Characteristics of sodium currents in rat geniculate ganglion neurons

Abstract

Geniculate ganglion (GG) cell bodies of chorda tympani (CT), greater superficial petrosal (GSP), and posterior auricular (PA) nerves transmit orofacial sensory information to the rostral nucleus of the solitary tract. We have used whole cell recording to investigate the characteristics of the Na⁺ channels in isolated Fluorogold-labeled GG neurons that innervate different peripheral receptive fields. GG neurons expressed two classes of Na⁺ channels, TTX sensitive (TTX-S) and TTX resistant (TTX-R). The majority of GG neurons expressed TTX-R currents of different amplitudes. TTX-R currents were relatively small in 60% of the neurons but were large in 12% of the sampled population. In a further 28% of the neurons, TTX completely abolished all Na⁺ currents. Application of TTX completely inhibited action potential generation in all CT and PA neurons but had little effect on the generation of action potentials in 40% of GSP neurons. Most CT, GSP, and PA neurons stained positively with IB₄, and 27% of the GSP neurons were capsaicin sensitive. The majority of IB₄-positive GSP neurons with large TTX-R Na⁺ currents responded to capsaicin, whereas IB₄-positive GSP neurons with small TTX-R Na⁺ currents were capsaicin insensitive. These data demonstrate the heterogeneity of GG neurons and indicate the existence of a subset of GSP neurons sensitive to capsaicin, usually associated with nociceptors. Since there are no reports of nociceptors in the GSP receptive field, the role of these capsaicin-sensitive neurons is not clear.

sensory information originating from receptors in the oral cavity is relayed to the brain stem via the trigeminal (Vth), facial (VIIth), and glossopharyngeal (IXth) nerves. The cell bodies of these afferent nerves are located in the trigeminal (TG), geniculate (GG), and petrosal (PG) ganglia, respectively. The lingual branch of the TG nerve is responsible for thermal, mechanical, and nociceptive sensation, and the chorda tympani (CT) branch of the facial nerve is responsible for chemosensation of the anterior two-thirds of the tongue. The lingual-tonsilar branch of the glossopharyngeal nerve transmits both chemosensory and somatosensory information from the posterior one-third of the tongue (Witt et al. 2003). The GG also contains sensory neurons that innervate the soft palate (greater superficial petrosal nerve, GSP) and skin of the external ear (posterior auricular nerve, PA) (Foley and DuBois 1943; Semba et al. 1984; Van Buskirk 1945).

Although investigators have studied the properties of TG and PG neurons, most reports of GG neuron function consist of in vivo extracellular recordings by investigators interested in taste-coding mechanisms (Boudreau et al. 1971; Lundy and Contreras 1999; Sollars and Hill 2005). Moreover, because these investigators have focused on the chemosensory response characteristics of the GG neurons, stimulation of peripheral receptive fields in the oral cavity was restricted to chemical stimuli. Therefore, investigators frequently describe CT and GSP afferent fibers as specifically responding to chemical stimulation of the oral and pharyngeal taste buds (see for example Arai et al. 2010), despite the fact that they have been demonstrated to respond to thermal and mechanical stimulation of their receptive fields (Breza et al. 2006; Ogawa et al. 1968; Sato et al. 1975; Shimatani et al. 2002). When investigators broaden the number of stimuli and determine the responses and receptive fields of nonchemoreceptive fibers of the CT, other groups of modality specific fibers emerge. In a cat study, the majority of the CT fibers were reported to be chemoresponsive, but other fibers responded specifically to thermal and mechanical stimuli (Robinson 1988, 1989; Smith et al. 2004). In addition, based on conduction velocity measures, the mechanoreceptive fibers were reported to be larger than the chemoresponsive fibers (Matsuo et al. 1995; Robinson 1988, 1989). Thus, contrary to the conclusions that CT neurons respond specifically to chemical stimulation, the afferent input to the GG is heterogeneous when based on responses to different sensory modalities. Moreover, the grouping of chemoresponsive GG neurons according to their responses to the basic taste qualities (Lundy and Contreras 1999; Sollars and Hill 2005) relates to the sensitivities of the taste receptors they innervate and not the inherent characteristic of the GG neurons.

In other well-characterized sensory ganglia, neurons have been classified into types based on soma size and peripheral fiber diameter that correlate with transmission of different sensory modalities and expression of combinations of voltage-gated ion channels (Nowycky 1992; Rush et al. 2007; Todd 2010). Initiation and propagation of action potentials requires inward Na⁺ currents, and a number of different types of Na⁺ currents have been described in sensory neurons of the trigeminal and dorsal root ganglia. These have been divided based on their responses to the Na⁺ channel blocker tetrodotoxin (TTX). Na⁺ currents have been broadly categorized as TTX-sensitive (TTX-S) and TTX-resistant (TTX-R) Na⁺ channels (Elliott and Elliott 1993). In contrast to the many detailed studies of Na⁺ currents in other sensory neurons, little is known about the characteristics of Na⁺ currents in GG neurons. In an earlier study we reported that GG action potentials were blocked after TTX application, indicating that all GG neurons Na⁺ channels were TTX sensitive (Grigaliunas et al. 2002). However, in this earlier study the receptive field of the GG neurons was not determined. Since GG neurons innervate taste receptors in different areas of the oral cavity as well as mechanoreceptors innervating the skin of the ear, it was not possible to correlate the characteristics of the Na⁺ currents with the type of peripheral innervation. As in our recent study of GG Ca²⁺ currents (Nakamura and Bradley 2011), we have now injected a retrograde label into the different receptive fields innervated by the GG neurons and demonstrated that GG neuron Na⁺ currents are more diverse than reported in the previous study.

METHODS

Fluorescent labeling.

Male Sprague-Dawley rats (Charles River Laboratories) ages 28–37 days were used in this study. All surgical procedures conformed to the National Institutes of Health “Guide for the Care and Use of Laboratory Animals” and were approved by the Animal Care and Use Committee of the University of Michigan. The ganglion cell labeling technique was similar to that described previously (King and Bradley 2000; Nakamura and Bradley 2011). In brief, rats were anesthetized with an intraperitoneal injection of a mixture of ketamine (80 mg/kg) and xylazine (16 mg/kg). The lower jaw was retracted and the tongue depressed. Under a dissecting microscope, a total volume of 10 μl of 3–5% Fluorogold (Fluorochrome) in H₂O was injected bilaterally over 4–8 sites either into the tip of the anterior tongue, the Geschmacksstreifen and middle portion of the soft palate, or just below skin of the ventral external auditory meatus and concha from a Hamilton microsyringe. Cotton swabs were used during injections to prevent the spread of the tracer that leaked from the injection sites to adjacent structures. The rats were then allowed to recover on a heating pad until ambulatory.

Dissection and dissociation of GG neurons.

Acutely dissociated GG neurons were prepared as described previously (King and Bradley 2000; Koga and Bradley 2000; Nakamura and Bradley 2011). Three to 12 days after fluorescent dye injection, the rats were reanesthetized with halothane and decapitated, and the head was secured in a stereotaxic bite bar. After removal of the superior skull, the forebrain was cut from brain stem just rostral to the cerebellum and excised from the skull. The facial nerves were identified at their exit from the brain stem, and the petrous portion of the temporal bones was removed to expose the GG. Both left and right ganglia were collected in oxygenated HEPES buffer containing 1.5 mg/ml trypsin (Sigma) and 2.5 mg/ml collagenase (type IA; Sigma) and incubated for 60 min at 37°C. The composition of the buffer was (in mM) 124 NaCl, 5 KCl, 5 MgCl₂, 10 sodium succinate, 15 dextrose, 15 HEPES, and 2 CaCl₂ (pH adjusted to 7.4 with NaOH). After incubation, ganglia were washed three times with HEPES buffer at room temperature and then triturated with a series of progressively decreasing-diameter, fire-polished Pasteur pipettes to produce a suspension of dissociated neurons. The resulting suspension was plated onto a 35-mm-diameter petri dish mounted on the stage of an inverted microscope (ECLIPSE TE300; Nikon) equipped with epifluorescence and phase-contrast optics. Dissociated neurons were kept for at least 30 min before electrophysiological recordings.

Whole cell patch-clamp recording.

Recordings were performed between 30 min and 7 h after plating. Fluorogold-labeled GG neurons were visualized by brief exposure to ultraviolet light. Patch electrodes were pulled using a two-stage puller (Narishige PP-83 puller) from borosilicate filament glass (1.5-mm outer diameter; World Precision Instruments). Electrode tip resistance was between 1.5 and 2.5 MΩ when filled with internal solution.

Electrodes were manipulated with a hydraulic three-axis micromanipulator (Narishige) under visual control. A gigaohm seal between the electrode and neuron was established in a HEPES buffer. Neurons were recorded in the standard whole cell patch-clamp configuration with an Axopatch 1D amplifier (Axon Instruments) for voltage-clamp and current-clamp experiments. Current- and voltage-clamp protocols, data acquisition, and analysis were performed using pCLAMP 8 software program (Axon Instruments). Signals were low-pass filtered at 2 kHz, digitized at 20–50 kHz (DigiData 1200; Axon Instruments), and stored on a personal computer. For voltage-clamp recordings, series resistance and cell membrane capacitance were electronically compensated. Leak subtraction was performed with an online P/4 or P/5 protocol using pCLAMP software. Measurements of cell capacitance were calculated by fitting a single-exponential curve to the uncompensated current trace from a voltage step from −60 to −70 mV. Recordings were made at room temperature.

For voltage-clamp experiments, the external solution used to isolate Na⁺ currents contained (in mM) 60 NaCl, 60 choline-Cl, 20 tetraethylammonium (TEA)-Cl, 5 KCl, 5 MgCl₂, 2 CaCl₂, 20 dextrose, and 10 HEPES (pH 7.4 adjusted to with NaOH). CdCl₂ (0.1 mM) was included to block Ca²⁺ currents. HEPES buffer was used as the external solution for current-clamp recordings. To determine TTX-R Na⁺ currents and action potentials, 0.3 μM TTX (Tocris Bioscience) was added to the external solution. For voltage-clamp experiments, electrodes were filled with an internal solution containing (in mM) 124 CsCl, 2 MgCl₂, 20 TEA-Cl, 10 HEPES, 3 EGTA, 4 Mg-ATP, and 0.3 Na-GTP (pH adjusted to 7.3 with CsOH). In some experiments (as noted in results), the internal solution contained (in mM) 130 CsF, 1 MgCl₂, 10 NaCl, 10 HEPES, and 11 EGTA (pH adjusted to 7.3 with CsOH). The internal solution for current-clamp recordings was composed of (in mM) 130 K-gluconate, 10 HEPES, 10 EGTA, 1 MgCl₂, 1 CaCl₂ and 2 Mg-ATP (buffered to pH 7.3 with KOH). The liquid junction potential of the internal solution used for current-clamp experiments (10 mV) was subtracted from the membrane potential values. The liquid junction potential for internal solution used for voltage-clamp experiments was not compensated.

In some experiments, neurons were incubated with 10 μg/ml isolectin B₄ (IB₄) conjugated to green fluorescein isothiocyanate (IB₄-FITC; Sigma) in the buffer for 10 min and then rinsed from it for at least 10 min before recordings. IB₄-FITC staining was visualized using a standard FITC filter on the inverted microscope. Only cells with an intense ring of stain around the plasma membrane were considered to be positive. At the end of experiments, to examine whether neurons respond to capsaicin, capsaicin (10 μM; Sigma) was applied at a holding potential of −60 mV in voltage-clamp configuration under pressure through a pipette positioned close to the cell using a Picospritzer (Genzen et al. 2001; Liu et al. 2006).

Data analysis.

Electrophysiological data were analyzed using Clampfit 8 (Axon Instruments) and Origin 6.1 software (OriginLab). The amplitude of the Na⁺ current at each test potential for current-voltage (I-V) plots was measured as the difference between the peak and the baseline values. The current density of Na⁺ currents was calculated by dividing the current amplitude (pA) by the cell capacitance (pF). TTX-S Na⁺ currents were obtained by digitally subtracting the TTX-R Na⁺ currents from the total Na⁺ currents. Neurons were considered sensitive to capsaicin if the evoked inward current magnitude was >100 pA and displayed an appropriate time course for onset and offset of the response. Average soma diameter was calculated from the mean of the longest and shortest axes of the neuron measured through an eyepiece micrometer. No changes in cell size were observed during recording.

Statistical analysis was conducted using the PASW statistics program (SPSS). Data are means ± SE, and statistical significances (P < 0.05) between groups were assessed using Student's t-test or one-way ANOVA, which was followed by a Scheffé's post hoc test for comparison of values.

RESULTS

Whole cell patch-clamp recordings were obtained from neurons that projected to the anterior tongue via the chorda tympani (CT neurons, n = 61), the soft palate via the greater superficial petrosal nerve (GSP neurons, n = 77); and to the inner surface of the ear via the posterior auricular branch of the facial nerve (PA neurons, n = 61). The Fluorogold-labeled neurons were readily identified under epifluorescent illumination, and the diameter of GG neurons was in the range of 23–32 μm. The average diameters of the CT, GSP, and PA neurons were 27.5 ± 0.02, 27.3 ± 0.02, and 29.0 ± 0.02 μm, respectively. An ANOVA indicated that the cell diameter was significantly different among CT, GSP, and PA neurons (ANOVA: F_2,193 = 35.3, P < 0.001). PA neurons were significantly larger than CT and GSP neurons (post hoc Scheffé's multiple comparison test: P < 0.001). Once a labeled neuron was identified, whole cell recordings were performed under bright-field illumination.

Voltage-activated Na⁺ currents in GG neurons.

Na⁺ currents from GG neurons were recorded under conditions that isolated Na⁺ currents (see methods). Figure 1 illustrates representative Na⁺ currents from dissociated GG neurons activated by a series of depolarizing step pulses from a holding potential of −90 mV to test potentials from −70 to 60 mV in 10-mV increments for 50 ms. The whole Na⁺ currents were first observed at a test potential of around −40 mV from a holding potential of −90 mV and reached the peak amplitude at −20 or −10 mV (Fig. 2A). The Na⁺ currents had rapidly activating and inactivating kinetics. We examined the composition of two components of the Na⁺ currents in GG neurons distinguished by their sensitivities to TTX. The TTX-R Na⁺ current was obtained by applying 0.3 μM TTX to the bath solution (Fig. 1, A–C, middle). Digital subtraction of the TTX-R component from the total Na⁺ current (Fig. 1, A–C, left) revealed the TTX-S component (Fig. 1, A–C, right). The majority of GG neurons expressed TTX-R currents. In 60% (45 of 76 neurons) of GG neurons, TTX caused a large reduction of Na⁺ currents, exhibiting relatively small TTX-R currents compared with TTX-S component amplitudes at each test potential (Fig. 1A). Relatively large TTX-R Na⁺ currents, larger current amplitudes than TTX-S Na⁺ currents, were observed in 12% (10 of 76 neurons) of GG neurons (Fig. 1B). In another 28% (21 of 76 neurons) of GG neurons, TTX completely abolished the Na⁺ currents (Fig. 1C).

Fig. 1.

Whole cell Na⁺ currents in geniculate ganglion (GG) neurons. A–C: representative examples of total (left), tetrodotoxin (TTX)-resistant (TTX-R; middle), and TTX-sensitive (TTX-S; right) Na⁺ current traces activated by depolarizing step pulses (50 ms) from a holding potential of −90 mV to test potentials from −70 to 60 mV in 10-mV increments as shown at bottom left in C. TTX-S Na⁺ currents were obtained by digitally subtracting the TTX-R Na⁺ currents from the total Na⁺ currents before TTX (0.3 μM) application. A: in 60% of GG neurons, TTX caused a large reduction of Na⁺ currents, but small TTX-R currents still remained. B: 12% of GG neurons expressed relatively large TTX-R currents. C: TTX completely abolished the Na⁺ currents in 28% of GG neurons.

Fig. 2.

Mean whole cell Na⁺ currents at different test potentials in different subpopulations of GG neurons. Peak current densities of total (A), TTX-R (B), and TTX-S (C) Na⁺ currents were plotted against membrane potentials. The total and TTX-S Na⁺ current densities for posterior auricular (PA) neurons displayed a prominent enhancement at test potentials of near −20 mV compared with chorda tympani (CT) and greater superficial petrosal (GSP) neurons, whereas the TTX-R Na⁺ current density was significantly larger in GSP neurons than in other subpopulations of GG neurons at test potentials greater than −10 mV. Data are means ± SE. *P < 0.05; **P < 0.01 indicate significant values by ANOVA. V_t, test potential.

TTX-R Na⁺ currents began to activate at the test potential of near −50 mV (Fig. 2B) in CT (n = 14), GSP (n = 15), and PA neurons (n = 26). The voltage of TTX-R currents reached their peaks at 0 mV in CT, GSP, and PA neurons. The maximal TTX-R currents were quite varied from neuron to neuron; however, I-V plots exhibited a significant enhancement of TTX-R current amplitude in GSP neurons compared with CT and PA neurons at the test potentials of −10 to 60 mV when compared at the same test potentials. For example, the peak amplitude of the TTX-R current was significantly larger in GSP neurons (3.62 ± 0.71 nA) than in CT (0.74 ± 0.11 nA) and PA neurons (1.12 ± 0.13 nA) at the test potential of 0 mV, at which the TTX-R currents reached peaks (ANOVA: F_2,52 = 15.8, P < 0.001; post hoc Scheffé's multiple comparison test: P < 0.001). The TTX-R current densities were also prominently larger in GSP neurons compared with CT and PA neurons at the test potentials from 0 to 60 mV (Fig. 2B). The TTX-R Na⁺ current density was significantly larger in GSP neurons (77.2 ± 12.2 pA/pF) than in CT (25.7 ± 3.3 pA/pF) and PA neurons (29.1 ± 3.2 pA/pF) at the test potentials of 0 mV (ANOVA: F_2,52 = 15.7, P < 0.001; post hoc Scheffé's multiple comparison test: P < 0.001).

TTX-S Na⁺ currents were observed in all GG neurons subgroups and started to activate at the test potential of near −40 mV in all CT (n = 24), GSP (n = 24), and PA neurons (n = 28), which was less depolarized compared with those for TTX-R Na⁺ currents (Fig. 2C). The peak TTX-S currents were at −20 mV for PA neurons and at −10 mV for CT and GSP neurons, which was slightly less hyperpolarized than those of TTX-R currents. I-V curves indicated a significant enhancement of TTX-S Na⁺ current amplitudes and current densities at test potentials from −30 to −10 mV in PA neurons compared with CT and GSP neurons when compared at the same test potentials. The peak amplitude of the TTX-S current was significantly larger in PA neurons (6.73 ± 0.38 nA) than in CT (4.44 ± 0.39 nA) and GSP neurons (4.80 ± 0.51 nA) at the test potential of −10 mV (ANOVA: F_2,73 = 8.30, P < 0.01; post hoc Scheffé's multiple comparison test: P < 0.01). The TTX-S Na⁺ current density was also significantly larger in PA neurons (188.4 ± 10.3 pA/pF) than in GSP neurons (143.6 ± 16.3 pA/pF) at the test potentials of −20 mV (ANOVA: F_2,73 = 5.98, P < 0.01; post hoc Scheffé's multiple comparison test: P < 0.01), but there was no significant difference between PA and CT neurons (164.3 ± 11.2 pA/pF, P = 0.19).

TTX-R Na⁺ currents could be separated into low- and high-threshold components (also referred to as persistent and slowly inactivating components, respectively) in some GG neurons, as previously reported in other ganglion neurons (Kwong et al. 2008; Scroggs 2010). In one set of experiments, TTX-R Na⁺ currents were obtained from 6 CT, 7 GSP, and 4 PA neurons with 50-ms voltage steps from −70 to 60 mV from a holding potential of −90 mV using an internal solution containing 130 mM CsF in the presence of 0.3 μM TTX (Fig. 3A). To isolate high-threshold components, the same protocol was applied but with a prepulse to −60 mV for 500 ms before the step pulses to inactivate the low-threshold components (Fig. 3B). Subtraction of the high-threshold components from the total TTX-R Na⁺ currents revealed the low-threshold component (Fig. 3C). Of the 17 GG neurons tested, 6 neurons (2 CT, 2 GSP, and 2 PA neurons) exhibited measurable but much smaller amounts of low-threshold components than those of the high-threshold components. The low-threshold components began to activate at a test potentials of −60 or −50 mV. In contrast, the high-threshold components started to activate at a test potentials of near −40 mV.

Fig. 3.

Separation of TTX-R Na⁺ current into low- and high-threshold components. A: total TTX-R Na⁺ currents evoked with 50-ms voltage steps from a holding potential of −90 mV to test potentials from −70 to 60 mV in 10-mV increments in the presence of 0.3 μM TTX. Fluoride was the primary anion in the internal solution in this set of experiments. B: the high-threshold components yielded by a 500-ms prepulse to −60 mV before voltage steps. C: the low-threshold components revealed by digital subtraction of the high-threshold components from the total TTX-R currents. Voltage protocols used to activate Na⁺ currents are shown below the current traces in A and B, respectively.

We compared the amplitude of the TTX-R and TTX-S Na⁺ current densities in CT, GSP, and PA neurons. The TTX-R (Fig. 4, A–C, left) and TTX-S Na⁺ current densities (Fig. 4, A–C, right) were estimated by measuring the Na⁺ currents evoked from a holding potential of −90 mV to a test potential of −10 mV and plotted as frequency distribution histograms. In all CT and most of the PA neurons, the TTX-R Na⁺ current densities were <60 pA/pF, whereas some of the GSP neurons (9 of 24 neurons, 38%) had relatively higher TTX-R current densities that were >60 pA/pF (Fig. 4B, left), which contribute to a remarkable enhancement of TTX-R Na²⁺ current at test potentials of around −10 mV in GSP neurons compared with CT and PA neurons as shown in the I-V plot of Fig. 2B. The TTX-R Na⁺ current density of <10 pA/pF was observed in 42% of CT and 38% of GSP neurons, whereas only 2 PA neurons (7%) had TTX-R currents of <10 pA/pF. All PA neurons had TTX-S currents that were >100 pA/pF, whereas ∼17% (4/24) of CT and 34% (8/24) of GSP neurons had TTX-S currents of <100 pA/pF (Fig. 4, A–C, right).

Fig. 4.

Frequency distributions of TTX-R and TTX-S Na⁺ current densities in CT, GSP, and PA neurons. TTX-R (left) and TTX-S (right) Na⁺ current densities were calculated by dividing the current amplitude (pA) at a test potential of −10 mV by the cell capacitance (pF) in 24 CT (A), 24 GSP (B), and 28 PA neurons (C). Whereas 38% (9 of 24 neurons) of GSP neurons expressed large TTX-R currents of >60 pA/pF, all CT and 96% (27 of 28 neurons) of PA neurons displayed small TTX-R currents of <60 pA/pF.

Effects of TTX on action potentials in GG neurons.

We used current-clamp recordings to further investigate the contribution of TTX-S and TTX-R Na⁺ currents to the generation of the spike firing of GG neuron subpopulations. Current-clamp recordings were performed on 19 CT, 20 GSP, and 15 PA neurons. Resting membrane potentials of all neurons ranged from −41 to 60 mV with a mean of −49 ± 0.2 mV (CT neurons), −49 ± 0.3 mV (GSP neurons), and −53 ± 0.2 mV (PA neurons), respectively. The resting membrane potential of PA neurons was slightly but significantly more positive than that of CT and GSP neurons [paired t-test: t(7) = 2.78, P < 0.05]. Action potentials were elicited by injection of 500-ms depolarizing current pulses in 0.1-nA increments (Fig. 5A, left).

Fig. 5.

Effect of TTX on action potentials in different subpopulations of GG neurons. Membrane potentials were recorded by injection of hyperpolarizing and depolarizing current pulses from −0.4 nA for 500 ms with 0.1-nA increments before (left) and after (middle) 0.3 μM TTX application. A: TTX eliminated the generation of action potentials in all CT (n = 19) and PA neurons (n = 15). TTX also completely abolished the spike firing in the majority of GSP neurons (12/20, 60%). The effect of TTX was reversed after washout (right). B: TTX had little effect on the generation of action potentials in 40% (8/20) of GSP neurons but increased the threshold of the spike firing. The firing returned partially after washout (right).

Application of 0.3 μM TTX completely inhibited the action potential generation in all CT and PA neurons (Fig. 5A, middle). The effect of TTX was completely reversible (Fig. 5A, right). In the majority of GSP neurons (12/20, 60%), TTX completely abolished the spike firing, whereas TTX had little effect on the generation of action potentials in 40% (8/20) of GSP neurons (Fig. 5B). However, the threshold of the spike firing was significantly increased from 0.23 ± 0.01 to 0.49 ± 0.04 nA in the presence of TTX (P < 0.05). This effect was also completely reversible (Fig. 5B, right). TTX had no effect during hyperpolarization in CT, GSP, and PA neurons.

TTX-R current characteristics and capsaicin sensitivity in IB₄-positive and IB₄-negative GG neurons.

Investigators have reported that small and medium-size dorsal root ganglion nociceptive neurons could be subdivided into two major classes based on molecular and anatomical criteria. One class has neuropeptides such as substance P and calcitonin-gene related neuropeptide (CGRP), whereas the other lacks these neuropeptides but contains cell surface glycoconjugates that bind IB₄ from the plant Griffonia simplicifolia (Nagy and Hunt 1982; Silverman and Kruger 1990b). Previous studies have revealed that IB₄-positive neurons have higher densities of TTX-R Na⁺ currents than IB₄-negative neurons in mice (Stucky and Lewin 1999) and rat dorsal root ganglion (Wu and Pan 2004). In addition, IB₄-positive reactivity has been shown in the majority of GG neurons (Silverman and Kruger 1990a). Therefore, to examine a possible relationship between TTX-R Na⁺ current expression and IB₄ binding in GG neurons, we incubated neurons with IB₄-FITC in external solution before recordings.

Figure 6A shows phase-contrast (a) and fluorescence images of the same field of GG neurons labeled with Fluorogold (b) and IB₄-FITC (c). The diameter of IB₄-positive neurons (28.0 ± 0.19 μm, n = 41) and IB₄-negative neurons (28.2 ± 0.37 μm, n = 15) was not significantly different. We evoked Na⁺ currents in IB₄-positive and IB₄-negative neurons in CT (n = 12), GSP (n = 26), and PA (n = 18) using the same protocols as shown in Fig. 1. At the end of Na⁺ current recordings, we further examined the sensitivity to capsaicin, which is used as a principal marker of nociceptive function for dorsal root ganglion neurons that respond to noxious heat (Caterina and Julius 2001), in CT, GSP, and PA neurons. Most of IB₄-positive GSP neurons that had relatively small TTX-R Na⁺ currents of <60 pA/pF at a test potential of 0 mV (11/12, 92%) did not respond to focal application of 10 μM capsaicin (Fig. 6B), whereas the majority (7/9, 78%) of IB₄-positive GSP neurons with large TTX-R Na⁺ currents of >60 pA/pF exhibited capsaicin-induced inward currents at a holding potential of −60 mV (Fig. 6C). On the other hand, all IB₄-negative GSP neurons expressed small TTX-R and large TTX-S currents and did not respond to capsaicin (Fig. 6D). All CT and PA neurons also did not respond to capsaicin regardless of amplitude of TTX-R currents.

Fig. 6.

TTX-R Na⁺ current expression and capsaicin sensitivity in isolectin B₄ (IB₄)-positive and -negative GG neurons. A: identification of IB₄-positive and IB₄-negative GG neurons: a, dissociated GG neurons viewed with bright-field illumination; b, same-field fluorescence image of Fluorogold-labeled neurons with ultraviolet excitation filter; c: same field of IB₄-labeled neurons with FITC. The Fluorogold-labeled, IB₄-positive neuron is indicated by the arrowheads. Scale bar, 50 μm. B–D: total (left) and TTX-R Na⁺ currents (middle) were obtained from IB₄-positive and IB₄-negative GG neurons using the same protocols as shown in Fig. 1. At completion of Na⁺ current recordings, 10 μM capsaicin was applied in voltage-clamp mode (right). B: example of an IB₄-positive GSP neuron with small TTX-R Na⁺ currents of <60 pA/pF at a test potential of 0 mV. This type of neuron did not respond to focal application of 10 μM capsaicin. C: example of an IB₄-positive GSP neuron with large TTX-R Na⁺ currents of >60 pA/pF. This type of neuron was capsaicin sensitive, and an inward current was elicited. D: example of an IB₄-negative GSP neuron. In this type of neuron, TTX caused a large reduction of total Na⁺ currents. These neurons were capsaicin insensitive. All CT and PA neurons with and without TTX-R currents also did not respond to capsaicin.

Figure 7 summarizes the comparison of mean total, TTX-R, and TTX-S Na⁺ current densities at a test potential of 0 mV in subgroups of GSP neurons categorized by expression of IB₄ and sensitivity to capsaicin. The TTX-S current densities were significantly smaller in IB₄-positive and capsaicin-sensitive [IB₄(+) Cap(+)] GSP neurons (73.3 ± 11.0 pA/pF, n = 7) than in IB₄-positive and capsaicin-insensitive [IB₄(+) Cap(−)] GSP neurons (143.0 ± 16.6 pA/pF, n = 14) (ANOVA: F_2,23 = 3.84, P < 0.05; post hoc Scheffé's multiple comparison test: P < 0.05). The TTX-S current density of IB₄(+) Cap(+) GSP neurons tended to be smaller than that of IB₄-negative and capsaicin-insensitive [IB₄(−) Cap(−)] GSP neurons (127.4 ± 18.4 pA/pF, n = 5); however, this difference was not significant (post hoc Scheffé's multiple comparison test: P = 0.26). On the other hand, the TTX-R current densities was significantly larger in IB₄(+) Cap(+) GSP neurons (93.2 ± 13.1 pA/pF, n = 7) than in IB₄(+) Cap(−) GSP neurons (33.4 ± 10.2 pA/pF, n = 14) and IB₄(−) Cap(−) GSP neurons (17.5 ± 9.5 pA/pF, n = 5) (ANOVA: F_2,23 = 8.06, P < 0.01; post hoc Scheffé's multiple comparison test: P < 0.01). There were no significant differences in the total Na⁺ current densities among the three subgroups of GSP neurons (ANOVA: F_2,23 = 0.46, P = 0.64).

Fig. 7.

Comparison of peak TTX-R Na⁺ current density among 3 subgroups of GSP neurons categorized by IB₄ immunoreactivity and capsaicin sensitivity. GSP neurons were subdivided into 3 groups: IB₄ positive and capsaicin sensitive [IB₄(+) Cap(+)], IB₄ positive and capsaicin insensitive [IB₄(+) Cap(−)], and IB₄ negative and capsaicin insensitive [IB₄(−) Cap(−)]. There were significant differences in TTX-R and TTX-S Na⁺ current densities between subgroups of GSP neurons (ANOVA: P < 0.01). Data are means ± SE. *P < 0.05; **P < 0.01 indicate significant values by ANOVA, followed by Scheffé's host hoc tests.

DISCUSSION

Our experiments, summarized in Table 1, show that the biophysical, anatomical, and Na⁺ channel expression of GG neurons are heterogeneous. PA neurons were larger and had a higher mean resting membrane potential than CT and GSP neurons. Action potential generation in all CT and PA neurons and 60% of the GSP was inhibited by TTX application. In a separate population of GSP neurons, TTX had little effect on depolarization-initiated action potential generation. Most CT, GSP, and PA neurons stained positively with IB₄, and 27% of the GSP neurons were capsaicin sensitive. These data, combined with heterogeneity of Ca²⁺ channel expression we recently reported (Nakamura and Bradley 2011), reveal the complex differences in the characteristics of GG ganglion neurons innervating different receptive fields.

Table 1.

Characteristics of Na²⁺ currents, IB₄ reactivity, and capsaicin responses in different subpopulations of GG neurons

	CT Neurons	GSP Neurons	PA Neurons
Cell diameter, μm	27.5 ± 0.02	27.3 ± 0.02	29.0 ± 0.02
RMP, mV	−49 ± 0.2	−49 ± 0.3	−53 ± 0.2
Effect of TTX on spike firing	Inhibited	60% inhibited, 40% unaffected	Inhibited
TTX-S current density (at −10 mV), pA/pF	164.3 ± 11.2	143.6 ± 16.3	188.4 ± 10.3
TTX-R current density (at 0 mV), pA/pF	25.7 ± 3.3	77.2 ± 12.2	29.1 ± 3.2
IB4-positive neurons, %	73.1	66.7	91.7
Capsaicin-sensitive neurons, %	0	26.9	0

Values are means ± SE in chorda tympani (CT), greater superficial petrosal (GSP), and posterior auricular (PA) neurons. RMP, resting membrane potential; TTX-S and TTX-R, tetrodotoxin sensitive and tetrodotoxin resistant; IB₄, isolectin B₄.

In other sensory ganglia, these differences in Na⁺ channel expression correlate with the different functional roles of the ganglion cells (Rush et al. 2007). For example, neurons in the dorsal root ganglion relay information from muscle and skeletal mechanoreceptors, cutaneous and subcutaneous mechanoreceptors, and nociceptors. In contrast, most GG neurons are generally considered to innervate chemoreceptors of the anterior two-thirds of the tongue and soft palate, although this conclusion suffers from a bias due to use of a limited number of stimuli to characterize their responses. There appears to be no information available on the response characteristics of the mechanoreceptors innervated by the PA nerve.

Despite the fact that the CT and GSP neurons innervate only taste buds, they differ in a number of other characteristics. Afferent fibers of GG neurons vary in size. Investigators have reported a bimodal distribution of CT axon diameters (Jang and Davis 1987). Axon diameters of the CT in rat and hamster have been measured after the parasympathetic efferent fiber component was eliminated by section of the central process (nervus intermedius). Seventy-nine percent of the afferent processes of the CT were myelinated fibers and only 30% were unmyelinated (Farbman and Hellekant 1978). The diameter of myelinated axons ranged from 1 to 5 μm, whereas that of unmyelinated axons ranged from 0.2 to 1.5 μm. Data from the hamster CT reveal a bimodal distribution of axon diameters with equal proportions of myelinated and unmyelinated fibers (Jang and Davis 1987). Thus there is considerable variability in the distribution of CT axon diameters. No information is available for GSP and PA axon diameters.

Investigators have report that the CT and GSP cell soma have a typical pseudounipolar morphology and vary between 15 and 25 μm in diameter (Grigaliunas et al. 2002; Kitamura et al. 1982; Nakamura and Bradley 2011). In the present study we confirm other reports that PA neuron somata are larger than those of CT and GSP neurons (Gomez 1978; Semba et al. 1984). A few larger neurons ranging from 30 to 40 μm in diameter are also reported in the GG (Gomez 1978). In ultrastructural studies of guinea pig, monkey, and human GG, the neurons are described as being similar to those in other sensory ganglia (Lieberman 1976) and are classified as either light or dark cells based on their appearance in electron micrographs (Kitamura et al. 1982; Nawar et al. 1980).

Na⁺ currents in GG neurons.

Although Na⁺ channels have been thoroughly characterized in dorsal root (Cummins et al. 1999; Roy and Narahashi 1992; Rush et al. 1998) and trigeminal sensory ganglia (Kim and Chung 1999; Liu et al. 2001), both of which innervate a variety of sensory receptors including nociceptors, the current study is the first study of voltage-gated Na⁺ channel in the GG. Two distinctly different TTX-sensitive Na⁺ currents differentiated by the amplitudes of TTX-S and TTX-R components were expressed in CT, GSP, and PA neurons. Although TTX-S Na⁺ currents were found in all GG neurons, the occurrence of TTX-R currents varied among different subpopulations of GG neurons. The majority of GG neurons expressed TTX-R currents, but the magnitude of the currents varied. For most neurons, TTX-R currents were of small amplitude, and large TTX-R currents were observed in relatively few neurons. Moreover, TTX-R current amplitudes differed among CT, GSP, and PA neurons (Table 1). The mean peak amplitude of the TTX-R current was significantly larger in GSP neurons than in CT and PA neurons. The TTX-R Na⁺ current density was also significantly larger in GSP neurons than in CT and PA neurons. TTX-S and TTX-R voltage-dependent Na⁺ currents also had different activation characteristics. TTX-S Na⁺ currents started to activate near −40 mV, and TTX-R Na⁺ currents began to activate near −50 mV.

Application of TTX completely inhibited the action potential generation in all CT and PA neurons and in the majority of GSP neurons. This result is similar to our previous report on action potential inhibition by TTX in cultured GG neurons (Grigaliunas et al. 2002). However, TTX had little effect on the generation of action potentials in a subgroup of GSP neurons (Table 1).

Comparison of these results with information on voltage-gated Na⁺ channels in other sensory ganglia is better served by results obtained on the petrosal ganglion, which has many similar functions to the GG. Like the GG, the petrosal ganglion innervates chemoreceptors of the tongue and carotid body as well as mechanoreceptive fibers of the tongue and pharynx (Gallego et al. 1987). Like the GG, a proportion of the petrosal ganglion neurons have TTX-R currents (Stea and Nurse 1992). However, when petrosal neurons innervating the chemoreceptors of the carotid body were isolated, almost all of them were TTX-S (Cummins et al. 2002). Presumably, the TTX-R channels are in the petrosal neurons innervating other receptors. CT and GSP chemoreceptor afferent neurons in the GG both have TTX-R currents, possibly related to the different functions of carotid body and taste bud chemoreceptors. No data are available on the Na⁺ channels of petrosal neurons innervating posterior tongue taste buds to compare with the data in the current study of anterior tongue chemoreceptor neurons.

Role of Na⁺ currents in action potential trains.

Voltage-gated Na⁺ currents are involved in the generation of action potentials, and the expression of the different types of Na⁺ channels found in the ganglion cells also have been demonstrated to extend along the axons (Rush et al. 2007). Investigators recording from CT fibers and GG neurons report spontaneous activity and high-frequency irregular discharge patterns of action potentials to chemical stimulation of the tongue (see for example Hellekant et al. 1997; Sollars and Hill 2005). The presence of TTX-S channels is important in the generation of spontaneous activity because of their activation-inactivation characteristics (Schild and Kunze 1997). The role of TTX-S and TTX-R channels in generation of action potential trains is not known, but it has been suggested that TTX-R Na⁺ currents are important in action potential duration and therefore influence discharge frequency (Bossu and Feltz 1984). It also has been suggested that due to the slow inactivation kinetics of TTX-R Na⁺, these channels may play a role in the adaptation pattern of the discharge (Elliott and Elliott 1993). Moreover, neurons with a large proportion of TTX-S Na⁺ channels that characteristically recover slowly from inactivation would be rapidly adapting (Elliott and Elliott 1993). Thus neuron discharge patterns in GG chemosensitive neurons result from both the characteristics of the receptor transduction process and Na⁺ channel expression.

Capsaicin-sensitive GSP neurons.

On the basis of sensitivity to capsaicin, GSP neurons can be divided into two subpopulations. The majority of the GSP neurons did not respond to capsaicin application, but almost 30% were capsaicin sensitive, presumably acting on TRPV1 receptors that have been demonstrated in GG neurons with immunohistochemistry (Katsura et al. 2006). In other ganglia, capsaicin sensitivity is a characteristic of nociceptive neurons (Caterina et al. 2000; Gold et al. 1996), and TRPV1 expression is usually associated with IB₄-positive neurons (Carlton and Hargett 2002; Liu et al. 2004). We found a similar sensitivity to capsaicin in GSP IB₄-positive GG neurons with large TTX-R Na⁺ currents, suggesting that a subpopulation of GSP neurons is nociceptive. However, there are no data to support the presence of GG nociceptors on the soft palate. To date most information on the response characteristics of GSP neurons has been derived from whole nerve recordings limited to gustatory, not nociceptive, stimulation of the palate (Harada and Smith 1992; Nejad 1986; Sollars and Hill 1998). Similarly, in the only study of GG GSP neuron responses (Sollars and Hill 2005), palatal stimuli were also limited to chemicals. Thus no evidence is available on the response of GSP neurons to mechanical, noxious, and thermal stimuli. However, one of the chemicals used in these studies is HCl as a representative sour taste stimulus. It is therefore possible that the TRPV1 channel may play a role in the acid-sensitive responses of the soft palate taste buds (Leffler et al. 2006; Liu et al. 2004). In support of this possibility, TRPV1-immunoreactive fibers have been described in soft palate taste buds (Kido et al. 2003), and application of capsaicin to the palate evokes a burning sensation and reflex secretion of saliva (Dunér-Engström et al. 1986).

In summary, we have shown that GG neurons with receptive fields in the tongue, soft palate, and external ear differ in a number of properties including soma size and types of voltage-gated Na⁺ channels. GG neurons innervating posterior ear skin were characteristically different from those innervating taste buds. We have shown that all the neurons investigated had both TTX-sensitive and -insensitive Na⁺ channels. In addition, neurons of the CT and GSP differed in sensitivity to TTX and capsaicin, suggesting that GSP neurons transmit both chemosensory and nociceptive information.

GRANTS

This work was supported by National Institute on Deafness and Other Communication Disorders Grant DC-000288 (to R. M. Bradley).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

S.N. performed experiments; S.N. and R.M.B. analyzed data; S.N. and R.M.B. interpreted results of experiments; S.N. prepared figures; R.M.B. conception and design of research; R.M.B. drafted manuscript; R.M.B. edited and revised manuscript; R.M.B. approved final version of manuscript.