Preclinical evaluation of marketed sodium channel blockers in a rat model of myotonia discloses promising antimyotonic drugs

Abstract

Although the sodium channel blocker mexiletine is considered the first-line drug in myotonia, some patients experiment adverse effects, while others do not gain any benefit. Other antimyotonic drugs are thus needed to offer mexiletine alternatives. In the present study, we used a previously-validated rat model of myotonia congenita to compare six marketed sodium channel blockers to mexiletine. Myotonia was induced in the rat by injection of anthracen-9-carboxylic acid, a muscle chloride channel blocker. The drugs were given orally and myotonia was evaluated by measuring the time of righting reflex. The drugs were also tested on sodium currents recorded in a cell line transfected with the human skeletal muscle sodium channel hNav1.4 using patch-clamp technique. In vivo, carbamazepine and propafenone showed antimyotonic activity at doses similar to mexiletine (ED50 close to 5mg/kg); flecainide and orphenadrine showed greater potency (ED50 near 1mg/kg); lubeluzole and riluzole were the more potent (ED50 near 0.1mg/kg). The antimyotonic activity of drugs in vivo was linearly correlated with their potency in blocking hNav1.4 channels in vitro. Deviation was observed for propafenone and carbamazepine, likely due to pharmacokinetics and multiple targets. The comparison of the antimyotonic dose calculated in rats with the current clinical dose in humans strongly suggests that all the tested drugs may be used safely for the treatment of human myotonia. Considering the limits of mexiletine tolerability and the occurrence of non-responders, this study proposes an arsenal of alternative drugs, which may prove useful to increase the quality of life of individuals suffering from non-dystrophic myotonia. Further clinical trials are warranted to confirm these results.

Keywords: Mexiletine; Non-dystrophic myotonia; Over-excitability; Patch-clamp; Rat model; Sodium channel blockers.

Fig. 1

In vivo antimyotonic effects of exploratory drugs in the rat model of myotonia. The time course of normalized TRR values was determined in 9AC-treated rats receiving various doses of propafenone, carbamazepine, flecainide, orphenadrine, lubeluzole, riluzole, and mexiletine. Each data point is the mean ± S.E.M. calculated from 3-to-4 rats. Statistical analysis of differences between drug doses was performed at each time point with ANOVA followed by ad-hoc Bonferroni's t-test and results are reported in Supplementary Fig. S1. The dose–response curve was constructed for all the drugs by reporting the value of normalized TRR (expressed as percentage of the TRR calculated in rat receiving drug vehicle alone), measured 10 min after drug administration (i.e. 30 min after 9AC injection). The dose–response curve was fitted with equation TRR = min + (100 min) / [1 + exp([Drug] / ED₅₀)^nH], where [Drug] is the drug dose, ED₅₀ value is the half-maximum efficient dose, nH is the slope factor, and min the curve baseline. Values of fit parameters ± the S.E. of the fit are given in Table 1.

Fig. 2

Effects of propafenone, carbamazepine, and carbamazepine-10,11-epoxide, on hNav1.4 sodium channel subtype. (A–C) Sodium currents were measured in HEK293 cells permanently transfected with the human skeletal muscle hNav1.4 isoform using the whole-cell patch-clamp method. Currents were elicited by depolarizing the cells for 20 ms at − 30 mV from a holding potential of − 120 mV, every 10 or 0.1 s (i.e. 0.1 Hz or 10 Hz stimulation frequency, respectively). Sodium current traces recorded in representative cells are shown in control condition (CTRL) and after acute application of 10 μM propafenone, 300 μM carbamazepine, or 300 μM carbamazepine-10,11-epoxide, at 0.1 and 10 Hz. (D) Concentration–response relationships were constructed at 0.1 and 10 Hz frequency stimulation for propafenone and carbamazepine. Each point is the mean ± S.E.M. from at least 3 cells. The relationships were fitted with equation [I_DRUG / I_CTRL = 1 / {1 + ([DRUG] / IC₅₀)^nH}], where IC₅₀ value is the half-maximum inhibitory concentration and nH is the slope factor. The nH values were comprised in between 0.8 and 1.2. The calculated IC₅₀ ± S.E. of the fit is reported in Table 2.

Fig. 3

Effects of exploratory drugs on hNav1.4 sodium channel subtype in myotonia-like conditions. Sodium currents were elicited in HEK293 cells permanently transfected with the human skeletal muscle hNav1.4 isoform using the whole-cell patch-clamp method. To mimic a myotonia condition, the cells were depolarized for 5 ms (action potential duration) at − 30 mV from a holding potential of − 90 mV (sarcolemma resting membrane potential), the stimulation frequency of 50 Hz (myotonic run frequency). (A) Sodium current traces recorded in a representative cell are shown in control condition at 0.1 Hz (CTRL) and 50 Hz (use-dependent block in control, UDB-C) and after acute application of 10 μM mexiletine at 0.1 Hz (tonic block exerted by the drug, TB) and 50 Hz (use-dependent block observed in the presence of drug, UDB-D). (B) Time course of sodium current peak amplitude reduction induced by 50 Hz stimulation in the absence of drug (UDB-C, triangles) and in the presence of 10 μM mexiletine (UDB-D, inverted triangles). The net reduction induced by the drug was obtained by subtraction of UDB-C to UDB-D and addition of TB. The relationships represent the means from 3 cells; S.E.M. is not reported for improving clarity of the graph. (C) Concentration–response relationships were constructed for all the exploratory drugs following the protocol shown in (B). Each point is the mean ± S.E.M. from at least 3 cells. The half-maximum inhibitory concentration (IC₅₀) values were approximated as the intercept of the relationships with 50% of (TB + UDB), and are reported in Table 2.

Fig. 4

Relationship between in vitro hNav1.4 sodium channel inhibition and in vivo antimyotonic activity in the rat model. The IC₅₀ values for sodium current inhibition calculated in vitro in myotonia-like conditions (calculated as in Fig. 3) are reported versus the ED₅₀ value for in vivo antimyotonic activity in the rat model (measured as in Fig. 1). For carbamazepine-10,11-epoxide (epo-CBZ), the ED₅₀ value was measured for carbamazepine. The linear regression considering all the drugs (but epo-CBZ) had a coefficient r² = 0.52 (not shown). The linear regression of all drugs except epo-CBZ, carbamazepine, and propafenone, shows an r² coefficient of 0.91.