Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Modulation of skeletal muscle function by -adrenoceptor agonists has long been described (Bowman and Nott, 1969); only recently, however, therapeutic use was proposed on the basis of experimental data, case reports, and rare clinical trials. In particular, the use of clenbuterol or salbutamol (albuterol) was proposed to counteract muscle atrophy that occurs in orthopedic patients, especially in elderly persons or patients in a malnutrition state, as well as in hereditary muscle dystrophies (Maltin et al., 1993; Hayes and Williams, 1998; Zeman et al., 2000; Herrera et al., 2001; Kissel et al., 2001). Potential benefits in these diseases are expected from the ₂-adrenoceptor-mediated anabolic effects of these drugs (Choo et al., 1992; Hinkle et al., 2002). Interestingly, ₂-agonists also successfully counteracted muscle paralysis in patients suffering from hyperkalemic periodic paralysis (HPP) (Wang and Clausen, 1976; Hanna et al., 1998). HPP is a hereditary muscle disorder caused by mutations in the gene encoding the sodium channel expressed exclusively in the skeletal muscle (Cannon, 2001). The genetic defect produces a susceptibility to sarcolemmal depolarization, which renders the fiber inexcitable and leads to frequent attacks of muscle weakness lasting 1 to 3 h. The ₂-agonists may be able to counteract the guilty membrane depolarization through activation of the electrogenic Na⁺-K⁺ pump (Clausen et al., 1993). Use of salbutamol, metaproterenol, and terbutaline in patients with HPP have been reported, but their relative efficacy in this disorder remains to be established. Differing from periodic paralysis are the myotonic syndromes that are characterized by sarcolemmal overexcitability, which are common to a number of hereditary diseases resulting from various genetic defects, including sodium channelopathies (Moxley, 2000; Cannon, 2001; Meola, 2002). Local anesthetic-like drugs, such as mexiletine, prove useful in myotonic patients, owing to use-dependent block of sodium channels (Moxley, 2000). To our knowledge, there have been no reports about the use of adrenergic agents in myotonia, most probably because sodium channel block by these drugs was not expected in skeletal muscle.

The ₂-subtype is the predominant -adrenergic receptor expressed in skeletal muscle (Liggett et al., 1988). This G protein-coupled receptor exerts its physiological function through phosphorylation of specific effectors, such as ion channels, by cyclic AMP-dependent protein kinase (PKA) (Yang and McElligott, 1989). For instance, ₂-agonists were shown to increase sodium channel activity in cardiac myocytes through the classic PKA-dependent pathway but also through a membrane-delimited pathway involving direct interaction between Gs protein and the channel (Matsuda et al., 1992; Lu et al., 1999). Nothing is known about possible similar mechanisms in the skeletal muscle.

In a previous study, we showed that two membrane-permeable analogs of cyclic AMP inhibited sodium currents (I_Na) in cell-attached patches of rat skeletal muscle fibers. The effect was not mimicked by externally-applied cAMP and persisted in the presence of the PKA inhibitor N-[2-(p-bromocinnamylamino)-ethyl]-5-isoquinoline-sulfonamide (H-89), indicating that cAMP acted within the cell to block skeletal muscle sodium channels independently of PKA activation (Desaphy et al., 1998). In the present study, we sought to determine whether ₂-adrenoceptor agonists might increase cyclic AMP level sufficiently to block sodium channels. We found that clenbuterol but not salbutamol inhibited I_Na in rat skeletal muscle fibers or in tsA201 cells expressing the human skeletal muscle sodium (hSkM1) channels. We also showed that sodium channel block by clenbuterol can affect action potential property in skeletal muscle fibers. Those effects were independent of ₂-adrenoceptor stimulation and did not involve PKA, calcium- and phospholipid-dependent protein kinase (PKC), or cyclic AMP. Thus, we propose that clenbuterol directly blocked sodium channels in a manner similar to local anesthetic drugs, and we defined some structural requirements in -agonists and antagonists for obtaining such an effect. Because of the differences in blocking muscle sodium channels, salbutamol should be safely used in periodic paralysis patients, whereas clenbuterol may be more indicated in patients suffering from myotonic syndromes.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

All experiments involving animals were conducted in accordance with the Italian Guidelines for the use of laboratory animals, which conform with the European Community Directive published in 1986 (86/609/EEC).

Sodium Current Measurement in Rat Skeletal Muscle Fibers. Fibers were enzymatically dissociated from flexor digitorum brevis muscles of adult rats, and sodium currents were recorded at room temperature in the cell-attached configuration of the patch-clamp method, as described previously (Desaphy et al., 1998a). Bath solution contained 145 mM CsCl, 5 mM EGTA, 1 mM MgCl₂, 10 mM HEPES, and 5 mM glucose. Pipette solution contained 150 mM NaCl, 1 mM MgCl₂, 1 mM CaCl₂, and 10 mM HEPES. Both solutions were buffered at pH 7.3. In these conditions, muscle fibers were depolarized near 0 mV such that the cell-attached patch potential was close to that held by the AxoPatch 1D amplifier (Axon Instruments, Union City, CA). The patch generally contained tens of sodium channels, allowing recording of macroscopic-current-like sodium currents (I_Na) with rapid onset and total inactivation in less than 3 ms. Patches with I_Na exhibiting >15% run-down in 20 min or anomalous activation and inactivation voltage-dependence were discarded from analysis (Desaphy et al., 1998a). Voltage-clamp protocols and data acquisition were performed using pCLAMP 6.0 software (Axon Instruments) through a digidata 1200 analog/digital interface. Current were low-pass filtered at 2 kHz (3 dB) by the amplifier four-pole Bessel filter and digitized at 10 to 20 kHz.

Sodium Current Measurement in tsA201 Cells. The tsA201 cells were cotransfected with 10 µg of plasmid DNA encoding the full-length hSkM1 cDNA and lower amount of a plasmid DNA encoding CD8 receptors using the calcium-phosphate precipitation method, as described previously (Desaphy et al., 2001). Successfully transfected cells were identified using Dynal microbeads coated with anti-CD8 antibody (Dynal A.S., Oslo, Norway). I_Na were recorded at room temperature using the whole-cell, patch-clamp method (Desaphy et al., 2001). Bath solution contained 150 mM NaCl, 4 mM KCl, 2 mM CaCl₂, 1 mM MgCl₂, 5 mM HEPES, and 5 mM glucose, and the pH was set to 7.4 with NaOH. The pipette solution contained 120 mM CsF, 10 mM CsCl, 10 mM NaCl, 5 mM EGTA, and 5 mM HEPES, and the pH was set to 7.2 with CsOH. Peak I_Na amplitudes ranging from 0.8 to 6 nA, stable series resistance errors less than 5 mV, and current run-down less than 5% within the experiment were our limiting criteria to consider the data for analysis.

Action Potential Measurement in Rat Skeletal Muscle Fibers. Action potentials were recorded in vitro in rat skeletal muscle fibers as described previously (Desaphy et al., 1998b). Briefly, the extensor digitorum longus muscles were dissected from animals under urethane anesthesia and fixed through tendons in a recording chamber containing a 95% O₂/5% CO₂-gassed physiological solution. Action potentials were elicited in current-clamp mode using two intracellular microelectrodes. The membrane potential was held at 80 mV by injecting a steady current, and 100-ms depolarizing current pulses of increasing amplitude were applied up to elicit first a single action potential (threshold) and then a train of action potentials.

Drugs and Chemicals. Salbutamol, clenbuterol, DL-propranolol, 8-(4-chlorophenylthio)adenosine 3',5'-cyclic monophosphate (CPT-cAMP), H-89, staurosporine, and okadaic acid were purchased from Sigma (Milan, Italy). QX-314 was a gift from Alomone Labs (Jerusalem, Israel). These compounds were directly dissolved in bath or pipette solution at the desired concentration, except for H-89 and staurosporine, which were first dissolved in dimethyl sulfoxide and then diluted in recording solutions. The final concentration of dimethyl sulfoxide did not exceed 0.1% and had no effect on sodium currents.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effects of ₂-Agonists on Sodium Currents in Rat Skeletal Muscle Fibers. Macroscopic-current-like sodium currents were elicited in cell-attached patches of rat skeletal muscle fibers by test pulses to the potential of 20 mV applied from a holding potential (hp) of 100 mV every 2 s (Desaphy et al., 1998a). Five minutes after seal formation, the peak I_Na amplitude was sufficiently stable to allow drug testing. As shown in Fig. 1A, bath application of 500 µM salbutamol for about 5 min had no effect on I_Na. In five cells, the peak I_Na amplitude in the presence of salbutamol was 97.4 ± 3.0% of control. Such an effect was not distinguishable from the spontaneous run-down observed in these experimental conditions (Desaphy et al., 1998a). In contrast to salbutamol, the other ₂-adrenoceptor agonist, clenbuterol (500 µM), reduced I_Na to 36.8 ± 3.0% of control in four muscle fibers (Fig. 1B). The effect of clenbuterol was quite fully reversible in 5 to 6 min, as shown in Fig. 1B. Normalized current-voltage relationships measured in three cells tested for clenbuterol effect are shown in Fig. 2A. Under control conditions, the current-voltage curve activated at 60 mV, peaked at 20 mV, and reached zero-current level at +70 mV. Clenbuterol reduced I_Na at all voltages and did not modify the voltage at which current amplitude was maximal. The voltage dependence of the activation curve was not modified by the drug (Fig. 2B), suggesting that no change in fiber V_m occurred in response to 500 µM clenbuterol. The midpoint potentials for activation were 40.9 mV in control and 42.8 mV in presence of clenbuterol. In contrast, clenbuterol shifted the voltage dependence of steady-state fast inactivation toward negative potentials, as assessed using a two-pulse protocol including a 200-ms conditioning pulse (Fig. 2C). The 8.5 mV shift of the half-maximum inactivation potential induced by the drug was larger than the spontaneous negative shift we generally observed in cell-attached patch recordings (i.e., usually 2 mV in 10 min) (Desaphy et al., 1998a). The effect of clenbuterol was dose-dependent because 1500 µM clenbuterol reduced peak I_Na to 10.7 ± 6.0% of control (n = 4, p < 0.001 versus. 500 µM). To further evaluate the role of PKA in current inhibition by clenbuterol, we applied the drug in presence of 10 µM H-89, a specific inhibitor of the kinase (Fig. 3). In 3 fibers, H-89 alone applied externally for 10 min had no effect on I_Na, whereas further application of 1500 µM clenbuterol still reduced peak current with potency similar to that observed in the absence of H-89.

View larger version (34K): [in this window] [in a new window]   Fig. 1.   Effects of 2-adrenoceptor agonists on sodium current amplitude in rat skeletal muscle fibers. Sodium currents were elicited in cell-attached patches by 25-ms depolarizing pulses to 20 mV applied every 2 s from a holding potential of 100 mV. Peak sodium current amplitude was reported as a function of time in control conditions (CTRL) and during external application of 500 µM salbutamol (A) or 500 µM clenbuterol (B). Ensemble average sodium currents were constructed from 10 consecutive traces recorded in absence (CTRL) and in presence of the drug.

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Effects of clenbuterol on sodium current voltage-dependence in rat skeletal muscle fibers. A, current-voltage relationships constructed before (CTRL) and after application of 500 µM clenbuterol (CLE). The patches were held at 110 mV and depolarized every 10 s to potentials ranging from 100 to + 70 mV, applied in 10-mV increments. Each data point is the mean ± S.E.M. from three patches. B, activation curves were constructed from current-voltage relationships by converting current (INa) to conductance (gNa) using the equation gNa = INa/(Vm  ENa), where Vm is the membrane potential and ENa is the equilibrium electrochemical potential for sodium ions, estimated to be +70 mV. Activation curves were fitted with the Boltzmann equation, gNa/gNa,max = 1/{1 + exp[(Vm  V1/2)/K]}, to determine the half-maximum activation potential (V1/2) and the slope factor (K). In control, the values of V1/2 and K along with the S.E. of the fit were 40.9 ± 0.9 mV and 7.7 ± 0.8 mV, respectively. With 500 µM clenbuterol, V1/2 was 42.8 ± 0.9 mV and K was 6.5 ± 0.8 mV. C, availability curves for sodium current were constructed using a standard double-pulse protocol. The patches were held at 110 mV and received a 200-ms conditioning prepulse ranging in amplitude from 140 to 20 mV followed by a test pulse to 20 mV. Data points were calculated as the mean ± S.E.M. of three patches and were reported as a function of the prepulse potential. The inactivation relationships were fitted with the Boltzmann equation, INa/INa,max = 1/{1 + exp[(Vm  Vh)/K]} to determine the half-maximum inactivation potential (Vh) and the slope factor (K). The values of Vh and K along with the S.E. of the fit were 89.6 ± 0.6 mV and 5.8 ± 0.6 mV in control conditions. With 500 µM clenbuterol, Vh was 98.1 ± 0.2 mV and K was 6.1 ± 0.2 mV.

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Effects of clenbuterol on sodium currents in rat skeletal muscle fibers in presence of the cyclic AMP-dependent protein kinase inhibitor, H-89. A, ensemble average sodium currents were constructed from 10 consecutive traces elicited from 100 to 20 mV in a cell-attached patch exposed to control bath solution (CTRL, dashed line), then to 10 µM H-89, and then to 10 µM H-89 + 1500 µM clenbuterol. B, the protocol described in A was repeated in three patches and data, normalized with respect to control current, were averaged as mean ± S.E.M., and reported together with average data obtained from three patches exposed to 1500 µM clenbuterol alone.

Effects of ₂-Adrenoceptor Agonists and Antagonists on Human Skeletal Muscle Sodium Currents Expressed in tsA201 Cells. Wild-type hSkM1 channels were transiently expressed in tsA201 cells, and the resulting I_Na were recorded with patch-clamp technique in the whole-cell configuration (Desaphy et al., 2001). Externally applied clenbuterol produced both tonic and use-dependent block of I_Na elicited by depolarizing pulses to 30 mV from an hp of 120 mV (Fig. 4). Tonic block was assayed 3 min after drug application by measuring the reduction of I_Na elicited at 0.1 Hz, whereas use-dependent block was further obtained by increasing stimulation frequency to 10 Hz. In the presence of 100 µM clenbuterol, I_Na was reduced to 40% (tonic block) and 20% (10-Hz block) of control current (Fig. 4A). The inhibitory effect of clenbuterol was dose-dependent, with IC₅₀ values of 76 µM for tonic block and 26 µM for 10 Hz-block (Fig. 4B). The Hill coefficients calculated from the fitting functions were close to unity, thereby indicating a 1:1 stoichiometry.

View larger version (17K): [in this window] [in a new window]   Fig. 4.   Dose-dependent effects of clenbuterol on human skeletal muscle sodium currents in tsA201cells. A, whole-cell INa were recorded in tsA201 cells transiently transfected with the hSkM1 channel. The cells were held at 120 mV and 25-ms test pulses were applied to 30 mV. Currents were recorded under control conditions, 3 min after application of 100 µM clenbuterol at a low frequency stimulation (0.1 Hz), and during high-frequency stimulation (10 Hz). B, dose-response relationships were constructed using the protocol described in A at both 0.1 Hz (tonic block) and 10 Hz (use-dependent block). Each data point was calculated as the mean ± S.E.M. from 4 to 13 cells. The relationships were fitted with the Hill binding function, Idrug/Icontrol = 1/{1 + ([drug]/IC50)nH}, to calculate the half-maximum inhibitory concentration (IC50), and the logistic slope factor (nH). For tonic block, the values of IC50 and nH together with the S.E. of the fit were 76.4 ± 5.0 µM and 1.17 ± 0.09, respectively. For use-dependent block (10 Hz), IC50 was 25.9 ± 4.9 µM and nH was 1.03 ± 0.21. Effect of 1 mM salbutamol (mean ± S.E.M., n = 3) obtained in the same experimental conditions is also reported for comparison.

View larger version (20K): [in this window] [in a new window]   Fig. 5.   No role for cyclic AMP, PKA, PKC, and -adrenergic receptor in the inhibitory effect of clenbuterol on human skeletal muscle sodium currents. A, whole-cell INa were recorded in tsA201 cells transiently transfected with hSkM1 channel. The cells were held at 120 mV and received a 25-ms depolarizing pulse to 30 mV every 10 s. Each bar represents the mean ± S.E.M. from the number of cells indicated on the left of the bar of the residual current (Idrug/Icontrol) measured 4 to 5 min. after external application of 500 µM CPT-cyclic AMP, 300 nM okadaic acid, 300 nM okadaic acid + 500 µM CPT-cyclic AMP, and 10 µM H-89. B, effects of 100 µM clenbuterol was measured as in A in patches containing control pipette solution alone, or supplemented with 10 µM H-89 or 1 µM staurosporine, or in the presence of 1 mM external nadolol.

View larger version (34K): [in this window] [in a new window]   Fig. 6.   State-dependent affinities of human skeletal muscle sodium channels for clenbuterol. A, voltage-dependence of INa availability in tsA201 cells transfected with hSkM1 channel. INa were evoked by a 20-ms test pulse to 30 mV after 50-ms conditioning prepulses to potentials ranging from 150 to 30 mV in 10-mV increments. Pulses were delivered at 10-s intervals and hp was 180 mV. The peak INa recorded during the test pulse was normalized with respect to the maximal INa, and means ± S.E.M. were calculated from n cells to be plotted against the prepulse potential. The relationship was determined in control conditions (CTRL) and in the presence of various concentrations of clenbuterol. Only the effects of 30 and 300 µM clenbuterol are shown. The relationships were fitted with the Boltzmann equation as in Fig. 2. The values of Vh and K along with the S.E. of the fit were 80.2 ± 0.3 mV and 6.3 ± 0.2 mV in CTRL (n = 13). In the presence of 30 µM clenbuterol, Vh was 89.2 ± 0.2 mV and K was 7.3 ± 0.3 mV (n = 10). In presence of 300 µM clenbuterol, Vh was 99.2 ± 0.3 mV and K was 7.9 ± 0.2 mV (n = 4). B, the affinity of clenbuterol for inactivated channels (KI) was estimated by plotting the half-maximum inactivation potential (Vh), determined as in A, as a function of clenbuterol concentration. Each data point was the mean ± S.E.M. from 4 to 33 cells. The relationship was fitted with the equation, Vh = KCTRL × ln(1/{1 + ([drug]/KI)}) + Vh,CTRL, where KCTRL and Vh,CTRL were the values of K and Vh measured in control conditions. The values of KI determined along with the S.E. of the fit was 18.8 ± 2.1 µM. C, recovery from inactivation and clenbuterol block of hSkM1 channels. The cells were held at 120 mV. A recovery pulse at the hp of increasing duration was included between two test pulses at 30 mV. The peak INa recorded during the second test pulse was normalized with respect to the peak INa recorded during the first test pulse and means ± S.E.M. were calculated from n cells to be plotted against the recovery time. The relationship determined in control conditions (CTRL) was fitted with a monoexponential function, I(t) = A0 + A1 × [1  exp(t/1)], whereas the relationship determined in presence of 100 µM clenbuterol was fitted with a two-exponential function, I(t) = A + A × [1  exp(t/1)] + A2 × [1  exp(t/2)], using the value of 1 determined in CTRL. Fit parameters with the S.E. of the fit were A0 = 0.27 ± 0.03, A1 = 1.25 ± 0.03, 1 = 2.02 ± 0.06 ms, A = 0.24 ± 0.02, A = 1.08 ± 0.02, A2 = 0.17 ± 0.01, and 2 = 11.2 ± 3.3 ms. D, dose-response curves for depolarized and closed channels in tsA201 cells. The affinity of clenbuterol for depolarized channels was determined by eliciting INa during a test pulse at 30 mV after a 1.5-s depolarization at 70 mV followed by a 35-ms recovery period at 120 mV. Peak INa measured in presence of clenbuterol was normalized with respect to control peak INa and each data point is the mean ± S.E.M. from at least four cells. The dose-response relationship was fitted using the Hill binding function, with IC50 = 30.1 ± 2.7 µM and nH = 0.91 ± 0.07. The affinity of clenbuterol for closed channels (KR) was determined by holding the cells at 180 mV and measuring the dose-response relationship at 0.1-Hz stimulation frequency. Each data point was calculated as the mean ± S.E.M. from 3 or 4 cells. The relationship was fitted using the Hill binding function, Idrug/Icontrol = 1/{1 + ([drug]/KR)nH}. The values of KR and nH together with the S.E. of the fit were 242.3 ± 15.0 µM and 1.06 ± 0.07, respectively.

View larger version (17K): [in this window] [in a new window]   Fig. 7.   Effects of the -adrenoceptor antagonist, propranolol, on human skeletal muscle sodium channels in tsA201 cells. Whole-cell INa were elicited in tsA201 cells transiently transfected with hSkM1 channels. The cells were held at 120 mV and received a 25 ms-long depolarizing pulse to 30 mV every 10 s. A, peak sodium current amplitude was reported as a function of time in control conditions (CTRL) and during external application of 1 mM propranolol. Inset are shown INa traces recorded before, during, and after (washout) application of propranolol. B, dose-response relationships were constructed using the protocol described in A at both 0.1 Hz (tonic block) and 10 Hz (use-dependent block). Each data point was calculated as the mean ± S.E.M. from three to five cells. The relationships were fitted with the Hill binding function, Idrug/Icontrol = 1/{1 + ([drug]/IC50)nH}, to calculate the half-maximum inhibitory concentration (IC50), and the logistic slope factor (nH). For tonic block, the values of IC50 and nH together with the S.E. of the fit were 68.9 ± 2.8 µM and 1.13 ± 0.05, respectively. For use-dependent block (10 Hz), IC50 was 7.9 ± 0.2 µM and nH was 1.03 ± 0.02. Effect of 1 mM nadolol (mean ± S.E.M., n = 4) obtained in the same experimental conditions is also reported for comparison.

View larger version (43K): [in this window] [in a new window]   Fig. 8.   Development of use-dependent block after internal diffusion of control pipette solution (CTRL, plain line), or pipette solution supplemented with 1 mM nadolol (), 1 mM salbutamol (), or 300 µM QX-314 (). The hSkM1-transfected tsA201 cells were held at 120 mV and received a 25-ms depolarizing pulse to 30 mV every 10 s to elicit INa. This protocol was applied about 5 min after achieving whole-cell configuration to allow pipette solution to diffuse well within the cell. Peak INa measured at each test pulse was normalized with respect to the first pulse INa. Each data point is the mean ± S.E.M. from five cells in each condition. The S.E.M. bars are omitted for CTRL, nadolol, and salbutamol to improve clarity.

Effects of ₂-Agonists and Antagonists on Action Potentials of Rat Skeletal Muscle Fibers. We looked at the effect of clenbuterol on action potential behavior in rat skeletal muscle fibers by means of two intracellular microelectrodes (Desaphy et al., 1998b). The membrane potential was clamped to 80 mV before to apply depolarizing currents of increasing amplitude up to elicit a single action potential and then a train with the maximal number of action potentials. After collection of data in control conditions, clenbuterol was applied to the muscle, and action potentials were recorded after a short delay of ~5 min. At the concentration of 3 µM, clenbuterol had no significant effect on the single action potential but reduced by ~50% the maximum number of spikes elicited (Fig. 9B). At 30 µM, clenbuterol reduced the amplitude of the single action potential to ~80% of control and completely inhibited action potential firing (Fig. 9A). In the presence of 300 µM clenbuterol, only 3 fibers of 7 were able to elicit a single action potential, which was ~65% of control amplitude (Fig. 9C). Thus the inhibitory effect of clenbuterol on action potential was dose-dependent and use-dependent, because the drug affected the number of spikes at lower concentrations than those required to affect the single action potential. As expected from patch-clamp data, the effect of clenbuterol on action potentials was also independent of ₂-adrenoceptor stimulation, because it persisted in presence of nadolol (Fig. 9D).

View larger version (31K): [in this window] [in a new window]   Fig. 9.   Effects of clenbuterol on action potentials in rat skeletal muscle fibers. Action potentials were recorded in rat muscle fibers using two-microelectrode current clamp method. A, representative single action potentials elicited by threshold current in absence (CTRL) and presence of 30 µM clenbuterol (CLE). B, representative train of action potentials elicited by subthreshold current in absence (CTRL) and presence of 3 µM clenbuterol (CLE). C, amplitude of single action potential elicited as in A (left) and maximum number of spikes obtained as in B (right), in the absence or presence of 3, 30, and 300 µM clenbuterol, are reported as means ± S.E.M. from n fibers of N rats, indicated in parenthesis as (N/n). D, amplitude of single action potential elicited as in A (left) and maximum number of spikes obtained as in B (right), were measured in control conditions (CTRL), in presence of 300 µM nadolol (NADO), and then in presence of 300 µM nadolol and 30 µM clenbuterol (NADO+CLE) and reported as percentage of control. For comparison, effect of 30 µM clenbuterol alone (CLE) is also reported. Each bar is the mean ± S.E.M. from n fibers of N rats, indicated as (N/n). Statistical differences were assessed with unpaired Student's t test (*, p < 0.001; **, p < 0.005).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Looking for potential modulation of skeletal muscle sodium channels by the -adrenergic signaling pathway using patch-clamp technique, we observed that the -adrenergic receptor agonist clenbuterol blocked I_Na in native rat skeletal muscle fibers or in tsA201 cells expressing the human skeletal muscle sodium channel. This effect was independent of -adrenoceptor modulation and rather resembled the sodium channel block by local anesthetic-like drugs, thereby suggesting direct binding of the drug to the channels. In contrast, the -agonist salbutamol had no effect on I_Na. Such observation may have important implications for the therapeutic use of these drugs. For example, this difference between the two drugs defines the rationale for the use of salbutamol in patients suffering from periodic paralysis (sarcolemmal inexcitability), whereas clenbuterol may be more indicated in patients presenting myotonic syndromes (sarcolemmal overexcitability).

Although membrane-permeable analogs of cyclic AMP inhibited sodium currents in rat muscle fibers (Desaphy et al., 1998a), the nucleotide CPT-cAMP had no effect on skeletal muscle sodium channels expressed in tsA201 cells, as reported by others (Bendahhou et al., 1995). As discussed elsewhere (Desaphy et al., 1998a), such a difference suggests that the heterologous system of expression lacks a component present in skeletal muscle fibers and responsible for the effect of cAMP on sodium channels. -Adrenergic stimulation with salbutamol was not able to increase cyclic AMP sufficiently to block sodium channels in our experimental conditions, thereby raising concerns about the physiological significance of sodium channel direct modulation by the nucleotide.

Nevertheless, the ₂-agonist clenbuterol produced a rapid and reversible block of rat and human skeletal muscle sodium channels. This effect persisted in presence of PKA or PKC inhibitors, was not mimicked by salbutamol, and was not antagonized by nadolol, a -adrenoceptor antagonist. Together, these data indicate that the inhibitory effect of clenbuterol on I_Na was independent of -adrenoceptor stimulation. On the other hand, the effect of clenbuterol on I_Na was very similar to that of local anesthetic drugs that bind and block sodium channels, including a 1:1 stoichiometry, voltage- and use-dependent properties, and negative shift of voltage-dependence of sodium channel availability (Ragsdale et al., 1996). Interestingly, use-dependent block of sodium channels was also observed in cardiomyocytes (Fischer et al., 2001). Such properties have been explained by the modulated receptor hypothesis that forecasts the drug binding dependence on channel state as a result of change in receptor affinity (Hille, 2001). Using specific voltage clamp protocols, we estimated the affinities for closed and inactivated channels to be ~240 and ~20 µM, respectively. For comparison, using the same expression system, a typical inactivated-channel blocker such as mexiletine showed closed-channel affinity of ~800 µM and inactivated-channel affinity of ~7 µM (Desaphy et al., 2001). It should be noted that higher clenbuterol concentrations were required to block I_Na in skeletal muscle fibers to the same extent as in tsA201 cells, although voltage-clamp protocols should be more favorable to block in the native system, where less negative hp and higher frequency of stimulation were used. Hypothetic causes for such include differences in receptor affinity between the rat and the human sodium channels or differences in intracellular medium (for muscle fiber) and experimental solutions between the two systems. Importantly, inhibition of muscle action potential firing was obtained in more physiologic conditions (e.g., hp = 80 mV) with clenbuterol concentrations lower than those required to block I_Na in cell-attached patches of muscle fibers, as described below.

The putative molecular receptor for local anesthetic-like drugs includes amino acids of the S6 segments of domains I, III, and IV that face the ion-conducting pore of voltage-gated sodium channel -subunits (Ragsdale et al., 1994; Nau et al., 1999; Wang et al., 2000; Yarov-Yarovoy et al., 2001; 2002). It was proposed that the two pharmacophore moieties of many local anesthetics, constituted of an uncharged aromatic ring and a charged tertiary amine, may bind to amino acid side chains through hydrophobic and cation- interactions, respectively (Ragsdale et al., 1994). Interestingly, clenbuterol also presents a hydrophobic ring at one extreme and an amine group at the other end (Table 1). The ring confers to clenbuterol a lipophilicity comparable with that of mexiletine, as evidenced by the Log P value. Moreover, the pK_a of clenbuterol is very similar to that of mexiletine, and drug molecules are mostly protonated at physiological pH. Thus, molecular structure, physicochemical properties, and sodium channel block feature of clenbuterol strongly suggest that the drug binds to the sodium channel at the local anesthetic receptor.

It has long been hypothesized that -adrenoceptor antagonists may exert part of their antiarrhythmic action by blocking directly cardiac sodium channels. Indeed direct interaction of -adrenoceptor antagonists, including propranolol, with sodium channels was proposed on the basis of ²²Na⁺ uptake measure in rat brain membrane (Matthews and Baker, 1982), cardiac action potential modulation (Ban et al., 1985; Courtney, 1990), and ³H-batrachotoxin-A 20--benzoate binding studies to rat cerebrocortical synaptosomes (Chidlow et al., 2000). The present study confirms inhibition of I_Na by propranolol using patch clamp technique. As clenbuterol, propranolol blocked human sodium channels in a use-dependent manner and shifted negatively the voltage dependence of channel availability (not shown). The IC₅₀ value for tonic block at a hp of 120 mV was similar to that of clenbuterol, whereas use-dependent block was three-fold more pronounced with the -antagonist. The structure of propranolol that includes a strongly lipophilic naphthalene moiety and a protonated amine fulfills the general structural requirements for sodium channel binding and block by local anesthetic-like drugs, as described above for clenbuterol.

In contrast to clenbuterol and propranolol, salbutamol and nadolol had no effect on I_Na. From Table 1, it seems that the two inactive compounds are characterized by the presence of two hydroxyl groups on the aromatic moiety that render them far less lipophilic compared with clenbuterol and propranolol. It can be hypothesized that the hydroxyl groups may impede the hydrophobic interaction between the aromatic moiety and the local anesthetic receptor. Interestingly, it should be noted that most of the ₂-agonists present two hydroxyl groups on their aromatic moieties, and that the atypical clenbuterol may be the unique ₂-agonist able to block sodium channels. On the other hand, nadolol is the unique -adrenoceptor antagonists with two hydroxyl groups on the aromatic moiety, suggesting that sodium channel blocking activity may be shared by many -antagonists, as suggested by previous studies (Matthews and Baker, 1982; Ban et al., 1985; Courtney, 1990; Chidlow et al., 2000).

Consistent with its blocking action on sodium channels, clenbuterol inhibited action potentials in skeletal muscle fibers. This effect was use-dependent because 3 µM clenbuterol drastically reduced the number of spikes without affecting the amplitude of a single action potential. Clenbuterol effect persisted in presence of the -adrenoceptor antagonist nadolol; thus, inhibition of action potential firing most probably resulted from direct block of sodium channels by the drug. Clenbuterol concentration can reach 1 to 2 ng/g in skeletal muscles of rats treated with 1 mg/kg body weight/day, which is the safe therapeutic clenbuterol dose in humans (Zeman et al., 2000; Von Deutsch et al., 2002). Such a concentration is quite lower than that we used to block depolarized channels, but caution should be used in comparing the results obtained in the heterologous system of expression with clinical data. Interestingly, the K_I value for clenbuterol is near that measured under the same experimental conditions for mexiletine (~7 µM; Desaphy et al., 2001) and flecainide (~15 µM; J.-F. Desaphy and D. Conte Camerino, unpublished observations), two drugs used with success in myotonic patients. Thus, higher doses of clenbuterol may affect tissue excitability; such an effect may occur especially in conditions of hyperexcitability owing to use-dependent block of sodium channels. Interestingly, such a mechanism of action was recently proposed for the beneficial effect of clenbuterol in various seizure models of experimental epilepsy (Fischer et al., 2001). At the skeletal muscle level, sarcolemmal hyperexcitability leads to myotonia, a condition of muscle stiffness shared by various genetic muscle diseases (Cannon, 2001; Moxley, 2000; Meola, 2002). On the basis of our results, it would be important to verify the therapeutic potential of clenbuterol in the myotonic syndromes. In particular, because of the possibility of combining antimyotonic activity with its well known anabolic action, clenbuterol might be remarkably indicated in the treatment of myotonic dystrophy, the most common hereditary disease of skeletal muscle, characterized by muscle wasting together with permanent or fluctuans myotonia (Meola, 2002). On the other hand, because sodium channel block may accentuate paralysis, clenbuterol should not be administrated to patients with HPP, whereas other ₂-agonists, such as salbutamol, have proven to be beneficial in those patients, most probably because ₂-adrenoceptor stimulation activates the Na,K-ATPase and consequently hyperpolarizes the muscle fiber (Wang and Clausen, 1976; Clausen et al., 1993; Hanna et al., 1998).