Inhibitory Effect of Eslicarbazepine Acetate and S-Licarbazepine on Nav1.5 Channels

Eslicarbazepine acetate (ESL) is a dibenzazepine anticonvulsant approved as adjunctive treatment for partial-onset epileptic seizures. Following first pass hydrolysis of ESL, S-licarbazepine (S-Lic) represents around 95% of circulating active metabolites. S-Lic is the main enantiomer responsible for anticonvulsant activity and this is proposed to be through the blockade of voltage-gated Na+ channels (VGSCs). ESL and S-Lic both have a voltage-dependent inhibitory effect on the Na+ current in N1E-115 neuroblastoma cells expressing neuronal VGSC subtypes including Nav1.1, Nav1.2, Nav1.3, Nav1.6, and Nav1.7. ESL has not been associated with cardiotoxicity in healthy volunteers, although a prolongation of the electrocardiographic PR interval has been observed, suggesting that ESL may also inhibit cardiac Nav1.5 isoform. However, this has not previously been studied. Here, we investigated the electrophysiological effects of ESL and S-Lic on Nav1.5 using whole-cell patch clamp recording. We interrogated two model systems: (1) MDA-MB-231 metastatic breast carcinoma cells, which endogenously express the “neonatal” Nav1.5 splice variant, and (2) HEK-293 cells stably over-expressing the “adult” Nav1.5 splice variant. We show that both ESL and S-Lic inhibit transient and persistent Na+ current, hyperpolarise the voltage-dependence of fast inactivation, and slow the recovery from channel inactivation. These findings highlight, for the first time, the potent inhibitory effects of ESL and S-Lic on the Nav1.5 isoform, suggesting a possible explanation for the prolonged PR interval observed in patients on ESL treatment. Given that numerous cancer cells have also been shown to express Nav1.5, and that VGSCs potentiate invasion and metastasis, this study also paves the way for future investigations into ESL and S-Lic as potential invasion inhibitors.


INTRODUCTION
Eslicarbazepine acetate (ESL) is a member of the dibenzazepine anticonvulsant family of compounds which also includes oxcarbazepine and carbamazepine (Almeida and Soares-da-Silva, 2007). ESL has been approved by the European Medicines Agency and the United States Federal Drug Administration as an adjunctive treatment for partial-onset epileptic seizures (Sperling et al., 2015). ESL is administered orally and rapidly undergoes first pass hydrolysis to two stereoisomeric metabolites, R-licarbazepine and S-licarbazepine (S-Lic; also known as eslicarbazepine; Figures 1A, B) (Almeida et al., 2005;Almeida et al., 2008;Perucca et al., 2011). S-Lic represents around 95% of circulating active metabolites following first pass hydrolysis of ESL and is the enantiomer responsible for anticonvulsant activity (Potschka et al., 2014;Sierra-Paredes et al., 2014). S-Lic also has improved blood brain barrier penetration compared to R-licarbazepine (Alves et al., 2008). Although S-Lic has been shown to inhibit T type Ca 2+ channels (Brady et al., 2011), its main activity is likely through blockade of voltage-gated Na + channels (VGSCs) (Hebeisen et al., 2015). ESL offers several clinical advantages over other older VGSC-inhibiting antiepileptic drugs, e.g. carbamazepine, phenytoin; it has a favourable safety profile (Brown and El-Mallakh, 2010;Hebeisen et al., 2015), reduced induction of hepatic cytochrome P450 enzymes (Galiana et al., 2017), low potential for drug-drug interactions (Falcao et al., 2012;Zaccara et al., 2015), and takes less time to reach a steady state plasma concentration (Bialer and Soares-da-Silva, 2012).
VGSCs are composed of a pore-forming a subunit in association with one or more auxiliary b subunits, the latter modulating channel gating and kinetics in addition to functioning as cell adhesion molecules (Catterall, 2014). There are nine a subunits (Na v 1.1-Na v 1.9), and four b subunits (b1-4) (Goldin et al., 2000;Brackenbury and Isom, 2011). In postnatal and adult CNS neurons, the predominant a subunits are the tetrodotoxinsensitive Na v 1.1, Na v 1.2, and Na v 1.6 isoforms (Van Wart and Matthews, 2006) and it is therefore on these that the VGSCinhibiting activity of ESL and S-Lic has been described. In the murine neuroblastoma N1E-115 cell line, which expresses Na v 1.1, Na v 1.2, Na v 1.3, Na v 1.6, and Na v 1.7, ESL and S-Lic both have a voltage-dependent inhibitory effect on the Na + current (Bonifacio et al., 2001;Hebeisen et al., 2015). In this cell model, S-Lic has no effect on the voltage-dependence of fast inactivation, but significantly hyperpolarises the voltage-dependence of slow inactivation (Hebeisen et al., 2015). S-Lic also has a lower affinity for VGSCs in the resting state than carbamazepine or oxcarbazepine, thus potentially improving its therapeutic window over firstand second-generation dibenzazepine compounds (Hebeisen et al., 2015). In acutely isolated murine hippocampal CA1 neurons, which express Na v 1.1, Na v 1.2 and Na v 1.6 (Westenbroek et al., 1989;Yu et al., 2006;Royeck et al., 2008), S-Lic significantly reduces the persistent Na + current, a very slowinactivating component~1% the size of the peak transient Na + current (Saint, 2008;Doeser et al., 2014). Moreover, in contrast to carbamazepine, this effect is maintained in the absence of b1 (Uebachs et al., 2010;Doeser et al., 2014).
In healthy volunteers, ESL has not been associated with cardiotoxicity and the QT interval remains unchanged on treatment (Vaz-Da-Silva et al., 2012). However, a prolongation of the PR interval has been observed (Vaz-Da-Silva et al., 2012), suggesting that caution should be exercised in patients with cardiac conduction abnormalities (Zaccara et al., 2015). Prolongation of the PR interval suggests that ESL may also inhibit the cardiac Na v 1.5 isoform, although this has not previously been studied. Na v 1.5 is not only responsible for the initial depolarisation of the cardiac action potential (George, 2005), but is also expressed in breast and colon carcinoma cells, where the persistent Na + current promotes invasion and Abbreviations: ESL, eslicarbazepine acetate; HEK-Na v 1.5, HEK-293 cells stably expressing Na v 1.5; I-V, current-voltage; k, slope factor; PSS, physiological saline solution; S-Lic, S-licarbazepin; T p , time to peak current; t f , fast time constant of inactivation; t s , slow time constant of inactivation; t r , time constant of recovery from inactivation; VGSC, voltage-gated Na + channel; V m , membrane potential; V h , holding potential; V peak , voltage at which current was maximal; V rev , reversal potential; V thres , threshold voltage for activation; V 1/2 , half-activation voltage.   (Roger et al., 2003;Fraser et al., 2005;House et al., 2010;Nelson et al., 2015a). Inhibition of Na v 1.5 with phenytoin or ranolazine decreases tumor growth, invasion and metastasis (Yang et al., 2012;Driffort et al., 2014;Nelson et al., 2015b). Thus, it is of interest to specifically understand the effect of ESL on the Na v 1.5 isoform.
In the present study we investigated the electrophysiological effects of ESL and S-Lic on Na v 1.5 [1] endogenously expressed in the MDA-MB-231 metastatic breast carcinoma cell line, and [2] stably over-expressed in HEK-293 cells. We show that both ESL and S-Lic inhibit transient and persistent Na + current, hyperpolarise the voltage-dependence of fast inactivation, and slow the recovery from channel inactivation. These findings highlight, for the first time, the potent inhibitory effects of ESL and S-Lic on the Na v 1.5 isoform.

MATERIALS AND METHODS
Pharmacology ESL (Tokyo Chemical Industry UK Ltd) was dissolved in DMSO to make a stock concentration of 67 mM. S-Lic (Tocris) was dissolved in DMSO to make a stock concentration of 300 mM. Both drugs were diluted to working concentrations of 100-300 µM in extracellular recording solution. The concentration of DMSO in the recording solution was 0.45% for ESL and 0.1% for S-Lic. Equal concentrations of DMSO were used in the control solutions. DMSO (0.45%) had no effect on the Na + current (Supplementary Figure 1).

Cell Culture
MDA-MB-231 cells and HEK-293 cells stably expressing Na v 1.5 (a gift from L. Isom, University of Michigan) were grown in Dulbecco's modified eagle medium supplemented with 5% FBS and 4 mM L-glutamine (Simon et al., 2020). Molecular identity of the MDA-MB-231 cells was confirmed by short tandem repeat analysis (Masters et al., 2001). Cells were confirmed as mycoplasmafree using the DAPI method (Uphoff et al., 1992). Cells were seeded onto glass coverslips 48 h before electrophysiological recording.

Electrophysiology
Plasma membrane Na + currents were recorded using the wholecell patch clamp technique, using methods described previously (Yang et al., 2012;Nelson et al., 2015a). Patch pipettes made of borosilicate glass were pulled using a P-97 pipette puller (Sutter Instrument) and fire-polished to a resistance of 3-5 MΩ when filled with intracellular recording solution. The extracellular recording solution for MDA-MB-231 cells contained (in mM): 144 NaCl, 5.4 KCl, 1 MgCl 2 , 2.5 CaCl 2 , 5.6 D-glucose, and 5 HEPES (adjusted to pH 7.2 with NaOH). For the extracellular recording solution for HEK-293 cells expressing Na v 1.5, the extracellular [Na + ] was reduced to account for the much larger Na + currents and contained (in mM): 60 NaCl, 84 Choline Cl, 5.4 KCl, 1 MgCl 2 , 2.5 CaCl 2 , 5.6 D-glucose, and 5 HEPES (adjusted to pH 7.2 with NaOH). The intracellular recording solution contained (in mM): 5 NaCl, 145 CsCl, 2 MgCl 2 , 1 CaCl 2 , 10 HEPES, 11 EGTA, (adjusted to pH 7.4 with CsOH) (Brackenbury and Djamgoz, 2006). Voltage clamp recordings were made at room temperature using a Multiclamp 700B or Axopatch 200B amplifier (Molecular Devices) compensating for series resistance by 40-60%. Currents were digitized using a Digidata interface (Molecular Devices), low pass filtered at 10 kHz, sampled at 50 kHz and analysed using pCLAMP 10.7 software (Molecular Devices). Leak current was subtracted using a P/6 protocol (Armstrong and Bezanilla, 1977). Extracellular recording solution ± drugs was applied to the recording bath at a rate of 1.5 ml/min using a ValveLink 4-channel gravity perfusion controller (AutoMate Scientific). Each new solution was allowed to equilibrate in the bath for~4 min following switching prior to recording at steady state.

Voltage Clamp Protocols
Cells were clamped at a holding potential of -120 mV or -80 mV for ≥250 ms, dependent on experiment (detailed in the Figure  legends). Five main voltage clamp protocols were used, as follows: 1. To assess the effect of drug perfusion and wash-out on peak current in real time, a simple one-step protocol was used where cells were held at -120 mV or -80 mV for 250 ms and then depolarised to -10 mV for 50 ms. 2. To assess the voltage-dependence of activation, cells were held at -120 mV for 250 ms and then depolarised to test potentials in 10 mV steps between -120 mV and +30 mV for 50 ms. The voltage of activation was taken as the most negative voltage which induced a visible transient inward current. 3. To assess the voltage-dependence of steady-state inactivation, cells were held at -120 mV for 250 ms followed by prepulses for 250 ms in 10 mV steps between -120 mV and +30 mV and a test pulse to -10 mV for 50 ms. 4. To assess recovery from fast inactivation, cells were held at -120 mV for 250 ms, and then depolarised twice to 0 mV for 25 ms, returning to -120 mV for the following intervals between depolarisations (in ms): 1, 2, 3, 5, 7, 10, 15, 20, 30, 40, 50, 70, 100, 150, 200, 250, 350, 500. In each case, the second current was normalized to the initial current and plotted against the interval time.

Curve Fitting and Data Analysis
To study the voltage-dependence of activation, current-voltage (I-V) relationships were converted to conductance using the following equation: , where G is conductance, I is current, V m is the membrane voltage and V rev is the reversal potential for Na + derived from the Nernst equation. Given the different recording solutions used, V rev for Na + was +85 mV for MDA-MB-231 cells and +63 mV for HEK-Na v 1.5 cells. The voltage-dependence of conductance and availability were normalized and fitted to a Boltzmann equation: where G max is the maximum conductance, V 1/2 is the voltage at which the channels are half activated/inactivated, V m is the membrane voltage and k is the slope factor. Recovery from inactivation data (I t /I t=0 ) were normalized, plotted against recovery time (D;t) and fitted to a single exponential function: t = A 1 + A 2 exp (-t/t 0 ), where A 1 and A 2 are the coefficients of decay of the time constant (t), t is time and t 0 is a time constant describing the time dependence of t. The time course of inactivation was fitted to a double exponential function: I = A f exp (-t/t f ) + A s exp (-t/t s ) + C, where A f and A s are maximal amplitudes of the slow and fast components of the current, t f and t s are the fast and slow decay time constants and C is the asymptote.

Statistical Analysis
Data are presented as mean and SEM unless stated otherwise. Statistical analysis was performed on the raw (non-normalized) data using GraphPad Prism 8.4.0. Pairwise statistical significance was determined with Student's paired t-tests. Multiple comparisons were made using ANOVA and Tukey post-hoc tests, unless stated otherwise. Results were considered significant at P < 0.05.

Effect of Eslicarbazepine Acetate and S-Licarbazepine on Transient and Persistent Na + Current
Several studies have clearly established the inhibition of neuronal VGSCs (Na v 1.1, Na v 1.2, Na v 1.3, Na v 1.6, Na v 1.7 and Na v 1.8) by ESL and its active metabolite S-Lic (Bonifacio et al., 2001;Doeser et al., 2014;Hebeisen et al., 2015;Soares-da-Silva et al., 2015). Given that ESL prolongs the PR interval (Vaz-Da-Silva et al., 2012), potentially via inhibiting the cardiac Na v 1.5 isoform, together with the interest in inhibiting Na v 1.5 in carcinoma cells to reduce invasion and metastasis (Driffort et al., 2014;Martin et al., 2015;Nelson et al., 2015b;Elajnaf et al., 2018;Djamgoz et al., 2019), it is also relevant to evaluate the electrophysiological effects of ESL and S-Lic on this isoform. We therefore evaluated the effect of both compounds on Na v 1.5 current properties using whole-cell patch clamp recording, employing a two-pronged approach: (1) recording Na v 1.5 currents endogenously expressed in the MDA-MB-231 breast cancer cell line (Roger et al., 2003;Fraser et al., 2005;Brackenbury et al., 2007), and (2) recording from Na v 1.5 stably over-expressed in HEK-293 cells (HEK-Na v 1.5) (Patino et al., 2011). Initially, we evaluated the effect of both compounds on the size of the peak Na + current in MDA-MB-231 cells. Na + currents were elicited by depolarising the membrane potential (V m ) to -10 mV from a holding potential (V h ) of -120 or -80 mV. Application of the prodrug ESL (300 mM) reversibly inhibited the transient Na + current by 49.6 ± 3.2% when the V h was -120 mV (P < 0.001; n = 13; ANOVA + Tukey test; Figures 2A, D). When V h was set to -80 mV, ESL (300 mM) reversibly inhibited the transient Na + current by 79.5 ± 4.5% (P < 0.001; n = 12; ANOVA + Tukey test; Figures 2C, E). We next assessed the effect of ESL in HEK-Na v 1.5 cells. Application of ESL (300 mM) inhibited Na v 1.5 current by 74.7 ± 4.3% when V h was -120 mV (P < 0.001; n = 12; Figures 2F, I) and by 90.5 ± 2.8% when V h was -80 mV (P < 0.001; n = 14; Figures 2H, J). However, the inhibition was only partially reversible (P < 0.001; n = 14; Figures Table 1). Together, these data suggest that ESL preferentially inhibited Na v 1.5 in the open or inactivated state, since the current inhibition was greater at more depolarised V h .
We next tested the effect of the active metabolite S-Lic. S-Lic (300 mM) inhibited the transient Na + current in MDA-MB-231 cells by 44.4 ± 6.1% when the V h was -120 mV (P < 0.001; n = 9; ANOVA + Tukey test; Figures 3A, D). When V h was set to -80 mV, S-Lic (300 µM) inhibited the transient Na + current by 73.6 ± 4.1% (P < 0.001; n = 10; ANOVA + Tukey test; Figures 3C, E). However, the inhibition caused by S-Lic (300 mM) was only partially reversible (P < 0.05; n = 10; ANOVA + Tukey test; Figures 3A, C-E). In HEK-Na v 1.5 cells, S-Lic (300 mM) inhibited Na v 1.5 current by 46.4 ± 3.9% when V h was -120 mV (P < 0.001; n = 13; ANOVA + Tukey test; Figures 3F, I) and by 74.0 ± 4.2% when V h was -80 mV (P < 0.001; n = 12; ANOVA + Tukey test; Figures 3H, J). Furthermore, the inhibition in HEK-Na v 1.5 cells was not reversible over the duration of the experiment. Application of S-Lic at a lower concentration (100 µM Table 1). Together, these data show that channel inhibition by S-Lic was also more effective at more depolarised V h . However, unlike ESL, channel blockade by S-Lic persisted after washout, suggesting higher target binding affinity for the active metabolite and/or greater trapping of the active metabolite in the cytoplasm.

Effect of Eslicarbazepine Acetate and S-Licarbazepine on Voltage Dependence of Activation and Inactivation
We next investigated the effect of ESL (300 µM) and S-Lic (300 µM) on the I-V relationship in MDA-MB-231 and HEK-Na v 1.5 cells. A V h of -120 mV was used for subsequent analyses to ensure that the elicited currents were sufficiently large for Representative persistent Na + currents in a HEK-Na v 1.5 cell elicited by a depolarisation from -120 to -10 mV. (H) Representative Na + currents in a HEK-Na v 1.5 cell elicited by a depolarisation from -80 to -10 mV. (I) Normalized Na + currents in HEK-Na v 1.5 cells elicited by a depolarisation from -120 to -10 mV. (J) Normalized Na + currents in HEK-Na v 1.5 cells elicited by a depolarisation from -80 to -10 mV. Results are mean + SEM. *P ≤ 0.05; ***P ≤ 0.001; one-way ANOVA with Tukey tests (n = 9-13). NS, not significant.  -D; Tables 1, 2). ESL also had no effect on the voltage at current peak in either cell line ( Figures 4A-D; Tables 1, 2). Although S-Lic had no effect on voltage at current peak in MDA-MB-231 cells, it was significantly hyperpolarised in HEK-Na v 1.5 cells from -18.0 ± 4.2 to -30.0 ± 5.6 mV (P < 0.001; n = 9; paired t test; Figures 4A-D; Tables 1, 2). ESL had no significant effect on the half-activation voltage (V½) or slope factor (k) for activation in MDA-MB-231 cells ( Figure 5A; Table 1). The activation k in HEK-Na v 1.5 cells was also unchanged but the activation V½ was significantly hyperpolarised by ESL from -39.4 ± 1.3 to -44.2 ± 1.8 mV (P < 0.05; n = 10; paired t test; Figure  5B; Table 1). S-Lic also had no significant effect on the activation V½ or k in MDA-MB-231 cells ( Figure 5C; Table 2). However, the V½ of activation in HEK-Na v 1.5 cells was significantly hyperpolarised from -32.8 ± 3.1 to -40.5 ± 3.4 mV (P < 0.01; n = 9; paired t test; Figure 5D; Table 2) and k changed from 5.9 ± 0.9 to 4.5 ± 1.1 mV (P < 0.05; n = 9; paired t test; Figure 5D; Table 2).

Effect of Eslicarbazepine Acetate and S-Licarbazepine on Activation and Inactivation Kinetics
We next studied the effect of both compounds on kinetics of activation and inactivation. In MDA-MB-231 cells, ESL (300 mM) significantly accelerated the time to peak current (T p ), upon depolarisation from -120 to -10 mV, from 2.1 ± 0.2 to 1.9 ± 0.2 ms (P < 0.01; n = 13; paired t test; Table 1). However, in HEK-Na v 1.5 cells, ESL significantly slowed T p from 1.4 ± 0.2 to 1.5 ± 0.2 ms (P < 0.001; n = 14; paired t test; Table 1). S-Lic (300 mM) had no significant effect on T p in MDA-MB-231 cells but significantly slowed T p in HEK-Na v 1.5 cells from 1.8 ± 0.5 to 2.3 ± 0.6 ms (P < 0.01; n = 13; paired t test; Table 2).
To study effects on inactivation kinetics, the current decay following depolarisation from -120 to -10 mV was fitted to a double exponential function to derive fast and slow time constants of inactivation (t f and t s ). Neither ESL nor S-Lic had any significant effect on t f or t s in MDA-MB-231 cells (Tables 1,  2). However, in HEK-Na v 1.5 cells, ESL significantly slowed t f from 0.9 ± 0.1 to 1.2 ± 0.1 ms (P < 0.001; n = 12; paired t test; Table 1) and slowed t s from 6.6 ± 0.8 to 20.8 ± 8.5 ms, although this was not statistically significant. S-Lic significantly slowed t f from 1.0 ± 0.04 to 1.3 ± 0.06 ms (P < 0.001; n = 11; paired t test; Table 2) and t s from 6.3 ± 0.5 to 7.3 ± 0.5 ms (P < 0.05; n = 11; paired t test; Table 2). In summary, both ESL and S-Lic elicited various effects on kinetics in MDA-MB-231 and HEK-Na v 1.5 cells, predominantly slowing activation and inactivation.

Effect of Eslicarbazepine Acetate and S-Licarbazepine on Recovery From Fast Inactivation
To investigate the effect of ESL and S-Lic on channel recovery from fast inactivation, we subjected cells to two depolarisations from V h of -120 to 0 mV, changing the interval between these in which the channels were held at -120 mV to facilitate recovery. Significance was determined by fitting a single exponential curve to the normalized current/time relationship and calculating the time constant (t r ). In MDA-MB-231 cells, ESL (300 mM) significantly slowed t r from 6.0 ± 0.5 to 8.7 ± 0.7 ms (P < 0.05; n = 10; paired t test; Figure 6A, Table 1). Similarly, in HEK-Na v 1.5 cells, ESL significantly slowed t r from 4.5 ± 0.4 to 7.1 ± 0.6 ms (P < 0.001; n = 10; paired t test; Figure 6B, Table 1). S-Lic (D) I-V plots of Na + currents in HEK-Na v 1.5 cells in PSS (black circles) and S-Lic (300 mM; red squares). Currents were elicited using 10 mV depolarising steps from -80 to +30 mV for 30 ms, from a holding potential of -120 mV. Results are mean ± SEM (n = 7-13).

DISCUSSION
In this study, we have shown that ESL and its active metabolite S-Lic inhibit the transient and persistent components of Na + current carried by Na v 1.5. We show broadly similar effects in MDA-MB-231 cells, which express endogenous Na v 1.5 (Roger et al., 2003;Fraser et al., 2005;Brackenbury et al., 2007), and in HEK-293 cells over-expressing Na v 1.5. Notably, both compounds were more effective when V h was set to -80 mV than at -120 mV, suggestive of depolarised state-dependent binding. In addition, the inhibitory effect of ESL was reversible whereas inhibition by S-Lic was less so. As regards voltage-dependence, both ESL and S-Lic shifted activation and steady-state inactivation curves, to varying extents in the two cell lines, in the direction of more negative voltages. ESL and S-Lic had various effects on activation and inactivation kinetics, generally slowing the rate of inactivation. Finally, recovery from fast inactivation of Na v 1.5 was significantly slowed by both ESL and S-Lic. To our knowledge, this is the first time that the effects of ESL and S-Lic have specifically been tested on the Na v 1.5 isoform. A strength of this study is that both the prodrug (ESL) and the active metabolite (S-Lic) were tested using two independent cell lines, one endogenously expressing Na v 1.5, the other stably overexpressing Na v 1.5. MDA-MB-231 cells also express Na v 1.7, although this isoform is estimated to be responsible for onlỹ 9% of the total VGSC current (Fraser et al., 2005;Brackenbury et al., 2007). MDA-MB-231 cells also express endogenous b1, b2, and b4 subunits (Chioni et al., 2009;Nelson et al., 2014;Bon et al., 2016). MDA-MB-231 cells predominantly express the developmentally regulated "neonatal" Na v 1.5 splice variant, which differs from the "adult" variant over-expressed in the HEK-Na v 1.5 cells by seven amino acids located in the extracellular linker between transmembrane segments 3 and 4 of domain 1 (Fraser et al., 2005;Brackenbury et al., 2007;Djamgoz et al., 2019). Notably, however, there were no consistent differences in effect of either ESL or S-Lic between the MDA-MB-231 and HEK-Na v 1.5 cells, suggesting that the neonatal vs. adult splicing event, and/or expression of (D) Activation and steady-state inactivation in HEK-Na v 1.5 cells in PSS (black circles) and S-Lic (300 mM; red squares). For activation, normalized conductance (G/G max ) was calculated from the current data and plotted as a function of voltage. For steady-state inactivation, normalized current (I/I max ), elicited by 50 ms test pulses at -10 mV following 250 ms conditioning voltage pulses between -120 and +30 mV, applied from a holding potential of -120 mV, was plotted as a function of the prepulse voltage. Results are mean ± SEM (n = 7-13). Activation and inactivation curves are fitted with Boltzmann functions.
Leslie et al.
Eslicarbazepine Effects on Nav1.5 endogenous b subunits, does not impact on sensitivity of Na v 1.5 to these compounds. This finding contrasts another report showing different sensitivity of the neonatal and adult Na v 1.5 splice variants to the amide local anaesthetics lidocaine and levobupivacaine (Elajnaf et al., 2018). Our findings suggest that the inhibitory effect of S-Lic on Na v 1.5 is less reversible than that of ESL. This may be explained by the differing chemical structures of the two molecules possibly enabling S-Lic to bind the target with higher affinity than ESL. Most VGSCtargeting anticonvulsants, including phenytoin, lamotrigine and carbamazepine, block the pore by binding via aromatic-aromatic interaction to a tyrosine and phenylalanine located in the S6 helix of domain 4 (Lipkind and Fozzard, 2010). However, S-Lic has been proposed to bind to a different site given that it was found to block the pore predominantly during slow inactivation (Hebeisen et al., 2015). Alternatively, the hydroxyl group present on S-Lic (but not ESL) may become deprotonated, potentially trapping it in the cytoplasm. The findings presented here broadly agree with in vitro concentrations used elsewhere to study effects of ESL and S-Lic on Na + currents. For example, using a V h of -80 mV, 300 µM ESL was shown to inhibit peak Na + current by 50% in N1E-115 neuroblastoma cells expressing Na v 1.1, Na v 1.2, Na v 1.3, Na v 1.6, and Na v 1.7 (Bonifacio et al., 2001). S-Lic (250 µM) also blocks peak Na + current by~50% in the same cell line (Hebeisen et al., 2015). In addition, S-Lic (300 µM) reduces persistent Na + current by~25% in acutely isolated murine hippocampal CA1 neurons expressing Na v 1.1, Na v 1.2, and Na v 1.6 (Westenbroek et al., 1989;Yu et al., 2006;Royeck et al., 2008;Doeser et al., 2014). Similar to the present study, ESL was shown to hyperpolarise the voltagedependence of steady-state inactivation in N1E-115 cells (Bonifacio et al., 2001). On the other hand, similar to our finding in HEK-Na v 1.5 cells, S-Lic has no effect on steady-state inactivation in N1E-115 cells (Hebeisen et al., 2015). Again, in agreement with our own findings for Na v 1.5, S-Lic slows recovery from inactivation in N1E-115 cells (Hebeisen et al., 2015). These observations suggest that the sensitivity of Na v 1.5 to ESL and S-Lic is broadly similar to that reported for neuronal VGSCs. In support of this, Na v 1.5 shares the same conserved residues proposed for Na v 1.2 to interact with ESL ( Figure 7) (Shaikh et al., 2014).
Notably, the concentrations used in this study are at or above those achieved in clinical use (e.g. ESL 1,200 mg once daily gives a peak plasma concentration of~100 µM) (Hebeisen et al., 2015). However, it has been argued that the relatively high concentrations which have been previously tested in vitro are clinically relevant given that S-Lic has a high (50:1) lipid:water partition co-efficient and thus would be expected to reside predominantly in the tissue membrane fraction in vivo (Bialer and Soares-da-Silva, 2012). Our (D) Recovery from inactivation in HEK-Na v 1.5 cells in PSS (black circles) and S-Lic (300 mM; red squares). The fraction recovered (I t /I c ) was determined by a 25 ms pulse to 0 mV (I c ), followed by a recovery pulse to -120 mV for 1-500 ms, and a subsequent 25 ms test pulse to 0 mV (I t ), applied from a holding potential of -120 mV, and plotted as a function of the recovery interval. Data are fitted with single exponential functions which are statistically different between control and drug treatments in all cases. Results are mean ± SEM (n = 7-10). study suggests that a clinically relevant plasma concentration (100 µM) would inhibit peak and persistent Na v 1.5 currents. Future work investigating the dose-dependent effects of ESL and S-Lic would be useful to aid clinical judgements.
The data presented here raise several implications for clinicians. The observed inhibition of Na v 1.5 is worthy of note when considering cardiac function in patients receiving ESL (Zaccara et al., 2015). Although the QT interval remains unchanged for individuals on ESL treatment, prolongation of the PR interval has been observed (Vaz- . Further work is required to establish whether the basis for this PR prolongation is indeed via Na v 1.5 inhibition. In addition, it would be of interest to investigate the efficacy of ESL and S-Lic in the context of heritable arrhythmogenic mutations in SCN5A, as well as the possible involvement of the b subunits (Brackenbury and Isom, 2008;Uebachs et al., 2010;Doeser et al., 2014;Rivaud et al., 2020). The findings presented here are also relevant in the context of Na v 1.5 expression in carcinoma cells (Fraser et al., 2014). Given that cancer cells have a relatively depolarised V m , it is likely that Na v 1.5 is mainly in the inactivated state with the persistent Na + current being functionally predominant (Yang and Brackenbury, 2013;Yang et al., 2020). Increasing evidence suggests that persistent Na + current carried by Na v 1.5 in cancer cells contributes to invasion and several studies have shown that other VGSC inhibitors reduce metastasis in preclinical models (Roger et al., 2003;Fraser et al., 2005;House et al., 2010;Yang et al., 2012;Driffort et al., 2014;Besson et al., 2015;Nelson et al., 2015a;Nelson et al., 2015b). Thus, use-dependent inhibition by ESL would ensure that channels in malignant cells are particularly targeted, raising the possibility that it could be used as an anti-metastatic agent (Martin et al., 2015). This study therefore paves the way for future investigations into ESL and S-Lic as potential invasion inhibitors.

DATA AVAILABILITY STATEMENT
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

AUTHOR CONTRIBUTIONS
TL, SC, and WB contributed to the conception and design of the work. TL, LB, and WB contributed to acquisition, analysis, and interpretation of data for the work. TL, SC, and WB contributed to drafting the work and revising it critically for important intellectual content. All authors contributed to the article and approved the submitted version.

SUPPLEMENTARY MATERIAL
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fphar.2020. 555047/full#supplementary-material FIGURE 7 | Clustal alignment of amino acid sequences of Na v 1.1-Na v 1.9 (SCN1A-SCN11A). ESL was proposed previously (Shaikh et al., 2014) to interact with the highlighted amino acids in Na v 1.2. An alignment of Na v 1. shows that the interacting amino acids highlighted in yellow are conserved between Na v 1.2 and Na v 1.5, along with most other isoforms. Asterisks indicate conserved residues. Colon indicates conservation between groups of strongly similar properties -scoring >0.5 in the Gonnet PAM 250 matrix. Period indicates conservation between groups of weakly similar properties -scoring ≤0.5 in the Gonnet PAM 250 matrix.