Articles, Neurobiology of Disease

Efficacy Loss of the Anticonvulsant Carbamazepine in Mice Lacking Sodium Channel β Subunits via Paradoxical Effects on Persistent Sodium Currents

Mischa Uebachs, Thoralf Opitz, Michel Royeck, Gesa Dickhof, Marie-Therese Horstmann, Lori L. Isom and Heinz Beck
Journal of Neuroscience 23 June 2010, 30 (25) 8489-8501; DOI: https://doi.org/10.1523/JNEUROSCI.1534-10.2010

Abstract

Neuronal excitability is critically determined by the properties of voltage-gated Na+ currents. Fast transient Na+ currents (INaT) mediate the fast upstroke of action potentials, whereas low-voltage-activated persistent Na+ currents (INaP) contribute to subthreshold excitation. Na+ channels are composed of a pore-forming α subunit and β subunits, which modify the biophysical properties of α subunits. We have examined the idea that the presence of β subunits also modifies the pharmacological properties of the Na+ channel complex using mice lacking either the β1 (Scn1b) or β2 (Scn2b) subunit. Classical effects of the anticonvulsant carbamazepine (CBZ), such as the use-dependent reduction of INaT and effects on INaT voltage dependence of inactivation, were unaltered in mice lacking β subunits. Surprisingly, CBZ induced a small but significant shift of the voltage dependence of activation of INaT and INaP to more hyperpolarized potentials. This novel CBZ effect on INaP was strongly enhanced in Scn1b null mice, leading to a pronounced increase of INaP within the subthreshold potential range, in particular at low CBZ concentrations of 10–30 μm. A combination of current-clamp and computational modeling studies revealed that this effect causes a complete loss of CBZ efficacy in reducing repetitive firing. Thus, β subunits modify not only the biophysical but also the pharmacological properties of Na+ channels, in particular with respect to INaP. Consequently, altered expression of β subunits in other neurological disorders may cause altered neuronal sensitivity to drugs targeting Na+ channels.

Introduction

Voltage-gated Na+ channels are present in all excitable cells and play a crucial role in action potential generation and propagation. They are composed of a pore-forming α subunit and one or two β subunits (Catterall, 1999). The β1 and β2 subunits are intriguing because they are capable of serving as cell adhesion molecules, regulate the surface expression and trafficking of channels, and modify the biophysical properties of the pore-forming α subunits (Isom, 2001). β1 subunits exhibit an extracellular Ig-like domain, which enables it to mediate diverse functions, including cell–cell interactions, axonal outgrowth, and neuronal migration (Brackenbury and Isom, 2008).

In addition, β1 subunits have been implicated in trafficking and targeting of channels to specific neuronal compartments (Brackenbury et al., 2010). β subunits also exert powerful modulatory effects on the biophysical properties of Na+ channel α subunits. These effects have been studied extensively by coexpression of pore-forming and accessory subunits in various expression systems. The functional effects of β subunit coexpression have proven quite diverse. Coexpression of β subunits modulates the activation and inactivation time course as well as the time course of recovery from fast inactivation of pore-forming subunits. Additionally, β subunits also affect the voltage dependence of activation and inactivation. These effects appear to be strongly dependent on the cellular model used for the experiments, indicating that the interaction of α and β subunits may be a strongly modulated process (Toib et al., 1998; Qu et al., 2001; Meadows et al., 2002; Xu et al., 2007). While there is debate regarding the precise nature of these effects, it is clear that β subunits powerfully modulate the biophysical properties and localization of Na+ channel α subunits.

In the present study, we have now elected to investigate how the presence of β subunits affects the pharmacology of Na+ channels. The Na+ channel complex is an important drug target, i.e., for drugs used in the treatment of both epilepsy and neuropathic pain. It is largely unclear how the subunit composition, and in particular the presence of β subunits, affects drug sensitivity of Na+ channel complexes. Intriguingly, a Scn1b mutation has been shown to affect the response of Na+ channels to phenytoin (Lucas et al., 2005). This suggests that the presence of β subunits may affect the response of the Na+ channel complex to anticonvulsant drugs. In the present study, we have therefore examined how ablating expression of Scn1b or Scn2b in native hippocampal neurons affects the CBZ response of Na+ channels and how this impacts anticonvulsant effects on neuronal excitability.

Materials and Methods

Scn1b- and Scn2b-null mice.

The generation of Scn1b or Scn2b-null mice has been previously described (Chen et al., 2002, 2004). Animals used in this study were bred from congenic Scn1b+/− or Scn2b+/− mice of generation N10 or higher bred on the C57BL/6 genetic background. Experiments were performed on slices or dissociated neurons from the hippocampus of male Scn1b or Scn2b null mice and their wild-type littermates. Experiments on Scn1b null mice and wild-type littermates were performed at postnatal days 12–20 (P12–P20) because Scn1b null mice die at approximately P20 (Chen et al., 2004). Recordings from Scn2b null animals and littermates were performed at P40–P60. All animal experiments were performed in accordance with the regulations of the local animal care and use committee.

Preparation of brain slices and dissociated hippocampal neurons.

Acutely dissociated CA1 and DG neurons were harvested from wild-type or littermate null mice. Animals were perfused through the heart under deep anesthesia (ketamine 100 mg/kg, xylazine 15 mg/kg) with ice-cold sucrose-based artificial CSF (ACSF) comprising (in mm): 60 NaCl, 110 sucrose, 26 NaHCO3, 2.5 KCl, 1.25 NaH2PO4, 5 MgCl2, 1 CaCl2, and 25 glucose, pH 7.4 adjusted with NaOH, osmolality 305 mOsm. Subsequently the brain was rapidly removed. Transverse hippocampal slices (400 μm) were prepared with a vibrating microslicer (Microm) and then transferred to a storage chamber filled with carbogenated sucrose ACSF (95% O2, 5% CO2). The chamber was gradually warmed to 36°C in a water bath and maintained at this temperature for ∼30 min. Finally, the slices were transferred into a chamber filled with ACSF containing the following (in mm): 125 NaCl, 2.5 KCl, 1 MgCl2, 2 CaCl2, 26 NaHCO3, 1.25 NaH2PO4, and 25 glucose, pH 7.4 adjusted with NaOH, osmolality 305 mOsm, where they were maintained at room temperature for an equilibration period of at least 60 min. Subsequently, they were either used for preparation of dissociated neurons (see below) or transferred to an interface chamber for sharp microelectrode recordings.

For dissociation, slices were retrieved from the storage chamber (up to 5 h after preparation) and transferred one at a time to a tube with 5 ml of saline containing the following (in mm): 145 methanesulfonic acid sodium salt (CH3SO3Na), 3 KCl, 1 MgCl2, 0.5 CaCl2, 10 HEPES, and 15 glucose, pH 7.4 adjusted with NaOH, osmolality 308 mOsm. Pronase (protease type XIV, 2 mg/ml, Sigma) was added to the oxygenated medium (100% O2). After an incubation period of 12 min at 36°C and a subsequent period of 10 min at room temperature, the slice was washed in Pronase-free saline of an identical composition. The CA1 or the DG region was dissected and triturated with fire-polished glass pipettes of decreasing aperture. The Petri dish containing the cell suspension was then mounted on the stage of an inverted microscope (Axiovert 100, Zeiss). Dissociated cells were allowed to settle for 5–10 min before recordings were attempted. Whole-cell recordings were performed only on morphologically identified neurons with a smooth surface and a three-dimensional contour.

Patch-clamp recordings.

Properties of INaT were examined in dissociated DG neurons, which were superfused with an extracellular solution containing the following (in mm): 40 CH3SO3Na, 90 tetraethylammonium (TEA), 1.6 CaCl2, 2 MgCl2, 10 HEPES, 5 4-aminopyridine, 0.2 CdCl2, and 15 glucose, pH 7.4 adjusted with NaOH, osmolality 315 mOsm. Patch pipettes (resistance: 3–4 MΩ) were filled with an intracellular solution containing the following (in mm): 110 CsF, 20 TEA, 2 MgCl2, 10 HEPES, 11 EGTA, 5 ATP, and 0.5 GTP, pH 7.2 adjusted with CsOH, osmolality 300 mOsm. Experiments on INaP were performed on dissociated CA1 neurons. Recordings were performed using an extracellular solution containing the following (in mm): 100 CH3SO3Na, 40 TEA, 2 CaCl2, 3 MgCl2, 10 HEPES, 5 4-aminopyridine, 0.2 CdCl2, and 25 glucose, pH 7.4 adjusted with NaOH, osmolality 315 mOsm. CBZ was applied by hydrostatic pressure via a nearby (50–100 μm) superfusion pipette. Solutions around the neurons were exchanged within <1 min. The intracellular solution for recordings of INaP contained (in mm) 110 CsF, 2 MgCl2, 10 HEPES, 11 EGTA, 2 ATP, and 0.5 GTP, pH 7.2 adjusted with CsOH, osmolality 300 mOsm. A liquid-junction potential of −10 mV was measured between the intracellular and extracellular solutions for both pairs of solutions and corrected so that data points were shifted in a depolarizing direction relative to the voltage axis. We obtained tight-seal whole-cell recordings with an EPC9 amplifier (HEKA; seal resistance: 1 GΩ, series resistance: <9 MΩ). Series resistance was compensated between 50% and 70%. Experiments in which the maximal residual voltage error exceeded 2.5 mV were rejected. Current signals were filtered at 10 kHz (3 dB, four-pole low-pass Bessel filter) and sampled at 20 kHz or more by an interface (ITC16; Instrutech). For analysis of INaT, we subtracted residual capacitance transients and leak conductance with a P/4 protocol and analyzed the Na+ current properties as described previously (Schaub et al., 2007). No P/4 protocol was applied for recordings of INaP. All voltage-clamp experiments were performed at room temperature (22 ± 1°C).

Sharp microelectrode recordings.

For intracellular sharp microelectrode recordings, 400-μm-thick hippocampal slices were placed in an interface chamber (35 ± 1°C) and perfused with oxygenated ACSF containing the following (in mm): 125 NaCl, 3.5 KCl, 1.25 NaH2PO4, 2 MgSO4, 2 CaCl2, 26 NaHCO3, and 15 d-glucose. Intracellular recordings were obtained using sharp glass microelectrodes containing 4 m K+-acetate (90–110 m). An active bridge circuit in the amplifier (Axoclamp 2B, Molecular Devices) allowed simultaneous injection of current and measurement of membrane potential. After stabilization of neurons, CBZ (100 μm) was applied by adding it to the ACSF. The effects of CBZ saturated 15 min after washin and were quantitatively evaluated 20 min after washin. Bridge balance was carefully adjusted before each measurement and experiments were discarded when it changed by >10%. The signals were filtered on-line at 10 kHz, digitized at a sampling rate of 100 kHz and stored on hard disk (pClamp 8, Molecular Devices).

Drugs.

CBZ (RBI/Sigma) was dissolved in ethanol. Identical concentrations of ethanol were present in all control and CBZ-containing solutions. All other chemicals were obtained from Sigma.

Voltage-clamp paradigms and data analysis.

The voltage dependence of activation and inactivation was determined using the protocol depicted in Figure 1A. Activation was assessed during an initial voltage jump to potentials between −100 and +10 mV (500 ms) (for current traces, see Fig. 1B, left). Inactivation was probed by examining the currents elicited by a subsequent voltage step (15 ms, 10 mV) (Fig. 1A; for current traces, see Fig. 1B, right). The conductance G(V) was calculated according to G(V) = I(V)/(V − VNa), where VNa is the Na+ reversal potential, V is the command potential, and I(V) is the peak current amplitude. G(V) then was fitted for each family of traces obtained from individual cells with the following Boltzmann equation: G(V) = A1 + [(A0 − A1)/(1 + e(V−V1/2)/k)], where A0 and A1 are Na+ conductances, V1/2 is the voltage where G(V) is half of maximal conductance (V1/2,act and V1/2,inact are half-maximal activation and inactivation, respectively), and k indicates the slope of the relation between channel activation or inactivation and membrane voltage.

Figure 1.
Figure 1.

Analysis of INaT properties. A–C, Analysis of the voltage dependence of activation and inactivation in a representative neuron. A, Voltage protocol used to analyze voltage dependence of INaT, with pulses to potentials ranging from −100 mV to +10 mV (500 ms) followed by a voltage step to −10 mV (15 ms). The break in the lines indicates a gap of 485 ms. B, The initial 500 ms voltage step to different potentials caused activation of INaT (leftmost panel), allowing construction of an activation curve. The subsequent voltage step to −10 mV allowed us to assess voltage-dependent inactivation by the preceding voltage step (representative family of traces in the rightmost panel). C, The peak current amplitudes were corrected for the driving force, and the resulting normalized conductances are depicted versus the potential of the first voltage step (squares and circles indicating the currents during the first and second voltage step, respectively). The data were then fitted with a Boltzmann equation (superimposed lines). Exemplary analysis is shown for the recordings shown in B. D–F, Analysis of the time course of recovery from inactivation. D, Voltage protocol used to analyze recovery from inactivation. An initial voltage step to −10 mV (15 ms) to inactivate INaT was followed by a voltage step of varying duration Δt to −80 mV to induce recovery from inactivation. The fraction of INaT recovered from inactivation was determined with a second voltage step to −10 mV. E, Representative recording of recovery from inactivation. The leftmost traces correspond to the currents elicited by the initial step to −10 mV. The rightmost family of traces reflect INaT elicited following the various different recovery intervals Δt and are depicted on a logarithmic time scale. F, Amplitudes are normalized to the peak amplitude INaT during the initial voltage step to −10 mV. A biexponential fit is superimposed as solid line. Exemplary analysis is shown for the recordings shown in E. For the quantitative average values for all experimental groups, see Table 1.

Double pulse experiments to analyze recovery from inactivation were performed according to established protocols as depicted in Figure 1D. A conditioning pulse from −80 to −10 mV (15 ms) was followed by a varying recovery period (1 to 8192 ms) at −80 mV and a subsequent test pulse to −10 mV (15 ms). Time constants of recovery τfast and τslow were extracted from the data points by fitting with a biexponential equation: I(t) = A0 + Afast(e−Δt/τfast) + Aslow(e−Δt/τslow), where I(t) is the normalized peak current amplitude following a recovery period of the duration Δt, Afast and Aslow are the relative amplitude contributions of the fast and slow recovery time constants τfast and τslow, and A0 is a constant offset. A nonlinear Levenberg–Marquardt algorithm was used for all fits. In all cases, fits were performed on the data points obtained from individual cells, and these parameters were then averaged. All results are presented as the mean ± SEM.

Current-clamp data analysis.

The voltage traces recorded in current-clamp mode with sharp microelectrodes were analyzed with Igor Pro 6.01 (Wavemetrics). We characterized action potential morphology using short current injections (3 ms). Action potentials were described quantitatively by determining their peak amplitude, the maximal rate of rise of the voltage trace, and the half-width measured at the voltage intermediate between action potential threshold and peak. The action potential threshold was determined as the voltage at which the slope of the voltage trace exceeded 15 mV/ms (Sekerli et al., 2004). Repetitive firing was examined using prolonged current injections (500 ms). Input–output relations were obtained with an automatic action potential detection routine programmed in Igor Pro 6.01. To assess the subthreshold excitation, we used defined current injections set to a value of 2× (Scn1b) or 1.5× (Scn2b) the threshold current injection. We determined the slope of the interspike interval as the average of the first derivation of the voltage trace within the interspike interval from the most hyperpolarized point to the spike threshold. For this analysis, we chose the interspike interval preceding the first action potential detected beginning 100 ms after the start of the recording. This interspike interval was selected to ensure that passive charging of the membrane capacitance had completed for all cells to avoid contamination of the interspike slope.

Computational modeling studies.

A model of a pyramidal neuron was adapted from Royeck et al. (2008) and contained biophysically realistic morphology (265 compartments) and 15 distributed voltage and/or Ca2+-dependent conductances (see supplemental Tables 1 and 2, available at www.jneurosci.org as supplemental material). Implementation was performed within the NEURON (Hines and Carnevale, 2001) modeling environment running on a dual core processor (2.39 GHz per core, 2 GB RAM) under Windows XP. Integration time steps were fixed at 0.025 ms. Full details of the model are available at MODELDB (https://senselab.med.yale.edu/, click on ModelDB, executable model available under accession code 125152).

Briefly, INaP was implemented as an activating and non-inactivating Na+ current as follows: iNaP = ḡNaP · m · (V − VNa), with V and VNa corresponding to the membrane potential and the Na+ equilibrium potential, respectively, and m describing the activation dynamics with the following equations: Embedded Image with m∞ and τm corresponding to the steady-state channel opening probability and the time constant of activation, respectively. The voltage dependence of activation was systematically varied in the model by altering the parameter ΔV1/2, which describes the shift in the voltage of half-maximal activation relative to a reference voltage of −52.3 mV. The INaP conductance was present within the soma/axon hillock and the axon initial segment, with ḡNaP at the soma/axon hillock being always 60% of ḡNap at the axon initial segment. We varied these conductances systematically without altering the ratio. For simplicity, we give values for ḡNaP at the axon initial segment throughout the paper.

Statistical analyses.

Statistical comparison was performed with a Student's t test at a significance level α of 0.05. Where appropriate, paired tests were used to compare the recordings before and during application of 100 μm CBZ. Unpaired tests were used for group comparisons of null mice to their wild-type littermates regarding channel properties and CBZ effects. The results of the statistical analysis are given throughout the results and figure legends and significance levels <0.05 are indicated in the figures with asterisks.

Results

Properties of INaT in DG neurons are unaltered in mice lacking β1 or β2 subunits

We first examined whether the properties of transient Na+ currents are altered in β subunit null mice compared to littermate control mice. We examined the voltage dependence of INaT using a standard protocol (Fig. 1A). Voltage steps ranging from −100 to +10 mV (500 ms) were applied, followed by a constant voltage step to −10 mV. The first of the two steps causes activation of INaT, allowing analysis of the voltage dependence of activation (Fig. 1B, left panel). The voltage dependence of inactivation was determined by measuring current amplitudes during the second, constant voltage step to −10 mV (Fig. 1A,B, rightmost panel). Peak amplitudes were converted to conductance, normalized, and plotted versus the voltage (Fig. 1C; black squares and circles correspond to activation and inactivation, respectively) and fitted with a Boltzmann equation superimposed on the data points (see Materials and Methods). This analysis revealed that genetic deletion of either Scn1b or Scn2b did not cause changes in the voltage dependence of activation or inactivation (see Table 1 for values for V1/2 and the slope factor k). Likewise, neither the maximal amplitude of INaT elicited following full recovery of Na+ currents from inactivation nor INaT current density was significantly altered (see Table 1). This finding is in contrast to previous studies showing consistently that Scn2b null neurons display reduced current density (Chen et al., 2002). However, we should note that our selection of cells with very low series resistance error (see Materials and Methods) introduces a considerable bias against including cells with large INaT, and therefore also a bias against detecting differences in the magnitude of this current.

View this table:
Table 1.

Properties of transient Na+ currents in Scn1b and Scn2b null mice and their littermate controls

We also examined time-dependent inactivation and recovery from inactivation. The inactivation of Na+ currents during voltage steps to −10 mV could be described by a monoexponential equation. The time-dependent inactivation was not different when comparing β subunit null mice with the appropriate wild-type controls (see Table 1). The recovery from inactivation was analyzed by applying a voltage step to −10 mV (15 ms). Subsequently, Na+ current availability was recovered by repolarizing to −80 mV for variable time intervals (Fig. 1D) (Δt: 1–8192 ms). Recovery of the INaT was assessed by a test pulse to −10 mV (Fig. 1D, representative current traces depicted in Fig. 1E). Recovery from inactivation was best fitted by a biexponential equation, as reported previously (Ellerkmann et al., 2001) (see Materials and Methods and superimposed line in Fig. 1F). We found no difference in the time constants and amplitude proportions describing the recovery from inactivation in β subunit null mice compared to littermate control animals (Table 1).

CBZ effects on peak INaT amplitude are unaltered by deletion of β subunits

We then asked whether the actions of CBZ on Na+ channels are modified by loss of either β1 or β2 subunits. First, we determined whether the effects of CBZ (100 μm) on the maximal current amplitude are altered. Maximal current amplitudes were obtained by stepping to −100 mV for 500 ms to induce recovery from fast inactivation and then eliciting Na+ currents by stepping to −10 mV (Fig. 2A). CBZ reduced the amplitude of INaT in all genotypes (Fig. 2B; black traces correspond to control recordings; red traces were recorded during CBZ application). The magnitude of inhibition ranged from ∼15–35% and was not significantly different when comparing wild-type and mutant mice (Fig. 2C). We should note that the time course of the decay of the current was accelerated by CBZ. When this phenomenon was quantified at command potentials of −10 mV, there was a significant reduction in τinact for all genotypes. This effect was seen over a large range of command potentials, and its magnitude was not significantly different in null animals compared to their wild-type littermates (supplemental Fig. 1, available at www.jneurosci.org as supplemental material; significant differences are indicated with asterisks).

Figure 2.
Figure 2.

CBZ reduces Na+ current amplitude, and this effect is unaltered in mice lacking either Scn1b or Scn2b. A, A voltage step to −10 mV (15 ms) was applied to elicit maximal Na+ currents. B, Representative currents before (black) and during (red) application of 100 μm CBZ are depicted for Scn1b wild-type and Scn2b wild-type (upper traces) and for Scn1b null and Scn2b null (lower traces). C, The fraction of channels blocked by CBZ was averaged for all cells. Neither loss of Scn1b [32.6 ± 5.0% vs 26.5 ± 3.3% for +/+ (n = 11) and −/− (n = 6) animals, respectively; n.s.] nor Scn2b [26.1 ± 4.4% vs 16.2 ± 3.5% for +/+ (n = 5) and −/− (n = 7) animals, respectively; n.s.] causes a significant reduction of the blocking effect of CBZ.

CBZ effects on the voltage dependence of activation of INaT are enhanced in mice lacking β2 subunits

The voltage dependence of activation was determined as described above (Fig. 1) in the absence and presence of CBZ. In both the Scn1b littermate controls (P12–P20) and the older (P40–P60) Scn2b littermate control mice, CBZ caused a significant shift of the voltage dependence of activation toward more hyperpolarized potentials. This shift was significantly more pronounced in Scn1b littermate control mice (Fig. 3A1,A2). This difference may reflect a developmental change in Na+ current pharmacology because Scn1b mice were recorded at earlier developmental time points (P12–P20) than Scn2b wild-type littermate control mice (P40–P60, see Materials and Methods). Indeed when Scn2b littermate control mice were recorded at earlier time points (P12–P20), the shift of the voltage dependence of activation induced by CBZ was comparable to Scn1b littermate controls (shift of V1/2,act = 5.71 ± 1.2 mV, n = 5, n.s. compared to Scn1b wild-type littermates). When we analyzed the effects of CBZ in Scn1b null mice (Fig. 3A3), we found no difference in the CBZ-induced shift of V1/2,act compared to the wild-type animals (Fig. 3, compare A1 and A3, shift of V1/2,act; see inset in A3). In contrast, loss of Scn2b resulted in a significantly larger shift of the voltage dependence of activation following application of CBZ compared to wild-type littermates (Fig. 3, compare A2 and A4; see inset of A4 for quantification). This finding indicates that the presence of β2 subunits modulates the effects of CBZ on voltage-dependent activation of INaT, although the extent of this modulation is very subtle (difference in CBZ-induced shift of V1/2,act of only ∼2 mV) (Fig. 3A4, inset).

Figure 3.
Figure 3.

CBZ shifts the voltage dependence of activation to more hyperpolarized potentials, and this effect is enhanced in mice lacking Scn2b. A, Voltage dependence of activation was analyzed according to Figure 1A–C in the presence and absence of CBZ (triangles and squares, respectively, in all panels). A1, A2, CBZ induced a strong hyperpolarizing shift in the voltage dependence of activation in the younger Scn1b wild-type mice (P12–P20, shift in V1/2,act of −8.2 ± 2.1 mV, n = 11, A1), and a less pronounced, but still significant, effect in the adult Scn2b wild-type mice (P40–P60, shift in V1/2,act of −1.9 ± 0.6 mV, n = 6, p < 0.05, A2). A3, A4, Comparison of the genotypes revealed a significant enhancement of the CBZ-induced left shift in Scn2b null compared to Scn2b wild-type mice (shift in V1/2,act of −3.8 ± 0.3 mV in Scn2b null mice, n = 7, p < 0.05, A4). Such an enhancement was not observed for Scn1b null compared to Scn1b wild-type mice (shift in V1/2,act of −6.9 ± 1.3 mV in Scn1b null, n = 5, n.s., A3, see insets in A3 and A4). B, Analysis of the voltage dependence of inactivation (see Fig. 1A–C for protocols and example traces) revealed a strong CBZ-induced hyperpolarizing shift of the voltage dependence of inactivation independent of expression of Scn1b (−17.4 ± 2.0 mV, n = 11 and −15.2 ± 2.0 mV, n = 6, for +/+ and −/− animals, respectively, n.s., B1) and Scn2b (−11.1 ± 2.1 mV, n = 5 and −11.0 ± 0.4 mV, n = 7, for +/+ and −/− animals, respectively, n.s., B2, see insets for quantification; open symbols correspond to Scn1b or Scn2b null mice and filled symbols to their wild-type littermates).

In contrast, the classical effects of CBZ on voltage-dependent inactivation (analyzed as depicted in Fig. 1) were identical in Scn1b null compared to wild-type littermate mice (Fig. 3B1), as well as when we compared Scn2b null to wild-type littermate mice (Fig. 3B2, shift in V1/2,inact depicted in the insets).

CBZ effects on recovery from inactivation of INaT are unaltered by deletion of β subunits

CBZ powerfully slows the time course of recovery from inactivation in dentate granule neurons. This effect is diminished or even abolished in most therapy refractory epilepsy patients and the pilocarpine model of epilepsy (Remy et al., 2003; Schaub et al., 2007). This loss is in the latter model associated with a transient loss of mRNA encoding both Scn1b and Scn2b (Ellerkmann et al., 2003). We therefore tested whether genetic deletion of either of these subunits in mice causes changes in the CBZ-induced slowing of the recovery from inactivation. The time course of recovery was analyzed as shown in Figure 1D. Application of CBZ caused pronounced slowing of the recovery from inactivation in both Scn1b and Scn2b control groups (Fig. 4A1,A2). Similar effects were observed for Scn1b and Scn2b null mice (Fig. 4A3,A4). When the time course of recovery from inactivation was described with a biexponential equation (see Materials and Methods), we did not find differences between Scn1b null mice and wild-type littermates or between Scn2b null mice and wild-type littermates regarding the CBZ-induced changes in time constants of recovery (τfast and τslow) (Fig. 4B,C) or the corresponding amplitude contributions (Fig. 4D). One conclusion from this finding is that loss of one of the β subunits alone is unlikely to account for loss of CBZ sensitivity of Na+ channels, which has been previously described in chronic experimental epilepsy (Remy et al., 2003).

Figure 4.
Figure 4.

Time course of recovery from inactivation is slower following CBZ application, and this effect is unaltered in mice lacking either β1 or β2 subunits. A, Analysis of the recovery behavior (see Fig. 1D–F for protocol and example traces) revealed a slower recovery from inactivation in the presence of 100 μm CBZ than in control recordings (triangles and squares, respectively, in all panels). This effect appeared to be similar in all genotypes (A1–A4). Biexponential fits are superimposed on the data points in all panels. B–D, Quantification of the biexponential fits: The effects of CBZ on τfast (B), τslow (C), and the amplitude proportion of τfast (D) were not significantly altered by genetic deletion of either Scn1b or Scn2b [τfast: 2.2 ± 0.3- and 2.0 ± 0.3-fold change for Scn1b wild-type (n = 10) and Scn1b null (n = 6) mice (n.s.) and 1.5 ± 0.1- and 1.7 ± 0.1-fold change for Scn2b wild-type (n = 8) and Scn2b null (n = 6) mice (n.s.), respectively].

CBZ effects on INaP in mice lacking β subunits

After this analysis of INaT, which showed a fast activation and inactivation, we examined CBZ effects on INaP. This current corresponds to a small fraction of the cellular Na+ current that does not inactivate during prolonged depolarization. INaP activates at more hyperpolarized potentials than INaT and thereby modulates the excitability of the membrane at subthreshold potentials. INaP was elicited with a slow voltage ramp (slope 50 mV/s) (Fig. 5A) to prevent activation of INaT. In the corresponding current traces, an inward current reflecting INaP could be observed in all cells (Fig. 5B1, trace a), and this current was reversibly reduced in magnitude by the application of 100 μm CBZ (Fig. 5B1, traces b and c). To isolate INaP from other ionic currents that might contribute to the ramp-induced currents, we applied 0.5 μm TTX at the end of all experiments. The trace recorded in the presence of TTX (Fig. 5B1, trace d) was then subtracted from all other traces to obtain pharmacologically separated INaP. The resulting current traces demonstrate a potent blocking effect of CBZ on the peak INaP (Fig. 5B2). The current traces were corrected for the driving force (see Materials and Methods), the resulting conductance traces are depicted versus the ramp voltage in Figure 5C, and were fitted with a Boltzmann equation (see Materials and Methods, fits superimposed on conductance traces). The average peak conductance derived from Boltzmann fits was not significantly different when comparing Scn1b or Scn2b null mice to the appropriate littermate controls (Fig. 5D). We did note an increased magnitude of INaP when comparing the littermate controls for Scn2b to the littermate controls for Scn1b. As Scn2b controls were older than the Scn1b controls (P40–P60 vs P12–P20), and the Scn2b wild-type littermates showed a similar small amplitude when recorded at the same age (P12–P20, gmax = 0.46 ± 0.18 nS, n = 5, n.s. compared to Scn1b wild-type littermates), this probably reflects a developmental increase in INaP (Fry, 2006). When we then considered the effects of CBZ on INaP, we found similar blocking effects of CBZ, with ∼50% reduction of INaP following CBZ application in all experimental groups (Fig. 5E).

Figure 5.
Figure 5.

CBZ reduces the amplitude of INaP independent of the β subunit expression. A, INaP was elicited by a slow voltage ramp (50 mV/s). B1, Representative current traces before (trace a), during (trace b), and after (trace c) washout of 100 μm CBZ, and following TTX application (trace d), are shown. B2, Currents obtained during TTX application are subtracted from all other traces, to isolate INaP. C, Currents were corrected for the driving force and the resulting conductance plotted versus the voltage. The voltage dependence of activation was fitted with a Boltzmann equation (superimposed black lines). D, Maximal conductances were averaged. Scn1b wild-type mice displayed a smaller conductance for INaP (0.43 ± 0.01 nS, n = 7) than Scn2b wild-type mice (0.89 ± 0.11 nS, n = 10, p < 0.05). Comparison of null mice (open bars) to their wild-type littermates (filled bars) revealed no genotype-dependent difference in INaP amplitude (0.49 ± 0.07 nS, n = 6 and 1.00 ± 0.16 nS, n = 9 for Scn1b null and Scn2b null mice, respectively). E, The amplitude reduction following CBZ application was similar [48.1 ± 1.8% and 58.6 ± 9.2% reduction for Scn1b wild-type littermates and Scn1b null mice (n.s.) and 61.9 ± 3.5% and 67.7 ± 4.3% reduction for Scn2b wild-type littermates and Scn2b null mice (n.s.), respectively] in all experimental groups (filled and open bars for wild-type and null animals, respectively).

The analysis of INaT had revealed a slight shift of the voltage of half-maximal activation to more hyperpolarized potential following application of CBZ. We wondered whether a similar effect could be observed for INaP. We determined the voltage dependence of activation by fitting INaP conductance–voltage relations as described in Figure 5C in the absence and presence of CBZ. INaP conductance was normalized to the peak conductance of each trace to visualize the voltage dependence of activation before and during application of CBZ and after washout. In these recordings, a large hyperpolarizing shift of INaP voltage dependence of activation following CBZ application was apparent (Fig. 6A). This shift was reminiscent of the shift observed for INaT (Fig. 3A) but was more pronounced. A quantitative analysis of this phenomenon in the different genotypes is depicted in Figure 6, B1 (Scn1b null and wild-type littermates) and B2 (Scn2b null and wild-type littermates). For generation of the voltage relations shown in this figure, voltage traces obtained from individual cells were first fitted separately. The parameters obtained were averaged and then used to construct voltage relations representative of the different experimental groups. This analysis clearly shows a shift toward more hyperpolarized potentials induced by CBZ, which is significantly larger in Scn1b null than in littermate wild-type mice (Fig. 6B1; ACSF: full lines, CBZ: dashed lines, Scn1b littermate control and null mice, black and red lines, respectively). In contrast to these results for Scn1b, the CBZ-induced hyperpolarizing shift was not significantly different when comparing Scn2b null to wild-type littermate mice (Fig. 6B2, shift in V1/2,act of INaP quantified in the insets of Fig. 6B1,B2).

Figure 6.
Figure 6.

CBZ shifts the voltage dependence of INaP to more hyperpolarized potentials, and this effect is strongly augmented by genetic deletion of Scn1b. A, To visualize the voltage dependence of activation, the conductance was normalized to the maximal conductance of INaP during the voltage ramp. This analysis reveals a pronounced left shift induced by CBZ. B, Activation curves calculated from the average fit parameters of all cells are depicted before (solid lines) and during (dashed lines) CBZ application for wild-type (black lines) and β subunit null (red lines) mice. Deletion of Scn1b (B1) but not Scn2b (B2) resulted in an increased CBZ effect [−12.3 ± 1.9 mV compared to −7.9 ± 0.6 mV shift for Scn1b null (n = 6) and Scn1b wild-type (n = 7) mice (p < 0.05) and −6.0 ± 0.6 mV compared to −4.3 ± 0.5 mV shift for Scn2b null (n = 9) and Scn2b wild-type (n = 10) mice (not significant), respectively] on the voltage dependence of activation (see insets).

The effect of CBZ on INaP was not fully reversible after washout in most cells. This could reflect incomplete washout of CBZ or a time-dependent change in the properties of INaP. To assess the magnitude of the latter change, we recorded INaP without applying CBZ for a time period equivalent to the pharmacological experiments. We found a hyperpolarizing shift of INaP (−3.1 ± 1.2 mV for Scn1b null mice, n = 3, and −5.1 ± 0.4 mV for Scn1b littermate wild-type controls, n = 4, n.s. when comparing genotypes). When we corrected the data presented in Figure 6B1 for this shift, we find corrected shifts in V1/2,act for INaP of −9.2 and −2.8 mV for Scn1b null mice and littermate controls, respectively. Thus, time-dependent changes in INaP cannot account for the effects of Scn1b deletion on CBZ sensitivity of INaP.

We then went on to examine whether the CBZ-induced hyperpolarizing shift in INaP activation indeed results in an increased inward current at membrane voltages close to the action potential threshold. To this end, we normalized the INaP conductance of each cell to the maximal conductance in the absence of CBZ and plotted the average normalized INaP conductance of Scn1b null and littermate wild-type mice in the absence (Fig. 7A1, continuous lines) and presence (Fig. 7A1, dashed lines; Scn1b null mice depicted in red) of CBZ. The curves obtained for Scn1b null mice clearly show that INaP is increased substantially by CBZ application in the threshold range (approximately −80 to −60 mV), while it is potently reduced at more depolarized potentials. The augmentation of INaP by CBZ is not apparent in Scn1b littermate control mice. This finding became more apparent when the CBZ-induced increase in INaP conductance was quantified by subtraction (Fig. 7B1; black lines and red lines correspond to Scn1b null mice and their wild-type littermates, respectively). INaP proved to be approximately threefold increased following CBZ application in Scn1b null mice, while this increase was much smaller in wild-type mice (Fig. 7C1). In contrast to the results for Scn1b, a CBZ-induced increase in INaP conductance was not seen for Scn2b null mice (Fig. 7A2–C2).

Figure 7.
Figure 7.

The hyperpolarizing shift of the voltage dependence of activation causes paradoxical effects of CBZ on INaP in Scn1b null mice. A, To demonstrate the net change in conductance resulting from the reduction in maximal amplitude (Fig. 5E) versus the shift in the voltage dependence (Fig. 6B1,B2), INaP conductance was normalized to the maximal conductance in the absence of CBZ. Activation curves calculated from the average fit parameters of all cells are depicted before (solid lines) and during (dashed lines) CBZ application for wild-type (black lines) and Scn1b (A1) or Scn2b (A2) null mice (red lines). This reveals an increase of conductance in the subthreshold range following application of CBZ in Scn1b null mice. B, The difference in the normalized conductance before and during application of CBZ was calculated to quantify the potentiating effects of CBZ at subthreshold voltages. The increase of INaP conductance induced by CBZ was significantly augmented (B1, n = 6, p < 0.05 in the range −73 to −67 mV) in the Scn1b null mice compared to wild-type mice (n = 7, see inset for full range; gray area indicates ±SEM for each data point; gray rectangle indicates the magnified range; note that the red trace does not lie within the magnified range and therefore is not depicted in B2 and C2 but is present in the insets) but not in Scn2b null mice (n = 9) compared to their wild-type littermates (n = 10, B2). C, The CBZ-induced increase of INaP conductance was normalized to the conductance in the absence of CBZ. This reveals a more than threefold augmentation of INaP conductance in the Scn1b null mice (C1; see inset for full range; gray rectangle indicates the magnified range).

We then tested whether these paradoxical effects of CBZ are also observed at lower CBZ concentrations (10–100 μm). These experiments revealed that the reduction of maximal INaP conductance increased dose dependently (Fig. 8A). Surprisingly, a significant hyperpolarizing shift in V1/2,act was present already at the lowest concentration of 10 μm CBZ (Fig. 8B). Concurrently, the concentration dependence of these two effects caused much larger paradoxical increase of INaP at low CBZ concentrations (Fig. 8C).

Figure 8.
Figure 8.

Concentration dependence of the paradoxical effect of CBZ in Scn1b null mice. Recordings identical to those described in Figure 5 were obtained before and during application of 10–100 μm CBZ. A, Reduction of the maximal INaP conductance following application of 10, 30, or 100 μm CBZ. B, Shift of the voltage dependence of activation induced by 10–100 μm CBZ. C, The CBZ effect expressed as fraction of the maximal conductance under control conditions (see Fig. 7) revealed larger paradoxical effects with lower concentrations of CBZ, compared to 100 μm CBZ.

CBZ effects on neuronal excitability in mice lacking β subunits

We next examined whether the paradoxical, CBZ-induced upregulation of INaP in Scn1b null mice has effects on neuronal excitability. We obtained sharp microelectrode recordings from CA1 pyramidal neurons of Scn1b null and littermate wild-type mice (see Materials and Methods), and assessed their response to 100 μm CBZ. The properties of single action potentials elicited by brief (3 ms) current injections showed only small differences between genotypes (with respect to action potential peak and half-width) (for data, see supplemental Table 3, available at www.jneurosci.org as supplemental material). They were not altered by application of CBZ (data not shown). When repetitive firing was elicited by prolonged (500 ms) current injections, firing frequency was markedly reduced by 100 μm CBZ in both the Scn1b and the Scn2b littermate control mice, as expected (Fig. 9A, left, and Fig. 9B1,B3). In Scn1b null mice, however, the firing frequency was not noticeably affected (Fig. 9A, right, and Fig. 9B2). In contrast to Scn1b null mice, Scn2b null mice displayed a pronounced CBZ response (Fig. 9B4) that seemed even larger than the Scn2b littermate control mice (Fig. 9B3). However, the magnitude of the CBZ effects was not different between Scn2b null and control mice.

Figure 9.
Figure 9.

Paradoxical CBZ-induced upregulation of INaP and effects on neuronal excitability. A, Current-clamp recordings from CA1 pyramidal neurons of Scn1b null (right panels) and littermate wild-type mice (left panels). Repetitive firing was elicited by prolonged (500 ms) current injections as indicated in ACSF (black voltage traces) and following washin of 100 μm CBZ (red voltage traces). B, Spike gain, depicted by plotting the mean firing rates versus the magnitude of the current injection steps. Analyses are shown for Scn1b null mice (n = 7, B2) and their littermate controls (n = 7, B1), as well as for Scn2b null mice (n = 5, B4) and littermate controls (n = 5, B3). The gain in the presence of CBZ (triangles) is compared to the gain in ACSF (squares) in all panels, revealing a significant reduction of the gain following application of CBZ in all experimental groups (p < 0.05, indicated by asterisks) except for the Scn1b null mice, where the effect of CBZ was completely abolished.

To examine whether a hyperpolarizing shift of the INaP voltage dependence can counteract the effects of reducing INaP conductance, we systematically varied both parameters in a computational model (Fig. 10A; see Materials and Methods and link to ModelDB). To reduce the degrees of freedom in the parameters varied in these simulations, we confined ourselves to varying parameters of INaP only. We then induced firing in the model via current injections of increasing magnitude [from 0.4 nA (lowermost traces) to 0.75 nA (uppermost traces) for each modeling condition (Fig. 10A)]. The example traces already reveal that a reduction of INaP causes decreased firing. Conversely, a more hyperpolarized voltage dependence of INaP activation (ΔV1/2 = −8 mV vs −4 mV) causes markedly increased firing (Fig. 10A, compare left and right panels). This relationship is plotted over a range of current injections for the examples shown in Figure 10A in Figure 10B1–B4 for different values of ΔV1/2 (indicated by grayscales in Fig. 10B1–B4 as indicated in B4). The interrelation of ΔV1/2 and INaP conductance is plotted in Figure 10C for an 800 pA current injection step. This plot indicates that even large reductions of INaP at ΔV1/2 = 0 mV can be easily counteracted by simultaneous negative shifts in the voltage dependence of activation of INaP of the order of magnitude observed in Scn1b null mice (approximately −9 mV).

Figure 10.
Figure 10.

Computational model—effects of varying maximal INaP conductance and voltage of half-maximal activation. A, Examples of somatic current injections of increasing amplitude (500 ms steps of 0.4 to 0.75 nA from lowermost to uppermost traces in all panels) for different maximal INaP conductances (A1–A4) and voltages of half-maximal activation of INaP (ΔV1/2 indicating relative shift of the V1/2 of activation vs a value of −52.3 mV). B, Gain plots of action potential frequencies versus the magnitude of the current injection step for the four different INaP maximal conductances (B1–B4) depicted in A. Darker grayscales indicate larger ΔV1/2, i.e., more hyperpolarized INaP voltage dependence of activation (see inset in B4). C, Interrelation of ΔV1/2 and INaP conductance for an 800 pA current injection step for a larger range of values. Different maximal INaP conductances are indicated as hatched lines with different grayscales as indicated in the legend.

How can the CBZ-induced paradoxical increase in INaP in the subthreshold voltage range explain that CBZ fails to affect repetitive firing in Scn1b null mice? We hypothesized that the increased magnitude of INaP in a subthreshold range might augment the rate at which the membrane potential re-depolarizes between individual spikes in a train, which would be expected to increase the spike rate. Indeed, an inspection of repetitive firing in Scn1b null mice revealed that CBZ strongly augmented the rate of interspike depolarization (Fig. 11A3, larger magnification in A4; red traces correspond to traces obtained after CBZ application). In Scn1b littermate control mice, no effects on the rate of interspike depolarization could be observed (Fig. 11A1, larger magnification in A2, quantification shown in Fig. 11B, black bars). As expected, neither Scn2b knock-out nor wild-type mice showed changes in the rate of interspike depolarization upon application of CBZ (Fig. 11B, white bars). These effects could be reproduced in the computational model. Systematically decreasing INaP conductance led to decreases in the interspike depolarization rate (Fig. 11C1–C3, compare light gray traces for ΔV1/2 = 0 mV). Over the range of INaP conductances examined, hyperpolarizing shifts in V1/2 led to opposing accelerations of interspike voltage (Fig. 11C1–C3, compare different ΔV1/2 values in each individual panel).

Figure 11.
Figure 11.

Effects of CBZ in the subthreshold voltage range during repetitive firing. A, Black traces were obtained in ACSF, while red traces were obtained following washin of CBZ. CBZ augments the rate of interspike depolarization in Scn1b null mice. (Fig. 10A3; larger magnifications of the area indicated by the gray boxes are superimposed in A4), but not in Scn1b wild-type mice (Fig. 10A1, larger magnification in A2). B, Quantification of the slope of the interspike depolarization (see Materials and Methods). The relative changes of the slope following application of 100 μm CBZ compared to the slope in ACSF are depicted. Lack of β1 (rightmost filled bar, n = 7, p < 0.01, indicated by asterisk) but not β2 (rightmost open bar, n = 3, n.s.) subunits results in a CBZ-induced acceleration of the depolarization during the interspike interval that was not observed in the wild-type littermate controls (leftmost bars, n = 5 for both groups, n.s.). C, The slope of the interspike depolarization following a simulated current injection of 800 pA is steeper with more hyperpolarized voltage of half-maximal activation (V1/2,act, darker grayscales indicate more hyperpolarized voltages). This holds true for different maximal INaP conductances (C1–C3). Zero point of the time axis in C is defined as the most hyperpolarized point of the voltage trace during the interspike interval.

Discussion

The major result of our study is that the sensitivity of voltage-gated Na+ channels to the anticonvulsant CBZ is strongly dependent on the presence of β1 subunits within the channel complex. The most striking finding was that CBZ shifts the voltage dependence of activation of INaP toward more hyperpolarized potentials. Genetic deletion of β1 significantly augments this leftward shift without altering other effects of CBZ. This novel effect is particularly intriguing because it increases INaP in the subthreshold voltage range, leading to a complete loss of CBZ efficacy in blocking repetitive firing. Thus, the presence of β1 subunits within the channel complex appears to be crucial in determining the effects of CBZ on cellular excitability. This finding may be relevant in neurological disorders exhibiting a loss of β subunit function.

Effects of CBZ on INaT and INaP

We found that deletion of either Scn1b or Scn2b did not affect classical CBZ effects on INaT (Ragsdale and Avoli, 1998), i.e., reduction of the maximal amplitude, slowing of recovery from inactivation, and a hyperpolarizing shift in the voltage dependence of inactivation. Scn1b or Scn2b downregulation is thus unlikely to underlie changes in use-dependent block of INaT by CBZ observed in animal models of temporal lobe epilepsy (Ellerkmann et al., 2003; Remy et al., 2003; Schaub et al., 2007). The only effect observed on INaT was that Scn2b but not Scn1b deletion slightly but significantly increased the CBZ-induced hyperpolarizing shift in the activation curve. The CBZ-induced hyperpolarizing shift of the voltage dependence of activation of INaT was large in young control mice aged P12–P20 (littermates for Scn1b or Scn2b null mice, ∼-7 mV), whereas it was much smaller in older (P40–P60) Scn2b control mice. The age dependence of this effect may explain why it was so far not observed in older rats (Ragsdale and Avoli, 1998; Remy et al., 2003; Schaub et al., 2007).

Very similar effects were seen for INaP, in which CBZ also shifted the activation curve toward more hyperpolarized potentials. In this case, the effects of deleting the β1 subunits were large: the CBZ-induced hyperpolarizing shift in the activation curve was nearly doubled in the absence of Scn1b but not altered in Scn2b null mice. Interestingly, the blocking effects of CBZ on the peak amplitude of INaP were not modified in Scn1b or Scn2b null mice. Thus, the presence of β1 subunits appears to selectively modify the effects of CBZ on the voltage dependence of INaP. As a consequence, CBZ application in Scn1b null mice resulted in a block of INaP at depolarized potentials, but increased INaP considerably at subthreshold voltages.

Functional consequences of CBZ effects on INaP

Our combined modeling and physiological data indicate that the hyperpolarizing shifts in INaP voltage dependence counteract other effects of CBZ, both on INaP maximal conductance (Figs. 5⇑⇑–8) and on INaT (Figs. 2⇑–4), most probably via an augmentation of the rate of interspike depolarization. The paradoxical increase of INaP in a confined subthreshold voltage range by CBZ may also be relevant for other aspects of subthreshold excitability of hippocampal pyramidal neurons, such as subthreshold resonance (Hu et al., 2002) and spike timing precision (Vervaeke et al., 2006).

The potential implications of this effect are intriguing. CBZ is used in the clinical treatment of many neurological diseases, like bipolar depression, multiple sclerosis, schizophrenia, myotonia, and dystonia, in addition to neuropathic pain and epilepsy (summarized in Rogawski and Löscher, 2004). Interestingly, β subunits are shown to be differentially regulated in animal models of neurological disorders. In experimental autoimmune encephalomyelitis, a model of multiple sclerosis, both Scn1b and Scn2b are transcriptionally downregulated (Nicot et al., 2003). Likewise, in some types of neurons, denervation appears to result in reduced expression of β1 subunits (Sashihara et al., 1996). Finally, in models of chronic temporal lobe epilepsy, both Scn1b and Scn2b mRNAs are downregulated (Gastaldi et al., 1998; Ellerkmann et al., 2003). In addition, a recent paper has shown that a functional null Scn1b mutation can result in a severe form of childhood epilepsy (Dravet syndrome) (Patino et al., 2009). These studies suggest that downregulation of β subunits is a pathological or adaptive response occurring in some common neurological diseases, and that this may affect the response to CBZ in these disorders.

Effect of β subunit deletion on properties of INaT and INaP

Our data also allow us to address the effects of deleting either β1 or β2 subunits on the biophysical properties of INaT and INaP. We found that neither the amplitude nor the voltage dependence of INaP was altered in mice lacking Scn1b or Scn2b, consistent with the data obtained in expression systems where only coexpression of β3 subunits (Scn3b) increased the amplitude of INaP (Qu et al., 2001). Regarding the properties of INaT, previous analyses of Scn2b null mice (Chen et al., 2002; Lopez-Santiago et al., 2006) or Scn1b/2b null mice (Grieco et al., 2005) described reduced INaT current densities in both central and peripheral neurons, most probably due to a reduced insertion of channels into the plasma membrane. In addition, a small, but measurable, hyperpolarizing shift of the voltage dependence of inactivation was observed in Scn2b null hippocampal pyramidal (Chen et al., 2002), while the kinetics of tetrodotoxin-sensitive current activation and inactivation were slowed in Scn2b null DRG nociceptors with no measurable effects on voltage dependence (Lopez-Santiago et al., 2006). Deletion of Scn1b did not alter the voltage dependence of activation or inactivation in hippocampal pyramidal neurons (Chen et al., 2004) or cerebellar granule cells (Brackenbury et al., 2010). In our experiments, we found no changes in biophysical properties of INaT in mice lacking either β1 or β2 subunits. The lack of difference in current density may be due to our strict selection criteria with regard to series resistance error, leading to rejection of cells with large currents (see Materials and Methods for exact criteria). The lack of difference in voltage dependence between Scn2b null and wild-type mice in our study may be explained by the fact that Chen et al. (2002) used F1 hybrids, and mice congenic on C57BL/6 were used in the present study. It should also be noted that β subunit effects are apparently strongly modulated by as yet unknown, tissue-specific factors. This is suggested by studies describing very varied effects of β subunit coexpression, depending on the expression system used for the study. For instance, coexpression of β2 with Nav1.2 subunits in oocytes results in a small but significant negative shift in the voltage dependence of inactivation, with no change in voltage dependence of activation (Isom et al., 1995). In contrast in HEK293 cells, coexpression of either β1, β2, or β3 with Nav1.2 subunits caused a shift of the voltage dependence of activation and inactivation toward more depolarized potentials when one of the three β subunits was expressed, whereas the recovery behavior was not affected (Qu et al., 2001; Xu et al., 2007). This diversity underscores that the effects of β subunit coexpression are likely modified by multiple signaling pathways and may explain why genetic deletion of β subunits has cell-type-specific results (Meadows and Isom, 2005).

In conclusion, our study describes a paradoxical effect of CBZ on INaP in Scn1b null mice caused by a CBZ-induced hyperpolarizing shift in voltage-dependent activation that results in a potent CBZ-induced increase in INaP at subthreshold voltages. As a consequence, the inhibitory effects of CBZ on repetitive neuronal firing are largely lost. These results suggest that the presence of Na+ channel β1 subunits may profoundly affect the response to anticonvulsant drugs in models of neurological disorders.

Footnotes

  • This work was supported by the Deutsche Forschungsgemeinschaft (SFB TR3, LE660/4-1), Nationales Genomforschungsnetzwerk NGFNplus EmiNet, EPICURE, and the BONFOR program of the University of Bonn Medical Center.

  • Correspondence should be addressed to Prof. Dr. Heinz Beck, Laboratory of Experimental Epileptology, Department of Epileptology, University of Bonn, Sigmund-Freud-Straße 25, 53105 Bonn, Germany. heinz.beck{at}ukb.uni-bonn.de

References

  1. ↵
    1. Brackenbury WJ,
    2. Isom LL
    (2008) Voltage-gated Na+ channels: potential for beta subunits as therapeutic targets. Expert Opin Ther Targets 12:1191–1203.
    OpenUrlCrossRefPubMed
  2. ↵
    1. Brackenbury WJ,
    2. Calhoun JD,
    3. Chen C,
    4. Miyazaki H,
    5. Nukina N,
    6. Oyama F,
    7. Ranscht B,
    8. Isom LL
    (2010) Functional reciprocity between Na+ channel Nav1.6 and beta1 subunits in the coordinated regulation of excitability and neurite outgrowth. Proc Natl Acad Sci U S A 107:2283–2288.
    OpenUrlAbstract/FREE Full Text
  3. ↵
    1. Catterall WA
    (1999) Molecular properties of brain sodium channels: an important target for anticonvulsant drugs. Adv Neurol 79:441–456.
    OpenUrlPubMed
  4. ↵
    1. Chen C,
    2. Bharucha V,
    3. Chen Y,
    4. Westenbroek RE,
    5. Brown A,
    6. Malhotra JD,
    7. Jones D,
    8. Avery C,
    9. Gillespie PJ 3rd.,
    10. Kazen-Gillespie KA,
    11. Kazarinova-Noyes K,
    12. Shrager P,
    13. Saunders TL,
    14. Macdonald RL,
    15. Ransom BR,
    16. Scheuer T,
    17. Catterall WA,
    18. Isom LL
    (2002) Reduced sodium channel density, altered voltage dependence of inactivation, and increased susceptibility to seizures in mice lacking sodium channel beta 2-subunits. Proc Natl Acad Sci U S A 99:17072–17077.
    OpenUrlAbstract/FREE Full Text
  5. ↵
    1. Chen C,
    2. Westenbroek RE,
    3. Xu X,
    4. Edwards CA,
    5. Sorenson DR,
    6. Chen Y,
    7. McEwen DP,
    8. O'Malley HA,
    9. Bharucha V,
    10. Meadows LS,
    11. Knudsen GA,
    12. Vilaythong A,
    13. Noebels JL,
    14. Saunders TL,
    15. Scheuer T,
    16. Shrager P,
    17. Catterall WA,
    18. Isom LL
    (2004) Mice lacking sodium channel β1 subunits display defects in neuronal excitability, sodium channel expression, and nodal architecture. J Neurosci 24:4030–4042.
    OpenUrlAbstract/FREE Full Text
  6. ↵
    1. Ellerkmann RK,
    2. Riazanski V,
    3. Elger CE,
    4. Urban BW,
    5. Beck H
    (2001) Slow recovery from inactivation regulates the availability of voltage-dependent Na+ channels in hippocampal granule cells, hilar neurons and basket cells. J Physiol 532:385–397.
    OpenUrlAbstract/FREE Full Text
  7. ↵
    1. Ellerkmann RK,
    2. Remy S,
    3. Chen J,
    4. Sochivko D,
    5. Elger CE,
    6. Urban BW,
    7. Becker A,
    8. Beck H
    (2003) Molecular and functional changes in voltage-dependent Na+ channels following pilocarpine-induced status epilepticus in rat dentate granule cells. Neuroscience 119:323–333.
    OpenUrlCrossRefPubMed
  8. ↵
    1. Fry M
    (2006) Developmental expression of Na+ currents in mouse Purkinje neurons. Eur J Neurosci 24:2557–2566.
    OpenUrlCrossRefPubMed
  9. ↵
    1. Gastaldi M,
    2. Robaglia-Schlupp A,
    3. Massacrier A,
    4. Planells R,
    5. Cau P
    (1998) mrna coding for voltage-gated sodium channel beta2 subunit in rat central nervous system: cellular distribution and changes following kainate-induced seizures. Neurosci Lett 249:53–56.
    OpenUrlCrossRefPubMed
  10. ↵
    1. Grieco TM,
    2. Malhotra JD,
    3. Chen C,
    4. Isom LL,
    5. Raman IM
    (2005) Open-channel block by the cytoplasmic tail of sodium channel beta4 as a mechanism for resurgent sodium current. Neuron 45:233–244.
    OpenUrlCrossRefPubMed
  11. ↵
    1. Hines ML,
    2. Carnevale NT
    (2001) Neuron: a tool for neuroscientists. Neuroscientist 7:123–135.
    OpenUrlAbstract/FREE Full Text
  12. ↵
    1. Hu H,
    2. Vervaeke K,
    3. Storm JF
    (2002) Two forms of electrical resonance at theta frequencies, generated by m-current, h-current and persistent Na+ current in rat hippocampal pyramidal cells. J Physiol 545:783–805.
    OpenUrlAbstract/FREE Full Text
  13. ↵
    1. Isom LL
    (2001) Sodium channel beta subunits: anything but auxiliary. Neuroscientist 7:42–54.
    OpenUrlAbstract/FREE Full Text
  14. ↵
    1. Isom LL,
    2. Scheuer T,
    3. Brownstein AB,
    4. Ragsdale DS,
    5. Murphy BJ,
    6. Catterall WA
    (1995) Functional co-expression of the beta 1 and type IIa alpha subunits of sodium channels in a mammalian cell line. J Biol Chem 270:3306–3312.
    OpenUrlAbstract/FREE Full Text
  15. ↵
    1. Lopez-Santiago LF,
    2. Pertin M,
    3. Morisod X,
    4. Chen C,
    5. Hong S,
    6. Wiley J,
    7. Decosterd I,
    8. Isom LL
    (2006) Sodium channel β2 subunits regulate tetrodotoxin-sensitive sodium channels in small dorsal root ganglion neurons and modulate the response to pain. J Neurosci 26:7984–7994.
    OpenUrlAbstract/FREE Full Text
  16. ↵
    1. Lucas PT,
    2. Meadows LS,
    3. Nicholls J,
    4. Ragsdale DS
    (2005) An epilepsy mutation in the beta1 subunit of the voltage-gated sodium channel results in reduced channel sensitivity to phenytoin. Epilepsy Res 64:77–84.
    OpenUrlCrossRefPubMed
  17. ↵
    1. Meadows LS,
    2. Isom LL
    (2005) Sodium channels as macromolecular complexes: implications for inherited arrhythmia syndromes. Cardiovasc Res 67:448–458.
    OpenUrlAbstract/FREE Full Text
  18. ↵
    1. Meadows LS,
    2. Chen YH,
    3. Powell AJ,
    4. Clare JJ,
    5. Ragsdale DS
    (2002) Functional modulation of human brain Nav1.3 sodium channels, expressed in mammalian cells, by auxiliary beta 1, beta 2 and beta 3 subunits. Neuroscience 114:745–753.
    OpenUrlCrossRefPubMed
  19. ↵
    1. Nicot A,
    2. Ratnakar PV,
    3. Ron Y,
    4. Chen CC,
    5. Elkabes S
    (2003) Regulation of gene expression in experimental autoimmune encephalomyelitis indicates early neuronal dysfunction. Brain 126:398–412.
    OpenUrlAbstract/FREE Full Text
  20. ↵
    1. Patino GA,
    2. Claes LRF,
    3. Lopez-Santiago LF,
    4. Slat EA,
    5. Dondeti RSR,
    6. Chen C,
    7. O'Malley HA,
    8. Gray CBB,
    9. Miyazaki H,
    10. Nukina N,
    11. Oyama F,
    12. De Jonghe P,
    13. Isom LL
    (2009) A functional null mutation of SCN1B in a patient with Dravet syndrome. J Neurosci 29:10764–10778.
    OpenUrlAbstract/FREE Full Text
  21. ↵
    1. Qu Y,
    2. Curtis R,
    3. Lawson D,
    4. Gilbride K,
    5. Ge P,
    6. DiStefano PS,
    7. Silos-Santiago I,
    8. Catterall WA,
    9. Scheuer T
    (2001) Differential modulation of sodium channel gating and persistent sodium currents by the beta1, beta2, and beta3 subunits. Mol Cell Neurosci 18:570–580.
    OpenUrlCrossRefPubMed
  22. ↵
    1. Ragsdale DS,
    2. Avoli M
    (1998) Sodium channels as molecular targets for antiepileptic drugs. Brain Res Brain Res Rev 26:16–28.
    OpenUrlCrossRefPubMed
  23. ↵
    1. Remy S,
    2. Gabriel S,
    3. Urban BW,
    4. Dietrich D,
    5. Lehmann TN,
    6. Elger CE,
    7. Heinemann U,
    8. Beck H
    (2003) A novel mechanism underlying drug resistance in chronic epilepsy. Ann Neurol 53:469–479.
    OpenUrlCrossRefPubMed
  24. ↵
    1. Rogawski MA,
    2. Löscher W
    (2004) The neurobiology of antiepileptic drugs for the treatment of nonepileptic conditions. Nat Med 10:685–692.
    OpenUrlCrossRefPubMed
  25. ↵
    1. Royeck M,
    2. Horstmann MT,
    3. Remy S,
    4. Reitze M,
    5. Yaari Y,
    6. Beck H
    (2008) Role of axonal Nav1.6 sodium channels in action potential initiation of ca1 pyramidal neurons. J Neurophysiol 100:2361–2380.
    OpenUrlAbstract/FREE Full Text
  26. ↵
    1. Sashihara S,
    2. Greer CA,
    3. Oh Y,
    4. Waxman SG
    (1996) Cell-specific differential expression of Na+-channel β1-subunit mRNA in the olfactory system during postnatal development and after denervation. J Neurosci 16:702–713.
    OpenUrlAbstract/FREE Full Text
  27. ↵
    1. Schaub C,
    2. Uebachs M,
    3. Beck H
    (2007) Diminished response of CA1 neurons to antiepileptic drugs in chronic epilepsy. Epilepsia 48:1339–1350.
    OpenUrlCrossRefPubMed
  28. ↵
    1. Sekerli M,
    2. Del Negro CA,
    3. Lee RH,
    4. Butera RJ
    (2004) Estimating action potential thresholds from neuronal time-series: new metrics and evaluation of methodologies. IEEE Trans Biomed Eng 51:1665–1672.
    OpenUrlCrossRefPubMed
  29. ↵
    1. Toib A,
    2. Lyakhov V,
    3. Marom S
    (1998) Interaction between duration of activity and time course of recovery from slow inactivation in mammalian brain Na+ channels. J Neurosci 18:1893–1903.
    OpenUrlAbstract/FREE Full Text
  30. ↵
    1. Vervaeke K,
    2. Hu H,
    3. Graham LJ,
    4. Storm JF
    (2006) Contrasting effects of the persistent Na+ current on neuronal excitability and spike timing. Neuron 49:257–270.
    OpenUrlCrossRefPubMed
  31. ↵
    1. Xu R,
    2. Thomas EA,
    3. Gazina EV,
    4. Richards KL,
    5. Quick M,
    6. Wallace RH,
    7. Harkin LA,
    8. Heron SE,
    9. Berkovic SF,
    10. Scheffer IE,
    11. Mulley JC,
    12. Petrou S
    (2007) Generalized epilepsy with febrile seizures plus-associated sodium channel beta1 subunit mutations severely reduce beta subunit-mediated modulation of sodium channel function. Neuroscience 148:164–174.
    OpenUrlCrossRefPubMed
Back to top

In this issue

Email
Print
View Full Page PDF
Citation Tools
Respond to this article
Efficacy Loss of the Anticonvulsant Carbamazepine in Mice Lacking Sodium Channel β Subunits via Paradoxical Effects on Persistent Sodium Currents
Mischa Uebachs, Thoralf Opitz, Michel Royeck, Gesa Dickhof, Marie-Therese Horstmann, Lori L. Isom, Heinz Beck
Journal of Neuroscience 23 June 2010, 30 (25) 8489-8501; DOI: 10.1523/JNEUROSCI.1534-10.2010
del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo

Jump to section

Responses to this article

Respond to this article

Jump to comment:

No eLetters have been published for this article.