INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

IT HAS BEEN SUGGESTED THAT the accumulation of K⁺ in the interstitium of skeletal muscle plays a role in muscle fatigue and pain during short-term exercise, helps control arterial flow to active muscle during exercise, and possibly contributes to the initial cardiac and ventilatory adjustments to exercise (7, 10). To date, only one study has described the response of muscle interstitial K⁺ (K⁺_i) to exercise in humans (10). Vyskocil et al. (10) inserted ion-selective microelectrodes into the brachioradialis muscle and reported K⁺_i values of 4-5 mmol/l at rest and a peak value of 9.5 mmol/l during 20 s of maximal, static forearm exercise. These values were consistent with those reported in earlier studies (1, 4) that used similar microelectrodes to measure resting and exercise-induced increases in K⁺_i in cat and rabbit muscle. However, other investigators have commented on the failure of the ion-selective microelectrode to accurately measure K⁺_i within contracting muscle (3), which is related to inadequate detection of an electrical signal (i.e., the K⁺ current) among the relatively large electrical background noise of active muscle, and there is only anecdotal evidence that such difficulties have since been encountered by other researchers. Since the study by Vyskocil et al. (10), there has been no further report of K⁺_i and its response to exercise in human skeletal muscle.

Microdialysis is a relatively new technique and provides another way of directly assessing chemical changes in muscle interstitium. The accurate measurement of such changes depends on the estimation of the recovery of a compound of interest by each microdialysis probe. This can be done by introducing a radioactive isotope of the compound of interest into the perfusate, measuring its loss from the microdialysis probe, and assuming that this loss equals the recovery of the compound of interest from the surrounding medium (8). The radioisotopes of K⁺ have very short half-lives and are, therefore, not suited to microdialysis and the measurement of K⁺_i. In contrast, an isotope of thallium, ²⁰¹Tl, might provide an internal reference for K⁺ because its hydrated ion radius (1.43 Å), which helps determine the passage of ions across membranes, is similar to that for K⁺ (1.33 Å). The transmembrane fluxes of Tl and K⁺ are also similar in whole animals as well as in an isolated muscle preparation (2, 5). Thus it seemed likely that their respective fluxes across a microdialysis membrane would also be similar. The half-life of ²⁰¹Tl (~73 h) also renders it suitable for microdialysis. Therefore, in an in vitro experiment we tested the hypothesis that the loss of ²⁰¹Tl from microdialysis probes would be positively related to and agree with K⁺ recovery by the same probes. Because the in vitro results supported this hypothesis, we then used microdialysis and ²⁰¹Tl to study K⁺_i in the human gastrocnemius medialis muscle during rest and exercise.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In vitro experiment. On 2 days, twelve single-lumen microdialysis probes were tested. Each probe consisted of 12-15 cm of microdialysis tubing (0.20 mm ID, 0.22 mm OD, molecular cutoff at 5-6 kDa; GFS 16-GFE 18; Gambro, Lund, Sweden). A length (~40 cm) of suture thread (15-mm OD; Vicryl, Ethicon, Denmark) was glued to the microdialysis tubing at both ends to provide stability when used in vivo. The microdialysis tubing and suture thread were then inserted and glued inside two hollow nylon tubes (0.5 mm ID, 0.6 mm OD; Portex, Hythe, Kent, UK) and adjusted so that the exposed length of the microdialysis tubing varied between 2.5 and 4 cm. The inlet and outlet lengths of tubing were 11 and 7 cm, respectively. The probes were placed in two 7 × 6.5 cm baths and suspended ~2.5 cm above the bottom surface. Approximately 400 ml of Ringer acetate solution were added to the baths and stirred slowly. The probes were perfused by using a microdialysis pump (CMA 102; CMA Microdialysis) at rates of 3, 4, 5, 6, 7, 8, and 10 µl/min for periods that would yield 150 µl of dialysate. The perfusate consisted of 0.9% NaCl, and to 100 ml of this salt solution 20 µl of ²⁰¹Tl (standard activity = 74 MBq; Amersham) was added, yielding, in 150 µl of perfusate, an initial radioactivity of 90,000 dpm. The dialysate was always weighed and discarded, but noted, if its weight was not within 5% of the expected value. ²⁰¹Tl activity in the perfusate and dialysate was measured by using an autogamma counter (Packard 5650). Potassium was measured by using an automated Na⁺/K⁺ analyzer (Radiometer KNA2). The relative loss of ²⁰¹Tl was calculated by using the following equation (8) The relative recovery of K⁺ was calculated as the difference between the concentration of K⁺ ([K⁺]) in the dialysate and perfusate divided by its concentration in the medium. All radioactivities and concentrations were corrected to 150 µl.

In vivo experiment. Three healthy men (age 27-32 yr) and one healthy woman (age 24 yr) volunteered to participate in the experiment and provided written, informed consent. All procedures were conducted in accordance with the statutes of the Declaration of Helsinki (1989) and approved by the ethics committee of the Copenhagen and Frederiksberg communities (project #KF 01-257/98). Two microdialysis probes (as described in In vitro experiment) were inserted into the gastrocnemius medialis muscle of both legs in all subjects. Before insertion, the skin, subcutaneous tissue, and fascia close to both the insertion and exit points were anesthetized with lidocaine (20 mg/ml). A sterile needle was then inserted into the muscle ~5 mm below the fascia and pushed in a direction approximately parallel to the muscle fibers to then exit the skin ~60 mm from the insertion site. The microdialysis probe was fed through the needle, and then the needle was withdrawn to leave the probe in the muscle. Ultrasound was used to confirm that the probe was placed in the muscle and not in the overlying subcutaneous tissue. In one subject, two probes were inserted into the overlying subcutaneous tissue. Therefore, probes in seven calves were used to measure muscle K⁺_i. Because of technical problems with three probes, the muscle data reported here are based on responses measured by a total of 11 probes.

Statistics. Repeated-measures ANOVA was used to test for differences in ²⁰¹Tl losses and K⁺_i across time, and Tukey's honestly significant difference test was used to locate any differences found. Linear regression was used to establish the relationship between ²⁰¹Tl loss and K⁺ recovery. The level of significance was set at P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In vitro experiment. Data obtained at perfusate flows of 10 µl/min were excluded because of significant perfusate loss (~30%) from the microdialysis probes. Figure 1 displays the scatterplot of ²⁰¹Tl loss vs. K⁺ recovery. A significant linear relationship was observed between these two variables (y = 0.027 + 0.961x), with both the slope and y-intercept not different from one and zero, respectively. ²⁰¹Tl losses and K⁺ recoveries ranged between 0.6 and 0.98. The level of agreement (i.e., relative recovery of K⁺  relative loss of ²⁰¹Tl) was, on average, not different from zero (Fig. 2) and was not systematically influenced by either perfusate flow or probe length (data not shown).

View larger version (13K): [in this window] [in a new window]   Fig. 1.   Relationship between the relative loss of thallium-201 (201Tl) and relative recovery of K+ in 12 microdialysis probes. Dashed lines, confidence intervals at the 95% level.

View larger version (9K): [in this window] [in a new window]   Fig. 2.   Relationship of level of agreement between relative recovery of K+ (RR) and relative loss of 201Tl (RL; i.e., RR  RL) in 12 microdialysis probes.

In vivo experiment. The relative loss of ²⁰¹Tl was used to estimate K⁺ recovery, and its response during rest and exercise is shown in Fig. 3. ²⁰¹Tl loss (i.e., perfusate  dialysate activity) is expressed relative to the ²⁰¹Tl activity in the perfusate and could range between zero (i.e., no ²⁰¹Tl loss) and one (i.e., maximum ²⁰¹Tl loss). During preexercise rest, ²⁰¹Tl loss was significantly different among all perfusate flows, ranging between ~0.3 and 0.7 at 8 and 2 µl/min, respectively. The average ²⁰¹Tl losses at 5 µl/min for the three collections during the 70-min preexercise period (0.42-0.45) were not different (P > 0.05) from each other. During exercise, when perfusate flows were also 5 µl/min, ²⁰¹Tl loss increased progressively (P < 0.05) with each increase in workload, from a low value of 0.53 ± 0.10 to a peak value of 0.63 ± 0.08. After exercise, ²⁰¹Tl loss declined to 0.51-0.53 ± 0.08, values that were higher than the preexercise values (P < 0.05) at the same perfusate flow.

View larger version (13K): [in this window] [in a new window]   Fig. 3.   Relative loss of 201Tl during rest, exercise, and recovery in human gastrocnemius medialis muscle (n = 7).

View larger version (17K): [in this window] [in a new window]   Fig. 4.   Dialysate and interstitial K+ during rest, exercise, and recovery in human gastrocnemius medialis muscle (n = 7).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

These results are the first to demonstrate the direct measurement of K⁺_i by using microdialysis in human skeletal muscle during rest and exercise. This is an important finding because of the possible roles that K⁺_i play in causing pain and fatigue during ischemic exercise and in contributing to the regulation of arterial flow during exercise, as well as its involvement in the muscle reflex that contributes to the cardiorespiratory adjustments to exercise (7, 10). Although direct measurements of muscle K⁺_i in humans were reported 15 yr ago, other investigators have been unable to reproduce these findings because of technical problems associated with the use of ion-selective microelectrodes in active skeletal muscle. These technical problems, which are related to the difficulty in measuring an electrical potential difference among the greater electrical activity of the active muscle, are circumvented when microdialysis is used, because its principle of operation is not electrical in nature.

During the initial periods of rest, K⁺_i values were stable (Fig. 4) and similar to human muscle interstitium values measured using microelectrodes (4.5 mmol/l; Ref. 10) and are consistent with K⁺_i values in the gastrocnemius muscle of the dog (3.3 mmol/l; Ref. 3), rabbit (4.7 mmol/l; Ref. 4), and cat (5.0 mmol/l; Ref. 4) measured using microelectrodes. The reduction in K⁺_i at the highest perfusate flow was probably due to a higher K⁺ uptake by the probe relative to the diffusion and possibly to release of K⁺ (by the muscle fibers) into the interstitium adjacent to the probe. This is because the absolute, not relative, recovery of K⁺ increased compared with that at lower flows, which is supported by the partial restoration of K⁺_i to initial resting values when the perfusate flow was then reduced to 5 µl/min. This data suggest that, at relatively low perfusate flows (5 µl/min), K⁺_i at rest is accurately measured with the use of microdialysis.

During exercise, K⁺_i increased to average values of 6.9-7.5 mmol/l, with peak values recorded at the highest workload (Fig. 4). These values are lower than the peak K⁺_i value of 9.45 mmol/l measured in human brachioradialis muscle during 20 s of maximal, static forearm exercise (10) and correspond with values reported by these same investigators during 20 s of submaximal, static contractions (i.e., 7-8 mmol/l) that were closer to the intensities used in the present study. The increase in K⁺_i across the first two workloads is also consistent with the data of Vyskocil et al. (10), which showed an increase in K⁺_i up to 7-8 mmol/l in response to "light" and "intermediate" efforts. The failure of K⁺_i to then significantly increase during the highest workload was unexpected, particularly given that the VAS scores (n = 4) and EMG (n = 1) increased proportionally across all three workloads, suggesting, although not demonstrating, that the force generated by the gastrocnemius muscle, rather than extraneous muscle groups, also increased. The data are, however, consistent with the findings of Saltin et al. (7), who showed that peak K⁺ levels in the venous effluent draining the quadriceps increased when the intensity of static knee extension was raised from 15 to 25% of maximum force (i.e., 5.2-5.7 mmol/l) but then failed to increase when the intensity was increased further to 50% of maximum force (i.e., 5.7 mmol/l). At a given frequency of action potentials, the K⁺_i will be influenced by the rate of K⁺ reuptake by muscle fibers, the rate of K⁺ efflux to the circulation, as well as water shifts into the interstitium (9). A relative increase in any one or more of these factors after the transition to the highest workload could help explain our finding. It is also possible that the additional force generated at the highest workload was caused by muscle fibers within gastrocnemius medialis muscle that were not sampled by the probes. Although this latter finding cannot be explained by the present data, it does counter the possibility that the exercise-induced rise in K⁺_i was caused by sarcolemmal damage and leakage of K⁺ to the interstitium as a result of the increased muscle force. If this were so, then K⁺_i would be expected to increase to a greater extent at the highest workload.

After exercise, K⁺_i decreased to values below those observed during exercise but was higher than those observed before exercise (Fig. 4). It is important to note that the postexercise K⁺_i values represented an average K⁺_i for the initial 6 min after exercise. Muscle K⁺_i measured with the use of microelectrodes returns to resting values within 1-7 min, depending on the exercise protocol and species used (3, 4, 10). Femoral venous K⁺ levels following incremental static knee-extensor exercise also take several minutes to return to resting levels after exercise (7). Therefore, that the recovery value in the present study (5.2 mmol/l) was higher than rest values is consistent with this other data and provides further support for the accuracy of the K⁺ measurements shown here.

A main problem with the measurement of K⁺_i by using microdialysis is accurately estimating the relative recovery of K⁺ from the interstitium by the probe. The in vitro results show that the relative loss of ²⁰¹Tl is both linearly related to, and agrees closely with, the relative recovery of K⁺ (Figs. 1 and 2). The accuracy with which ²⁰¹Tl loss could represent K⁺ recovery in vivo was tested at rest by changing K⁺ recovery through adjustments in perfusate flow. Changes in the relative loss of ²⁰¹Tl from 0.4 to 0.7 during the initial periods of rest exerted no effect on K⁺_i (Figs. 3 and 4). Assuming that K⁺_i remained stable across this period, this demonstrates that the changes in K⁺ recovery were accurately measured by ²⁰¹Tl loss at perfusate flows of 2 and 5 µl/min and supports the use of ²⁰¹Tl as an internal reference for K⁺.

During exercise when perfusate flow was fixed at 5 µl/min, the ²⁰¹Tl loss progressively increased with each increase in workload and then declined after exercise. This is consistent with other data that showed an exercise-induced increase in ethanol loss from microdialysis probes that might be due to pressure oscillations in response to muscle contraction and relaxation (6). Such oscillations would be expected to increase as the muscle force increases and would help explain the exercise effect on ²⁰¹Tl loss in the present study. In addition, ²⁰¹Tl loss would also depend on the interstitial volume, and this probably increased from rest to exercise and as the exercise intensity increased (9). This might also explain why ²⁰¹Tl loss was higher during rest after exercise, as interstitial water content can remain elevated for at least 30 min after exercise (9). High ²⁰¹Tl loss after exercise cannot be attributed to mechanical events only present during exercise, and it is also not likely to be due to postexercise hyperemia, because blood flow per se has no effect on probe recovery (6).

A primary limitation of the method proposed here is that the measurement of K⁺_i was constrained to periods of 6-12 min at the highest, acceptable perfusate flow (i.e., 5 µl/min). These measurement periods need to be reduced considerably to enable the study of the role of K⁺_i in the vascular, cardiac, and ventilatory adjustments to exercise. However, unlike ion-selective electrodes, collection of dialysate is advantageous when more than one compound needs to be assessed. Considerable care must be taken to ensure that the microdialysis membrane lies entirely within the skeletal muscle. In one subject, we inserted microdialysis probes in the subcutaneous tissue. Resting K⁺_i varied between 2.5 and 3.5 mmol/l, and K⁺_i and ²⁰¹Tl loss were both lower and unchanged during exercise compared with muscle responses. Therefore, when the measurement of muscle responses is attempted, both K⁺_i and ²⁰¹Tl loss would be reduced in proportion to the amount of membrane sitting within the subcutaneous tissue.

In conclusion, the findings of the present study and their similarity to limited data in the literature demonstrate that K⁺_i can be accurately measured during rest and exercise by using microdialysis and ²⁰¹Tl as an internal reference. This provides the basis of a technique for the study of K⁺_i and its role in a variety of exercise responses.