Abstract
Voltage-gated sodium (NaV) channels are responsible for the initiation and conduction of action potentials within primary afferents. The nine NaV channel isoforms recognized in mammals are often functionally divided into tetrodotoxin (TTX)-sensitive (TTX-s) channels (NaV1.1–NaV1.4, NaV1.6–NaV1.7) that are blocked by nanomolar concentrations and TTX-resistant (TTX-r) channels (NaV1.8 and NaV1.9) inhibited by millimolar concentrations, with NaV1.5 having an intermediate toxin sensitivity. For small-diameter primary afferent neurons, it is unclear to what extent different NaV channel isoforms are distributed along the peripheral and central branches of their bifurcated axons. To determine the relative contribution of TTX-s and TTX-r channels to action potential conduction in different axonal compartments, we investigated the effects of TTX on C-fiber-mediated compound action potentials (C-CAPs) of proximal and distal peripheral nerve segments and dorsal roots from mice and pigtail monkeys (Macaca nemestrina). In the dorsal roots and proximal peripheral nerves of mice and nonhuman primates, TTX reduced the C-CAP amplitude to 16% of the baseline. In contrast, >30% of the C-CAP was resistant to TTX in distal peripheral branches of monkeys and WT and NaV1.9−/− mice. In nerves from NaV1.8−/− mice, TTX-r C-CAPs could not be detected. These data indicate that NaV1.8 is the primary isoform underlying TTX-r conduction in distal axons of somatosensory C-fibers. Furthermore, there is a differential spatial distribution of NaV1.8 within C-fiber axons, being functionally more prominent in the most distal axons and terminal regions. The enrichment of NaV1.8 in distal axons may provide a useful target in the treatment of pain of peripheral origin.
SIGNIFICANCE STATEMENT It is unclear whether individual sodium channel isoforms exert differential roles in action potential conduction along the axonal membrane of nociceptive, unmyelinated peripheral nerve fibers, but clarifying the role of sodium channel subtypes in different axonal segments may be useful for the development of novel analgesic strategies. Here, we provide evidence from mice and nonhuman primates that a substantial portion of the C-fiber compound action potential in distal peripheral nerves, but not proximal nerves or dorsal roots, is resistant to tetrodotoxin and that, in mice, this effect is mediated solely by voltage-gated sodium channel 1.8 (NaV1.8). The functional prominence of NaV1.8 within the axonal compartment immediately proximal to its termination may affect strategies targeting pain of peripheral origin.
Introduction
Noxious stimuli are transduced at nociceptor terminals and the resulting action potential discharge is transmitted by axons to the dorsal spinal cord. For the detection of tissue-damaging stimuli in the external environment, nerve terminals express select sets of ion channels, receptors, and neuropeptides. The family of voltage-gated sodium (NaV) ion channels is responsible for the generation and propagation of action potentials (Ahern et al., 2016). In nociceptors, the predominant NaV channel currents originate from NaV1.7, which is sensitive to tetrodotoxin (TTX-s), NaV1.8, and NaV1.9, both of which are TTX-resistant (TTX-r). NaV1.8 expression appears to be important for action potential generation in small, unmyelinated C-fiber neurons (Renganathan et al., 2001). The relatively hyperpolarized activation range of NaV1.9 allows it to pass a persistent current, but a role for NaV1.9 in mediating action potential conduction remains a matter of debate (Priest et al., 2005). The NaV1.5 isoform is also relatively resistant to blockade by TTX (IC50 ∼2 μm). NaV1.5 can deliver a large transient current in neonatal and early postnatal dorsal root ganglion (DRG) neurons, but its expression falls in adult DRG neurons (Renganathan et al., 2002), although a splice variant is reported in adult mouse DRG (Kerr et al., 2007).
Within DRG neurons, the distribution and functional availability of NaV channel isoforms differs between somal and axonal compartments. Axonally targeted NaV1.8 mRNA (Ruangsri et al., 2011) and protein (Gold et al., 2003) in peripheral nerves can be significantly higher compared to the soma. Functional studies suggest a differential spatial distribution of NaV channel isoforms in somatosensory neurons. Electrophysiological recordings from isolated DRG neuronal somata implicate TTX-r isoforms as the primary mediators of inward sodium current during an action potential (Scholz et al., 1998; Blair and Bean, 2002). In contrast, assays of axonal conduction by means of TTX indicate that, for the majority of mammalian C-fibers, action potential propagation relies primarily on TTX-s NaV channel isoforms (Villière and McLachlan, 1996; Farrag et al., 2002; Zimmermann et al., 2007; De Col et al., 2008). Application of TTX to mammalian peripheral nerves (Yoshida and Matsuda, 1979; Steffens et al., 2001) and dorsal roots (Pinto et al., 2008) results either in overt blockade of C-fiber conduction or a profound slowing of the C-fiber response and a decrease in amplitude. Typically, the compound C-fiber amplitude is reduced to 20–30% of its value before TTX, although values of up to 80% (mean 47%) have been reported in samples of human nerve taken at biopsy or after amputation (Quasthoff et al., 1995). The requirement of TTX-s NaV channels for axonal conduction in the majority of C-fiber axons manifests functionally in the spinal dorsal horn, where the likelihood of transmitter release in response to C-fiber stimulation is either reduced (Jeftinija, 1994) or largely abolished (Pinto et al., 2008) by TTX. Although NaV1.8 is postulated to serve a role in sensory nerve terminals by remaining available for activation despite cooling, the role of TTX-r NaV channel isoforms in axonal C-fiber conduction is less clear.
We studied C-fiber-mediated compound action potentials (C-CAPs) along different segments of axonal membrane (distal peripheral nerve, proximal peripheral nerve, and dorsal root) in samples from mice and compared the results to those obtained from nonhuman primates. We discovered that a substantially greater portion of the C-CAP was resistant to TTX in distal compared with proximal nerves. Studies in NaV1.9 and NaV1.8 knock-out mice showed that the TTX-r component of the C-CAP is mediated solely by NaV1.8. These findings demonstrate that the TTX-resistant NaV1.8 isoform not only plays a role in action potential generation in peripheral terminals, but can also support action potential conduction in distal peripheral unmyelinated axons.
Materials and Methods
Study approval.
The Animal Care and Use Committee of Johns Hopkins University (JHU) approved the experiments on nervous tissue from mice (C57BL6, NaV1.9−/−) and nonhuman primates. The Ethics Committee of the Regional Government (Karlsruhe, Baden-Wuerttemberg, Germany) approved experiments involving C57BL6 and NaV1.8−/− mice performed at the University Heidelberg Medical Faculty Mannheim.
Mouse nerve dissection.
Adult inbred C57BL6 mice of both sexes and weighing 25–35 g were obtained from Harlan and Janvier. NaV1.8−/− mice (Akopian et al., 1999) were bred at the University of Erlangen–Nuremberg (Germany) and animals of both sexes weighing 25–35 g (∼5–6 weeks) were used. As part of an ongoing research project, NaV1.9−/− mice were generated by inserting a Neo-STOP minigene upstream of SCN11A, thereby reducing expression by ∼80%. Mice were anesthetized with isoflurane (5%; Midwest Veterinary Supply) and killed by cervical dislocation. Dorsal roots together with proximal (i.e., sural, saphenous, peroneal, and tibial; Fig. 1A) nerve segments and distal peripheral nerve segments (Fig. 1B) were dissected from the hind limbs of mice and used within ∼24 h (JHU) or 0–2 h (Mannheim) of harvesting. The proximal nerve segments were dissected free over a 15–25 mm length immediately distal to their point of division at the trifurcation of the sciatic nerve at the knee (Fig. 1A). Distal nerve segments were harvested from the corium side of a skin flap of the dorsal hindpaw over a length of ∼10–15 mm immediately proximal to the tips of the toes so that only one branch innervating the toe was used for recording (Fig. 1B).
Primate nerve dissection.
Dorsal root and peripheral nerve materials were acquired postmortem from male and female adult pigtail monkeys (Macaca nemestrina) that were bred at JHU and had been part of a control cohort in an ongoing SIV research project. Proximal peripheral (i.e., median and ulnar nerves), distal peripheral nerves (i.e., digital nerves), and dorsal roots were dissected and used within ∼48 h after harvesting. Tissues not used immediately were stored in synthetic interstitial fluid (see below for details) at room temperature, which was changed after 24 h.
Compound action potential recordings.
Isolated nerve segments were maintained in a synthetic interstitial fluid composed of either of the following (in mm): 118 NaCl, 3.2 KCl, 1.5 CaCl2, 1 MgCl2, 6 HEPES, 20 Na+ gluconate, and 5.6 d-glucose, pH 7.4 (bubbled with 100% O2) or 107.7 NaCl, 3.5 KCl, 0.69 MgSO4, 26.2 NaCO3, 1.67 NaH2PO4, 1.5 CaCl2, 9.64 Na+ gluconate, 5.5 d-glucose, and 7.6 sucrose, pH 7.4 (bubbled with 95% O2, 5% CO2) and subsequently desheathed. C-CAPs were recorded as described previously (Lang et al., 2008; Freysoldt et al., 2009; Carr et al., 2010; Sittl et al., 2012). Briefly, each end of the desheathed nerve segment was drawn into a glass pipette in an organ bath and embedded in petroleum jelly to establish high resistance electrical seals (Fig. 1C). Pairs of silver wire electrodes were placed inside each pipette and in the bath. One pair of electrodes was used for constant current stimulation. The other electrode pair recorded extracellular signals (Fig. 1D). The distance between the sites of stimulation and recording varied between 3 and 8 mm. The organ bath (∼1 ml) was perfused continuously with physiological solution at a flow rate of 2 ml/min. For experiments performed in Mannheim, the temperature of the perfusion solution was controlled with an inline Peltier device. All experiments at JHU were conducted at room temperature. Constant current electrical stimulation (Linear Stimulus Isolator, A395; WPI) was delivered to the nerve with the silver wire inside the stimulating pipette, serving as the anode. Current pulses were applied at either 0.3 or 1 Hz. Extracellular signals were amplified (NeuroLog NL905; Digitimer), filtered (low-pass 5 kHz; high-pass, 0.1 Hz), digitized, and stored to disk. Stimulus intensity and data acquisition were controlled using either QTRAC software (Digitimer) or DAPSYS version 8.0 (http://www.dapsys.net).
Experimental protocol.
Constant current pulses of fixed amplitude were used to assess conduction in A-fibers (0.1 ms) and C-fibers (1 ms). Under control conditions, CAP responses were recorded until stable amplitude and latency values were established (∼5–20 min) at 23 ± 2°C. For JHU experiments using nerves from nonhuman primates and WT and NaV1.9−/− mice, TTX (1 μm) was applied for at least 5 min or until CAP amplitude had stabilized. This was followed by coapplication of TTX and A803467 (5 μm), a selective NaV1.8 blocker (Jarvis et al., 2007), for at least 5 min or until CAP amplitude had stabilized. TTX and A803467 were subsequently washed out until the C-CAP amplitude had at least partially recovered indicating viability of the preparation. Lidocaine (12 mm) was applied directly into the recording chamber to identify the electrical stimulus artifact. The concentration of A803467 used was ∼35× the IC50 reported to block TTX-r current in rat DRG neurons (Jarvis et al., 2007). In a separate series of experiments performed in Mannheim, nerve segments from WT and NaV1.8−/− mice were examined. In the control period, nerve segments were maintained at 23 ± 2°C until the latency and amplitude of the C-fiber volley were stable. The tissue was then warmed to 32 ± 2°C and held at this temperature for >5 min or until the recording was stable again. TTX (500 nm) was subsequently added to the perfusing solution for 5 min, after which the temperature was cooled to 23 ± 2°C for an additional 5 min and then rewarmed to 32 ± 2°C. TTX was washed out for ∼30 min until C-CAP amplitude reached a plateau and no further increase in the amplitude was observed. In all experiments, except those involving NaV1.8−/− mice, lidocaine (1 mm) was added to the perfusing solution at the end of the protocol for 5 min at 32 ± 2°C to confirm that the recorded compound action potential signal was mediated by NaV channels. The TTX concentrations (1 μm and 500 nm) used in these studies are 100–1000× above the IC50 required to block TTX-s NaV channel isoforms and 40,000–60,000 × below the IC50 (40–60 mm) reported to block TTX-r isoforms NaV1.8 and NaV1.9 (Catterall et al., 2005), but only a factor of 4 lower than the IC50 for NaV1.5 (Renganathan et al., 2002).
Data analysis.
For the studies at JHU, from the end of each incubation period (baseline and superfusion with TTX and TTX + A803467), 10 traces of C-CAP recordings were averaged in DAPSYS and data from this average were copied into Excel (Microsoft Office 2010) to determine the peak-to-peak amplitude of the averaged C-CAP waveform (see inset in Fig. 1D). For the studies at Mannheim, the peak-to-peak amplitude (see inset in Fig. 1D) and latency were determined for CAP responses to each electrical stimulus during the recording by QTRAC software. Post hoc data analysis was performed using custom routines written in IgorPro (Wavemetrics). CAP latency and peak-to-peak amplitude were determined by averaging values across three to five sequential sweeps at time points corresponding to the end of the control period, during incubation with TTX, after washing, and during lidocaine. Area under the curve (AUC) of the C-fiber CAP was determined offline by integrating the CAP record over a fixed 25 ms time window beginning 2–3 ms before the C-fiber peak, as determined by inspection.
Statistics.
Statistical analyses were performed with Statistica software using nonparametric or parametric tests where appropriate, as described in detail in the Results section. For group comparisons, data are presented as mean ± SEM. p-values < 0.05 were regarded as significant.
Chemicals.
Stock solution aliquots of TTX (Tocris Bioscience and Sigma-Aldrich) were prepared in PBS and stored at −20°C. Lidocaine hydrochloride (Samuel Perkins or Sigma-Aldrich) was stored as a stock solution in distilled water at room temperature. Stock solutions were diluted to the desired concentration in physiological solution on the day of the experiment. A803467 (Tocris Bioscience) was initially dissolved in DMSO and serially diluted into physiological solution.
Results
One-fourth of the C-CAP signal in distal axonal segments is TTX resistant
We used CAP recordings in combination with pharmacological agents to examine functionally whether there are spatial differences in the contribution of NaV channel isoforms to conduction along peripheral axons. In particular, we investigated the effects of TTX (1 μm) and a combination of TTX (1 μm) and A803467 (5 μm), a NaV1.8-specific blocker (Jarvis et al., 2007), on conduction in unmyelinated C-fibers in dorsal roots and proximal and distal (digital) segments of peripheral nerve. The C-CAP response to supramaximal electrical stimuli was monitored during superfusion with the different compounds and examples of recordings from different nerve segments from WT mice are shown in Figure 2. In all preparations, TTX (1 μm) slowed C-fiber conduction as evidenced by an increase in response latency of the C-CAP. In distal nerve segments (Fig. 2A), a considerable fraction of the control C-CAP amplitude was still apparent at the end of the TTX superfusion period. In contrast, TTX largely suppressed the C-CAP amplitude in proximal nerve segments and dorsal roots (Fig. 2B,C). During washout, C-CAP recovered partially and subsequent application of lidocaine completely blocked axonal conduction. As can be seen from the example traces, C-CAP amplitude varied considerably between recordings. To quantify effects across different experiments, C-CAP amplitudes were normalized to baseline C-CAP amplitude. For the recordings in Figure 2, A–C, the normalized C-CAP in the presence of TTX was 25.7% for distal nerve, but only 4.0% and 6.5% for the proximal nerve and dorsal root, respectively. The effect of TTX on C-CAP amplitude differed significantly between neuronal compartments (Fig. 2D, Kruskal–Wallis ANOVA, H(2,44) = 8.16, p = 0.017). During exposure to TTX, C-CAP amplitude was significantly larger in distal nerves (24.7 ± 5.5%, n = 13) than in both dorsal roots (5.8 ± 1.2%, n = 15) and proximal nerve segments (7.1 ± 1.6%, n = 16, p < 0.05, post hoc multiple comparisons). In distal nerve preparations, but not in dorsal root or proximal nerve preparations, the amplitude of the remaining C-CAP in the presence of TTX varied substantially. In some distal nerves, but not in others, the C-CAP was completely abolished by TTX. Similar to TTX alone, the effect of TTX combined with A803467 differed significantly between different nerve segments (Kruskal–Wallis ANOVA, H(2,37) = 9.60, p = 0.008), with the TTX-r C-CAP amplitude being significantly larger in distal nerves than dorsal roots (p = 0.006, post hoc multiple comparisons). Only within the distal nerve segment did the incubation with combined TTX and A803467 further decrease the C-CAP amplitude (25.2 ± 5.1%) compared with TTX alone (29.3 ± 5.5%, p < 0.05, Wilcoxon matched pairs, n = 11), but this effect was rather small.
To test whether the effects of TTX or combined TTX and A803467 also differed between neuronal compartments in primates, recordings were made from dorsal roots, proximal and distal (digital) nerves from pigtail monkeys. Similar to recordings from WT mice, TTX slowed the C-CAP in all nerve segments. In the distal nerve segments (Fig. 3A), the C-CAP was reduced to 21.5% of its control amplitude by TTX (1 μm). In proximal nerve segments (Fig. 3B), TTX reduced the C-CAP amplitude to 10% (i.e., by 90%). As in WT mice, primate dorsal roots and proximal nerves did not differ in their response to TTX and data from these tissues were therefore combined for group comparisons (Fig. 3C). After incubation with TTX or combined TTX and A803467, the C-CAP amplitude resistant to blockade was significantly larger in distal than proximal nerves (TTX: 24.2 ± 4.2%, n = 27, vs 9.4 ± 1.3%, n = 31, p = 0.03, Mann–Whitney U test; TTX/A803467: 25.6 ± 4.4%, n = 22, vs 10.2 ± 1.5%, n = 26, p = 0.01, Mann–Whitney U test). In distal, but not proximal, nerve segments, exposure to combined TTX and A803467 reduced C-CAP amplitude compared with TTX alone (TTX vs TTX/A803467: 29.7 ± 4.2% vs 25.6 ± 4.4%, n = 22, p = 0.0041, Wilcoxon matched pairs), but the effect was again small. Together, the data from mice and nonhuman primates indicate that conduction in a considerable fraction of unmyelinated C-fibers in distal nerves can be mediated by TTX-r NaV channel isoforms.
Axonal TTX-r C-CAP is present in NaV1.9−/− mice
In adult mammals, somatosensory C-fiber neurons express the TTX-resistant NaV channel isoforms NaV1.8 and NaV1.9. To determine the contribution of NaV1.9 to TTX-resistant axonal conduction in C-fibers, C-CAPs were recorded from distal and proximal nerve segments of NaV1.9−/− mice (Fig. 4A,B). The effect of 1 μm TTX on axonal C-fiber conduction in NaV1.9−/− mice was indistinguishable from its effect on peripheral nerves from WT mice and nonhuman primates. In proximal nerve segments, C-CAP was largely blocked by TTX (Fig. 4B), whereas 40% of the C-CAP in distal nerve segments was TTX resistant (Fig. 4A). Analysis of the group data (Fig. 4C) revealed that the amplitudes of the remaining C-CAP during both TTX and combined TTX and A803467 were significantly larger in distal compared with proximal nerve segments (TTX: 41.5 ± 7.4%, n = 10, vs 6.8 ± 2.0%, n = 7, p = 0.022, Mann–Whitney U test; TTX plus A803467: 42.5 ± 5.8%, n = 9, vs 8.4 ± 1.3%, n = 5, p = 0.022, Mann–Whitney U test). In proximal and distal nerve segments, incubation with TTX and A803467 did not further decrease the C-CAP amplitude (proximal: 9.5 ± 1.4% vs 8.4 ± 1.3, n = 5, p = 0.14, Wilcoxon matched pairs; distal: 46.1 ± 6.4% vs 42.5 ± 5.8%, (n = 10, p = 0.086, Wilcoxon matched pairs).
C-CAP amplitude is increased by cooling, primarily in distal axonal segments
In all nerve segments (proximal sural and saphenous as well as distal nerve segments), 500 nm TTX blocked axonal conduction in A-fibers. The A-CAP amplitude was reduced by TTX to 4 ± 2% of its control value (data not shown; 1-way ANOVA, F(1,30) = 16.81, p < 0.01). Changes in temperature and application of 500 nm TTX were used to characterize TTX-r axonal conduction in peripheral C-fibers. In WT mice, cooling from 32°C to 23°C increased the C-CAP amplitude both before (Fig. 5A, control, repeated-measures ANOVA, with “cooling” as within-subject factor, F(1,24) = 28, p < 0.01) and during application of TTX (500 nm; Fig. 5B, repeated-measures ANOVA, F(1,24) = 12.87, p < 0.01). There was no difference in the relative increase in amplitude upon cooling before and during TTX treatment (Fig. 5C; 2-way ANOVA with “treatment” and “nerve segment” as independent variables; treatment F(1,48) = 1.27, p = 0.26; nerve segment F(2,48) = 1.83, p = 0.17). In contrast to cooling, TTX (500 nm) significantly reduced the C-CAP amplitude in all nerves (Fig. 5A,B) both at 32°C (2-way ANOVA with “treatment” and “nerve segment” as independent variables; treatment F(1,48) = 42.86, p < 0.01) and 23°C (2-way ANOVA F(1,48) = 35.3, p < 0.01). Markedly, for nerves from WT mice, there were prominent differences in the absolute C-CAP amplitude across different nerve segments (Fig. 5A). Accordingly, to make comparisons between nerves for the effects of TTX and temperature, C-CAP amplitudes were normalized to their control values at 32°C (Fig. 6A–C). The normalized amplitude of the TTX-r C-CAP for proximal sural (Fig. 6A) and saphenous (Fig. 6B) nerves was 14% at 32°C and 10–16% at 23°C. In distal nerve segments, 27% of the C-CAP amplitude was TTX-r at 32°C and 57% was TTX-r at 23°C (Fig. 6C). Kruskal–Wallis ANOVA indicated a statistically significant difference in TTX-r amplitude between different nerve segments at 23°C (Fig. 6D; H(2,27) = 8.96, p < 0.05), a feature that was not apparent at 32°C (Fig. 6D; H(2,27) = 4.76, p = 0.09). A nonparametric test was used in this analysis because the data violated the assumption of equal variances. To confirm that electrically evoked TTX-r CAPs were mediated by NaV channels, the broad-spectrum NaV blocker lidocaine (1 mm) was applied at the end of each experiment (Fig. 6A–C). Lidocaine blocked C-fiber conduction in all nerve segments, with the amplitude of the C-CAP reduced to 1–3% of its control value (2-way ANOVA with “treatment” and “nerve segment” as independent variables; treatment F(1,46) = 54.75, p < 0.01). The conduction velocity of the C-CAP under control conditions did not differ in sural, saphenous, or distal nerve segments from WT mice (data not shown; 1-way ANOVA F(2,19) = 2.97, p = 0.08).
Axonal TTX-r C-fiber conduction is absent in NaV1.8−/− mice
Transgenic mice lacking NaV1.8 were used to determine the contribution of NaV1.8 to TTX-r CAPs recorded in peripheral nerve segments. In addition, because NaV1.8 resists inactivation during cooling (Zimmermann et al., 2007), the involvement of NaV1.8 in the cooling induced increase in C-CAP amplitude (Fig. 5) was also assessed.
In sural, saphenous, and distal nerve segments from NaV1.8−/− mice, 500 nm TTX blocked the electrically evoked C-CAP at 32°C (2-way ANOVA with “treatment” and “nerve segment” as independent variables; treatment F(1,54) = 62.91, p < 0.01) and this effect could not be rescued by cooling to 23°C (Fig. 7A–C; 2-way ANOVA with “treatment” and “nerve segment” as independent variables; treatment F(1,54) = 60.45, p < 0.01); that is, there was no detectable axonal TTX-r C-CAP. This observation suggests that NaV1.8 is solely responsible for the TTX-r CAP signal in peripheral mouse nerve. In contrast, the increase in C-CAP amplitude upon cooling was independent of NaV1.8 (cf. WT control in Fig. 6A–C with NaV1.8−/− in Fig. 7A–C). That is, under control conditions, before TTX, the amplitude of C-CAPs from NaV1.8−/− nerves increased during cooling from 32°C to 23°C (Fig. 7D, repeated-measures ANOVA with “cooling” as within-subject factor, F(1,27) = 14.63, p < 0.01) in a manner comparable to WT nerves (Fig. 5A). For all nerves from NaV1.8−/− mice, blockade of C-fiber conduction by 500 nm TTX was partially reversed by washout. There were no differences in the conduction velocity of the C-CAP under control conditions between different nerve segments from NaV1.8−/− mice (data not shown; 1-way ANOVA F(2,25) = 1.68, p = 0.21).
To explore the possibility that block of C-CAPs in nerves from adult NaV1.8−/− mice might be due to the absence of a subset of C-fiber axons, the absolute amplitude of the C-CAP (Buchthal and Rosenfalck, 1966; Dyck et al., 1971; Tackmann et al., 1976) and the AUC (Stys et al., 1992) of the C-CAP were used as indices of axon number. However, there were no statistically significant differences in either C-CAP amplitude (Fig. 7E; 2-way ANOVA; genotype F(1,51) = 0.76, p = 0.39) or AUC (Fig. 7F; 2-way ANOVA; genotype F(1,50) = 0.44, p = 0.51) between WT and NaV1.8−/− mice.
Discussion
The results demonstrate that C-CAPs in dorsal roots and proximal nerve segments of mice and nonhuman primates were largely abolished by TTX (0.5–1 μm), whereas in distal nerves, a more considerable portion of the C-fiber signal was resistant to TTX. These findings indicate that C-fiber conduction in proximal nerves and dorsal roots depends largely on TTX-s NaV channel isoforms, whereas in distal axonal segments, more C-fibers can support TTX-r action potential conduction. Using transgenic mice, NaV1.8 was identified as the sole isoform responsible for axonal TTX-r action potential conduction in C-fibers and the distribution of NaV1.8 was site dependent, with this channel being available functionally at distal axon sites involved in action potential initiation as well as conduction. In contrast, NaV1.9 does not contribute substantially to the ability to generate and conduct a TTX-resistant axonal action potential.
Previous studies have shown that TTX-r NaV channels are functional in peripheral terminals of unmyelinated afferent nerve fibers (Brock et al., 1998; Brock et al., 2001; Carr et al., 2002; Zimmermann et al., 2007). The present study extends this concept by showing that NaV1.8 underlies TTX-r action potential conduction in C-fiber axons and that TTX-r conduction is enriched in the distal peripheral axonal branches of pseudo-unipolar DRG C-type neurons. Although immunohistochemical techniques indicate that NaV1.8 and NaV1.9 appear to cluster along the course of unmyelinated axons (Fjell et al., 2000; Rush et al., 2005) and at nerve terminals (Black et al., 2002), no variation in signal intensity along the sciatic nerve has been reported. In neurites of cultured mouse DRG neurons, fluorescently tagged NaV1.8 clustered in lipid rafts along the neuronal membrane but were not enriched in the terminal regions (Pristerà et al., 2012). It is currently unclear how the observed spatial differences in our functional assay of NaV1.8 might occur. The preferential availability of NaV1.8 in distal axon segments and terminals may involve mechanisms related to local protein translation (Jiménez-Díaz et al., 2008; Thakor et al., 2009; Obara and Hunt, 2014), channel trafficking, membrane insertion (for reviews, see Swanwick et al., 2010; Bao, 2015), and regulation of channel function, for example, through phosphorylation and NaV channel β-subunits (for review, see Ahern et al., 2016). Our observation that TTX-r channels support conduction in only a limited number of unmyelinated axons in proximal nerves or dorsal roots is consistent with previous functional observations in rats indicating that TTX-r NaV channel isoforms alone were not sufficient for synaptic transmission of nociceptive input within the dorsal horn (Pinto et al., 2008). Together, previous and present findings suggest that TTX-r NaV channel isoforms are functionally relevant in peripheral terminals and distal axons of unmyelinated fibers.
Several factors could contribute to the observation of a larger TTX-r CAP in distal nerve segments. For example, more prominent diffusion barriers could limit access of TTX to unmyelinated fibers in distal nerve segments. Although all nerves were desheathed to remove the perineurium/epineurium, which is known to impair access of TTX to peripheral axons (Hackel et al., 2012), we consider hampered toxin diffusion unlikely because TTX fully blocked A-fibers in all preparations and similarly abolished C-CAPs in distal nerves from NaV1.8−/− animals (Fig. 7C), demonstrating that TTX had access to unmyelinated axons.
We did not observe a biologically significant effect of A803467 on TTX-r CAPs in distal nerves. Currently, we cannot explain this limited effect because A803467 has been shown previously to block transient and resurgent TTX-r current in isolated rat DRG neurons (Jarvis et al., 2007; Tan et al., 2014). However, the state-dependent affinity for inactivated NaV1.8 led to effects of A803467 on TTX-r current in DRG neurons that were dependent on membrane potential (Jarvis et al., 2007), a parameter not controlled in our axonal recordings. Furthermore, the efficacy of A803467 on C-CAPs may be larger after peripheral nerve injury or inflammation. In models of chronic pain, but not in naive rats, A803467 effectively reduced nociceptive behavior and inhibited spontaneous and evoked activity in wide dynamic range dorsal horn neurons (Jarvis et al., 2007; McGaraughty et al., 2008; Joshi et al., 2009; Liu et al., 2014; Rahman and Dickenson, 2015).
Our results with NaV1.8−/− nerves support the notion that TTX-r-resistant C-CAPs are mediated by NaV1.8. Like its effect in WT animals, TTX only partially reduced C-CAPs in peripheral nerves from NaV1.9−/− animals, suggesting that NaV1.9 contributes minimally to TTX-r C-CAPs. Interestingly, the TTX-resistant C-CAP in sural and saphenous nerves from NaV1.9−/− mice appeared to be larger than in WT (∼20% vs 40%; cf. Figs. 3C, 4C), a result that may indicate compensatory overexpression of NaV1.8 in distal nerves of such animals.
During cooling, axonal action potentials slow in conduction speed and increase in amplitude (Hodgkin and Katz, 1949; Swadlow et al., 1981; Stys et al., 1992; Sittl et al., 2012; Fig. 5). The reduction in conduction speed is attributed to a slowing of the activation time constant of NaV channels (Frankenhaeuser and Moore, 1963). The increase in CAP amplitude with cooling is also attributed primarily to changes in NaV kinetics. Cooling slows the rate of both NaV activation and inactivation with the Q10 for inactivation (Q10 = 2.8) exceeding that for activation (Frankenhaeuser and Moore, 1963; Kimura and Meves, 1979; Collins and Rojas, 1982; Schwarz and Eikhof, 1987) (Q10 = 1.8), prolonging NaV channel opening time and thus action potential amplitude. In addition, membrane resistance increases substantially during cooling (Reid and Flonta, 2002) primarily due to the temperature-dependent closure of two-pore-domain potassium channels (Maingret et al., 2000; Kang et al., 2005; Schneider et al., 2014). The resulting increase in resistance amplifies the effect of NaV-channel mediated current on voltage.
Here, we confirm that cooling increased the amplitude and AUC of C-CAPs in nerves from WT and NaV1.8−/− mice (Figs. 6, 7). Therefore, a component of temperature-dependent changes in CAP shape occurs independently of NaV1.8. The pronounced effect of cooling on distal nerve segments of WT tissue might therefore suggest an enrichment of both NaV1.8 and perhaps background potassium currents in the most distal axonal compartment.
A possible explanation for the observation that conduction in all C-fiber axons is sensitive to TTX in NaV1.8−/− mice is that select populations of DRG neurons are impaired developmentally, thereby leading to axonal loss. Specifically, a reduction in the number of IB4-positive and CGRP-positive DRG neurons might be expected based upon diphtheria ablation in NaV1.8-expressing neurons (Abrahamsen et al., 2008). However, an analysis of C-CAP amplitude and AUC as indices of axon number suggests that the number of C-fiber axons did not differ significantly between WT and NaV1.8−/− mice (Fig. 7E,F). Similarly, in the original description of this NaV1.8-null transgenic line, DRG neuronal counts and immounohistochemical ratios were both taken to indicate no neuronal loss after global deletion of the channel (Akopian et al., 1999). Although our indices of axon number (C-CAP amplitude and AUC) indicated no loss of axons in NaV1.8−/− nerves, it should be noted that these indices varied depending upon the nerve tested (Fig. 7E,F). Compared with WT, NaV1.8−/− mice showed larger C-CAP amplitude and AUC for sural nerve segments, but lower values for saphenous and distal nerve segments. This apparent inconsistency across nerves may suggest that loss of NaV1.8 results in different compensatory mechanisms in different nerves and is perhaps dependent on the proportion of sensory and sympathetic fibers or the ratio of C- to A-fibers. Immunohistochemical studies using NaV1.8-specific antibodies (Amaya et al., 2000) and studies using a NaV1.8-Cre mouse line (Abrahamsen et al., 2008; Shields et al., 2012) showed colocalization of NaV1.8 and the A-fiber-associated heavy-chain neurofilament marker NF200. Similarly, single-cell analysis revealed NaV1.8 expression in large-diameter DRG neurons overlapping with NF200 expression (Ho and O'Leary, 2011).
We observed that NaV1.8 was functional in distal peripheral nerves of mice (Fig. 2), consistent with the proposal that this channel likely plays an important role in distal axons of unmyelinated afferents. Behaviorally, NaV1.8-deficient mice display a delayed onset of inflammatory hyperalgesia after complete Freund's adjuvant (CFA) and reduced sensitivity to painful mechanical and cold stimuli (Akopian et al., 1999; Zimmermann et al., 2007). Antisense oligodeoxynucleotide targeting NaV1.8, but not NaV1.9, increased paw withdrawal thresholds in rats after intraplantar CFA (Yu et al., 2011). Spinal nerve injury induced an upregulation of NaV1.8 in uninjured fibers of the sciatic nerve (Gold et al., 2003) and knockdown of NaV1.8 reversed signs of neuropathic pain in animal models (Lai et al., 2002; Yang et al., 2014) and reduced spontaneous activity in dissociated DRG neurons after spinal cord injury (Yang et al., 2014). In humans, NaV1.8-gain-of-function mutations have been observed in patients with painful distal neuropathy (Faber et al., 2012) and NaV1.8 was significantly increased in painful lingual nerve injury neuromas (Bird et al., 2013). Collectively, these studies suggest that NaV1.8 plays a crucial role in peripheral nociceptive signaling. The results from our studies indicate that NaV1.8 is not only functional in peripheral, receptive terminals, but also in distal axons of mice and nonhuman primates. Accordingly, therapeutic targeting of NaV1.8 in peripheral tissues (e.g., at sites of injury) may provide a strategy for the relief of pain of peripheral origin while avoiding central side effects.
Footnotes
This work was supported by the National Institutes of Health (Grant R01NS097221 to J.L.M., F.B., and M.R.; Grant P40 OD013117 to RJ Adams; and F32 Fellowship Grant F32DA036991 to A.H.K.) and the German Research Society (DFG Grant SFB1158/1-TP 01 to M.S. and Grant SFB1158/1-TP 04 to R.C.). We thank the Blaustein Pain Research and Education Fund, the Brain Science Institute and the Neurosurgery Pain Research Institute at the Johns Hopkins University for their support of this study; the National Center for Research Resources and the National Center for Advancing Translational Sciences (NCATS) of the National Institutes of Health (Grant 1UL1TR001079) for statistical analysis; and R.A. Meyer for critically reading a previous version of this manuscript.
The authors declare no competing financial interests.
- Correspondence should be addressed to either of the following: Richard Carr, Department of Anesthesiology, Medical Faculty Mannheim, University of Heidelberg, Ludolf-Krehl Str. 13-17, 68167 Mannheim, Germany. Richard.Carr{at}medma.uni-heidelberg.de; or Matthias Ringkamp, Department of Neurosurgery, School of Medicine, Johns Hopkins University, Meyer 5-109, 600 N Wolfe St, Baltimore, MD 21287. platelet{at}jhmi.edu