Mexiletine is a class 1 antiarrhythmic drug with neuroprotective effects in models of brain ischemia due to inhibition of brain sodium channels. We compared effects of R-mexiletine on wild-type (WT) and mutant rat brain (rbIIA) and heart (rh1) sodium channel alpha subunits transiently expressed in tsA-201 cells. R-mexiletine induced tonic and frequency-dependent block of both brain and heart channels. The affinity of resting channels was determined as the limiting minimum affinity reached at strongly negative holding potentials (-140 mV or –160 mV). R-mexiletine bound with a 25-fold (brain) or 35-fold (heart) higher affinity to depolarized channels. Affinities for both resting and depolarized channels for R-mexiletine block were approximately two-fold higher for heart than for brain channels. Mutations in transmembrane segment IVS6 of heart (rhF1762A) and brain (rbF1764A and rbY1771A) channels, which reduce block by other local anesthetics, also reduced block of inactivated channels and use-dependent block by R-mexiletine. Unlike previous local anesthetics studied, the strongest effect was observed for mutation rbY1771A. A comparison of mutations of the homologous phenylalanine residue in brain and heart channels showed striking differences in the effects of the mutations. rbF1764A reduced drug block by drastically slowing R-mexiletine inhibition of depolarized channels. rh1762A reduced R-mexiletine block by increasing the rate of unbinding from depolarized or resting channels. Thus, rbF1764/rhF1762 is a critical determinant of local anesthetic block in both brain and heart sodium channels, but its role differs in the two channel isoforms.

Introduction

Inhibitors of voltage-gated sodium channels are widely used clinically. Blockers of cardiac sodium channels are potent antiarrhythmics, whereas the inhibition of neuronal sodium channels is useful for local anesthesia and treatment of epilepsy (Butterworth and Strichartz, 1990; Caron and Libersa, 1997; Hondeghem and Katzung, 1984; Catterall, 1987; Ragsdale et al., 1991). Moreover, inherited myotonias and cardiac arrhythmias caused by mutations in skeletal muscle and cardiac sodium channels (Mitrovi'c et al., 1995) can be effectively treated by class I antiarrhythmics (De Luca et al., 1995). Thus, different sodium channel subtypes expressed in neurons, cardiac or skeletal muscle cells can be modulated by molecules of similar chemical structures. These overlapping actions bear risks and benefits. On the one hand, local anesthetics inadvertently injected into a blood vessel can cause severe cardiac arrhythmias. On the other hand, some cardiac antiarrhythmics, including mexiletine, also penetrate the blood-brain barrier, and have interesting neuroprotective properties (e.g. Stys and Lesiuk, 1996).

Voltage-gated sodium channels are heteromultimeric proteins consisting of a principal a subunit of 360 kD, as well as a b1 subunit of 36 kD and, in the brain, a b2 subunit of 33 kD. The a subunit consists of four homologous transmembrane domains (I-IV), each containing six transmembrane a-helical segments, termed S1 through S6 (Catterall, 1992; Fozzard and Hanck, 1996). The principal electrophysiological functions are mediated by the a subunit; the b subunits have only minor effects when the channels are heterologously expressed in mammalian cells (Isom et al., 1995). The rat brain type IIA sodium channel a subunit is a principal isoform expressed in the brain (Gordon et al., 1987; Beckh et al., 1989) and its heterologous expression in mammalian cells yields sodium currents with physiological and pharmacological properties that are similar to those observed in rat brain neurons (West et al., 1992; Ragsdale et al., 1991). The rH1 sodium channel a subunit is the primary isoform expressed in the heart (Rogart et al., 1989; Kallen et al., 1990) and expression of this isoform in Xenopus oocytes or mammalian cells yields channels with physiological and pharmacological properties characteristic of heart sodium channels (Cribbs et al., 1990; White et al., 1991; Qu et al., 1994; Qu et al., 1995).

A study in which alanines were substituted for each amino acid in transmembrane segment IVS6 identified mutations of two amino acid residues in the rbIIA sodium channel, F1764A and Y1771A, that reduced block by the local anesthetic etidocaine (Ragsdale et al., 1994). Mutation of these residues also reduced block by a range of local anesthetic, anticonvulsant and anti-arrhythmic compounds to varying degrees and with different rank order potencies, depending on the structure of the particular blocker in rbIIA neuronal sodium channels (Ragsdale et al., 1996) and m 1 skeletal muscle channels (Wright et al., 1998; Wang et al., 1998). Mutation of the residue homologous to rbIIA F1764 in the rH1 sodium channel, F1762, to ala also resulted in loss of block of the rH1 channel by the quaternary lidocaine analogue, QX314 in the rH1 (Qu et al., 1995). That study showed that rH1 F1762/rbIIA F1764 was a critical residue for local anesthetic block in both channel backgrounds.

In the experiments described here, we compared the effects of R-mexiletine on cloned rat brain and heart sodium channel a subunits heterologously expressed in a common cellular background. We also examined the effects of F1764A and Y1771A mutations in the rbIIA channel isoform and compared the effects of mutation of phe rhF1762/rbF1764 in both heart and brain backgrounds. We show that heart channels have an intrinsically higher affinity for block by R-mexiletine than do brain channels when expressed in the same cellular background. In contrast, lidocaine has a similar intrinsic affinity for skeletal muscle and heart muscle sodium channels (Wright et al., 1997). Mutations rbF1764A and rbY1771A both potently reduced R-mexiletine affinity but with different specificity than for other local anesthetics. Surprisingly, although mutations of the homolous phe residue in brain and heart channels (rhF1762A/rbF1764A) both reduce block by R-mexiletine, the kinetic details of their effects on are strikingly different. Thus, although this phenylalanine is essential for local anesthetic block of both heart and brain sodium channel isoforms, its role in that block is fundamentally different in the two channel backgrounds. This detailed comparison of effects of mutations of a single homologous amino acid in different channel backgrounds yields novel insights into the role of this residue in heart and brain channels.

Material and Methods

Cell maintenance and transient transfection for electrophysiological recording. tsA-201 cells which are a subclone of HEK293 cells expressing SV40 t-antigen, were a kind gift of Dr. Robert Dubridge (Cell Genesys, Foster City, CA). Cells were maintained in DMEM/F12 media (Gibco/Life Technologies) supplemented with 10% fetal bovine serum (Hyclone), 25 units/ml penicillin and 25 m g/ml streptomycin (Sigma). An EcoRV fragment containing mutants F1764A and Y1771A of RIIA in pVA2580 (Ragsdale et al., 1996) was subcloned into pCDM8 containing the remainder of the rIIA sodium channel a subunit as described (Linford et al., 1998). rH1 (Rogart et al., 1989) and rH1 mutant F1762A in pCDM8 were described previously (Qu et al., 1996). tsA-201 cells were transiently transfected with WT or mutant a subunits and a vector encoding the human CD8 cell surface protein (EBO-pCD-leu2; American Type Culture, Rockville, MD) for cell recognition as described (Margolskee et al., 1993). Successfully transfected cells were labeled to recognize them for recording using anti-CD8-coated polystyrene microspheres (Dynabeads M-450 CD8, Dynal, Great Neck, NY) as described (Jurman et al., 1994).

Electrophysiological recording. Sodium currents were recorded from transiently transfected tsA-201 cells in the whole cell voltage clamp configuration (Hamill et al., 1981) at 22^o C as described (Qu et al., 1996). The extracellular solution contained (in mM): 140 NaCl, 5 CsCl, 1.8 CaCl₂, 1.0 MgCl₂, 10 glucose, and 10 HEPES (pH = 7.4, adjusted with NaOH). The intracellular solution contained 90 CsF, 50 CsCl, 10 CsEGTA, 10 NaF, 2 MgCl₂ and 10 HEPES (pH=7.4 adjusted with CsOH). Recording pipettes had resistances of 0.8 to 1.8 MOhms when filled with intracellular solution. The cells were bathed with the effluent of a gravity-driven "sewer-pipe" perfusion system consisting of a series of parallel tubes with each tube containing either control solution or a solution containing R-mexiletine. Mexiletine is a racemate of S(+)- and R(-) enantiomers, which have differential effects on sodium channels (De Luca et al., 1995). In this study only the (-) enantiomer of R-mexiletine was used. The compound was synthesized at the Department of Pharmaceutical Chemistry at Boehringer Ingelheim KG (Ingelheim, Germany). R-mexiletine was dissolved in extracellular solution at the highest concentration to be used in an experiment and diluted with extracellular solution to each of the other concentrations used. Solutions were changed by translating the array of tubes so that the tube containing the appropriate concentration was bathing the cell. The entire petri dish was continuously perfused with control extracellular solution. Solution changes were complete within 2 s. Currents were recorded using an Axopatch 200B patch clamp amplifier (Axon Instruments, Foster City, CA). Voltage clamp commands were delivered and currents recorded using PClamp 6 controlling a Digidata 1200 interface (Axon Instruments). Whole cell capacitance was compensated using the internal voltage clamp circuitry and approximately 80% of series resistance was compensated. Residual linear leakage and capacitance were subtracted using a P/4 protocol when appropriate (Bezanilla and Armstrong, 1977). Data analysis and curve fitting were performed using Sigma Plot (SPSS, Chicago) or Prism (Graph-Pad Software, San Diego). All data points are the means of 3 to 6 experiments and grouped data are reported as ± SD.

Unless otherwise indicated, holding and test potentials were -100 and 0 mV for rbRIIA WT and F1764A, -120 and 0 mV for Y1771A, and -120 and -20 mV for rH1 heart sodium channels.

Results

Mutations F1764A and F1771A inhibit tonic and phasic block of brain Na+ channels by R-mexiletine. The principal characteristics of block by R-mexiletine are illustrated in Fig. 1A. tsA-201 cells expressing rbIIA sodium channel a subunits were voltage clamped at a holding potential of -100 mV, and the control current trace was recorded in response to a 10 ms depolarization to 0 mV. The cell was then exposed to R-mexiletine (100 mM) in the absence of pulses. This was followed by a 5 Hz train of 100 pulses (10 ms duration) to 0 mV. The first and one hundredth pulse of the train are shown. Approximately 20% of the current was tonically blocked by this concentration of R-mexiletine. The pulse train resulted in approximately 45% block in the steady state at this concentration and pulse frequency. Thus, repetitive depolarization causes greater sensitivity to block by R-mexiletine, as is observed for many local anesthetic compounds (Hille, 1977; Hondeghem and Katzung, 1984). Concentration-response curves were obtained from similar experiments with IC50 values of 305 mM and 165 mM for tonic and phasic block, respectively (Fig. 1B). Thus, block increased approximately 2-fold during a 5 Hz train of this type.

Two mutants in transmembrane segment IVS6 of the rat brain sodium channel a subunit, F1764A and Y1771A, diminish block by diverse local anesthetic, antiarrhythmic, and antiepileptic compounds and are proposed to contribute to their binding sites to varying degrees (Ragsdale et al., 1994; Ragsdale et al., 1996). Therefore, we have examined R-mexiletine block of these mutant channels, as illustrated for 300 mM R-mexiletine in Figs. 1C,E. Even at this 3-fold higher drug concentration compared to WT (Fig. 1A), phasic block during the train is reduced for F1764A and virtually abolished for Y1771A. This is reflected in the IC50 values of 610 µM and 410 mM for tonic and phasic block, respectively, of F1764A (Fig. 1D), and of 717 µM and 616 mM for Y1771A (Fig. 1F). Thus, in comparison to WT, tonic block is disrupted by 2-fold and 2.4-fold in mutants F1764A and Y1771A, respectively. Similarly, phasic block measured at 5 Hz is reduced by 2.5-fold and 3.7-fold in F1764A and Y1771A, respectively.

The degree of phasic block becomes greater as stimulation frequency is increased from 1 to 25 Hz (Fig. 2A), indicating that increased frequency of depolarization strongly favors drug block. Fig. 2B shows examples of R-mexiletine block of WT (closed circles), F1764A (squares), and Y1771A (triangles) channels during 25 Hz trains of depolarizations. Whereas almost 70% of the current is blocked in the WT channels, only 27%is blocked in F1764A and block develops much more slowly during the train. Even at this frequency, phasic block of Y1771A is virtually nonexistent. Use-dependent block results from preferential block of depolarized channels and is strongly influenced by the rate of block during each depolarization and the rate of unblock during intervening repolarizations. In order to understand the mechanism by which phasic block is disrupted in F1764A and Y1771A, each of these factors was examined.

Disruption of voltage-dependent block in mutants F1764A and Y1771A. We first investigated the kinetics of R-mexiletine binding to depolarized sodium channels. To measure the time course of binding to depolarized channels, prepulses to 0 mV of variable duration (0.5 - 1000 ms) were applied in the absence or presence of increasing concentrations of R-mexiletine followed by a test pulse. Prepulse and test pulses were separated by an interval of 45 ms at the holding potential of -100 mV. This interval was long enough to allow unblocked channels to recover completely from inactivation (unpublished results), but short enough so that R-mexiletine unbinding was minimal (<10%, see Fig. 7A). Data values in the presence of the compound were normalized to the controls. In WT, as well as F1764A channels, the data points could be best fitted by double-exponential functions (Fig. 3A,B). Block of Y1771A mutant channels was so slow that relatively little block developed, even after 1 s of depolarization (Fig. 3C). The small magnitude of the decay prevented fitting it reproducibly. The slow, weak block of Y1771A is consistent with the lack of phasic block (see Fig. 2B).

As most channels are inactivated for depolarizations longer than 5 ms to 0 mV, the slow time constant of block for WT and F1764A can be attributed to R-mexiletine binding to inactivated channels. The most likely interpretation of the faster component is that it reflects rapid binding of R-mexiletine via the open sodium channel. Consistent with this idea, the time constant and magnitude of the exponential component represented by the fast time constant were also concentration-dependent. They corresponded roughly to the period of time over which channels opened. However, the exact time course and magnitude of this component was not resolved precisely enough in these experiments to determine its nature in detail. Like the slow component, its magnitude was greatly reduced in the mutant channels (Fig. 3B,C).

We also examined drug block at potentials where most channels are inactivated but activation is minimal. WT and mutant channels were depolarized to -40 mV (-60 mV for Y1771A) for 1 s to allow drug to bind. At this potential more than 98% of channels were inactivated, thus allowing us to study the effect of R-mexiletine on inactivated channels. The membrane potential was then returned to the holding potential of –100 mV (-120 mV for Y1771A) for 10 ms followed by a 10 ms test pulse to 0 mV. A more negative potential was used for Y1771A to compensate for its more negative voltage dependence of inactivation when expressed in tsA-201 cells (Weiser and Scheuer, unpublished observations). In control experiments virtually all of the channels that were inactivated during the –40 mV prepulse recovered from inactivation during the 10 ms repolarization period, whereas only 2% of the R-mexiletine-blocked channels recovered from drug block (see Fig. 7A below). Concentration-response curves obtained using this protocol showed that R-mexiletine inhibited WT channels with an IC50 of 20.7 µM (Fig. 4). Inactivated mutant channels had reduced affinity for block by R-mexiletine with IC50 values of 156 mM and 309 mM for F1764A and Y1771A channels, respectively (Fig. 4). Thus, the mutation F1764A disrupted block of the inactivated state by 7.5 fold, and the mutation Y1771A caused a 14.9-fold disruption of inactivated state block. For WT channels, the ratio between the tonic block affinity and this measurement of affinity for inactivated channels was 14.7. In contrast, for mutant F1764A this ratio was only 3.9 and for mutant Y1771A the ratio was 2.3. Thus, these mutations affect R-mexiletine affinity for inactivated channels more than they affect affinity for resting channels.

Compounds that bind preferentially to depolarized and inactivated channels generally shift the voltage-dependence of steady-state inactivation. This is illustrated in Fig. 5 for R-mexiletine block of WT channels. We applied test pulses to 0 mV after 1s prepulses to different potentials in the absence or presence of various concentrations of R-mexiletine (Fig. 5A). To demonstrate the concentration-dependent shift in inactivation curves, the same data were normalized to the maximum value obtained with the Boltzmann fits (Fig. 5B). For these WT channels, increasing concentrations of mexiletine progressively shifted the steady-state inactivation curves to more negative potentials, consistent with R-mexiletine preferently binding to the inactivated state of the channel as measured directly above. Since preferential binding to inactivated channels was reduced in mutants F1764A and Y1771A, inactivation curves would be expected to be less strongly shifted as a function of drug concentration. Consistent with this expectation, steady-state inactivation curves were shifted less in the presence of 300 m M R-mexiletine in comparison to WT (Fig. 5C). Similar results were observed at other R-mexiletine concentrations (Table 1).

Affinity of resting WT and mutant rat brain Na+ channels for R-mexiletine. The previous experiments show that R-mexiletine binds with high affinity to inactivated brain Na+ channels. R-mexiletine also induced tonic inhibition of sodium currents with lower apparent affinity (see Fig. 1). This reduced affinity could result from a high affinity for the small number of inactivated sodium channels at the holding potential of -100 mV. Tonic block that is entirely due to stabilization of residual inactivated channels should continue to decrease with further hyperpolarization as that inactivation is reversed. Therefore, we asked whether increased holding potential hyperpolarization reduces tonic block by R-mexiletine (Fig. 6). Cells were voltage clamped for 20 s to –80 mV (squares), -100  mV (triangles), -120  mV (inverted triangles), and -140  mV (diamonds) for WT (Fig. 6A) and F1764A (Fig. 6B), or –100  mV (squares), -120  mV (triangles), -140  mV (inverted triangles), and -160 mV (diamonds) for Y1771A (Fig. 6C), prior to a test pulse to 0 mV. Concentration-response curves for block by R-mexiletine were determined at each holding potential. For each construct, the curves for holding potentials more negative than -120 mV superimpose, indicating that R-mexiletine affinity approaches a limiting value with hyperpolarization that indicates the IC50 for resting channels. The limiting IC50 value for resting WT channels (533 µM at –140 mV, Table 1) was less than for F1764A (813  µM) or Y1771A (655 µM) channels, indicating that the affinity of resting channels for R-mexiletine was reduced in the mutants. The dependence of the IC50 for R-mexiletine block on the holding potential (compare diamonds with squares in Fig. 6) was strongest in WT channels, less pronounced for F1764A, and almost absent for Y1771A mutant channels.

Recovery from use-dependent block by R-mexiletine. In addition to increased affinity for depolarized channels, the degree of use-dependent block during a train of depolarizations is also determined by the rate of recovery from such block between the depolarizations. Therefore, we analyzed the kinetics of recovery from use-dependent block by R-mexiletine. Use-dependent inhibition was induced by a conditioning train of 10 depolarizing pulses at 10 Hz frequency in the presence of 300 µM R-mexiletine. The recovery was characterized by applying test pulses after recovery intervals of variable duration (1-5000 ms). The data obtained by this procedure could be fitted with the sum of two exponential functions, with the time-constant of the faster component describing the recovery from inactivation of unblocked channels, and the slower time constant, the recovery of the mexiletine-blocked fraction of channels. For WT channels (Fig. 7A), the slow time-constant due to drug unbinding was 571 ms, indicating that mexiletine left its binding sites relatively rapidly. For F1764A channels (Fig. 7B), the total amount of use-dependent block induced by the train was reduced, in comparison to WT. However, the time-constant of drug unbinding was even slightly slower than in the WT channels (858 ms). Little phasic block of Y1771A mutant channels could be developed, even at high stimulation frequencies (Fig. 2B). Therefore, recovery was not analyzed for this mutant. These results indicate that the reduced use-dependent block in mutant F1764A is caused by reduced drug binding during each depolarization and not from accelerated unbinding between depolarizations.

Block of WT and F1762A mutant rat heart sodium channels by R-mexiletine. Block of sodium channels formed by expression of rat heart a subunits by R-mexiletine was also studied and compared to brain channels. Tonic and phasic block were initially examined during a train of 100 depolarizations using a protocol analogous to that of Fig. 1 (Fig. 8A). For these experiments, the holding and test potentials were 20 mV negative to those used for brain to account for the more negative activation and inactivation properties of the rat heart channel (Fozzard and Hanck, 1996; Weiser and Scheuer, unpublished observations). Concentration-response curves (Fig. 8B) obtained using this protocol gave IC50 values of of 165 and 52 µM for tonic (closed circles) and phasic (open circles) block, respectively. These values were lower than those for WT brain channels (Fig. 1A, C; dotted lines in Fig. 8B), suggesting a higher affinity of R-mexiletine for heart sodium channels.

We also examined the F1762A mutant of the heart channel, which is homologous to the brain F1764A mutation (Qu et al., 1995). As with its brain counterpart, tonic (closed circles) and phasic (open circles) inhibition by R-mexiletine were reduced for this mutant (Fig. 8C) with IC50 values of 748 µM and 619 µM, respectively (Fig. 8D). In contrast to the WT channels, the affinity for the mutant heart channel measured in this way was actually lower than for its F1764A brain counterpart (dotted lines in Fig. 8D).

As with brain channels, the degree of phasic block increased with stimulation frequency from 5% at 1 Hz to 85% at 25 Hz. However, for the equivalent concentration and frequency, greater block was reached for heart than for brain sodium channels (compare Fig. 9A, heart, with Fig. 2A, brain). This is consistent with the higher sensitivity of heart channels to phasic block by R-mexiletine.

Phasic inhibition (Fig. 9B) of mutant F1762A (squares) was much less pronounced than for the WT heart channel (circles). A 25 Hz stimulus train in the presence of 100 µM mexiletine induced only about 20% inhibition in the F1762A mutant compared to 85% in the WT. Despite the large disruption by this mutant, the rates of development of use-dependent block were quite similar for rH WT and F1762 (compare the dashed line in Fig. 9B which is the fit to the F1762A data scaled to be similar in magnitude to the WT data). For the 100 µM R-mexiletine concentration shown, the stimulation-dependent decline of peak current responses could be fitted with single exponential functions with similar time constants for the WT and mutant channels (on-rate WT: 0.39/pulse; on-rate for F1762A: 0.42/pulse). This contrasts dramatically with the F1764A mutant brain channel where the development of phasic block during a train in the mutant channel was dramatically slower than WT (Fig. 2B).

Voltage-dependent inhibition of WT and F1762A mutant heart sodium channels. The difference in the effects of mutants on the kinetics of use-dependent block in the brain and heart channels suggested that the mutants might have different effects on the detailed kinetics of R-mexiletine action. To investigate the onset of inhibition in more detail, we applied depolarizing prepulses of various lengths prior to a test pulse, a protocol analogous to that of Fig. 3. For a given concentration, the degree of depolarization-dependent block was greater for the WT rH1 channel (Fig. 10A) in comparison to the rbIIA channel (Fig. 3A), consistent with the increased block of the heart channel during trains of pulses presented in Fig. 9. As for brain channels, data could be best fitted by double exponentials. Significant binding to the heart channel occurs after most channels have inactivated, and suggests that mexiletine can bind to inactivated channels. The onset of block in response to depolarization in mutant F1762A showed that the final level of block at the end of a 1s depolarization was greatly reduced for a given concentration (Fig. 10B), but the kinetics of the development of block were not dramatically affected. This is demonstrated in Fig. 10D where the fits to the onset data for rh1 WT and F1762A are normalized and superimposed, showing their similar time courses. This similarity contrasts with the drastically different time courses observed for rbIIA WT and mutant F1764A (Fig. 10C).

As for the brain channel, we recorded concentration-response curves after depolarizations that inactivated most channels, but produced minimal activation. WT- or F1762A-expressing cells were depolarized to -60 mV for 1 s followed by a 10 ms return to the holding potential. The degree of inactivation induced by the conditioning pulse was then assayed using a test pulse to -20 mV. R-mexiletine inhibited WT channels at -60 mV with an IC50 of 9.4 µM, whereas 149 µM were necessary for half-maximum block of F1762A channels (Fig. 11A). Thus, the IC50 values for the inhibition of WT heart or brain channels differed by a factor of about 2, whereas the concentrations for half-maximum block of the rH F1762A and rbIIA F1764A were similar (156 versus 149 µM). The amount of inactivation at the prepulse potential was greater than 97% for each channel type. The similar amounts of inactivation for the different channels allows a direct comparison of the IC50 values obtained with the prepulse experiments. Thus, the affinity of R-mexiletine for depolarized heart WT channels is about two-fold higher than for brain WT channels, and is disrupted by 7.2- and 16.5-fold in rb F1764A and rh F1762A, respectively.

The shift of steady-state inactivation curves towards more negative potentials was also observed for R-mexiletine-blocked heart sodium channels (Fig. 11B, circles). For a given concentration, the effect was more pronounced for rhWT channels than for their rbWT counterparts (see Table 1). As was the case for the rb F1764A mutant, the shifts were reduced in the rh F1762A channels (Fig. 11B, squares; Fig. 11C). In contrast to WT channels, the shifts were comparable for rh F1762A and rb F1764A mutants at the same R-mexiletine concentration (compare Figs. 5C and 11C; Table 1).

Affinity of resting WT and mutant rat heart Na⁺ channels for R-mexiletine. We also analyzed the concentration dependence of tonic inhibition of heart sodium channels at increasingly negative potentials. For rhWT channels, tonic inhibition by R-mexiletine was similar at holding potentials of –140 mV and –160 mV (Fig. 12A; inverted triangles, diamonds). Fits to these curves gave an IC50 of 165 mM, approximately twice the affinity of the rbWT channel. Higher affinities were measured at holding potentials of –120 mV and –100 mV (Fig. 12A, triangles, squares) as expected if the affinity for inactivated channels is higher than for resting channels. The affinity for resting channels was reduced in mutant rhF1762A at -140 mV and -160 mV(Fig. 12B, inverted triangles, diamonds), and low affinity block was also observed at -120 mV (Fig. 12B, triangles), consistent with the reduced affinity of the inactivated state. Fits of the curves at –140 and –160 mV yielded an IC50 of 748 mM, similar to the rbF1764A value of 614 mM. Thus, resting mutant rbF1764A brain and rhF1762A heart channels had similar affinity for R-mexiletine, whereas WT heart channels had an approximately 2-fold higher affinity (Table 1).

Recovery of heart sodium channels from use-dependent block by R-mexiletine is accelerated in mutant rhF1762A. Frequency-dependent block was induced by a train of 10 conditioning pulses to -20 mV (10 Hz frequency). Recovery was assayed by a test pulse to -20 mV after repolarizations to –120 mV of increasing duration. Approximately 75% of rhWT channels were blocked by such a train in the presence of 100 mM R-mexiletine, as indicated by the fraction of channels recovering with the slower exponential slow time course at the holding potential. For rhWT channels, the time constant for this recovery was 529 ms (Fig. 13A). Frequency-dependent block of rhF1762A channels in the presence of 300 mM R-mexiletine produced only approximately 25% block using this protocol, consistent with the reduced affinity of the depolarized channel in this mutant (Fig. 9). Recovery of blocked channels occurred with a time constant of 168 ms, 3-fold faster than the rhWT channel (Fig. 13B). Thus, accelerated recovery from drug block on repolarization in mutant F1762A must contribute substantially to the reduced frequency-dependent block observed in this mutant.

Channel-specific amino acid residues in the IVS6 transmembrane segment of the rat heart sodium channel slow recovery from use-dependent block by R-mexiletine. The quaternary local anesthetic QX314 blocks WT heart, but not brain, Na+ channels from the extracellular medium (Baumgarten et al., 1991). Mutations in the extracellular portion of transmembrane segment IVS6 can provide an extracellular access pathway in the brain channel (Ragsdale et al., 1994). Conversely, chimeric mutants rhLTT-FVS and rhT1775V, which change rat heart residues L1752, T1755 and T1756 to their brain counterparts, occlude such a pathway in the heart channel (Qu et al., 1995). These mutants that prevent extracellular access of local anesthetics also slow recovery from use-dependent block by QX314 (Qu et al., 1995), presumably because the drug molecule leaves its binding site more slowly via the extracellular opening of the pore of the mutant channels. To examine whether this extracellular access pathway is important for recovery from use-dependent block by R-mexiletine, we compared the recovery of rhWT channels with mutant rhLTT-FVS and rhT1775V. The rhLTT-FVS and rhT1755V mutants behaved comparably to the rhWT counterpart in terms of their affinities for R-mexiletine block of resting and depolarized channels (Table 1). The primary effect of these mutations was on recovery from use-dependent block by R-mexiletine (Fig. 14). For both rhLTT-FVS and rhT1755V, recovery from use-dependent block was slightly, but significantly slower than for rhWT channels. The slow time constant of recovery was 779 ms for rhLTT-FVS, 784 ms for rhT1755V (Fig. 14, dotted line), and 529 ms for rhWT. These results show that the difference in R-mexiletine block of brain and heart channels is not caused by the differences in these amino acid residues which control extracellular access. Differences in release from the local anesthetic receptor site may be primarily responsible for the different rates of recovery.

Discussion

R-Mexiletine block of native and heterologously expressed heart sodium channels is similar. R-Mexiletine has been described as a fast-onset, use-dependent blocker of heart sodium channels. Using rat ventricular myocytes, Yatani and Akaike (1985) found an IC50 of 28 µM for tonic inhibition. Those data were obtained using a holding potential of -80 mV, which caused approximately 50% channel inactivation under their experimental conditions. The time-constant for recovery from block at a holding potential of -90 mV was 370 ms. In the same preparation, Ono et al. (1994) reported half-maximum inhibition of depolarized channels to be 15 µM. Heterologous expression of WT human heart sodium channels resulted in steady-state block of depolarized channels with an IC50 of 15 µM (Wang et al., 1997). These data are consistent with the IC50 of 9.4 µM that we obtained for R-mexiletine block of expressed rat heart sodium channels under depolarized conditions in the present study. Thus, the data on rat heart sodium channel a subunits presented here are in good agreement with those obtained previously.

Higher intrinsic affinity for block of heart versus brain sodium channels. We examined the affinity of the resting and of the inactivated states of heart and brain channels for R-mexiletine. The best estimate of affinity for resting channels is to determine block of peak current using increasingly negative holding potentials. Block by R-mexiletine reaches a limiting affinity which is not reduced as the holding potential is made more negative (Figs. 6 and 12; Table 1). This limiting IC50 is 533 µM in the brain channel and 321 µM in the heart channel. Thus, when elicited from the fully resting state, the heart sodium channel is approximately 1.6 times more sensitive to block than the brain channel. Likewise, our best estimate for block of depolarized channels obtained using a depolarizing prepulse (Figs. 4 and 11; Table 1) indicates an approximately 2-fold higher sensitivity of depolarized heart channels to block by R-mexiletine than of brain channels. These findings contrast with a recent report comparing lidocaine sensitivity for resting brain and heart sodium channels indicating that there is little intrinsic difference in sensitivity of the isoforms to block of resting channels by lidocaine (Wright et al., 1997). Evidently, R-mexiletine does have approximately two-folder higher intrinsic affinity for heart sodium channels. This preferential block of heart sodium channels likely contributes to the therapeutic effect of mexiletine on cardiac arrhythmias.

Interactions with rbF1764/rHF1762 are responsible for the difference in affinity for brain and heart sodium channels. The greater affinity of heart sodium channels for block by R-mexiletine was largely abolished by the mutations F1764A and F1762A. The affinity of the inactivated states of these two mutants were not significantly different (K_I=149 µM or 156 µM, respectively, Table 1), and IC50 values at different potentials were also similar for these two mutants. These results suggest that interactions with this key phenylalanine are primarily responsible for the higher affinity of heart sodium channels for R-mexiletine.

Studies of effects of other local anesthetics on brain or skeletal muscle isoforms of sodium channels have demonstrated that an additional stable blocked state is present in the heart sodium channel that is absent in brain and skeletal muscle isoforms (Zamponi et al., 1993; Gingrich et al., 1993). If R-mexiletine block of the heart channel exhibits a similar stable blocked state that is absent in the brain isoform, it could be this state that is disrupted in the heart mutant F1762A.

Common mechanism of R-mexiletine block of sodium channels. Despite the different affinities for brain and heart sodium channel a subunits, the effects on WT channels of both subtypes are qualitatively similar. In both channel types R-mexiletine caused use-dependent block, shifted steady-state inactivation curves toward more negative potentials, and had higher affinities for depolarized, inactivated channels in comparison to hyperpolarized resting ones. Furthermore, onset of R-mexiletine block during depolarization occurred in rapid and slower exponential phases representing binding to open and inactivated states of both channel types. These observations suggest that the basic mechanism of R-mexiletine block of brain and heart sodium channels is similar.

The literature contains considerable discussion of whether mexiletine binds to the open or inactivated state of depolarized channels (e.g. Ono et al., 1995; Wang et al., 1997). Our data demonstrate a clear biexponential onset of R-mexiletine block of depolarized channels. The most straightforward explananation of such a time course is that the fast and slow components of block represent R-mexiletine binding to open and inactivated channels, respectively. The slower exponential phase of block of depolarized sodium channels occurs largely at times when the vast majority of sodium channels are inactivated. Clearly, R-mexiletine can bind to inactivated channels. The time constant of the fast component was not well resolved in our experiments but its time course overlaps the period when channels are open and and its rate increases with increasing concentration. Furthermore, the fraction of block attributable to the fast time constant increases with concentration (Figs. 3 and 10; Table 1). This is expected if R-mexiletine can bind to either open or inactivated states of depolarized channels but it can achieve that block more rapidly if the channels are open.

Homologous mutations F1764A and F1762A affect R-mexiletine block in different ways. In spite of the similarities in steady-state block of the mutant brain and heart channels, the mutations at F1764/1762 disrupted R-mexiletine block in strikingly different ways. Mutation F1764A of the brain channel had a pronounced effect on the rate of association of the drug with the depolarized channel (Figs. 3 and 10C). After drug block had developed, there was little effect of the mutation on drug unbinding from depolarized or hyperpolarized channels (Fig. 7). Thus, the major effects of replacing this phe with ala was to greatly slow R-mexiletine binding to the mutant brain channel; once bound, the stability of the complex was comparable to WT. Change of the analogous amino acid in the heart sodium channel had quite a different effect. For this mutant, the kinetics of drug association with the heart channel were comparable to association with the WT channel (Fig. 10D). However, drug dissociation from the depolarized (Fig. 10B) and hyperpolarized mutant channel (Fig. 13) was much faster than from the WT channel. Thus, in the brain channel the presence of the WT phe at position 1764 facilitates development of the blocked state but has small effects on the dissociation of the complex. In the heart sodium channel, the homologous phe at position 1762 stabilizes the drug in its binding site and prevents unbinding. This is reflected in a 3.5-fold increase in IC50 for binding to the resting F1762A mutant channel at –160 mV, a decreased level of steady-state block at -20 mV (Fig. 10), a 15-fold increase in IC50 at -60 mV (Fig. 11A) in the mutant channel, and a 3-fold decrease in the time constant for drug unbinding from channels after repolarization to –120 mV (Fig. 13). The greater effects on depolarized as opposed to hyperpolarized channels indicate that the F1762A mutation preferentially disrupts R-mexiletine binding to depolarized states of the channel.

Common and divergent effects of IVS6 mutations on block of brain sodium channels by R-mexiletine and other local anesthetic/antiarrhythmic compounds. Each point mutation in transmembrane segment IVS6 that was known to reduce block by other local anesthetic, anticonvulsant and antiarrhythmic drugs, also reduced use-dependent R-mexiletine block of rbIIA sodium channels (Ragsdale et al., 1994; Qu et al., 1995). Use- and voltage-dependent block of brain sodium channels by etidocaine is reduced in mutants F1764A and Y1771A, and F1764 is more important for binding than Y1771 (Ragsdale et al., 1994). Although these two mutations also reduced the affinity for R-mexiletine, Y1771A reduced R-mexiletine block much more profoundly than F1764A. Similarly, (Ragsdale et al., 1996) found that effects of a range of sodium channel inhibitors on mutants F1764A and Y1771A were all reduced compared to WT channels. However, rank orders of potencies for use-dependent block were different for the tested compounds. Block by lidocaine was reduced more by mutant F1764A than by mutant Y1771A, whereas, for flecainide, the efficacy of the mutations was reversed. Thus, sodium channel modulators of different chemical structure may interact in different ways with the local anesthetic receptor site in segment IVS6 of the sodium channel.

Overall this study demonstrates that sodium channel blockers like mexiletine can have different affinities for binding to the local anesthetic receptor site brain and heart sodium channel subtypes. Mutations of equivalent amino acids can have strongly different effects on the kinetics of channel inhibition in different subtypes. These findings suggest different roles for equivalent amino acids in the different subtypes and provide a rationale for the development of tissue-specific sodium channel blockers with defined kinetic properties.

Acknowledgements

The major portion of this study was performed during a sabbatical leave of T. Weiser from Boehringer Ingelheim. T.W. would like to express his gratitude to Dr. N. Mayer, Prof. R. Hammer and Prof. B. Wetzel (Boehringer Ingelheim) for offering him this opportunity and for critical discussions of the work. The authors also thank Dr. M. Grauert for his assessment of the purity of the enantiomer(s) of mexiletine and Ms. Elizabeth M. Sharp for assistance with molecular biology.

References

These experiments were supported by NIH Program Project Grant P01 HL44948 (W.A.C., Principal Investigator).
 
 

TABLE 1

Block of WT and mutant brain and heart channels by R-mexiletine.
 

Channel Type	rb WT	rb F1764A	rb Y1771A	rh WT	rh F1762A
IC50 Tonic block (µM)a	305	614	718	165	748
Phasic block (µM)a	148	410	617	52	618
Tau recovery (ms)b	571	858	n.d.	529	168
h¥ shift (mV)c
30 µM	n.d.	n.d.	n.d.	-3.2	n.d.
100 µM	-7.5	-2.0	n.d.	-9.0	n.d.
300 µM	-13.1	-5.8	-2.2	-17.3	-6.2
1000 µM	-20.9	-11.6	-8.3	n.d.	-12.4
IC50 Tonic block (µM)d
-80 mV	155	531	n.d.	n.d.	n.d.
-100 mV	394	714	532	107	806
-120 mV	498	803	629	245	1261
-140 mV	533	813	655	321	1468
Tau onset of block (ms)e
10 µM	326	n.d.	n.d.	878	n.d.
30 µM	152	2475	n.d.	266	128
100 µM	52	2598	n.d.	85	187
300 µM	24	1139	n.d.	44	56
1000 µM	n.d.	549	n.d.	n.d.	65
IC50 depolarized block (µM)f	20.7	156	309	9.4	149

1. Tonic and phasic block measured as described in the legend to Fig. 1 for brain channels and Fig. 8 for heart channels.
2. Time constant for recovery from block induced by a train of 10 depolarizations to 0 mV as described in the legend to Fig. 7 for brain channels and Fig. 13 for heart channels.
3. Shift in the half inactivation voltage in the presence of the indicated R-mexiletine concentration determined as described in the legends to Figs. 5 and 11.
4. Tonic block determined at different holding potentials as described in the legends to Fig. 6 and 12.
5. Time constant for onset of block at 0 mV determined as described in the legends to Figs. 3 and 10.
6. IC50 for steady-state block at a holding potential of –40 mV (brain) or –60 mV (heart).

Fig. 1. Mutations in the IVS6 region of brain type sodium channel a subunits differently influence tonic and phasic block by mexiletine. A single pulse to 0 mV was applied in the absence of drug. Then R-mexiletine was washed into the bath for 30 s to 3 min without pulsing. Control experiments showed that equilibrium R-mexiletene block was reached within seconds. Finally, a 5 Hz train of 100 depolarizations (10 ms duration) to 0 mV was applied to voltage-clamped tsA-201 cells that had been transfected with rat brain sodium channel a subunits. A,C,E, Current traces evoked by the single pulse in drug-free solution (control) and by pulse 1 and pulse 100 of the train in the presence of 100 µM (A) or 300 µM R-mexiletine (C and E). The inhibition in response to pulse 1 is referred to as tonic block, and to pulse 100 is referred to as phasic block. B,D,F, Concentration-response curves for tonic (closed circles) and phasic (open circles) block of rbWT (B), rb F1764A (D) and rbY1771A (F). The solid lines are fits to the data for each construct with IC50 values of 305 and 148 µM for tonic and phasic block of rbWT channels, respectively, (B), 614 and 410 µM for rbF1764A (D), and 718 and 617 µM for Y1771A (F). Dotted lines in D and F represent the fitted curves for rbWT channels taken from B.

Fig. 2. Frequency-dependent inhibition of rhWT and mutant sodium channels. A, Trains of 15 pulses to 0 mV (10 ms duration) were applied in the presence of 300 µM mexiletine. The fraction of current blocked by trains of different frequencies was determined as [1-(peak current evoked by pulse 15/ peak current evoked by pulse 1)]. B, Trains of 25 Hz depolarizations were applied to cells expressing rbWT(circles), rbF1764A(squares), and rbY1771A(triangles) channels. Peak current normalized to peak current in pulse 1 is plotted versus pulse number.

Fig. 3. Onset of mexiletine block during depolarizations. Onset of R-mexiletine block of sodium currents in cells expressing rbIIA WT (A), F1764A (B), and Y1771A (C) sodium channels. Currents were elicited by test pulses to 0 mV that had been preceded by conditioning pulses to 0 mV of variable duration (0-1000 ms) to develop drug block followed by a 45 ms return to the holding potential of –100 mV (-120 mV for Y1771A) to allow drug-free channels to recover from inactivation. Peak amplitudes in the presence of R-mexiletine were normalized to those determined with the same protocol in control. Double-exponential functions were fit to the data points. The time constants of the fast components were assumed to reflect the rate of inhibition via the open channels, whereas the slow components describes the rate of block of inactivated channels. The block of rbY1771A channels was too small to be fit accurately, and the data points were connected by straight lines. Concentrations of R-mexiletine were 10 µM (circles), 30 µM (triangles), 100 µM (inverted triangles), 300 µM (squares), and 1000 µM (diamonds), respectively. Note that the range of concentrations was different for rbWT and mutant channels.

Fig 4. Mutants rbF1764A and rbY1771A reduce block of depolarized sodium channels. The inhibition of inactivated sodium channels was investigated using a 1 s depolarizing prepulse to -40 mV (-60 mV for rbY1771A) to inactivate most channels and allow R-mexiletine to bind to the inactivated channels. The membrane potential was then returned to the holding potential of -100 mV (-120 mV for Y1771A) for 10 ms to allow recovery of drug-free channels to recover from inactivation. A test pulse to 0 mV was then applied. R-mexiletine block of peak test pulse current is plotted as a function of R-mexiletine concentration for WT (circles), rbF1764A (squares), and rbY1771A (triangles) channels. The solid lines are fits of a logistic equation to the data with IC50 values of 20.7 µM (WT), 156 mM (rbF1764A) and 309 mM (rbY1771A).

Fig. 5. Hyperpolarizing shifts of steady-state availability curves in response to R-mexiletine. Steady-state availability was investigated using 1 s prepulses to a variable potential followed by a test pulse to 0 mV. A, Response of a typical cell in control (circles) and in the presence of 100 µM (triangles), 300 µM (inverted triangles) and 1000 µM (diamonds) R-mexiletine normalized to the largest currents in control. The solid lines are fits of a Boltzmann relationship to the data points. B, Each curve from (A) has been normalized to its largest value to emphasize the concentration-dependent shift in voltage dependence. C, The shift of inactivation curves is reduced in mutants rbF1764A and rbY1771A, compared to rbWT channels. The bar graph shows the shift of the midpoints of inactivation curves in response to 300 µM R-mexiletine for the three rIIA constructs.

Fig. 6. Limiting potency of R-mexiletine at negative holding potentials. Concentration-response curves for block by R-mexiletine were determined using different holding potentials for block of rbWT (A), rbF1764A (B) and rbY1771A (C). For A and B the holding potentials tested were -80 mV (squares), -100 mV (triangles), -120 mV (inverted triangles), and –140 mV (diamonds). The potentials in (C) were -100 mV (squares), -120 mV (triangles), -140 mV (inverted triangls), and –160 mV (diamonds). For all constructs, block saturated at more hyperpolarized potentials.

Fig. 7. Recovery from use-dependent block for rbWT and rbF1764A mutant channels. Use-dependent inhibition by 300 µM mexiletine was generated by applying a 10 Hz conditioning train of 10 pulses to 0 mV. The membrane potential was then returned to –100 mV and a test pulse to 0 mV was applied after a recovery interval of variable duration. The rates of recovery at –100 mV for WT (A) and mutant rbF1764A (B) were generated by plotting the peak test pulse current normalized to its value after 5000 ms recovery as a function of recovery interval. The data were fitted by double-exponential functions, where the faster time constant reflected the recovery from inactivation of unblocked channels, and the slower, the recovery from block by mexiletine. The time constants for recovery from use-dependent block by R-mexiletine were 571 and 858 ms for rbWT and rbF1764A channels respectively.

Fig. 8. Tonic and phasic inhibition of WT and mutant rat heart sodium channels. A and C, Current recorded during test pulses to –20 mV in control, and during the first and one hundredth pulse of 5 Hz trains in a cell expressing rhWT channels (A) and rhF1762A (C) channels. The protocol was the same as that of Fig. 1 except that holding and test potentials were –120 mV and –20 mV, respectively, to account for the more negative voltage dependence of the heart channel. B and D, Concentration-response curves for tonic (closed circles) and phasic (open circles) block of rhWT (B) and rhF1762A (D) by R-mexiletine determined from pulse 1 and pulse 100 of trains as described in the legend to Fig. 1. The solid lines are fits of a logistic equation with IC50 values for rhWT (B) of 165 µM and 52 µM for tonic and phasic block, respectively, and for rhF1762A (D) of 748 µM and 618 µM. The dotted lines are the fit curves to the analogous rbWT results from Fig. 1B and from rbF1764A from Fig. 1D.

Fig. 9. Frequency-dependent inhibition of rhWT and rhF1762A mutant sodium channels. A, Cells expressing rhWT sodium channels were stimulated with trains of 15 depolarizations (10 ms duration) to –20 mV from a holding potential of –120 mV at the indicated frequencies in the presence of 100 µM mexiletine. The current evoked by pulse 15 is plotted as a fraction of the peak current evoked by pulse 1 (the tonically blocked channel) for each stimulus frequency. B, The pulsewise development of frequency-dependent block using 5 Hz stimulation was for WT (circles) and rhF1762A (squares) channels. The solid lines are exponential fits to the data with rate constants of 0.39/pulse for WT and 0.42/pulse for rhF1762A. For better comparison of pulse-dependent development of block the rhF1762A curve was scaled vertically to match the magnitude of block in rhWT (dotted line).

Fig. 10. Depolarization-dependent block of heart sodium chanels by R-mexiletine. A and B, The time course of onset of R-mexiletine block during depolarization was studied using a prepulse to –20 mV of variable duration followed by a test pulse to the same potential. A 45 ms interval was present between prepulse and test pulse to allow of non-drug-bound channels to recover from fast inactivation. Peak test pulse current in the presence of 10 mM (circles), 30 mM (triangles), 100 mM (inverted triangles), 300 mM (squares) and 1000 mM (diamonds) R-mexiletine was normalized to control values obtained using the same protocol and plotted versus prepulse duration for cells expressing rhWT (A) or rhF1762A (B) sodium channels. Note the different range of concentrations for rhWT and rhF1762A. C and D, Comparison of the onset of block by 300 µM R-mexiletine for rbWT and rbF1764A (C; see Fig. 3A,B), and rhWT and rhF1762A channels (D), respectively. Shown are fits of two exponentials to the data points. The first and last points of the curves were set to 1 and 0, respectively, to compare the kinetics of development of block. The mutant F1764A had a much stronger effect on the onset of block (C), compared to the the equivalent heart construct, rhF1762A (D).

Fig. 11. Mutant rhF1762A reduces block of depolarized sodium channels. A, The inhibition of inactivated sodium channels was investigated using a 1 s depolarizing prepulse to –60 mV. The membrane potential was then returned to the holding potential of –120 mV for 10 ms, and a test pulse to –20 mV was then applied. R-Mexiletine block of peak test pulse current is plotted as a function of R-mexiletine concentration for rhWT (circles) and rhF1762A (squares). The solid lines are fits of a logistic equation to the data with IC50 values of 9.4 µM (rhWT) and 149 µM (rhF1762A), respectively. The dotted lines represent logistic fits for the corresponding brain channels (see Fig. 4), with IC50 values of 20.7 µM (rbWT), and 156 µM (rbF1764A). B, Shift in half inactivation voltage, V_0.5, as a function of R-mexiletine concentration in rhWT (circles) and rhF1762A (squares). C, Shift in half inactivation voltage in response to 300 µM R-mexiletine.

Fig. 12. Holding potential-dependence of concentration-response curves at strongly hyperpolarized potentials to determine the limiting affinity of resting channels at negative potentials. The indicated concentrations of R-mexiletine were applied to rhWT (A) or rhF1762A (B) channels at holding potentials of -100 mV (squares), -120 mV (triangles), -140 mV (inverted triangles) and -160 mV (diamonds). The concentration-response curve for each channel type reaches a limiting affinity at the most negative potentials.

Fig. 13. A,B, Recovery from use-dependent block of rhWT and rhF1762A sodium channels. Use-dependent inhibition by 100 or 300 µM mexiletine was generated by applying a 10 Hz conditioning train of 10 pulses to 0 mV. The membrane potential was then returned to -120 mV and a test pulse to -20 mV was applied after a recovery interval of variable duration. The rates of recovery at -120 mV for WT (A) and mutant rbF1762A (B) were generated by plotting the peak test pulse current normalized to its value after 5000 ms recovery as a function of recovery interval. The data were fitted by double-exponential functions, where the faster time constant reflected the recovery from inactivation of unblocked channels, and the slower, the recovery from block by mexiletine. This slower time constant was 529 ms for rhWT, but only 168 ms for rhF1762A.

Fig. 14. Recovery from inactivation is slowed in mutant rhT1755V. The pulse protocol was identical to that of Fig. 13. Fits of a double exponential function to the recovery data gave a time constant for the slow component of 784 ms. The fit to the recovery data from rhWT channels is shown for comparison (dotted line).

index terms

sodium channels, local anesthetic, antiarrhythmic, heart, brain, channel blocker