensitive (B) and ouabain-insensitive 86 Rb + uptake (C) in rat EDL Animals did not run (control) or performed 6 min of intermittent treadmill running and were killed at 15 or 9 min following the cessation of exercise (6 min Run + 15 min Rest, 6 min Run + 9 min Rest). Dissected EDL muscles were treated as described in the legend to Fig. 5. Data are means with S.D.; n = 8 26 muscles. No significant differences were found. Intracellular K + content Rest * Run Control Rest * Run Ouabain Figure 7. Effect of ouabain on exercise-induced changes in intracellular Na + (A) andk + content (B) in rat soleus Animals were injected intraperitoneally with saline containing ouabain (2 nmol (g body weight) 1, Ouabain) and rested for 3 min. Control animals received no injections. Animals then performed 6 min of intermittent treadmill running (Run) or rested for a matched duration (Rest). Immediately following the final running interval, animals were killed, the soleus muscle dissected and measured for intracellular Na + and K + content as described in the legend to Fig. 2. Data are means with S.D.; n = 7 11 muscles. P <.1 versus control Rest, P <.1 versus control Run, P <.1 versus Ouabain Rest.

10 1258 K. T. Murphy and others Exp Physiol pp running animals (165 ± 47 c.p.m. ml 1, P =.26, all n = 8 versus 8 animals). Discussion The present study describes a novel technique for the precise quantification of the changes in Na + and K + contents occurring with in vivo exercise in rat skeletal Intracellular Na + content Intracellular K + content Rest Run Control Rest * Run Ouabain Figure 8. Effect of ouabain on exercise-induced changes in intracellular Na + (A) andk + content (B) in rat EDL Animals were injected intraperitoneally with saline containing ouabain (2 nmol (g body weight) 1, Ouabain) and rested for 3 min. Control animals received no injections. Animals then performed 6 min of intermittent treadmill running (Run) or rested for a matched duration (Rest). Immediately following the final running interval, animals were killed, the EDL muscle dissected and measured for intracellular Na + and K + content as described in the legend to Fig. 2. Data are means with S.D.; n = 7 11 muscles. P <.5 versus control Rest, P <.5 versus control Run. muscle, and finds that in the soleus muscle, in vivo exercise causes an increase in Na + K + pump activity and is accompanied by an up to 8% elevation in intracellular Na + content that, surprisingly, is partly sustained for at least 15 min postexercise (38% above control levels). This exercise-induced elevation in intracellular Na + content is exacerbated by pretreatment with ouabain, as well as by K + depletion, indicating that it is likely to involve inadequate Na + clearance by the Na + K + pumps. In contrast, neither Na + K + pump activity nor intracellular Na + content is significantly increased with exercise in the EDL, presumably because during running this muscle is less active than the soleus. Thus, recordings of the motor unit firing frequency in spontaneously active rats show that in EDL, the motor units are only active for 5 22% of every 24 h period, whereas soleus motor units are active for 22 35% of the time (Hennig & Lømo, 1985). During locomotion, EMG activity in rat soleus (an ankle extensor) begins just before foot contact with the ground and continues until immediately before the foot lifts off again. In contrast, in EDL (a toe extensor), the EMG activity begins after the foot has left the ground and continues until just after the foot touches down again (Nicolopoulos- Stournaras & Iles, 1983; James et al. 1995). In general, during running, slow oxidative and fast oxidative fibres appear to provide the majority of the force produced by the leg muscles, with a smaller contribution by fast glycolytic fibres (Sullivan & Armstrong, 1978). Electromyogram recordings in rat soleus muscles (slow-twitch oxidative fibres) also showed a markedly higher activity than in the tibialis anterior muscle (fast-twitch fibres; Roy et al. 1991). Accordingly, it can be expected that during running, the gain of Na + and loss of K + will be larger in soleus than in EDL muscles. Exercise increases intracellular Na + content in the soleus In the present study, 6 min of running at a treadmill speed of 12 m min 1 led to an 8% increase in the intracellular Na + content and an equivalent reduction in the intracellular K + content of soleus muscles. Conversely, there was very little effect of exercise on intracellular Na + and K + contents in the EDL. Likewise, an activitydependent increase in Na + K + pump activity following exercise was found in the soleus, but not in EDL. The exercise-induced changes in intracellular ion content in the working muscle cells are most probably caused by the influx of Na + and efflux of K + associated with the action potentials, combined with insufficient activation of the Na + K + pumps. The importance of this balance between passive Na + K + fluxesduringactionpotentialsandthe active counter-transport of the ions for cellular Na + K + homeostasis is supported by the experiments showing that intraperitoneal injection with ouabain, as well as

11 Exp Physiol pp Running exercise and Na + K + exchange in rat muscle 1259 Table 1. Effect of 1 days K + depletion on exercise-induced changes in intracellular Na + and K + contents in rat soleus and EDL Intracellular Na + content Intracellular K + content (μmol (g wet weight) 1 ) (μmol (g wet weight) 1 ) Muscle Treatment Control Run Control Run Soleus K + supplementation 13.2 ± ± ± ± 3. K + depletion 36.8 ± ± ± ± 2.3 EDL K + supplementation 11.1 ± ± ± ± 3. K + depletion 38.3 ± ± ± ± 3.5 Animals were fed a normal K + -supplemented diet or a K + -depleted diet for 1 days. Animals did not run (Control) or were killed immediately following 6 min intermittent running at a treadmill speed of 12 m min 1 (Run). Dissected soleus and EDL muscles were mounted at resting length and washed out four times for 15 min in ice-cold Na + -free Tris sucrose buffer. Following washout, muscles were blotted, tendons removed, and the muscles weighed and taken for flame photometric analysis of Na + and K + content. Values for Na + content were multiplied by 1.22 and 1.33 for soleus and EDL, respectively, to correct for the loss of intracellular Na + during the washout. Data are means ± S.D.; n = 4 muscles. P <.1 versus K + -supplemented control; P <.1 versus K + -supplemented Run; P <.1 versus K + -depleted control. Table 2. EDL Effect of exercise on in vitro [ 3 H]ouabain binding site content in rat soleus and [ 3 H]Ouabain binding [ 3 H]Ouabain binding (pmol (g wet weight) 1 ) (pmol (g dry weight) 1 ) Soleus EDL Soleus EDL Control 623 ± ± ± ± min Run 53 ± ± ± ± min Run 585 ± ± ± ± min Run 69 ± 69 7 ± ± ± min Run + 15 min Rest 63 ± ± ± ± min Run + 9 min Rest 663 ± ± ± ± 124 Animals did not run (control) or performed 1, 3 or 6 min of intermittent treadmill running and were killed either immediately (1 min Run, 3 min Run, 6 min Run) or at 15 or 9 min following the cessation of exercise (6 min Run + 15 min Rest, 6 min Run + 9 min Rest). Dissected soleus and EDL muscles were rapidly frozen in liquid N 2 for later determination of in vitro [ 3 H]ouabain binding using the standard vanadate-facilitated method. Muscle [ 3 H]ouabain binding was expressed per gram wet weight or per gram dry weight. Data are means ± S.D.; n = 5 28 muscles. No significant differences were found. downregulation of the population of Na + K + pumps induced by prior K + depletion, markedly increased the elevationinintracellularna + content with exercise. Taken together, these observations indicate that the clearance of intracellular Na + during exercise depends on the capacity of the Na + K + pumps. Insufficient Na + K + pump activity has previously been reported during exercise in human studies, in which a large net K + loss was found during the first few minutes of exercise (Medbø & Sejersted, 199; Verburg et al. 1999). Interestingly, in one of these studies, although net K + loss ceased after 2 min of exercise, indicating increased Na + K + pump activity, it reappeared after 3 min of exercise, which was taken to indicate insufficient further Na + K + pump activation (Verburg et al. 1999). During recovery from 1 3 s of 6 Hz in vitro electrical stimulation of isolated muscles, however, intracellular Na + has been observed to decline to the control level within 1 min and, if muscles were allowed to rest, intracellular Na + was further reduced in the next 9 min to below 7% of the control level (Nielsen & Clausen, 1997). This decrease in intracellular Na + content is thought to be due to an excitationinduced increase in Na + K + pump activity. Since an increase in Na + K + pump activity was also observed following exercise in the present study, a fast reduction in intracellular Na + content could be expected during recovery in the soleus muscle of Run animals. Surprisingly, however, intracellular Na + content in the soleus muscle was still elevated by 38% in rats that were allowed to rest for 15 min following the cessation of exercise. These results demonstrate that in vivo, the postexercise recovery of intracellular Na + content is slow, taking at least 15 min to reach control levels. The differing response of intracellular Na + following in vitro electrical stimulation

12 126 K. T. Murphy and others Exp Physiol pp compared with in vivo exercise may reflect differences in the magnitude of contraction-induced Na + K + pump activation. While, based on rates of Na + efflux, Na + K + pump activity in the first 5 s following 1 and 3 s of 6 Hz in vitro electrical stimulation was increased by 3 and 7% above resting levels, respectively (Nielsen & Clausen, 1997), the present study found only a 3% increase in Na + K + pump activity following in vivo exercise. Although the measurement of Na + K + pump activity following exercise was made at 15 min recovery, it is unlikely that Na + K + pump activity in the first minute following exercise would have been as elevated above control values as earlier observed for the relative increases of 3 and 7% in Na + K + pump activity observed in isolated muscles exposed to electrical stimulation in vitro (for details, see Nielsen & Clausen, 1997). It therefore appears that a greater Na + K + pump activation induced by in vitro electrical stimulation compared with in vivo exercise may contribute to the faster decline in intracellular Na + content observed following repeated muscle contractions. This may reflect the greater stimulation frequency (6 Hz) in the in vitro experiments (Roy et al. 1991; Nielsen & Clausen, 1997) and therefore greater excitation-induced Na + K + pump activation (Clausen, 23) with in vitro electrical stimulation. It is also likely that the clearance of calcitonin-gene related peptide (CGRP) released from the nerve endings is smaller in vitro compared with in vivo, where CGRP may be cleared via the circulation. This would lead to more pronounced Na + K + pump activation with in vitro electrical stimulation (Andersen & Clausen, 1993). Clearly, further research needs to investigate the mechanisms responsible for the differing Na + K + pump activation and therefore regulation of intracellular Na + between in vivo exercise and in vitro electrical stimulation. Measurements of 22 Na + influxdemonstratedthatthe prolonged elevation of intracellular Na + content after exercise could not be related to an increased passive Na + permeability of the muscle fibres induced by the work. In accordance with this, 2 min incubation with bumetanide had no effect on the intracellular Na + content in either control muscles or muscles from exercising rats, indicating that the Na + Cl cotransport activity of the muscle fibres was unaffected by exercise (Dørup & Clausen, 1996). Importantly, the lack of change in 22 Na + influx also indicates that the exercise protocol does not induce significant muscle damage (Gissel & Clausen, 1999). Treadmill running has been shown to increase muscle cell volume (Komulainen & Vihko, 1994). However, in the presentstudy,weobservednosignificanteffectsofexercise on muscle wet weight, indicating that possible changes in water content had recovered during the four 15 min washout periods in the cold. Therefore, the exerciseinduced changes in intracellular Na + and K + content are unlikely to reflect differences in cell volume. Exercise increases Na + K + pump activity in the soleus In vitro electrical stimulation increases Na + K + pump activity when measured by the rates of ouabain-sensitive 86 Rb + uptake and Na + efflux (Everts & Clausen, 1994; Nielsen & Clausen, 1997). This is in keeping with the acceleration of muscular K + re-uptake following exercise (Medbø & Sejersted, 199). The present study is the first to investigate the effect of in vivo exercise on Na + K + pump activity measured using ouabain-sensitive 86 Rb + uptake. In soleus muscles taken from rats that ran for 6 min and were then rested for 15 min, we found a 3% increase in Na + K + pump activity that returned to control levels at 9 min postexercise. This increase in Na + K + pump activity accounts for the 21% increase in total 86 Rb + uptake and, indeed, there is no effect of running on ouabain-insensitive 86 Rb + uptake. In the EDL, exercise has no effect on the total, ouabain-sensitive or ouabaininsensitive 86 Rb + uptake. Measurements of [ 3 H]ouabain binding revealed no effect of exercise on the abundance of the major (α 2 ) isoform of the Na + K + pumps in the outer membrane of the muscle fibres. This was the case both when [ 3 H]ouabain binding was determined in isolated muscles using 2 h of incubation with ouabain and when ouabain binding was performed in vivo, where the binding took only 15 min and the risk of possible trafficking of Na + K + pumps between the outer membrane and some intracellular pool of Na + K + pumps is therefore greatly reduced. Thus, an elevated abundance of Na + K + pumps seems not to be involved in the exerciseinduced increase in Na + K + pump activity after exercise. Conversely, although the measurements of intracellular Na + content do not allow for a precise calculation of the intracellular Na + concentration, the increase in intracellular Na + content indicates that increased cellular Na + concentration contributed to the Na + K + pump activation (Everts & Clausen, 1992; Nielsen & Clausen, 1997). Indeed, the Na + K + pump activity (ouabainsensitive 86 Rb + uptake) and intracellular Na + content data show similar trends, being elevated at 15 min following the cessation of 6 min of exercise and returning to control levels at 9 min following exercise. Since intracellular Na + at the cessation of 6 min exercise was higher than at 15 min following the exercise (Fig. 2), the correlation between intracellular Na + and Na + K + pump activity indicates that the activity of the Na + K + pump immediately following exercise in the soleus muscle must have been even higher than that observed after 15 min rest. Perspectives During exercise, K + lost from working muscles accumulates in the extracellular compartment. The