OUP user menu

Functional profiles of SCN9A variants in dorsal root ganglion neurons and superior cervical ganglion neurons correlate with autonomic symptoms in small fibre neuropathy

Chongyang Han, Janneke G. J. Hoeijmakers, Shujun Liu, Monique M. Gerrits, Rene H. M. te Morsche, Giuseppe Lauria, Sulayman D. Dib-Hajj, Joost P. H. Drenth, Catharina G. Faber, Ingemar S. J. Merkies, Stephen G. Waxman
DOI: http://dx.doi.org/10.1093/brain/aws187 2613-2628 First published online: 22 July 2012

Summary

Patients with small fibre neuropathy typically manifest pain in distal extremities and severe autonomic dysfunction. However, occasionally patients present with minimal autonomic symptoms. The basis for this phenotypic difference is not understood. Sodium channel Nav1.7, encoded by the SCN9A gene, is preferentially expressed in the peripheral nervous system within sensory dorsal root ganglion and sympathetic ganglion neurons and their small diameter peripheral axons. We recently reported missense substitutions in SCN9A that encode functional Nav1.7 variants in 28% of patients with biopsy-confirmed small fibre neuropathy. Two patients with biopsy-confirmed small fibre neuropathy manifested minimal autonomic dysfunction unlike the other six patients in this series, and both of these patients carry the Nav1.7/R185H variant, presenting the opportunity to compare variants associated with extreme ends of a spectrum from minimal to severe autonomic dysfunction. Herein, we show by voltage-clamp that R185H variant channels enhance resurgent currents within dorsal root ganglion neurons and show by current-clamp that R185H renders dorsal root ganglion neurons hyperexcitable. We also show that in contrast, R185H variant channels do not produce detectable changes when studied by voltage-clamp within sympathetic neurons of the superior cervical ganglion, and have no effect on the excitability of these cells. As a comparator, we studied the Nav1.7 variant I739V, identified in three patients with small fibre neuropathy characterized by severe autonomic dysfunction as well as neuropathic pain, and show that this variant impairs channel slow inactivation within both dorsal root ganglion and superior cervical ganglion neurons, and renders dorsal root ganglion neurons hyperexcitable and superior cervical ganglion neurons hypoexcitable. Thus, we show that R185H, from patients with minimal autonomic dysfunction, does not produce detectable changes in the properties of sympathetic ganglion neurons, while I739V, from patients with severe autonomic dysfunction, has a profound effect on excitability of sympathetic ganglion neurons.

  • sodium channel
  • pain
  • voltage-clamp
  • current-clamp

Introduction

Small fibre neuropathy, a disorder typically manifested as neuropathic pain in distal extremities and profound autonomic symptoms, is characterized by degeneration of epidermal nerve endings of thinly myelinated and unmyelinated nerve fibres (Stewart et al., 1992; Gorson and Ropper, 1995; Holland et al., 1998; Lacomis, 2002; Lauria, 2005; Devigili et al., 2008). Small fibre neuropathy has been reported to be idiopathic, i.e. does not have an identifiable cause, in 24–93% of patients in various series (Lacomis, 2002; Devigili et al., 2008; Bednarik et al., 2009). Recently, we reported variants of voltage-gated sodium channel Nav1.7, encoded by the SCN9A gene, in 28% of patients with biopsy-confirmed idiopathic small fibre neuropathy (Faber et al., 2012).

Voltage-gated sodium channel Nav1.7, encoded by the SCN9A gene, is abundantly present in dorsal root ganglion (DRG) and sympathetic ganglion neurons (Toledo-Aral et al., 1997; Rush et al., 2006) and their axons (Persson et al., 2010). Nav1.7 opens in response to small depolarizations close to resting potential (Cummins et al., 1998). Gain-of-function mutations in Nav1.7 that enhance activation and impair fast inactivation cause the heritable painful disorders inherited erythromelalgia (Dib-Hajj et al., 2005, 2010) and paroxysmal extreme pain disorder (Fertleman et al., 2006). The channel variants from patients with Nav1.7 channelopathy-associated small fibre neuropathy that have been investigated functionally thus far show gain-of-function properties, impaired fast-inactivation, impaired slow-inactivation or enhanced resurgent current and render DRG neurons hyperexcitable, a change that is thought to underlie the neuropathic pain reported by patients carrying these variants (Faber et al., 2012; Han et al., 2012).

Despite the debilitating nature of autonomic symptoms in small fibre neuropathy, the molecular basis for autonomic dysfunction in small fibre neuropathy is not well understood. Most patients with Nav1.7 channelopathy-associated small fibre neuropathy experience profound autonomic dysfunction in addition to neuropathic pain (Faber et al., 2012). However, small fibre neuropathy can sometimes present a picture of pain and sensory dysfunction with only minimal autonomic dysfunction. Two of the eight patients described in the previous paper by Faber et al. (2012) reported minimal autonomic symptomatology, and both carried the R185H variant Nav1.7 channel. Although all the variants studied thus far have produced hyperexcitability within DRG neurons, consistent with the pain phenotype in these patients, the effects of these variants on sympathetic neurons, which might correlate with autonomic dysfunction, have not been investigated. Reasoning that R185H channels might affect DRG and sympathetic neurons in different ways, we assessed the biophysical effects of this variant by voltage-clamp in both DRG and sympathetic ganglion neurons [superior cervical ganglia (SCG)] and examined the functional effects of this channel on firing properties of neurons from DRG and SCG. As a comparator, we assessed Nav1.7/I739V, the variant channel associated with the most severe autonomic symptoms in our series (Faber et al., 2012).

Materials and methods

Patients

The patients, three of whom were evaluated as part of a series reported by Faber et al. (2012), and two additional patients evaluated subsequently at Maastricht University Medical Centre neurological clinic, were diagnosed with small fibre neuropathy. This study was approved by medical ethics committees at Yale University and Maastricht University Medical Centre. All aspects of the study were explained and written informed consent obtained prior to study initiation.

Clinical characterization

Skin biopsy

Punch biopsy (10 cm above lateral malleolus) specimens were fixed (2% paraformaldehyde–lysine–sodium periodate, 4°C), cryoprotected and stored (−80°C) in 20% glycerol before sectioning (50 μm). Numbers of individual nerve fibres crossing dermal–epidermal junctions were analysed in each of three sections, immunostained with polyclonal rabbit anti-protein gene product 9.5 antibody (PGP9.5; Ultraclone), by bright-field microscopy (Olympus BX50 stereology workstation, PlanApo oil objective 40×/numerical aperture = 1.0). Linear quantification of intraepidermal nerve fibre density (IENF/mm) was compared with age- and gender-adjusted normative values (Lauria et al., 2010).

Quantitative sensory testing

Quantitative sensory testing was performed according to previous guidelines (Shy et al., 2003) with a TSA-2001 instrument (Medoc). Thresholds were assessed at the dorsum of both feet and thenar eminences, using ascending/descending (warm/cool) thermal ramp stimuli. Heat pain modality was also examined. Results were compared with normative values (Yarnitsky and Sprecher, 1994), and were considered abnormal for Z-values > 2.5. Sensory modalities were classified as abnormal if results of both method-of-limits and method-of-levels were abnormal (Hoitsma et al., 2003).

SCN9A sequence analysis

Genomic DNA isolation and amplification/analysis of SCN9A-coding exons and flanking intronic sequences, and exons encoding 5′- and 3′-untranslated sequences within complementary DNA, were described previously (Drenth et al., 2005). Exon sequences were compared with reference Nav1.7 complementary DNA (NM 002977.3) (Klugbauer et al., 1995) to identify variations using Alamut mutation interpretation software (Interactive-Biosoftware). DNA from 1000 Caucasian control subjects (2000 chromosomes) from the same geographical region as the index patient was analysed for c.554G>A and c.2215A>G substitutions by PCR and high-resolution melting curve analysis. Primer pairs 3′-ATGGTGGTTGTATTCTTTTCA-5′/3′-AACACTGTGCTGCCTGAG-5′ and 3′-TTTTATTGTAATGGATCCTTTTGTAG-5′/3′-ATTGGGTGGTGTTCCATAGC-5′, and fluorescent dye EvaGreen® (Biotium) were used to amplify the variant exons; presence of variant alleles was determined from melting curves (65°C to 95°C; ramp rate 0.1°C/10 s; CFX96™ Real-Time PCR Detection System), analysed using Precision Melt Analysis™ software (Bio-Rad). We interrogated two online databases (the 1000 genome project; www.1000genomes.org) and the Exome Variant Database (http://evs.gs.washington.edu/EVS) for the presence of detected variants in control populations.

Functional analysis

Transfection of dorsal root ganglion and superior cervical ganglion neurons and voltage-clamp recordings

All protocols for care and sacrifice of animals were approved by the Veterans Administration Connecticut IACUC. Since initial voltage-clamp assessment did not reveal biophysical changes in properties of R185H channels expressed within HEK293 cells, voltage-clamp analysis was performed after biolistic (Helios® Gene Gun, Bio-Rad) transfection of DRG cultured from 4 - to 8-week-old Nav1.8-null mice (Stirling et al., 2005) as described previously (Dib-Hajj et al., 2009). Wild-type or R185H plasmids were mixed with green fluorescent protein (GFP) plasmid DNA (channel:GFP ratio 3:1) using 1.0 µm gold particles, discharged into cells at a pressure of ∼120 psi (∼827 kPa). Absence of Nav1.8 tetrodotoxin-resistant current in these DRG neurons facilitated analysis of wild-type and R185H Nav1.7 channels, which were rendered tetrodotoxin resistant (Herzog et al., 2003).

Adult Nav1.8-null mice or rats did not yield SCG neurons that could be reproducibly transfected with sodium channel constructs. Therefore, we used rat pups as the source for SCG neurons for transfection. SCG were harvested from 0- to 5-day-old Sprague–Dawley rats, and neurons were isolated as previously described (Rush et al., 2006). SCG neurons were maintained under standard tissue culture conditions for ∼8 h before biolistic transfections. Tetrodotoxin resistant versions of wild-type, R185H or I739V plasmids were mixed with GFP plasmid DNA (channel:GFP ratio 3:1) using 1.0 µm gold particles, discharged into cells at a pressure of ∼240 psi (∼1654 kPa). Since tetrodotoxin resistant sodium currents are absent in SCG neurons, the tetrodotoxin resistant (Herzog et al., 2003) versions of Nav1.7 wild-type, R185H or I739V facilitated the functional analysis of these channels.

Whole-cell patch-clamp recordings in DRG or SCG neurons were obtained in voltage-clamp mode at room temperature (∼21°C), 40–48 h after transfection using 0.8–1.5 MΩ electrodes and EPC-9 amplifier and Pulse 8.5 (HEKA). Cells transfected with wild-type or variant Nav1.7 channels were selected using identical criteria, which required that cells be GFP-fluorescent, isolated, round with 20–30 µm diameter and one or fewer neurites (to reduce space-clamp artefact), smooth membranes, peak current >1 nA. Voltage-errors were minimized using 80–90% series resistance-compensation and linear leak subtraction; capacitance artefact was cancelled using computer-controlled circuitry. The pipette solution contained (mM): 140 CsF/1 EGTA/10 NaCl/10 HEPES/pH 7.3 with CsOH (adjusted to 315 mmol/L with dextrose). The bath solution was (mM) 140 NaCl/3 KCl/1 MgCl2/1 CaCl2/20 TEA-Cl/5 CsCl/0.1 CdCl2/0.0003 tetrodotoxin/10 HEPES, pH 7.3 with NaOH (adjusted to 320 mmol/L with dextrose).

Recordings were started 5 min after establishing whole-cell configuration. DRG or SCG neurons were held at −100 mV and stepped to a range of potentials (−80 to +60 mV in 5 mV increments) for 100 ms. Peak inward currents (I) were plotted as a function of depolarization potential to generate I–V curves. Activation curves were obtained by converting I to conductance (G) at each voltage (V) using the equation G = I/(V − Vrev), where Vrev is the reversal potential which was determined for each cell individually. Activation curves were then fit with Boltzmann functions in the form of G = Gmax/{1 + exp[(V1/2,act − V)/k]}, where Gmax is the maximal sodium conductance, V1/2,act is the potential at which activation is half-maximal, V is the test potential and k is the slope factor. Steady-state fast-inactivation was examined using a series of 500 ms prepulses from −150 to 0 mV followed by test pulses to −10 mV. Steady-state slow-inactivation was assessed using a 20 ms pulse to −10 mV after a 30 s prepulse to potentials ranging from −130 to 10 mV followed by a 100 ms pulse to −120 mV to remove fast-inactivation. Peak inward currents obtained from steady-state fast inactivation and slow inactivation protocols were normalized to the maximal peak current (Imax) and fit with Boltzmann functions: Embedded Image where V represents the inactivating prepulse potential, V1/2,inact represents the midpoint of the inactivation curve and Rin is the fraction of channels that are resistant to inactivation. To assess deactivation, cells were held at −100 mV and tail currents generated by 0.5 ms depolarization to −20 mV followed by a series of repolarizations (−100 to −40 mV). Deactivation kinetics was measured using single exponential fits to estimate current decay time at potentials from −100 to −40 mV.

Resurgent currents were assessed with a two-step protocol that initially depolarized the membrane to +30 mV for 20 ms before testing for resurgent sodium currents by hyperpolarizing the membrane potential in −5 mV increments from 0 to −80 mV for 100 ms, then returning to the holding potential of −100 mV. Recordings for resurgent current reveal two components: transient peak current evoked by the 30 mV depolarization and the following resurgent current by repolarization. The peak current may display a small outward current at the end of 20 ms recording, which is due to leakage. The relative amplitude of the resurgent current was calculated as a percentage of the peak current evoked by a +30 mV depolarization. Quantification and analysis of resurgent sodium current were carried out as described previously (Jarecki et al., 2010). Under our recording conditions, we were able to reliably detect resurgent currents with amplitude >100 pA and could detect resurgent currents as small as 50 pA if they displayed typical waveform in cells with low noise.

Transfection of dorsal root ganglion and superior cervical ganglion neurons: current-clamp recordings

SCG neurons were isolated and cultured from 0 - to 5-day-old Sprague–Dawley rats as described previously (Rush et al., 2006). DRG neurons from animals of the same age were cultured using a protocol as described by Dib-hajj et al. (2009). Wild-type Nav1.7 R, R185H, or I739V variant channels and GFP constructs (channel:GFP ratio 5:1) were electroporated into DRG or SCG neurons using Rat Neuron Nucleofector® Solution (Lonza) as described previously (Dib-Hajj et al., 2005). Whole-cell current-clamp recordings were obtained from transfected small diameter (<25 µm) DRG or SCG neurons with robust GFP fluorescence, within 40–48 h by using an EPC-9 amplifier and Pulse 8.5 (HEKA). Electrodes (1–3 MΩ) were filled with pipette solution (mM): 140 KCl/0.5 EGTA/5 HEPES/3 Mg-ATP/pH 7.3 with KOH (adjusted to 315 mmol/L with dextrose). The extracellular solution contained (mM): 140 NaCl/3 KCl/2 MgCl2/2 CaCl2/10 HEPES, pH 7.3 with NaOH (adjusted to 320 mmol/L with dextrose). Whole-cell configuration was obtained in voltage-clamp mode before proceeding to current-clamp. Cells with stable resting membrane potentials were used for data collection by PulseFit 8.74 (HEKA Electronics) software. Threshold was determined by the first action potential elicited by a series of depolarizing current injections (5 pA increments). Responses to sustained depolarization were assessed using 500 ms depolarization (25 pA increments for DRG or 10 pA for SCG).

Data analysis

Electrophysiological data were analysed using PulseFit 8.74 (HEKA Electronics) and Origin 8.5.1 (Microcal), and were presented as means ± standard error. Statistical significance was determined by unpaired Student’s t-test or two proportions Z-test (comparison of proportion of cells producing resurgent currents). For multi-group statistical analysis, we used one-way ANOVA followed by Tukey post hoc test or Kruskal–Wallis followed by Dunn procedure depending on whether data showed a normal distribution.

Results

Clinical description

Patients with variant c.554G>A; p.R185H

Patient 1

A 54-year-old male was referred to the neurology outpatient clinic because of unpleasant paraesthesias in his feet and hands. The paraesthesias started in his feet at the age of 24 years and had a gradual progressive course. At age 52 years, the patient developed burning pain in his feet, especially in the morning. He also complained of ‘electric current’ in his soles and reddening of his feet during exercise. These complaints interfered with walking. There were no other autonomic symptoms. Short-term treatment with pregabalin and amitriptyline was ineffective. Similar complaints were reported in a brother. The patient’s grandfather had difficulty in walking at advanced age. He was reported to have diminished heat sensation in his hands, and experienced painless burns.

Neurological examination was unremarkable with no signs of large fibre involvement. Extensive laboratory investigations and nerve conduction studies revealed no abnormalities, whereas quantitative sensory testing showed abnormal temperature thresholds for warmth and cold sensation. Skin biopsy showed markedly reduced intraepidermal nerve fibre density (1.0/mm), compared with age- and gender-specific normative values (≥3.5/mm). The patient was diagnosed as having small fibre neuropathy and DNA analysis showed a variant in Nav1.7: c.554G>A; p.R185H.

Patient 2

This 24-year-old female presented with severe pain in both feet. Her symptoms began at the age of 23 years with tingling in both feet, and gradually increased to involve the legs and hands, and became severe. Pain was described as maximal during rest, especially in the evening and during the nights, leading to a sleep disturbance. Drinking alcohol was reported to aggravate the pain. This patient reported only minor autonomic symptoms of occasional dizziness and dry mouth. Medical history was unremarkable, except for migraine since the age of 12 years. The family history was negative. Treatment with acetaminophen, anti-depressants, anti-convulsants, mexiletine and opioids did not provide relief.

Neurological examination demonstrated severe allodynia at the feet. There were no skin colour abnormalities. No signs of large fibre involvement were found. Extensive laboratory investigations showed no abnormalities. Nerve conduction studies were normal. Quantitative sensory testing showed abnormal temperature thresholds for warmth and cold sensation. Intraepidermal nerve fibre density was reduced (4.9/mm), compared to age- and gender-matched controls (≥8.4/mm). The patient was diagnosed as having small fibre neuropathy and DNA analysis showed a variant in Nav1.7: c.554G>A; p.R185H.

Patients with variant c.2215A>G, p.I739V

Patient 3

This 51-year-old female, previously described in detail by Han et al. (2012), presented with a 37-year history of episodic burning pain, flushing and itching of the face, lower legs and feet, triggered by exercise or rising temperature, warm water or lying under a blanket and relieved by cooling. The patient had severe dysautonomic symptoms, which included dry mouth and eyes, blurred vision, orthostatic dizziness, alternating constipation and diarrhoea, hyperhydrosis, heart rhythm palpitations and episodic swallowing difficulties. Treatment with amitriptyline did not provide relief. Physical examination demonstrated patchy allodynia at the feet and lower legs in a symmetrical pattern, and purple coloured skin of the hands. Family history revealed similar symptoms in two sons and a sister, who did not consent to DNA analysis. DNA was obtained from an unaffected brother, and did not carry the Nav1.7 variant found in the patient.

Blood and chest X-ray examinations were normal. Normal nerve conduction studies excluded large fibre involvement. Temperature thresholds were abnormal as measured by quantitative sensory testing (Han et al., 2012). A skin biopsy revealed that intraepidermal nerve fibre density in the lower leg was 3.4/mm, which was lower than the corresponding reported normative values (4.3/mm). The patient was diagnosed as having small fibre neuropathy, and DNA analysis showed a variant in Nav1.7: c.2215A>G; p.I739V.

Patient 4

This 66-year-old female with an unremarkable family history presented with sensory disturbances and severe autonomic dysfunction leading to disability. Her complaints began at the age of 64 years with numbness of the toes of the right foot. Six months later the numbness expanded to both feet and lower legs, making it difficult to maintain balance and the hands became affected. The patient experienced penetrating, stabbing pain, cramps and a feeling of ‘electric shocks’ in the feet, aggravated by exercise, walking or cold temperature. Resting or an electric heating blanket gave some relief. It became impossible to move without aids, due to excruciating pain.

The patient also suffers from severe dysautonomic symptoms (cardiac dysrhythmias, blood pressure fluctuations, diarrhoea, urge incontinence, abdominal discomfort, hot flashes and hyperhidrosis). Because of arrhythmias during the night, the patient used continuous positive airway pressure. The patient is totally disabled and homebound due to these symptoms.

Carbamazepine, duloxetine and tramadol did not provide any relief, and were subsequently stopped. Pregabalin caused unacceptable side effects. Gabapentin, morphine and transcutaneous electrical nerve stimulation provided partial pain relief.

Physical examination showed allodynia and numbness in a stocking-glove distribution, with normal muscle strength and tendon reflexes. Laboratory tests, nerve conduction studies and a chest X-ray revealed no abnormalities. Quantitative sensory testing showed abnormal temperature thresholds for warmth and cold sensation. Skin biopsy showed an intraepidermal nerve fibre density of 2.3/mm, which is reduced compared to normative values (fifth percentile: 3.2/mm). The patient was diagnosed as having small fibre neuropathy, and DNA analysis showed a variant in Nav1.7: c.2215A>G; p.I739V.

Patient 5

This 72-year-old female was referred to the neurological outpatient clinic because of a 5-year history of orthostatic intolerance, numbness and tingling of the feet and severe pain of the feet and ankles. Orthostatic dizziness, worst in the morning, interfered with getting out of bed or standing from a chair. She also complained of severe palpitations, for which the cardiologist could find no other explanation, and episodes of swelling and red discoloration of the feet that interfered with walking. She described the pain as burning, and noted that cooling of the feet or morphine, prescribed by a physician, tended to relieve it.

Physical examination showed erythema and pitting oedema of both feet and ankles. Light touch sense and vibration sense were abnormal up to the knees. Both Achilles tendon reflexes were absent. Nerve conduction studies revealed signs of a severe axonal sensorimotor polyneuropathy. Quantitative sensory testing showed abnormal temperature thresholds for cold sensation in the right foot, despite being incomplete due to severe pain during examination. Skin biopsy showed an intraepidermal nerve fibre density of 4.8/mm, which is normal compared to normative values (fifth percentile: 2.2/mm). The patient was diagnosed as having peripheral polyneuropathy with small and large fibre involvement, and DNA analysis revealed a variant in Nav1.7: c.2215A>G; p.I739V.

Molecular genetic analysis

Sequence analysis of SCN9A-coding exons from Patients 1 and 2 demonstrated a G to A substitution (c.554 G>A) in both patients. This variant substitutes arginine 185 with histidine (R185H) in the linker between DI/S2 and DI/S3. R185 is highly conserved in all human voltage-gated sodium channels (Fig. 1A). The c.554 G>A substitution (rs73969684) has been reported as a single nucleotide polymorphism with allele frequency of 0.6% (heterozygote frequency of 1.2%) in the 1000 Genomes Project; and with 1.2% heterozygote frequency (55 heterozygotes among 4700 individuals) reported in the Exome Variant Database (http://evs.gs.washington.edu/EVS). Screening a panel of 1000 ethnically matched (Dutch nationals of European ancestry) control population, we report the c.554 G>A in 0.4% of 1000 control subjects (0.4% heterozygote frequency, 0.2% allele frequency).

Figure 1

Schematic of a voltage-gated sodium channel showing the locations of the R185H (open circle) and I739V (solid circle) substitutions and the aligned sequences for the linker between Nav1 DIS2 and DIS3 (A) and for the relevant part of Nav1 DIIS1 (B). R185 (bold) and I739 (bold) are both conserved in all known human voltage-gated sodium channels.

As reported previously (Han et al., 2012), sequencing of SCN9A-coding exons from Patient 3 revealed the c.2215A>G, Nav1.7/I739V substitution. This substitution was also found in Patients 4 and 5. The I739 residue, located within the first transmembrane segment in domain II, is conserved in all human voltage-gated sodium channels (Fig. 1B), and in all mammalian Nav1.7 orthologues reported to date (Han et al., 2012). Sequencing of DNA from an asymptomatic brother of Patient 3 did not reveal any SCN9A variants. The c.2215A>G substitution in exon 13 of SCN9A has not been reported as a variant in the 1000 Genome Database (release 9 September 2011), but was reported with 0.5% frequency in the general population (25 heterozygotes among 4255 individuals) in the Exome Variant Database, and was found in 14 (1.4%) of 1000 Dutch control subjects (all heterozygous; 0.7% of 2000 chromosomes).

R185H enhances resurgent currents and increases excitability in dorsal root ganglion neurons; I739V impairs slow-inactivation and increases excitability in ganglion neurons

Voltage-clamp analysis

Figure 2A shows representative Nav1.7 sodium currents recorded from DRG neurons (from Nav1.8 null mice) expressing wild-type channels, and Fig. 2B from neurons expressing R185H variant channels. Peak current densities were not significantly different (wild-type: 523 ± 53 pA/pF, n = 48; R185H: 543 ± 59 pA/pF, n = 46). As Fig. 2C shows, the voltage-dependence of channel activation was not significantly different between wild-type and R185H. The activation midpoints determined from fitting the data with a Boltzmann function were: −20.8 ± 1.0 mV (n = 17) for wild-type and −22.5 ± 1.2 mV (n = 18) for R185H. The midpoints of fast-inactivation (from Boltzmann function fits of data) were not significantly different between wild-type (−71.3 ± 1.8 mV, n = 15) and R185H channels (−72.1 ± 1.4 mV, n = 16) (Fig. 2D). Kinetics of deactivation were estimated from measurements of current decay at potentials from −100 to −40 mV after briefly activating the channels at −20 mV for 0.5 ms. Figure 2E shows the rates of current decay for wild-type and R185H variant channels. The deactivation rates of R185H channels were not significantly different from those of wild-type channels across all voltages tested. The voltage-dependence of slow-inactivation (Fig. 2F) was not significantly different between R185H and wild-type channels. When fitted with a Boltzmann function, the midpoints of slow-inactivation curves were −54.5 ± 2.1 mV (n = 13) for wild-type and −53.6 ± 1.9 mV (n = 14) for R185H channels, and the component of non-inactivating channels measured at 10 mV was not significantly different between wild-type and R185H channels (wild-type: 15.6 ± 1.9%, n = 13; R185H: 15.3 ± 1.4%, n = 14).

Figure 2

Voltage-clamp analysis of wild-type (WT) and R185H channels in DRG neurons. Representative current traces recorded from DRG neurons from Nav1.8-null mice expressing wild-type (A) or R185H (B), evoked by voltage steps (100 ms) from −80 to 60 mV in 5-mV increments, from a holding potential of −100 mV. (C) Normalized peak current–voltage relationship curves for wild-type and R185H channels. (D) Comparison of voltage-dependent activation and steady-state fast-inactivation for wild-type and R185H channels. R185H does not alter activation or fast-inactivation. (E) Comparison of time constants for deactivation. R185H displays similar deactivation rates as wild-type channels. (F) Comparison of steady-state slow-inactivation curves between wild-type and R185H variant channels. R185H does not alter steady-state slow-inactivation. (G and H) Representative resurgent current traces recorded from DRG neurons expressing wild-type (G) or R185H (H) channels. Inset bar graph showing a higher percentage of DRG neurons expressing R185H (9 of 28 cells, 32%) (red) compared to cells expressing wild-type channels (3 of 29 cells, 10%) (grey) produce resurgent currents. *P < 0.05.

To assess resurgent currents, DRG neurons were depolarized to +30 mV for 20 ms from a holding potential of −100 mV followed by series of hyperpolarizations from 0 mV to −80 mV for 100 ms. Figure 2G and H show representative resurgent currents recorded from DRG neurons expressing wild-type or R185H channels, respectively. As described previously (Jarecki et al., 2010), not all small DRG neurons produce resurgent current. In this study, we found that the percentage of DRG neurons that generate resurgent current was significantly higher for cells expressing R185H (9 out 28 cells, 32%, P < 0.05) than for DRG neurons expressing wild-type channels (3 out 29 cells, 10%) (Fig. 2H, inset). The amplitudes of resurgent current were 5.3% ± 0.6% (n = 3) and 6.6% ± 1.3% (n = 9) for DRG neurons expressing wild-type or R185H channels, respectively.

As a comparator we assessed the I739V variant, which was identified in our previous series (Faber et al., 2012) in the patient with the most severe autonomic dysfunction and was subsequently found in two additional patients with small fibre neuropathy that included disabling autonomic symptoms (Patients 3, 4 and 5). As reported previously (Han et al., 2012), the I739V variant impaired slow-inactivation, depolarizing the slow-inactivation midpoint by 5.6 mV, and significantly increased the non-inactivating component at 10 mV in DRG neurons. I739V did not alter activation, steady-state fast-inactivation, kinetics of deactivation or resurgent currents in DRG neurons.

Current-clamp analysis: both R185H and I739V render dorsal root ganglion neurons hyperexcitable

Patients 1 and 2, who carried the R185H variant, experienced pain, displayed abnormal temperature thresholds and exhibited reduced intraepidermal nerve fibre density, all symptoms of DRG neuron dysfunction. We therefore assessed the effect of the R185H channels on DRG neuron excitability, using current-clamp recording after expressing wild-type or R185H variant channels in small DRG neurons from rat pups. Patients 3, 4 and 5, who carried the I739V variant, also experienced pain, displayed abnormal temperature thresholds and exhibited reduced intraepidermal nerve fibre density. Therefore, I739V channels were also transfected into DRG neurons (0–5 days old) as a comparator for assessment by current-clamp recording (Fig. 3).

Figure 3

R185H and I739V both render DRG neurons hyperexcitable. (A) Representative traces from a DRG neuron (0 - to 5-day-old rat) expressing wild-type (WT) channels, showing sub-threshold response to 95 pA current injection and subsequent action potentials evoked by injections of 100 pA (current threshold for this neuron) and 115 pA. (B) Representative traces from a DRG neuron expressing R185H variant channels, showing a lower current threshold (65 pA for this neuron) for action potential generation. (C) Representative traces from a DRG neuron expressing I739V variant channels, showing a lower current threshold (55 pA for this neuron) for action potential generation. (D) R185H does not alter resting membrane potential (RMP) of DRG neurons, but I739V significantly depolarizes RMP by 4.7 mV. ***P < 0.001 versus wild-type. (E) Comparison of current threshold among DRG neurons expressing wild-type, R185H or I739V channels. Current thresholds of DRG neurons are significantly reduced after expression R185H or I739V variant channels. ***P < 0.001 versus wild-type. (F) Comparison of action potential amplitude among DRG neurons expressing wild-type, R185H and I739V channels. R185H and I739V do not have an effect on action potential amplitude of DRG neurons.

Input resistance was not significantly different among DRG neurons expressing wild-type channels (1.0 ± 0.1 GΩ, n = 35), R185H variant channels (1.1 ± 0.1 GΩ, n = 38) and I739V variant channels (1.1 ± 0.1 GΩ, n = 30). The resting membrane potential of DRG neurons expressing R185H variant channels (−55.3 ± 0.7 mV, n = 38) was not significantly different from that of DRG neurons expressing wild-type channels (−56.9 ± 0.8 mV, n = 35), whereas I739V variant channels depolarized resting membrane potential of DRG neurons by 4.7 mV (−52.2 ± 0.6 mV, n = 30, P < 0.001) (Fig. 3D).

To assess the effect of variant channels on current threshold of DRG neurons, we injected series of depolarization currents to determine the current threshold for producing the first all-or-none action potential. DRG neurons expressing R185H and DRG neurons expressing I739V variant channels both displayed significantly reduced current threshold compared to DRG neurons expressing wild-type channels. Figure 3A shows traces from a representative DRG neuron expressing wild-type channels. For this neuron, the current threshold to generate first action potential was 100 pA, and a 95 pA current injection only evoked a small, graded membrane potential depolarization. Figure 3B shows recordings from a representative DRG neuron expressing R185H channels, in which a lower current injection of 65 pA could produce an overshoot action potential. Figure 3C shows recordings from a representative DRG neuron expressing I739V channels, in which a 55 pA current input could elicit an action potential. As indicated in Fig. 3E, average current threshold was significantly decreased for DRG neurons expressing R185H variant channels (63 ± 7 pA, n = 38, P < 0.001) and I739V variant channels (61 ± 7 pA, n = 30, P < 0.001) compared with DRG neurons expressing wild-type channels (111 ± 11 pA, n = 35). In contrast, action potential amplitude was not significantly different among DRG neurons expressing wild-type (114.8 ± 1.6 mV, n = 35), R185H (114.8 ± 1.5 mV, n = 38) and I739V channels (111.7 ± 1.8 mV, n = 30) (Fig. 3F).

We next investigated the responses of DRG neurons expressing wild-type, R185H and I739V channels to series of 500 ms sustained depolarizations. Compared to DRG neurons expressing wild-type channels, both DRG neurons expressing R185H channels and DRG neurons expressing I739V channels generated more action potentials in response to sustained depolarizing stimuli. Figure 4A shows the responses from three representative neurons which expressed wild-type, R185H or I739V channels, respectively, evoked by 500 ms current steps at one, two and three times the current threshold. DRG neurons expressing wild-type channels produced only two or three action potentials in response to current inputs at two or three times the current threshold, whereas DRG neurons expressing R185H channels and DRG neurons expressing I739V channels both produced more action potentials. As summarized in Fig. 4B, DRG neurons expressing R185H and I739V channels generated significantly more action potentials, compared with DRG neurons expressing wild-type channels, at all stimulus intensities >75 pA.

Figure 4

R185H and I739V both increase repetitive firing in DRG neurons. (A) Representative responses of DRG neurons (0 - to 5-day-old rat) expressing wild-type (WT), R185H and I739V channels, respectively, to 500 ms depolarizing current steps that are one, two and three times (top, middle and bottom traces, respectively) the current threshold for action potential generation. (B) Comparison of mean action potential (AP) spike numbers among DRG neurons expressing wild-type, R185H and I739V channels across a range of 500 ms step current injections from 25 to 500 pA. *P < 0.05 versus wild-type. R185H and I739V variant channels both increase spike numbers compared to wild-type.

R185H does not alter biophysical properties of Nav1.7, while I739V impairs slow-inactivation in superior cervical ganglion neurons

Sodium channel Nav1.7 is known to be expressed in both DRG and SCG neurons (Toledo-Aral et al., 1997; Rush et al., 2006). The two patients carrying the R185H variant displayed minimal autonomic symptoms. In contrast, Patients 3, 4 and 5, who carried the I739V variant, displayed severe autonomic dysfunction. We therefore asked whether the variant R185H or I739V channels change the biophysical properties of Nav1.7 within sympathetic ganglion neurons. To address these questions we transfected wild-type, R185H or I739V channels into SCG neurons of rat pups by biolistic transfections and performed voltage-clamp analysis.

Figure 5A–C shows representative Nav1.7 sodium currents recorded from small SCG neurons (diameter ≤25 µm) expressing wild-type channels, R185H variant, or I739V variant channels, respectively. Peak current densities were not significantly different among cells transfected with wild-type (708 ± 115 pA/pF, n = 36), R185H (694 ± 110 pA/pF, n = 26) and I739V channels (650 ± 90 pA/pF, n = 34). As the I–V curves in Fig. 5D demonstrate, the voltage-dependence of activation was not significantly different among wild-type, R185H and I739V. The activation midpoints determined from fitting the data with a Boltzmann function were −23.7 ± 1.6 mV (n = 22) for wild-type, −22.0 ± 1.6 mV (n = 14) for R185H and −23.2 ± 1.9 mV (n = 18) for I739V. The midpoints of fast-inactivation (from Boltzmann function fits of data) were not significantly different among wild-type (−75.2 ± 1.3 mV, n = 22), R185H (−72.8 ± 1.4 mV, n = 14) and I739V channels (−74.1 ± 1.6 mV, n = 18) (Fig. 5E). The rates of current decay for wild-type, R185H and I739V variant channels, as shown in Fig. 5F, were not significantly different across all voltages tested.

Figure 5

Voltage-clamp analysis of wild-type (WT), R185H and I739V channels in SCG neurons. Representative current traces recorded from SCG neurons (0 - to 5-day-old rat) expressing wild-type (A), R185H (B) or I739V channels (C), evoked by voltage steps (100 ms) from −80 mV to 60 mV in 5-mV increments, from a holding potential of −100 mV. (D) Normalized peak current–voltage relationship curves for wild-type, R185H and I739V channels. (E) Comparison of voltage-dependent activation and steady-state fast-inactivation for wild-type, R185H and I739V channels. R185H and I739V do not alter activation or fast-inactivation. (F) Comparison of time constants for deactivation. R185H and I739V both display similar deactivation rates as wild-type channels. (G) Comparison of steady-state slow-inactivation curves among wild-type, R185H and I739V variant channels. R185H does not alter steady-state slow-inactivation. I739V shifts the steady-state slow-inactivation to the right by 6.1 mV and increases the non-inactivated current at 10 mV. (H–J) SCG neurons expressing wild-type (H), R185H (I) or I739V (J) channels do not produce resurgent currents.

Slow-inactivation was impaired in I739V, but not R185H channels within the SCG cell background. The voltage-dependence of slow-inactivation (Fig. 5G) was not significantly different between R185H and wild-type channels within SCG neurons. The midpoints of slow-inactivation were −57.8 ± 1.3 mV (n = 13) for wild-type and −54.6 ± 1.7 mV (n = 13) for R185H channels. In contrast, the voltage-dependence of slow-inactivation was significantly impaired in I739V channels compared with wild-type channels, and the midpoint of slow-inactivation (from Boltzmann function fits of data) was −51.7 ± 1.9 mV (n = 12, P < 0.05) for I739V channels, ∼6 mV shifted in the depolarizing direction. The component of non-inactivating channels measured at 10 mV was not significantly different between wild-type and R185H channels (wild-type: 10.4 ± 1.3%, n = 13; R185H: 11.1 ± 1.5%, n = 13), whereas for I739V channels the non-inactivated component was significantly increased (17.4 ± 1.4%, n = 12, P < 0.01) (Fig. 5G). Resurgent current protocols did not evoke resurgent currents in SCG neurons transfected with either wild-type, R185H or I739V variant channels (Fig. 5H–J).

R185H does not alter excitability of superior cervical ganglion neurons whereas I739V renders superior cervical ganglion neurons hypoexcitable

To assess the effect of the R185H and I739V variant channels on SCG neuron excitability, we performed current-clamp recording after expressing wild-type, R185H and I739V variant channels in SCG neurons from rat pups. We assessed cells displaying strong green fluorescence, which provided a marker for transfected cells (Dib-Hajj et al., 2009), and confirmed the presence of the variant channels within SCG neurons by voltage-clamp.

Input resistance was not significantly different among SCG neurons expressing wild-type (1.2 ± 0.1 GΩ, n = 28), R185H (1.1 ± 0.1 GΩ, n = 18) and I739V channels (1.0 ± 0.1 GΩ, n = 23). The resting membrane potential of SCG neurons expressing R185H variant channels (−52.1 ± 0.9 mV, n = 18) was not significantly different from that of SCG neurons expressing wild-type channels (−50.8 ± 0.6 mV, n = 28). In contrast, I739V channels produced a significant depolarizing shift of 4 mV in resting membrane potential of SCG neurons (−46.8 ± 0.8 mV, n = 23, P < 0.001) (Fig. 6D).

Figure 6

R185H does not alter SCG neurons excitability, but I739V renders SCG neurons (0 - to 5-day-old rat) hypoexcitable. (A) Representative traces from a SCG neuron expressing wild-type (WT) channels show sub-threshold responses to 10 pA current injection and subsequent all-or-none action potentials elicited by current inputs of 15 pA and 25 pA. (B) Representative traces from a SCG neuron expressing R185H channels, showing similar current threshold for action potential generation as SCG neuron expressing wild-type channels. (C) Representative traces from a SCG neuron expressing I739V channels, showing that a higher current injection of 30 pA is required to generate the first all-or-none action potential. (D) R185H does not alter the RMP of SCG neurons, but I739V causes a significantly depolarizating shift in RMP by 4.0 mV. ***P < 0.001 versus wild-type. (E) Average current threshold of SCG neurons expressing R185H channels is not significantly different from SCG neurons expressing wild-type channels, whereas it is significantly increased after expression of I739V variant channels. **P < 0.01 versus wild-type. (F) SCG neurons expressing R185H channels display similar action potential amplitude as neurons expressing wild-type channels, whereas the amplitude is significantly smaller after expression I739V channels. ***P < 0.001 versus wild-type.

Figure 7

I739V attenuates repetitive firing in SCG neurons, while R185H does not affect repetitive firing. (A) Representative responses of SCG neurons (0 - to 5-day-old rat) expressing wild-type (WT), R185H and I739V channels, respectively, to 500 ms depolarizing current steps that are one, two and three times (top, middle and bottom traces, respectively) the current threshold for action potential (AP) generation. (B) Comparison of mean action potential spike numbers between SCG neurons expressing wild-type, R185H and I739V channels across a range of 500 ms step current injections from 10 to 200 pA. Compared to wild-type, I739V attenuates spike numbers, while R185H does not. *P < 0.05 versus wild-type.

While, as described earlier, the R185H and I739V variants both reduced the current threshold in DRG neurons, the effects of these two variants on the current threshold of SCG neurons were dramatically different. Figure 6A show traces recorded from a representative SCG neuron transfected with wild-type channels. In contrast to the high threshold of DRG neurons expressing wild-type channels, SCG neurons expressing wild-type channels displayed a much lower current threshold; for this SCG neuron, 15 pA current injection evoked the first all-or-none action potential. As shown in Fig. 6B, recordings from a representative SCG neuron that expressed R185H channels showed that the same depolarizing current stimulus (15 pA) could evoke an all-or-none action potential. Figure 6C shows typical traces recorded from a SCG neuron expressing I739V channels, where a larger current injection of 30 pA was required to produce the first overshooting action potential. As shown in Fig. 6E, the average current threshold of SCG neurons expressing R185H variant channels (18.6 ± 2.4 pA, n = 18) was not significantly different from that of SCG neurons expressing wild-type channels (17.9 ± 1.5 pA, n = 28). In contrast, I739V channels (27.8 ± 2.8 pA, n = 23, P < 0.01) significantly increased the current threshold of SCG neurons. The effects of R185H and I739V channels on the action potential amplitude of SCG were markedly different. As shown in Fig. 6F, the action potential amplitude of SCG neurons expressing R185H channels (100.5 ± 3.8 pA, n = 18) was not significantly different from that of SCG neurons expressing wild-type channels (98.2 ± 2.6 pA, n = 28). In contrast, the action potential amplitude of SCG neurons (81.0 ± 3.8 pA, n = 23, P < 0.001) was significantly reduced following expression of I739V channels.

As with DRG neurons, we also investigated the responses of SCG neurons to series of 500 ms sustained depolarizations after expressing wild-type, R185H and I739V channels. Because current threshold was lower in SCG neurons, the span of injected currents was narrowed to 10–200 pA for these cells. Figure 7A shows responses from three representative SCG neurons which expressed wild-type, R185H or I739V channels, respectively, to 500 ms current steps at one, two and three times the current threshold. SCG neurons expressing R185H channels displayed a pattern of repetitive action potentials similar to that in SCG neurons expressing wild-type channels and, in response to stimuli at three times the threshold, produced repetitive firing (seven action potentials in 500 ms) that was sustained throughout the 500 ms depolarizing stimulus. In contrast, SCG neurons expressing I739V channels generated fewer action potentials, and could not sustain firing throughout the 500 ms stimulus. Figure 7B summarizes the different effects of the R185H and I739V variants on the response of SCG neurons to prolonged stimuli, and shows that in SCG neurons, the R185H variant does not affect repetitive firing, whereas the I739V variant attenuates it. Taken together, these results show that R185H has no effects on the excitability of SCG neurons, while in contrast, I739V renders SCG neurons hypoexcitable.

Discussion

While small fibre neuropathy usually presents a clinical picture of pain, sensory loss and autonomic symptoms, occasionally patients display pain and sensory loss with only minimal autonomic dysfunction. We have reported the presence of functional Nav1.7 variants in 8 of a series of 28 patients with biopsy-confirmed small fibre neuropathy, with all except two of these patients manifesting autonomic dysfunction (Faber et al., 2012). One of the patients in this series (Patient 3 in this study) with the most severe autonomic symptoms, and two additional patients reported here (Patients 4 and 5) with severe autonomic dysfunction as well as limb pain, carry the single amino acid substitution I739V. In contrast, two unrelated patients with pain but with minimal autonomic symptoms (Patients 1 and 2) carry the single amino acid substitution R185H, presenting the opportunity to compare variants associated with the extreme ends of a spectrum from minimal to severe autonomic dysfunction. In this study, we hypothesized that the difference in autonomic symptoms in patients carrying the two different Nav1.7 variants reflects a neuron type-specific effect of the variant channels. Because primary sensory (DRG) and sympathetic ganglion neurons express different ensembles of channels (Rush et al., 2006), we hypothesized that R185H and I739V channels might differentially alter excitability in these two neuronal types. We show here that the R185H variant affects biophysical properties of the channels by enhancing resurgent current within DRG, but not SCG neurons, while the I739V variant produces a similar change in channel function of impaired slow-inactivation, in both cell types (Tables 1 and 2). We also show here that I739V renders DRG neurons hyperexcitable while rendering SCG neurons hypoexcitable, consistent with the clinical picture of pain accompanied by autonomic dysfunction in patients carrying this variant. In contrast, we show that the R185H variant channel renders DRG neurons hyperexcitable without affecting the excitability of SCG neurons, consistent with the clinical picture predominated by pain with only minimal autonomic dysfunction in patients carrying this variant. Our data suggest a correlation between effects of these two Nav1.7 variants on sensory and sympathetic neurons and the clinical manifestations in patients carrying these variants.

The conservation of the R185 and the I739 residues among Nav1.7 mammalian orthologues and the human channel paralogues is consistent with an important role for these residues, and with altered gating of the R185H and I739V channel variants. The I739 residue is highly conserved within the middle of DII/S1, and the I739V substitution impairs slow-inactivation (Han et al., 2012). Missense substitutions of residues corresponding to R185 in DI/S2-3 of Nav1.2 and Nav1.5 have been linked to epilepsy and LQT3 excitability disorders, respectively. Nav1.5/R190Q has been identified in a patient with LQT3 cardiac disorder (Chung et al., 2007), although the effect of this substitution on the gating properties of Nav1.5 have not been determined. Nav1.2/R187W has been identified in a patient with seizures and the mutant channel has been shown to inactivate more slowly than wild-type channels (Sugawara et al., 2001). Our observations show that R185H enhances resurgent currents. Changes produced by both I739V and R185H variant channels are predicted to induce a hyperexcitability phenotype in DRG neurons.

The expression of Nav1.7/I739V in neonatal DRG neurons (this study) depolarizes the resting membrane potential, lowers threshold for action potential firing and increase repetitive firing in response to suprathreshold stimuli, effects that are similar to those we recently reported in adult DRG neurons (Han et al., 2012). Expression of R185H channels in neonatal DRG neurons also increases excitability of these cells. Although the enhanced resurgent currents in the R185H variants would be predicted to increase repetitive firing (Raman and Bean, 1997; Raman et al., 1997), our biophysical profiling of the R185H channels does not provide an explanation for the reduced action potential threshold in DRG neurons transfected with this channel variant. We detected resurgent currents in 32% of DRG neurons transfected with R185H, while most of these cells were hyperexcitable. It is possible that differences in age and species of the expression platform in assessment of resurgent currents (voltage-clamp in DRG neurons from 4 - to 8-week-old Nav1.8−/− mice) and excitability (current-clamp in DRG neurons from 0 - to 5-day-old rats) may have contributed to the discrepancy. Alternatively, under our recording conditions, we were able to reliably detect resurgent currents with amplitude >100 pA and could detect resurgent currents as small as 50 pA if they displayed typical wave form in cells with low noise, but we could not reliably detect smaller resurgent currents. Thus, the possibility exists that the R185H variant induced small but functional resurgent currents in cells where we could not detect them, so that the estimate of resurgent current presence within 32% of cells may be an under-estimate. Finally, it is possible that the R185H variant alters some aspect of channel function that was not assessed by our recording protocol. Irrespective of the underlying explanation, the effects of I739V and R185H channels on the excitability of small DRG neurons, which includes nociceptors, are consistent with the severe pain experienced by the patients carrying these substitutions.

Expression of these two variant Nav1.7 channels in neonatal SCG neurons resulted in different effects on neuronal firing. Similar to DRG neurons, expression of I739V channels in SCG neurons depolarizes the resting membrane potential. However, I739V expression in SCG neurons increases threshold for action potential firing and reduces overshoot amplitude and number of action potentials fired in response to a sustained stimulus, thus rendering SCG neurons hypoexcitable. The differential effect of I739V, which renders DRG neurons hyperexcitable and SCG neurons hypoexcitable, is similar to the effect of the inherited erythromelalgia mutation L858H, which we have previously described (Rush et al., 2006). Both I739V and L858H depolarize resting membrane potential of DRG and SCG neurons. This depolarization at rest appears to differentially affect the excitability of these neurons because of the presence of Nav1.8, which is relatively resistant to inactivation by depolarization (Akopian et al., 1996, 1999; Sangameswaran et al., 1996), in DRG neurons and its absence in SCG neurons (Rush et al., 2006). In support of this hypothesis, we have previously shown that firing of SCG neurons expressing mutant Nav1.7/L858H channels can be rescued by the co-expression of Nav1.8 (Rush et al., 2006).

R185H channels did not produce resurgent currents or any other changes in sodium currents when expressed in SCG neurons and did not alter the resting membrane potential of either DRG or SCG neurons, and thus would not be expected to alter the excitability of SCG neurons. Indeed, R185H expression did not lower the threshold or increase repetitive firing in response to sustained stimuli in SCG neurons. Enhanced resurgent current by the R185H channels within DRG neurons is consistent with the hyperexcitability of DRG neurons, as has recently been shown for Nav1.7 mutations from patients with paroxysmal extreme pain disorder (Jarecki et al., 2010). Our demonstration of resurgent currents in 32% of DRG neurons expressing the R185H channel is consistent with previous studies which reported that roughly half of small DRG neurons transfected with Nav1.6 channels (Cummins et al., 2005) or Nav1.7 channels (Jarecki et al., 2010) produce resurgent currents. The enhanced resurgent current produced by the R185H channels in DRG neurons may underlie increased firing of these neurons, since it has been shown that resurgent sodium currents contribute to the high frequency firing of Purkinje neurons (Raman and Bean, 1997; Raman et al., 1997). Importantly, resurgent current has been shown to be critically dependent on cell background and the same sodium channel that produces a robust resurgent current in one neuronal type may not generate this current in a different neuronal type (Raman and Bean, 1997; Raman et al., 1997; Cummins et al., 2005). Our data suggest that SCG neurons do not support the production of resurgent current, unlike DRG neurons, a difference that would contribute to the differential effect of R185H on these two neuronal types.

Although our results demonstrate the R185H and I739V variants in multiple patients with idiopathic small fibre neuropathy, we would stress that a causative role has not been definitively demonstrated. In contrast to rare sodium channel variants associated with disease (some unique to one family), the R185H and I739V variants both appear to be present in the general population at frequencies that may be higher than the frequency of idiopathic small fibre neuropathy. Therefore, some unaffected individuals may be heterozygous for these variants and have the same genotype as the patients. Importantly, clinical manifestations of idiopathic small fibre neuropathy are usually not recognizable until adulthood, and there can be variability in age of clinical onset for patients carrying the same variant, as illustrated for the patients carrying the I739V variant reported here and by patients carrying other Nav1.7 variants (Estacion et al., 2011). These considerations raise the possibility that these variants may in fact be ‘risk factors’ that are not sufficient to cause disease on their own, but may contribute to development of idiopathic small fibre neuropathy in combination with other genetic or environmental factors/stressors.

We show here that while the I739V channels depolarize resting membrane potential in both cell types, they render DRG neurons hyperexcitable and SCG neurons hypoexcitable. The opposite functional effects of the I739V on neuronal excitability are likely to be caused by the selective expression of Nav1.8 in DRG but not SCG neurons. In contrast, R185H enhances resurgent current and renders DRG neurons hyperexcitable, but does not alter excitability of SCG neurons, perhaps due to inability of these neurons to produce the resurgent current. Our observations on the differential effects of the R185H and I739V Nav1.7 variants on DRG and SCG neurons provide a demonstration that sodium channel variants can have a spectrum of cell-background-dependent effects in different types of neurons. Our results suggest that these differential effects in different types of neurons can contribute to clinical phenotype, in this case explaining the presence of profound autonomic dysfunction in patients carrying the I739V variant, and minimal autonomic dysfunction in patients carrying the R185H variant.

View this table:
Table 1

Summary functional characterization of Nav1.7 small fibre neuropathy variants in DRG neuronsa

Neuronal typeDRG
Channel variantVoltage-clampCurrent-clamp
Resurgent currentSlow-inactivationRMPThresholdRepetitive firing
R185HIncreasedNo changeNo changeReducedIncreased
I739VNo change*Impaired*DepolarizedReducedIncreased
  • a Compared to wild-type channels expressed in DRG neurons.

  • *Han et al. (2012).

  • RMP = resting membrane potential.

View this table:
Table 2

Summary functional characterization of Nav1.7 small fibre neuropathy variants in SCG neuronsa

Neuronal typeSCG
Channel variantVoltage-clampCurrent-clamp
Resurgent currentSlow-inactivationRMPThresholdRepetitive firing
R185HNot detectableNo changeNo changeNo changeNo change
I739VNot detectableImpairedDepolarizedIncreasedDecreased
  • a Compared to wild-type channels expressed in SCG neurons.

  • RMP = resting membrane potential.

Funding

This work was supported in part by grants from the Rehabilitation Research Service and Medical Research Service, Department of Veterans Affairs and The Erythromelalgia Association (S.G.W.), the ‘profileringsfonds’ University Hospital Maastricht (C.G.F. and I.S.J.M.). The Centre for Neuroscience and Regeneration Research is a Collaboration of the Paralyzed Veterans of America with Yale University.

Acknowledgements

We thank Els K. Vanhoutte, Lynda Tyrrell, Ying Sun and Palak Shah for their assistance with this project.

Abbreviations
DRG
dorsal root ganglion
GFP
green fluorescent protein
SCG
superior cervical ganglion

References

View Abstract