A Single Sodium Channel Mutation Produces Hyper- or Hypoexcitability in Different Types of Neurons*

Because different ensembles of channels are present within DRG and SCG neurons, we hypothesized that the same sodium channel mutation might have different effects on excitability in these two neuronal types. Here we test this hypothesis for one of the first Na_V1.7 erythermalgia mutations to be characterized, L858H.²^,⁷ We show that although the L858H mutation produces a depolarizing shift in resting membrane potential (RMP) in both cell types, it renders DRG neurons hyperexcitable and SCG neurons hypoexcitable. We demonstrate that the opposing functional effects of this mutation are a result of the selective presence of another sodium channel, Na_V1.8, in DRG, but not SCG, neurons. These results suggest a contribution of Na_V1.7 mutant channels to the sympathetic dysfunction that has been reported in erythermalgia. More generally, these observations show that a single mutation can cause functionally opposing changes in the different types of neurons in which the gene is expressed.

Results

L858H Mutation Produces Depolarization of RMP and Decreases Action Potential Threshold in DRG Neurons

Figure 1 shows the effect of the Na_V1.7/L858H mutation (L858H) on firing threshold in DRG neurons, a cell type that is known to express Na_V1.7 (refs. 8, 9). DRG neurons expressing wild-type (WT) Na_V1.7 produced robust overshooting action potentials in response to stepwise current inputs. The representative cell shown in figure 1A, with a RMP of approximately –51 mV, produced subthreshold responses to 50- to 130-pA current injections and required a ≥135-pA input to elicit an all-or-none action potential that arose at a voltage threshold of approximately –15 mV.

Figure 1 L858H renders DRG neurons hyperexcitable. (A and B) Action potentials were evoked from small (≤ 25 µm in diameter) DRG neurons by using depolarizing current injections from the RMP. V_m, membrane potential. (A) Representative traces from a cell expressing WT Na_V1.7 show subthreshold responses to 50- to 130-pA current injections and subsequent all-or-none action potentials evoked by injections of 135 pA (current threshold for this neuron) and 155 pA. (B) In contrast, in a cell expressing L858H, action potentials were evoked by a 60-pA current injection. The voltage for take-off of the all-or-none action potential (approximately –14.5 mV, dashed line) was similar for the neurons in A and B. (C) L858H causes a depolarizing shift in the RMP of DRG neurons. DRG neurons expressing WT Na_V1.7 had an average RMP of –50.1 ± 0.9 mV (n = 20), whereas those expressing L858H mutant channels had a significantly (*, P < 0.001) depolarized RMP of –44.9 ± 1.1 (n = 25). (D) The average current threshold for action potential firing of DRG neurons expressing WT Na_V1.7 channels was 120.6 ± 23.9 pA (n = 20), whereas that of neurons expressing L858H mutant channels was significantly (*, P < 0.01) reduced to 69.2 ± 9.8 pA (n = 25). (E) Action potential overshoot in cells expressing WT Na_V1.7 channels (67.8 ± 3.0 mV, n = 20) was not significantly different from that in cells expressing L858H mutant channels (64.4 ± 2.6 mV, n = 20; P > 0.05). The voltage of action potential take-off was unchanged (WT, –14.5 ± 1.2 mV, n = 20; L858H, –14.5 ± 1.3 mV, n = 25; P > 0.05). n.s., not significant.

In contrast, the RMP is approximately –46 mV, i.e., is depolarized by ≈5 mV, in a representative DRG neuron expressing L858H mutant channels, and this cell required a much lower current input of only 60 pA for firing (figure 1B). For the entire population of cells studied, the average RMP of DRG cells expressing L858H was significantly more depolarized (–44.9 ± 1.1 mV, n = 25) than that of cells expressing WT Na_V1.7 (–50.1 ± 0.9 mV, n = 20) (P < 0.001; figure 1C). The current threshold was significantly decreased, by >40% (P < 0.01), in cells expressing L858H channels (69.2 ± 9.8 pA, n = 25) compared with WT Na_V1.7 (120.6 ± 23.9 pA, n = 20) (figure 1D). Despite the depolarized RMP and changes in current threshold, action potential overshoot was not significantly different in cells expressing WT Na_V1.7 (67.8 ± 3.0 mV, n = 20) or L858H channels (64.4 ± 2.6 mV, n = 20) (P > 0.05; figure 1E).

L858H Mutation Produces Depolarization of RMP but Increases Action Potential Threshold in SCG Neurons

We next investigated the functional effects of L858H mutant channels in SCG neurons, which are also known to express Na_V1.7 (refs. 8, 9). Figure 2A shows subthreshold responses evoked by 15- to 20-pA current injections from a representative SCG neuron expressing WT Na_V1.7 channels, which required current injections of ≥25 pA to reach the threshold of approximately –20 mV and fire all-or-none action potentials, from the RMP of –47 mV.

Figure 2 L858H renders SCG neurons hypoexcitable. (A and B) Action potentials were evoked by using depolarizing current injections from resting potential. (A) Representative traces from a cell expressing the WT channel show subthreshold responses to 15- to 20-pA current injections and subsequent all-or-none action potentials evoked by injections of ≥25 pA. (B) In contrast, in a cell expressing the L858H channel, action potentials required a ≥70-pA current injection. The voltage for take-off (dashed line) of the all-or-none action potential was unchanged. (C) L858H channels caused a depolarizing shift in the RMP of SCG neurons. SCG neurons expressing WT channels had an average RMP of –46.3 ± 0.8 mV (n = 15), whereas those expressing L858H had a significantly (P < 0.001) depolarized RMP of –41.6 ± 0.8 (n = 17). (D) The average current threshold for action potential firing of SCG neurons expressing WT channels was 22.7 ± 3.6 pA (n = 15), whereas that of neurons expressing L858H channels was significantly (*, P < 0.01) increased to 42.9 ± pA (n = 17). (E) Action potential overshoot in cells expressing WT channels (47.8 ± 3.4 mV, n = 15) was significantly larger (*, P < 0.001) than that in cells expressing L858H (23.8 ± 4.7 mV, n = 20). The voltage of action potential take-off was unchanged (WT, –23.1 ± 1.2 mV, n = 15; L858H, –19.8 ± 1.3 mV, n = 17; *, P > 0.05).

SCG neurons expressing L858H channels showed a depolarization of the RMP by ≈5 mV, similarly to DRG neurons, but the effects of L858H on excitability were markedly different. Figure 2B shows a representative SCG neuron where only subthreshold responses were seen with a <70-pA current injection, inputs that produced action potentials with WT Na_V1.7 channel expression. When the threshold was reached (approximately –16 mV) with a 70-pA input, the neuron generated an action potential, but it was attenuated, with substantially reduced overshoot. The average RMP of SCG neurons expressing L858H channels was significantly more depolarized (–41.6 ± 0.8 mV, n = 17) than that of cells expressing WT Na_V1.7 channels (–46.3 ± 0.8 mV, n = 15) (P < 0.001; figure 2C). For the entire population of SCG cells studied, current threshold was significantly increased by ≈88% (P < 0.01) in cells expressing L858H (42.9 ± 6.3 pA, n = 17) compared with WT Na_V1.7 channels (22.7 ± 3.6 pA, n = 15) (figure 2D). In contrast to DRG neurons where action potential overshoot was maintained after expression of L858H, action potential overshoot was significantly reduced by ≈50% (P < 0.001) in SCG neurons expressing L858H (23.8 ± 4.7 mV, n = 20) compared with WT channels (47.8 ± 3.4 mV, n = 15) (figure 2E). Thus, in contrast to DRG neurons where L858H mutant channels reduce threshold for single action potentials, expression of L858H in SCG neurons has an opposite effect, increasing threshold.

L858H Mutation Enhances Repetitive Firing in DRG Neurons

Previous studies have shown that ≈50% of DRG neurons fire repetitively in response to sustained depolarizing stimuli.³^,¹⁰^–¹³ In this study, 65% (13 of 20) of DRG neurons expressing WT Na_V1.7 fired two or more action potentials in response to prolonged stimulation. Figure 3A shows a representative DRG neuron that fired one action potential in response to a 950-ms input of 100 pA but was capable of firing multiple action potentials with a higher current injection of 250 pA (figure 3, inset). With expression of L858H, a higher proportion of DRG neurons (88%, or 21 of 24) fired two or more action potentials in response to 950-ms current injections. Figure 3B shows a representative DRG neuron expressing L858H, which responded to a 100-pA input with five action potentials, i.e., a higher frequency than for WT Na_V1.7 channels. For the entire population of DRG neurons studied, the firing frequency evoked with 50- and 100-pA inputs was increased by ≈550% (P < 0.05) and ≈280% (P < 0.05), respectively, in neurons with L858H compared with WT Na_V1.7 channels (figure 3C).

Figure 3 The L858H mutation increases firing frequency in DRG and decreases firing frequency in SCG neurons. (A) Representative DRG neuron expressing WT Na_V1.7 fires a single action potential in response to a 950-ms input of 100 pA from the RMP of this neuron (approximately –50 mV). (*Inset*) The same neuron fires multiple action potentials in response to a 250-pA stimulus. (B) Representative DRG neuron expressing L858H fires five action potentials in response to a 100-pA current injection from the RMP of this neuron (approximately –42 mV). (C) For the entire population of DRG neurons studied, the firing frequency evoked by 50-pA current stimuli was 0.32 ± 0.13 Hz after transfection with WT channels (n = 20) and 2.06 ± 0.79 Hz after transfection with L858H (n = 24; *, P < 0.05), and the firing frequency evoked by 100-pA stimuli was 0.89 ± 0.28 Hz after transfection with WT and 3.37 ± 1.13 Hz after transfection with L858H (*, P < 0.05). (D) Representative SCG neuron expressing WT Na_V1.7 fires six action potentials in response to a 950-ms input of 40 pA from the RMP (approximately –45 mV). (E) Representative SCG neuron expressing L858H fires only two action potentials in response to a 100-pA current injection from the RMP (approximately –40 mV). (*Inset*) When the cell was held at –60 mV to overcome the depolarization of the RMP caused by L858H, it produced four action potentials with an identical stimulus. (F) For the entire population of SCG neurons studied, the firing frequency evoked by 30-pA stimuli was 5.33 ± 1.5 Hz after transfection with WT channels (n = 14) and 0.63 ± 0.01 Hz after transfection with L858H channels (n = 15; P < 0.05). The firing frequency evoked by 40-pA stimuli was 7.05 ± 1.86 Hz after transfection with WT and 1.96 ± 1.0 Hz after transfection with L858H channels (*, P < 0.05).

L858H Mutation Attenuates Repetitive Firing in SCG Neurons

SCG neurons also fired repetitively in response to prolonged stimuli, but, as with threshold, the effect of the L858H mutation was opposite to that in DRG neurons. Ninety-three percent (13 of 14) of SCG neurons expressing WT Na_V1.7 channels produced multiple action potentials in response to prolonged stimulation. Figure 3D shows a representative SCG neuron expressing WT Na_V1.7, where six action potentials were produced in response to a 950-ms current injection of 40 pA. In contrast, an identical stimulus in a SCG neuron expressing L858H evoked only two action potentials, with substantially reduced overshoot (figure 3E). Interestingly, when this cell was held at –60 mV (a maneuver that reversed the depolarization induced by L858H), multiple overshooting action potentials could be evoked (figure 3E, inset); thus, the intrinsic ability of this neuron to fire repetitive and full-scale action potentials was not impaired. Compared with SCG neurons expressing WT Na_V1.7, a decreased proportion (53%, or 8 of 15) of SCG neurons expressing L858H produced two or more action potentials in response to prolonged stimulation. The firing frequency of SCG neurons with L858H channels in response to 950-ms inputs of 30 and 40 pA was substantially reduced by 88% (P < 0.02) and 72% (P < 0.03), respectively, compared with SCG neurons with WT Na_V1.7 channels (figure 3F). Attenuated repetitive firing appeared to be caused by the depolarization of RMP by L858H because, in 91% (10 of 11) of the cells tested, multiple action potential firing could be restored by hyperpolarizing to a holding potential of –60 mV.

SCG Neurons Express Na_V1.7 but Not Na_V1.8

DRG neurons express Na_V1.8 as well as Na_V1.7 sodium channels,⁸^,⁹^,¹⁴^–¹⁶ and action potential generation in these cells involves sequential activation of Na_V1.7 and then Na_V1.8 (refs. 10, 17, 18). SCG neurons normally express Na_V1.7 (refs. 8, 9), but the full complement of sodium channel isoforms within SCG neurons has not been previously defined. We therefore identified the sodium channel isoforms present in SCG neurons and compared them with DRG neurons using multiplex PCR and restriction enzyme polymorphism analysis.¹⁹ Restriction analysis of DRG (figure 4A, lanes 1–9) demonstrated Na_V1.1 (lane 2), Na_V1.6 (lane 6), Na_V1.7 (lane 7), Na_V1.8 (lane 8), and Na_V1.9 (lane 9) in the cDNA pool, in agreement with previous results.¹⁹ In contrast, profiling of the SCG products (figure 4A, lanes 10–18) demonstrated Na_V1.3 (lane 13), Na_V1.6 (lane 15), and Na_V1.7 (lane 16). We confirmed the presence of Na_V1.7 (figure 4Ba) and Na_V1.8 (figure 4Bb) protein in adult rat DRG and postnatal day 2 (P2) rat DRG neurons in culture (figure 4Bc and d) by immunocytochemistry using isoform-specific antibodies. Consistent with published studies on sodium currents in SCG neurons,²⁰^–²² we observed that adult rat SCG neurons express Na_V1.7 (figure 4Ca), but not Na_V1.8 (figure 4Cb), in native tissue and that Na_V1.8 expression is not induced in P2 rat SCG neurons under culture conditions (figure 4Cc and d).

Selective Presence of Na_V1.8 within DRG, but Not SCG, Neurons Contributes to Opposing Effects of the L858H Mutation in These Two Cell Types

Having demonstrated that L858H produces hyperexcitability in DRG neurons and hypoexcitability in SCG neurons, we hypothesized that this difference was caused, at least in part, by the presence of Na_V1.8 in DRG neurons ¹⁴^–¹⁶ and its absence in SCG neurons. Na_V1.8 channels have depolarized voltage-dependence of activation and inactivation¹⁴^,¹⁵^,²³^,²⁴ and thus allow DRG neurons to fire action potentials even when depolarized.¹⁰ We tested this hypothesis by coexpressing Na_V1.8 together with L858H in SCG neurons and examining the effects on firing behavior. Figure 5A shows representative suprathreshold action potentials from SCG cells expressing WT Na_V1.7 (blue), L858H (red), and L858H plus Na_V1.8 (green). As before, expression of L858H channels depolarized the RMP and reduced action potential overshoot. However, when Na_V1.8 was coexpressed with L858H, action potential overshoot was restored, even though the depolarization of the RMP induced by L858H persisted. For the population of cells studied, the depolarization of the RMP by ≈5mV (–41.6 ± 0.76 mV, n = 17 with L858H) was maintained with coexpression of Na_V1.8 (–40.5 ± 1.01 mV, n = 17) (P > 0.05; figure 5B). However, current threshold for firing was reduced when Na_V1.8 was coexpressed with L858H (P < 0.05; figure 5C). In addition, action potential overshoot was restored when Na_V1.8 was coexpressed with L858H (P < 0.05; figure 5D). These results show that the presence or absence of Na_V1.8 is a major determinant of the functional effects of this mutation.

Figure 5 Coexpression of L858H and Na_V1.8 channels rescues electrogenic properties in SCG neurons. When Na_V1.8 was coexpressed with L858H, current threshold and action potential overshoot were restored, although the depolarization of the RMP induced by L858H persisted. (A) Suprathreshold action potentials recorded from representative SCG neurons transfected with WT (blue), L858H (red), and L858H plus Na_V1.8 (green) channels. (B) Depolarized RMP in cells with L858H channels (–41.6 ± 0.76 mV, n = 17) was maintained with coexpression of Na_V1.8 (–40.5 ± 1.01 mV, n = 17; P > 0.05). n.s., not significant. (C) Current threshold for action potential firing was reduced from 42.9 ± 6.3 pA (n = 17) for L858H to 26.8 ± 4.3 pA (n = 17) for L858H coexpressed with Na_V1.8 (*, P < 0.05). (D) Action potential overshoot in SCG neurons with L858H channel (23.8 ± 4.7 mV, n = 17) was increased when Na_V1.8 was coexpressed with L858H (41.5 ± 4.6 mV, n = 17; *, P < 0.05).

Discussion

The L858H erythermalgia mutation results in a single amino acid substitution in the domain II S4–S5 linker within Na_V1.7 (refs. 2, 7), a sodium channel that is preferentially expressed in DRG and sympathetic ganglia neurons.⁸^,⁹^,¹⁶^,²⁵ Our experiments show that L858H produces hyperexcitability (decreased threshold and enhanced repetitive firing) within DRG neurons and hypoexcitability (increased threshold and attenuated repetitive firing) within sympathetic ganglion neurons. The latter observation provides a molecular basis for the sympathetic dysfunction that has been reported ²⁶^,²⁷ in erythermalgia. Moreover, this observation suggests the more general hypothesis that other ion channel mutations may have differing physiological effects in different cell types in which the channel is normally expressed. Although opposing functional effects of the same mutation in different types of neurons may at first seem paradoxical, we demonstrate that it is caused by different cell backgrounds, including different repertoires of other ion channels within the two types of neurons.

The L858H mutation produces a hyperpolarizing shift in activation, slows deactivation, and increases the ramp response of Na_V1.7 to small stimuli.⁷ We observed that L858H produces a depolarization, of ≈5 mV, in the RMP of both DRG and SCG neurons. A similar depolarization has been observed with a mutation of the adjacent residue in the Na_V1.4 muscle sodium channel that is associated with muscle weakness.²⁸^,²⁹ The depolarization may be a result of increased window currents predicted from voltage-clamp analysis of L858H to be present between –80 and –35 mV and largest between –60 and –45 mV, close to resting potential⁷ because of the hyperpolarizing shift in activation, as for the Na_V1.4 mutation.²⁸^,²⁹

We hypothesized that L858H produces hyperexcitability in DRG neurons and hypoexcitability in SCG neurons because of the selective expression of Na_V1.8 sodium channels in DRG and its absence in SCG neurons. The majority of nociceptive DRG neurons express Na_V1.8 (refs. 14, 30), which contributes most of the current underlying the action potential upstroke.¹⁰^,¹⁸ Because it has depolarized voltage-dependence of activation [V_1/2 = –16 to –21 mV (refs. 14, 15, 23, 24)] and inactivation [V_1/2 = –30 mV (refs. 14, 15, 23, 24)] compared with other sodium channels, Na_V1.8 permits DRG neurons to generate action potentials and sustain repetitive firing when depolarized.¹⁰^,¹² This finding led us to predict that Na_V1.7 mutations can produce hyperexcitability of DRG neurons because these cells also express Na_V1.8 channels that are, because of depolarized inactivation voltage-dependence, still available for activation.

We observed a similar depolarization, of ≈5 mV, in SCG neurons after the expression of L858H, but in these neurons the mutation produced hypoexcitability, i.e., increased threshold, attenuated repetitive firing, and reduced action potential amplitude. Na_V1.3, Na_V1.6, and Na_V1.7, but not Na_V1.8, are present in SCG neurons (figure 4). Because of the relatively hyperpolarized steady-state inactivation of Na_V1.3, Na_V1.6, and Na_V1.7 in neurons [V_1/2 values between –65 and –78 mV (refs. 31, 32)], the L858H-induced depolarizing shift in the RMP would be expected to inactivate sodium channels within SCG neurons, with a resultant decrease in excitability and action potential amplitude. Consistently with this, when we held SCG neurons expressing the L858H mutation at potentials hyperpolarized compared with their RMP, excitability and action potential amplitude increased (figure 3E).

To test the hypothesis that the selective presence within DRG neurons of Na_V1.8, with its depolarized V_1/2 of activation and inactivation,¹⁴^,¹⁵^,²³^,²⁴ contributes to the opposing functional effects of L858H in DRG and SCG neurons, we expressed Na_V1.8 within SCG neurons, where it is not normally present. In agreement with this hypothesis, the coexpression of Na_V1.8 with L858H tended to restore action potential threshold and amplitude toward the values seen with WT Na_V1.7, protecting against the hypoexcitability conferred by L858H in the absence of Na_V1.8, even though L858H produced a depolarizing shift of ≈5 mV in the RMP. Although we cannot exclude an additional contribution of differential expression of other molecules, such as potassium channels, to the opposing changes in excitability in DRG and SCG neurons expressing L858H, our results show that Na_V1.8, which is selectively expressed within DRG neurons, is a major contributor to this effect.

Our results show that the same sodium channel mutation can produce hyperexcitability in one type of neuron in which the channel gene is normally expressed, while producing hypoexcitability in another type of neuron where the gene is also expressed. Thus, the effects of sodium channel mutations on neuronal function should not be considered as unidirectional or predictable on the basis of the changes in channel function per se: they also depend on the cell background in which the mutation is expressed. More generally, we suggest the possibility that mutations of other ion channels, e.g., other sodium channel isoforms or calcium or potassium channels, may have different functional effects in different types of cells.

Materials and Methods

SCG and DRG Cultures

As described in ref. 33, we isolated SCG from deeply anesthetized (ketamine/xylazine, 80:10 mg/kg, i.p.) 1- to 5-day-old Sprague–Dawley rats, washed them with cold Hanks’ balanced salt solution (HBSS) (Ca²⁺- and Mg²⁺-free), incubated them for 40 min at 37°C in HBSS (Ca²⁺- and Mg²⁺-free) containing 0.2% trypsin (Worthington), washed them twice in warm Leibovitz’s L-15 medium (Invitrogen), and triturated them with a fire-polished Pasteur pipette in Leibovitz’s L-15 medium containing 0.75 mg/ml BSA/trypsin inhibitor. SCG cells were pelleted by low-speed centrifugation, resuspended in modified Leibovitz’s L-15 medium supplemented with 1 µg/ml nerve growth factor (Alomone, Jerusalem), 5% rat serum, 38 mM glucose, 24 mM sodium bicarbonate, and penicillin/streptomycin (each 50 units/ml), plated on 12-mm circular coverslips precoated with poly(D-lysine)/laminin (BD Biosciences, Franklin Lakes, NJ) and incubated at 37°C in 5% CO². DRG neurons from age-matched animals were cultured as described in ref. 34.

Cell Culture Immunocytochemistry

We incubated cultured DRG and SCG neurons sequentially in complete saline solution, 4% paraformaldehyde in 0.14 M Sorensen’s buffer (pH 7.4) for 10 min, PBS, blocking solution (PBS containing 5% normal goat serum, 2% BSA, and 0.1% Triton X-100) for 30 min, and primary Ab [rabbit anti-Na_V1.7, 6 µg/ml, and rabbit anti-Na_V1.8, 3.2 µg/ml (Alomone)] in blocking solution overnight at 4°C. The next day, we incubated the coverslips sequentially in PBS, goat anti-rabbit IgG–Cy3 secondary Ab (0.5 µg/ml, Amersham Pharmacia), and PBS. Coverslips were mounted on slides with Aqua-Poly/Mount and examined with a Nikon E800 microscope equipped with epifluorescent optics by using METAVUE software (Universal Imaging, Downingtown, PA).

Tissue Immunocytochemistry

Adult Sprague–Dawley rats were deeply anesthetized (80 mg/kg ketamine/10 mg/kg xylazine, i.p.) and perfused with PBS and then 4% paraformaldehyde in 0.14 M Sorensen’s buffer. SCG and DRG were excised, rinsed with PBS, and cryoprotected in 30% sucrose in PBS overnight at 4°C (ref. 35). Cryosections (10 µm) of DRG and SCG were incubated sequentially in (i) PBS containing 5% normal goat serum, 2% BSA, and 0.1% Triton X-100 for 30 min; (ii) primary Ab [rabbit anti-Na_V1.7, 6 µg/ml (Alomone), and rabbit anti-Na_V1.8, 0.3 µg/ml (ref. 35)] overnight at 4°C; (iii) PBS, six times for 5 min each time; (iv) goat anti-rabbit IgG–biotin (1:250); (v) PBS, six times for 5 min each time; (vi) avidin–horseradish peroxidase (1:250); (vii) PBS, six times for 5 min each time; (viii) 0.4% diaminobenzidine and 0.003% hydrogen peroxide in PBS for 7 min; and (ix) PBS containing 0.02% sodium azide and coverslipped with Aqua-Poly/Mount.

Reverse Transcription–Multiplex PCR

We synthesized first-strand cDNA from total cellular RNA isolated from L4–L5 DRG and SCG dissected from adult Sprague–Dawley rats, using the RNeasy Mini Kit (Qiagen, Valencia, CA).¹⁹ We amplified fragments of sodium channel templates in the cDNA pool by multiplex PCR using four forward and three reverse primers (F1–F4 and R1–R3) designed against highly conserved sequences in domain 1 of a-subunits.¹⁹ We investigated the presence of specific sodium channel templates by digesting 1/20th of the volume of the multiplex amplicons in a 10-µl final volume with restriction enzymes that produce distinct restriction products, as described in ref. 19.

Transfection of DRG and SCG Neurons and Current-Clamp Recordings

We transfected DRG and SCG neurons as described in ref. 3, using Rat Neuron Nucleofector Solution (Amaxa, Gaithersburg, MD) and a channel/GFP ratio of 5:1, with WT tetrodotoxin-resistant Na_V1.7 (ref. 36) or L858H mutant derivative,⁷ and, in a cotransfection assay, we combined the L858H mutant channel and WT Na_V1.8 channel to transfect SCG neurons. Transfected DRG and SCG neurons were incubated in DMEM (Ca²⁺- and Mg²⁺-free) plus 10% FCS (for 5 min at 37°C) to allow recovery, diluted in regular culture medium (for DRG, DMEM/FCS; for SCG, modified Leibovitz’s L-15) supplemented with nerve growth factor and glial cell-derived neurotrophic factor (50 ng/ml), plated on precoated 12-mm circular coverslips, and incubated at 37°C in 5% CO².

Whole-cell current-clamp recordings were made from transfected small-diameter (≤25 µm) DRG or SCG neurons with robust GFP fluorescence, within 24–60 h, at room temperature (21–25°C) by using an Axopatch 200B amplifier (Axon Instruments, Union City, CA). The pipette solution contained 140 mM KCl, 0.5 mM EGTA, 5 mM Hepes, and 3 mM Mg·ATP (pH 7.3), adjusted to 315 mosM/liter with glucose. External solution contained 140 mM NaCl, 3 mM KCl, 2 mM MgCl², 2 mM CaCl², and 10 mM Hepes (pH 7.3), adjusted to 310 mosM/liter with glucose. Pipette potential was set to zero before seal formation without correction for liquid junction potentials. We canceled capacity transients before switching to current-clamp mode and compensated for series resistance (3–6 MΩ) by ≈70%. Traces were acquired from cells with stable RMP, excluding cells where RMP changed by >10%, by using CLAMPEX 8.1 software (Axon Instruments), filtered at 5 kHz, at a sampling rate of 20 kHz. Input resistances, measured by recording voltage changes evoked by injection of hyperpolarizing current, were not significantly different between groups. We measured action potential threshold at the beginning of the sharp upward rise of the action potential and determined current threshold by a series of depolarizing currents from 0 to ≈200 pA in 5-pA increments. We assessed repetitive firing by recording responses to sustained (950 ms) injection of depolarizing current.

Statistical Analysis

Data are presented as mean ± SEM. Data were analyzed by using CLAMPFIT 8.2 (Axon Instruments) and ORIGIN 6.1 (Microcal Software, Northampton, MA) software. Statistical significance was determined by using Student’s t test, where we assumed that the apparent Gaussian nature of the data sets would be extended to the population.

Notes

1 Waxman SG, Dib-Hajj S. 2005. Trends Mol Med 11: 555–562.
2 Yang Y, Wang Y, Li S, Xu Z, Li H, Ma L, Fan J, et al. 2004. J Med Genet 41: 171–174.
3 Dib-Hajj SD, Rush AM, Cummins TR, Hisama FM, Novella S, Tyrrell L, Marshall L, Waxman SG. 2005. Brain 128: 1847–1854.
4 Drenth JP, Te Morsche RH, Guillet G, Taieb A, Kirby RL, Jansen JB. 2005. J Invest Dermatol 124: 1333–1338.
5 Michiels JJ, Te Morsche RH, Jansen JB, Drenth JP. 2005. Arch Neurol (Chicago) 62: 1587–1590.
6 Han C, Rush A, Dib-Hajj S, Li S, Xu Z, Wang Y, Tyrrell L, Wang X, Yang Y, Waxman S. 2006. Ann Neurol 59: 553–558.
7 Cummins TR, Dib-Hajj SD, Waxman SG. 2004. J Neurosci 24: 8232–8236.
8 Toledo-Aral JJ, Moss BL, He ZJ, Koszowski AG, Whisenand T, Levinson SR, Wolf JJ, Silossantiago I, Halegoua S, Mandel G. 1997. Proc Natl Acad Sci USA 94: 1527–1532.
9 Sangameswaran L, Fish LM, Koch BD, Rabert DK, Delgado SG, Ilnicka M, Jakeman LB, et al. 1997. J Biol Chem 272: 14805–14809.
10 Renganathan M, Cummins TR, Waxman SG. 2001. J Neurophysiol 86: 629–640.
11 Waddell PJ, Lawson SN. 1990. Neuroscience 36: 811–822.
12 Blair NT, Bean BP. 2003. J Neurosci 23: 10338–10350.
13 Zhang XF, Zhu CZ, Thimmapaya R, Choi WS, Honore P, Scott VE, Kroeger PE, Sullivan JP, Faltynek CR, Gopalakrishnan M, Shieh CC. 2004. Brain Res 1009: 147–158.
14 Akopian AN, Sivilotti L, Wood JN. 1996. Nature 379: 257–262.
15 Sangameswaran L, Delgado SG, Fish LM, Koch BD, Jakeman LB, Stewart GR, Sze P, Hunter JC, Eglen RM, Herman RC. 1996. J Biol Chem 271: 5953–5956.
16 Black JA, Dib-Hajj S, McNabola K, Jeste S, Rizzo MA, Kocsis JD, Waxman SG. 1996. Brain Res Mol Brain Res 43: 117–131.
17 Cummins TR, Howe JR, Waxman SG. 1998. J Neurosci 18: 9607–9619.
18 Blair NT, Bean BP. 2002. J Neurosci 22: 10277–10290.
19 Dib-Hajj SD, Tyrrell L, Black JA, Waxman SG. 1998. Proc Natl Acad Sci USA 95: 8963–8968.
20 Freschi JE. 1983. J Neurophysiol 50: 1460–1478.
21 Nerbonne JM, Gurney AM. 1989. J Neurosci 9: 3272–3286.
22 Schofield GG, Ikeda SR. 1988. Pflugers Arch 411: 481–490.
23 Cummins TR, Waxman SG. 1997. J Neurosci 17: 3503–3514.
24 Sleeper AA, Cummins TR, Dib-Hajj SD, Hormuzdiar W, Tyrrell L, Waxman SG, Black JA. 2000. J Neurosci 20: 7279–7289.
25 Felts PA, Yokoyama S, Dib-Hajj S, Black JA, Waxman SG. 1997. Brain Res Mol Brain Res 45: 71–82.
26 Davis MD, Sandroni P, Rooke TW, Low PA. 2003. Arch Dermatol 139: 1337–1343.
27 Mork C, Kalgaard OM, Kvernebo K. 2002. J Invest Dermatol 118: 699–703.
28 Lehmann-Horn F, Kuther G, Ricker K, Grafe P, Ballanyi K, Rudel R. 1987. Muscle Nerve 10: 363–374.
29 Cummins TR, Zhou J, Sigworth FJ, Ukomadu C, Stephan M, Ptacek LJ, Agnew WS. 1993. Neuron 10: 667–678.
30 Djouhri L, Fang X, Okuse K, Wood JN, Berry CM, Lawson S. 2003. J Physiol 550: 739–752.
31 Cummins TR, Aglieco F, Renganathan M, Herzog RI, Dib-Hajj SD, Waxman SG. 2001. J Neurosci 21: 5952–5961.
32 Herzog RI, Cummins TR, Ghassemi F, Dib-Hajj SD, Waxman SG. 2003. J Physiol 551: 741–750.
33 Higgins D, Lein P, Osterhout D, Johnson M. 1991. In Culturing Nerve Cells, eds. Banker G., Goslin, K. (MIT Press, Cambridge, MA), pp. 177–205.
34 Rizzo MA, Kocsis JD, Waxman SG. 1994. J Neurophysiol 72: 2796–2815.
35 Black JA, Dib-Hajj S, Baker D, Newcombe J, Cuzner ML, Waxman SG. 2000. Proc Natl Acad Sci USA 97: 11598–11602.
36 Herzog RI, Cummins TR, Ghassemi F, Dib-Hajj SD, Waxman SG. 2003. J Physiol 551: 741–750.

A Single Sodium Channel Mutation Produces Hyper- or Hypoexcitability in Different Types of Neurons *