INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

One of the principal taste organs on the mammalian tongue is the set of fungiform papillae and resident taste buds innervated by nerve fibers from two main sensory ganglia (Mistretta 1991; Mistretta and Hill 1995). The trigeminal ganglion, via the lingual branch of the mandibular nerve root innervates both the lingual epithelium that lies between fungiform papillae and the papilla epithelium that surrounds taste bud cells but not taste buds per se. In contrast, the geniculate ganglion, via the chorda tympani nerve branch, provides innervation only to the taste bud. Thus in this taste organ, nerve fibers from the geniculate ganglion travel to and innervate cells in a highly circumscribed location, whereas fibers from the trigeminal ganglion essentially encompass a geniculate-derived nerve basket to innervate surrounding papilla and lingual epithelium (Miller 1974; Whitehead et al. 1985). Although the lingual receptors are not the only peripheral target organs for trigeminal and geniculate ganglia (Brodal 1981), neurites from these ganglia have contiguous but distinct receptive fields within the fungiform papillae that require precise developmental controls to establish appropriate innervation patterns.

The trigeminal and geniculate ganglia not only have distinctive target organs but also have different embryonic derivations (Graham and Begbie 2000). The trigeminal placode is neurogenic and forms from CNS tissue near the midbrain-hindbrain junction. The geniculate placode is epibranchial, forming between the branchial arches in close proximity to neural crest and pharyngeal endoderm (Begbie et al. 1999). Although trigeminal and geniculate ganglion function is integral to formation of taste circuits and to gustatory and related somatosensory sensation, there is little known about electrophysiological differentiation of these ganglion cells. To begin to understand early ganglion cell function and the factors that regulate emerging function, we have studied basic neural properties of embryonic trigeminal and geniculate ganglia in rat.

In general, immature ganglion neurons have electrophysiological properties that include action potentials with a relatively long duration, a smaller amplitude, and a resistance to the sodium channel blocking agent, tetrodotoxin (Spitzer and Ribera 1998; Vincent et al. 2000). There follows a gradual maturation to briefer, sodium-dependent action potentials. However, in adult mammals, ganglia typically have heterogeneous subpopulations of neurons with varying action potential, ion channel and neuropeptide characteristics (De Armentia et al. 2000; Hall et al. 1997). In dorsal root ganglia (DRG), functional subpopulations have been related to specific peripheral receptor type (Fedulova et al. 1998; Koeber et al. 1988), but initial neuropeptide expression in sensory subtypes apparently emerges in advance of connections with receptor targets (Hall et al. 1997). On the other hand, neurotrophins in target tissues have known roles in regulating differentiation of neuron morphology, ion channel expression, and neurotransmitters (Huang and Reichardt 2001). Electrical stimulation is also important in developmental expression of neuropeptides and signaling pathways in embryonic rat DRG cells (Ai et al. 1998; Fields 1998). Furthermore patterned electrical stimulation versus chronic depolarization differentially alters ion channel activity and gene expression in embryonic petrosal ganglion (Brosenitsch and Katz 2001), suggesting potential for receptor input to alter ganglion neuron properties.

Although there are some basic neural properties that characterize neurons as immature and relatively homogeneous (Spitzer and Ribera 1998), the extent to which these generalizations apply to all sensory ganglion neurons and the age at which ganglia demonstrate divergent neural properties are not clear. Both the trigeminal (Liu and Simon 1996; Liu et al. 2001) and geniculate (King and Bradley 2000; Koga and Bradley 2000) ganglia have been studied in postnatal or adult ages to discern whether neurons projecting to defined receptive fields have distinctive biophysical characteristics that correlate with neuron subtypes. Action potential properties in postnatal geniculate neurons that are primarily gustatory differ in some aspects from those that are primarily somatosensory (King and Bradley 2000). In adult rat, trigeminal neurons are subtyped based on responses to nociceptive chemicals, and action potential and ion channel properties (Liu and Simon 1996; Liu et al. 2001).

Discovering how and when the gradual process of neurophysiological differentiation is regulated by intrinsic and extrinsic factors, including interactions with target organs, requires basic knowledge of neurophysiological development in pre- and postnatal ganglia. There are no experimental data comparing embryonic trigeminal or geniculate ganglion neural properties. Therefore it is not clear how intrinsic ganglion neuron development, emerging neural activity from developing sense organs, and molecular factors in the target environment might interact to effect the basic neurobiology of these two sensory ganglia and their participation in formation of taste and taste-related sensory circuits.

To study the trigeminal and geniculate ganglia, we used an explant system of ganglia maintained in culture for several days and made whole cell recordings from single ganglion neurons. Explant cultures were used to retain a biological environment more similar to that in vivo than with dissociated neurons. Ganglia were dissected at gestational day 16, which is an embryonic stage when the cell soma extend neurites to provide robust innervation of the early fungiform papilla (Farbman and Mbiene 1991; Mbiene and Mistretta 1997) but is a period in advance of taste bud formation. To sustain ganglion neurons in the explant cultures, we added the appropriate neurotrophin to maximally support neuron survival (Al Hadlaq et al. 2001a; Davies 1997). Thus trigeminal explants were maintained with nerve growth factor (NGF) and geniculate explants with brain-derived neurotrophic factor (BDNF). Furthermore NGF is a principal neurotrophin in lingual receptive fields of the trigeminal ganglion, whereas BDNF is in circumscribed target fields of geniculate ganglion neurons (Nosrat and Olson 1995; Nosrat et al. 2001).

Our experiments had two main objectives. First, to understand neurophysiological function over time in culture, we tested the hypothesis that basic cell membrane and action potential properties for each ganglion type remain stable over days in vitro. Second, with basic information about the explant system in hand, we tested the hypothesis that trigeminal and geniculate ganglion cells cultured at E16 and maintained with optimal neurotrophin support, have similar neural properties. Early reports of these studies have appeared in abstracts (Grigaliunas et al. 1999, 2000).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Ganglion cultures

Timed-pregnant Sprague Dawley rats at gestational day 16 (day 0, vaginal plug detected) were deeply anesthetized with an intraperitoneal injection of pentobarbital sodium (50 mg/kg body wt), which effectively anesthetizes the embryos also. Embryos were removed and placed in cold (4°C), Earle's balanced salt solution with gentamicin sulfate (20 µg/ml) and 20 mM HEPES buffer (pH 7.4). Embryo heads were hemisected in a sagittal plane and ganglia were exposed from the medial side of each half. Trigeminal and geniculate ganglia were dissected and explanted onto glass coverslips covered by a matrix deposited by cultured bovine corneal endothelial cells (MacCallum et al. 1982). The relatively large trigeminal ganglia were cut in half before explanting, and the two halves were placed on one coverslip. Two individual geniculate ganglia were placed on one coverslip. Coverslips with ganglia were placed in a culture dish in medium containing a 1:1 mixture of Dulbecco's modified Eagle's minimum essential medium and Ham's nutrient mixture F12, 1% fetal bovine serum, 20 µg/ml gentamicin sulfate, and 2% B27 culture supplement (GIBCO, Life Technologies). The medium was supplemented with 10 ng/ml NGF and 10 ng/ml BDNF (Alomone Labs, Jerusalem, Israel) for trigeminal and geniculate ganglia, respectively, to optimally sustain cells and promote neurite outgrowth. Explanted ganglia were maintained in an incubator in humidified 5% CO₂ in air, at 37°C, for up to 10 days. The culture medium was changed every second day.

Solutions and drugs

For whole cell recording, coverslips with cultured ganglia were removed from the culture dish, placed in a petri dish, and superfused (3 ml/min) with oxygenated solution containing (in mM) 124 NaCl, 5 KCl, 5 MgCl₂, 10 Na-succinate, 15 dextrose, 15 HEPES, and 2 CaCl₂ (Du and Bradley 1996). The pH was adjusted to 7.4 with NaOH. In experiments with Na⁺-free solution, Na⁺ was substituted by choline to maintain osmolality, and pH was adjusted with KOH. The sodium channel blocker, tetrodotoxin (TTX; Sigma, St. Louis, MO), was kept in citrate buffer (pH 4.3) at 20°C and diluted in the external bath solution to the required concentration before each experiment. The L-type calcium channel blocker, nifedipine (Sigma), was dissolved in dimethylsulfoxide (20 mM stock solution) before adding to the bath solution. Verapamil hydrochloride (Tocris Cookson, Ballwin, MO) was added directly to the external solution. Specific calcium channel blockers, -conotoxin-GVIA and -conotoxin-MVIIA (N-type channels) and -agatoxin-TK (P/Q-type channels; Alomone Labs), were dissolved in a solution containing 0.1% bovine serum albumin (BSA), 100 mM NaCl, 10 mM Tris (pH 7.5), and 1 mM EDTA and stored at 20°C. All blockers were diluted to required concentrations in the bath solution and were applied under pressure through a pipette positioned close to the cell, using a Picospritzer. During experiments with toxins, 0.01% BSA was added to the bath solution, according to manufacturer's recommendations (Alomone Labs).

Electrophysiological recordings

Pipettes were pulled in two stages from 1.5 mm OD borosilicate filament glass (WPI, MTW150F-4) using a Narishige PP-83 electrode puller and were filled with a solution containing (in mM) 130 K-gluconate, 10 HEPES, 10 EGTA, 1 MgCl₂, 1 CaCl₂, and 2 ATP; pH was adjusted to 7.2 with KOH. Electrodes had tip resistances of 6-8 M. All experiments were performed at room temperature (20-22°C).

Coverslips with cultured ganglion explants were placed in a petri dish mounted on the stage of an inverted microscope, equipped with Nomarski optics and epifluorescent illumination. Neurons for whole cell recording were selected among cells migrating from the central explant core and electrodes were positioned with a three-coordinate, hydraulic manipulator under visual control. Whole cell recordings in current- and voltage-clamp modes were made using an Axoclamp-2A amplifier (Axon Instruments). The measured liquid junction potential (10 mV) between pipette filling and bath solutions was subtracted from the recorded membrane voltages (Neher 1992). Criteria for successful recording were a minimum 10-min recording time, with a stable resting membrane potential more negative than 40 mV, an action potential amplitude of 70 mV or greater, and 100 M or higher input resistance.

To examine morphology of neuronal processes, more than 30 trigeminal and 30 geniculate neurons were filled with 5% Lucifer yellow CH (dipotassium salt, Sigma), added to the pipette solution. Recordings were made to confirm neuron function, and cells were photographed to analyze the neurite branch pattern.

Data analysis

Data were acquired and analyzed using pCLAMP software (Axon Instruments). Input resistance was estimated from current-clamp recordings of the voltage response to 400-ms-long hyperpolarizing, 25-pA current steps. Membrane time constant was measured by fitting a single exponential function to the charging transient of the same recording. Membrane capacitance was calculated by dividing the time constant by input resistance. To evaluate action potential properties, a short (3 ms) 300-pA current step protocol was used for excitation. Action potential variables analyzed were: amplitude, from resting membrane to peak potential; half-duration, spike duration at half-amplitude; decay time, duration of the falling phase between 90 and 10% of spike amplitude; maximal right and left slopes, slope at the steepest point of spike rising or falling phases. Threshold of excitation was determined from recordings after application of an increasing series of depolarizing, 400-ms current steps at 25 pA. Afterhyperpolarization (AHP) amplitude was measured from the level of resting membrane potential to the lowest point, and AHP duration was defined as recovery time from the lowest point to the level of resting membrane potential. Average soma diameter was calculated from the mean of the longest and shortest axes of the cell measured with an eyepiece micrometer. No changes in the cell size were observed during recording.

Data are presented as means ± SD for passive membrane and action potential properties. Differences across data for days in culture for either trigeminal or geniculate ganglia were tested using ANOVA with Scheffe's post hoc tests. Differences in electrophysiological properties between trigeminal and geniculate ganglia were evaluated using the Student's t-test. Reported differences were significant at P  0.05, but actual P values are included in the text.

In addition to the parametric statistical tests reported here, we tested for differences in neuron properties across days in culture for the trigeminal or geniculate ganglion with the Kruskal-Wallis nonparametric test and for differences between ganglia with the Mann-Whitney U test. Results for statistical significance did not differ with use of parametric or nonparametric measures.

Scanning electron microscopy

A sample of ganglion explants (n = 4 trigeminal, 3 geniculate) was prepared for examining high-resolution culture topography with scanning electron microscopy. Cultures were rinsed briefly in phosphate buffered saline and fixed in 2% paraformaldehyde and 2.5% glutaraldehyde in 0.025 M cacodylate buffer (pH 7.3) for 6 h at 4°C. Cultures were then rinsed in buffer and postfixed for 1 h each in a sequence of 1% OsO₄, 1% tannic acid, and 1% OsO₄. Dehydration was through an ascending series of alcohols, and alcohol displacement was accomplished with three changes of hexamethyldisilazane (HMDS). Residual HMDS was evaporated in a fume hood overnight before mounting cultures on specimen stubs for light sputter coating with gold/palladium.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Cellular characteristics of ganglion explants

Cultured trigeminal and geniculate ganglion explants were characterized by a core of cell soma from which a dense mat of neurites extended at 30 h in culture (Fig. 1A). From the central mass of cells, individual neurons began to migrate within 30 h of culture (Fig. 1B), and migration continued over days in culture although a central mass of neurons was retained. Ganglion neurons were clearly identified by a round to oval shape and neurite extensions (Fig. 1C). The explant perimeter had a fine network of processes (Fig. 1D), extending more than 400 µm from ganglion soma. Cultured ganglion neurons retained characteristics of actively growing cells, with neurite processes that formed varied growth-cone morphologies (Fig. 1E).

View larger version (154K): [in this window] [in a new window]   Fig. 1. A-E: scanning electron micrographs of an E16 trigeminal explant after 3 days in culture. A: trigeminal ganglion explant at low power illustrates central core of neuronal soma and dense mat of extended neurites. B: neurons migrate from the explant over days in culture. Scale = 150 µm. C: neuronal soma are ovoid or round in shape with a smooth surface and are easily distinguished in explants. Scale = 25 µm. D: perimeter of explant has fine terminations of growing neurites. Scale = 100 µm. E: growth cones of varying shapes are observed at neurite tips. Scale = 10 µm. Bottom right: graph of soma diameter (means ± SD) in E16 trigeminal and geniculate ganglion explants after 3 through 10 days in culture. Numbers of trigeminal neurons studied for each of days 3-10 were 16, 19, 24, 43, 55, 51, 26, 20; for geniculate, numbers for days 4-10 were 18, 20, 19, 28, 22, 20, 23. Diameters of trigeminal and geniculate neuronal soma increase gradually over 10 days in culture.

Soma diameter, measured for all recorded neurons, was 29 ± 4 (SD) µm (range = 18-40 µm) for trigeminal (n = 254) and 27 ± 5 µm (range = 16-44 µm) for geniculate (n = 150) ganglion neurons. A small, gradual but statistically significant (P < 0.001) increase in average soma diameter across days in culture was observed for neurons in both types of ganglia (Fig. 1, bottom right). Average trigeminal soma diameter increased from 25 ± 3.0 µm at day 3 to 32 ± 2.8 µm at day 10; and, geniculate soma from 22 ± 4.3 µm at day 4 to 29 ± 4.6 µm at day 10. Our measures of ganglion cell size overlap in range with data on adult neurons. Trigeminal soma diameters in adult rat range from 20 to 51 µm (Janigro et al. 1997; Kim and Chung 1999; Liu and Simon 1996). Acutely isolated, adult rat geniculate ganglion neurons have a reported average soma diameter of 26-30 µm (King and Bradley 2000; Koga and Bradley 2000). Cell survival in trigeminal and geniculate explants was not obviously altered with time in culture, and indeed counts of geniculate neurons in explants demonstrate that there is no decrease in total number of neurons from 1 to 12 days in culture with BDNF (Al Hadlaq et al. 2001a).

Trigeminal and geniculate neurons filled with Lucifer yellow at time of recording exhibited a densely filled soma and fine processes (Fig. 2). Branch characteristics of neurites ranged from simple, bipolar and pseudounipolar extensions to more complex, multipolar configurations. Neurite patterns for geniculate ganglion neurons were generally pseudounipolar (Fig. 2, gg) and more homogeneous than in trigeminal neurons which exhibited multipolar and bipolar branches (Fig. 2, tg).

View larger version (104K): [in this window] [in a new window]   Fig. 2. Photomicrographs of trigeminal (tg) and geniculate (gg) neurons filled with Lucifer yellow from the recording pipette, photographed under epifluorescent illumination. Trigeminal neurons had bipolar, pseudounipolar or multipolar neurite extension patterns, whereas all geniculate neurons had a pseudounipolar pattern with a long primary extension. Scale = 100 µm.

Neurophysiological properties of ganglion cells in explants cultured for 3-10 days

Whole cell recordings were made from 254 trigeminal and 150 geniculate ganglion neurons to characterize their passive membrane and action potential properties. To determine whether neurophysiological properties changed during several days in vitro, electrophysiological characteristics were measured from ganglion neurons in culture for 3-10 days for trigeminal (neuron numbers at each day = 16, 19, 24, 43, 55, 51, 26 and 20) and 4-10 days for geniculate (neuron numbers at each day = 18, 20, 19, 28, 22, 20 and 23). Results are organized to present data for trigeminal neurons across days in culture first, followed by data for geniculate neurons.

TRIGEMINAL GANGLION. Passive membrane properties. Resting membrane potential (from 60 ± 5 to 60 ± 3 mV over 3-10 days) and membrane time constant (from 31 ± 14 to 29 ± 15 ms) did not alter during 3-10 days in culture for trigeminal neurons (Fig. 3). However, a decrease in input resistance, from 552 (±227) to 313 (±154) M (P = 0.002) and increase in capacitance, from 57 (±13) to 96 (±30) pF (P < 0.001) were observed (Fig. 3), perhaps reflecting the gradual increase in average soma diameter (Fig. 1, graph). Indeed, there was an inverse correlation between soma size and input resistance (Pearson's r = 0.93, P = 0.001) and a positive correlation between soma size and capacitance (Pearson's r = 0.88, P = 0.004).

View larger version (29K): [in this window] [in a new window]   Fig. 3. Graphs of passive membrane properties for trigeminal () and geniculate () neurons over 10 days in culture. Data are presented as means and SD. Numbers of trigeminal neurons recorded in cultures at each of days 3-10 were 16, 19, 24, 43, 55, 51, 26, 20; for geniculate, numbers for days 4-10 were 18, 20, 19, 28, 22, 20, 23. These numbers apply to all graphs in Figs. 3-5. Resting membrane potential for trigeminal neurons did not alter over days in culture, whereas for geniculate ganglion neurons the membrane potential was more negative at later times. Average membrane time constant, measured by fitting a single exponential function to the charging transient of each recording, was stable over days for both trigeminal and geniculate neurons. For both ganglion types, also, there was a decrease in input resistance, determined from voltage changes in response to a series of hyperpolarizing current steps, and an increase in capacitance, determined from measures of time constant and input resistance. The magnitude of reported differences in properties over time in culture and specific days when differences were observed are discussed in the text of RESULTS.

View larger version (31K): [in this window] [in a new window]   Fig. 4. Left: whole cell recordings from trigeminal and geniculate neurons in response to a series of hyperpolarizing and depolarizing pulses shown at the bottom of the figure. The recordings illustrate generation of single (trigeminal) or multiple (geniculate) action potentials at threshold currents and a lower current threshold to generate the action potential for the geniculate neuron (after 6 current steps of 25 pA each) vs. trigeminal (after 10 current steps). Whereas trigeminal neurons always produced a single action potential at threshold, 35% of geniculate neurons had multiple action potentials. Right: graphs of action potential amplitude and thresholds of excitation (mV; pA) for trigeminal and geniculate neurons in culture (means ± SD). Action potential amplitude generated at threshold in trigeminal or geniculate neurons, measured from resting membrane to peak potential, did not alter over days in culture (1st graph). Threshold of excitation in mV at action potential generation (measured at the beginning of the action potential rise) was stable over days in trigeminal, but decreased in geniculate neurons (2nd graph). However, the current in pA to reach threshold, measured in response to a series of depolarizing current injections in 25-pA steps altered over days in both trigeminal and geniculate neurons (3rd graph). The magnitude of reported differences in properties over time in culture, and specific days when differences were observed, are discussed in RESULTS.

View larger version (30K): [in this window] [in a new window]   Fig. 5. Graphs of action potential (AP) and afterhyperpolarization (AHP) properties for trigeminal and geniculate neurons during 3-10 days in culture (means ± SD). AP half-duration, a measure of spike duration at half-amplitude, decreased in trigeminal and geniculate neurons, and in both ganglion types, the maximum left and right slopes of the AP increased in magnitude, contributing to sharper spikes. AP decay time, the duration of the falling phase of the action potential between 90 and 10% of spike amplitude, decreased somewhat in trigeminal but not in geniculate neurons. AHP amplitude, measured from resting membrane to lowest recorded potential, did not alter over days in culture for trigeminal neurons but decreased in magnitude for geniculate. In contrast, the AHP duration, from the lowest potential to level of the resting membrane potential, increased in trigeminal but did not alter in geniculate neurons. The magnitude of reported differences in properties over time in culture, and specific days when differences were observed, are discussed in RESULTS.

GENICULATE GANGLION. Passive membrane properties. The resting membrane potential of geniculate neurons became slightly more negative (from 55 ± 3 to 58 ± 4 mV, days 4-10; P = 0.007) during days in culture (Fig. 3). However, with post hoc analysis, no differences between specific days were found. As in trigeminal cells, the average membrane time constant was stable (32 ± 14 to 31 ± 14 ms; Fig. 3). A decrease in input resistance (691 ± 297 to 487 ± 186 M; P = 0.001) and increase in membrane capacitance (48 ± 20 to 64 ± 14 pF; P < 0.001) were similar to data trends in trigeminal neurons (Fig. 3) and probably reflect the increase in cell soma size. Soma size correlated inversely with input resistance (Pearson's r = 0.96, P = 0.001) and positively with capacitance (Pearson's r = 0.96, P = 0.001).

Comparisons between trigeminal and geniculate ganglion cells

To study differences in neurophysiological properties between embryonic trigeminal and geniculate neurons, data for each ganglion were pooled across culture days 5-8, when neurophysiological characteristics were relatively stable. Comparisons of passive membrane and action potential properties between ganglia are presented in Table 2.

Geniculate neurons were smaller in diameter (27 µm geniculate; 29 µm trigeminal) and had a slightly more positive resting membrane potential (57 mV geniculate; 59 mV trigeminal) than trigeminal ganglion cells. Whereas the time constant did not differ between the two ganglia (30 ms geniculate; 27 ms trigeminal), the higher input resistance (525 M geniculate; 392 M trigeminal) and lower capacitance (59 pF geniculate; 69 pF trigeminal) of geniculate cells corresponds with their smaller soma size.

Geniculate neurons also demonstrated an action potential of smaller amplitude (99 mV geniculate; 105 mV trigeminal) and shorter duration (2.2 ms geniculate; 3.1 ms trigeminal) than trigeminal. Average maximal left (145 V/s geniculate; 127 V/s trigeminal) and right slopes (47 V/s geniculate; 40 V/s trigeminal) of the action potential were steeper in geniculate neurons. Furthermore, geniculate neurons had a lower threshold of excitation (39 mV geniculate; 36 mV trigeminal) and needed less current (49 pA geniculate; 128 pA trigeminal) to reach threshold of excitation compared with trigeminal. Whereas 35% of geniculate cells were multiple spiking, all trigeminal cells generated just a single action potential at threshold level. In general, comparisons of action potential properties suggest that geniculate neurons are more excitable than trigeminal at this embryonic stage.

AHP characteristics were not investigated in detail, but both ganglion types had long AHPs. However, geniculate neurons had an AHP that was smaller in amplitude (9 mV geniculate; 11 mV trigeminal) and shorter in duration (88 ms geniculate; 135 ms trigeminal) than trigeminal cells. This would suggest a more rapid recovery for subsequent depolarization in keeping with the multiple spiking characteristics of many geniculate cells.

In summary, virtually all membrane and action potential characteristics differed substantially between embryonic trigeminal and geniculate ganglion neurons. To learn whether subgroups of neurons were present within either ganglion type based on electrophysiology, frequency histograms were generated for each neurophysiological property from either trigeminal or geniculate recordings. No distributions were seen to suggest obvious subpopulations of neurons in E16 ganglia (data not shown).

Ionic currents in trigeminal and geniculate neurons

Observed differences in passive membrane and action potential properties between embryonic trigeminal and geniculate ganglion neurons suggest that different ionic currents are expressed in these two ganglia. We used several ion channel blockers to determine and compare sodium and calcium currents in trigeminal and geniculate ganglion cells. Potassium currents were not studied in detail. However, all trigeminal and geniculate cells had a time-dependent "depolarizing sag" during hyperpolarization, reflecting the activation of an inward rectifying current (Figs. 4 and 6). This current, presumably caused by mixed cation conductance, was reversibly abolished after the application of 1-3 mM Cs⁺ in eight experiments each with trigeminal and geniculate neurons (Fig. 6).

View larger version (17K): [in this window] [in a new window]   Fig. 6. Whole cell voltage (top) and current (bottom) recordings from a trigeminal ganglion neuron in control bath solution before and during application of Cs+. A pronounced inward rectifier response to 150- and 200-pA hyperpolarizing current is seen in the voltage "sag" in the control condition and is eliminated with application of Cs+. The voltage sag, which reflects activation of an inward rectifying current, was eliminated in all recordings from 8 trigeminal and 8 geniculate neurons, after application of 1-3 mM Cs+.

SODIUM CURRENTS. To compare currents involved in generation of action potentials in cultured embryonic trigeminal and geniculate neurons, we examined effects of the specific sodium channel blocker, TTX, in recordings from 14 geniculate and 18 trigeminal neurons. Concentrations of 0.3-3 µM TTX for geniculate and 3-10 µM for trigeminal cells were used. In all geniculate neurons application of TTX abolished the generation of action potentials (Fig. 7). This effect was completely reversible.

View larger version (35K): [in this window] [in a new window]   Fig. 7. Whole cell recordings from geniculate and trigeminal neurons in control bathing solution and after application of TTX and in control and sodium-free bathing solutions (trigeminal only, at bottom of figure). TTX (0.3-3 µM) eliminated the ability to generate an action potential in all 14 recorded geniculate neurons. However, in all 18 recorded trigeminal neurons even higher concentrations of TTX (3-10 µM) did not eliminate the action potential, but only increased action potential threshold. A Na+-free solution did eliminate action potential generation in all 8 recorded trigeminal ganglion neurons; therefore sodium-sensitive currents do participate in action potential generation in trigeminal neurons.

CALCIUM CURRENTS. All E16 trigeminal and most geniculate (95%) neurons have an inflection (or "hump") in the falling phase of the action potential. However, these inflections have a different shape in the two types of neurons. Maximal slopes of "upper" and "lower" levels of the action potential falling phase usually are comparable in trigeminal neurons (Fig. 8). In contrast, for geniculate neurons the slope closer to the peak of the action potential usually is considerably steeper than at the lower level (Fig. 8). The first-order derivative of these slopes clearly illustrates this difference (Fig. 8, bottom). Because currents through voltage-gated calcium channels contribute to formation of this inflection (Gallego 1983; Schild et al. 1994), we investigated the effect of voltage-gated calcium channel blockers on action potential properties of trigeminal and geniculate neurons.

View larger version (12K): [in this window] [in a new window]   Fig. 8. Top: whole cell recordings of action potentials from trigeminal and geniculate neurons illustrate the inflection or "hump" on the repolarizing phase of the action potential that was typically observed in both ganglion types. Bold and fine arrows on the trigeminal action potential indicate steep slopes on either side of the inflection. On the geniculate action potential, however, there is a steeper slope in the early repolarizing phase (bold arrow) than at the later phase (finer arrow). Bottom: first-order derivatives of the action potential data more clearly illustrate differences in the inflections between trigeminal and geniculate neurons. The action potentials were evoked in response to a short current step as described in METHODS.

View larger version (14K): [in this window] [in a new window]   Fig. 9. Effects of L-type calcium channel blockers, verapamil and nifedipine, on the repolarizing inflection of the action potential in trigeminal and geniculate neurons. Superimposed recordings of action potentials before and after application of verapamil illustrate similar effects between ganglia. In all 10 trigeminal and 4 geniculate recordings, verapamil reduced or eliminated the repolarization inflection. In contrast, the more specific blocker, nifedipine, has a greater effect on geniculate (n = 4 neurons) than on trigeminal (n = 4) action potentials. The action potentials were evoked in response to a short current step as described in METHODS.

View larger version (18K): [in this window] [in a new window]   Fig. 10. Superimposed whole cell recordings to illustrate effects of the N-type calcium channel blocker, -conotoxin-MVIIA (top), and sequential application of the N-type (-conotoxin-GVIA) and P/Q-type (-agatoxin-TK) channel blockers (bottom) on the action potential inflection in trigeminal and geniculate neurons. For all recorded neurons (13 trigeminal and 7 geniculate) 0.5-2 µM -conotoxin-MVIIA reduced the inflection. Application of the N-type, calcium channel blocker, -conotoxin-GVIA at 1 µM also reduced the action potential inflection in all 13 trigeminal and 5 geniculate neurons. These results indicate a contribution from conotoxin-sensitive N-type calcium channels to the potential inflection in both ganglion types. After blocking N-type channels with -conotoxin-GVIA in 9 trigeminal and 3 geniculate recordings and then applying 200 nM of the P/Q-type channel blocker -agatoxin-TK, a further effect on the action potential inflection was observed. This suggests some participation of P/Q-type calcium channels in action potential repolarization in embryonic trigeminal and geniculate neurons.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In explant cultures of embryonic trigeminal and geniculate ganglia, neurons extend neurites with complex terminal extensions over several days, and a number of electrophysiological characteristics gradually alter. Over a period of 5-8 days, neural properties are relatively stable. Therefore this system allows neurophysiological comparisons between different embryonic ganglia in an in vitro environment that supports biological function. The data demonstrate that explants of trigeminal and geniculate ganglia have very different passive membrane and action potential properties at a stage well in advance of differentiation of their respective target sensory organs. This indicates that differences in neural characteristics emerge without electrophysiological input from differentiated receptors.

By E16, in vivo neurons from both ganglia have extended neurites to densely innervate the fungiform papillae on anterior tongue (Mbiene and Mistretta 1997). Although geniculate neurons do not yet encounter specialized target organs, because taste buds have not developed, trigeminal and geniculate neurites already occupy essentially separate tissue locations in papillae (Farbman and Mbiene 1991). Therefore these neurons have differentiated in numerous ways from ganglion emergence at E10 in rat, and time of peak neuron production in the trigeminal at E14 and geniculate at E15 (Altman and Bayer 1982), to E16 in rat. Although further functional and morphological differentiation is expected during the lengthy remaining period of embryonic and postnatal development of tongue and taste target organs (Mistretta 1991; Mistretta and Hill 1995), by E16, these two sensory ganglia already have very different neurophysiological characteristics when studied in an explant system.

To compare functional properties between explanted embryonic trigeminal and geniculate ganglia, it was essential to maintain maximal numbers of neurons. Therefore we added specific neurotrophins to the culture medium for each ganglion type: NGF maximally supports rodent trigeminal ganglia at later embryonic stages, after E13 in mouse (Buchman and Davies 1993; Davies 1997), and the NGF receptor TrkA mRNA or the TrkA protein is the predominant trk receptor in the mouse trigeminal from E13.5 (Wyatt and Davies 1993; Huang et al. 1999). BDNF is optimal for survival of the geniculate ganglion at E16 in rat (Al Hadlaq et al. 2001a), and most or all cells express the BDNF receptor, TrkB mRNA, in rat geniculate at E16-18 (Ernfors et al. 1992). If NGF was used to support both neuron types at E16, only about 11% of the "original" population of geniculate ganglion neurons would survive for recording. If BDNF was used for both ganglia, only about 10-25% of trigeminal neurons would survive over days in explant (based on extrapolation from mouse data at E14, Paul and Davies 1995; rat data at E14, Ibanez et al. 1993). These maximally supportive neurotrophins also are those localized at high concentration in the innervating fields of the respective ganglia (Nosrat and Olson 1995; Nosrat et al. 2001). On the other hand, mRNA for NGF is not within mouse trigeminal ganglion at E15.5 (Schecterson and Bothwell 1992), whereas data conflict for BDNF mRNA in the geniculate ganglion on reported presence (E15.5 mouse) (Schecterson and Bothwell 1992) or absence of expression (E16-18 rat) (Ernfors et al. 1992). By extrapolation from mouse data, in the rat trigeminal ganglion, E16 would very probably be a period of programmed cell death, presumably due to competition for limited amounts of NGF (Figueiredo et al. 2001). It is likely, therefore, that naturally occurring cell death is prevented in our explant systems.

Use of other neurotrophins in the explant medium would predictably alter ganglion neurophysiological properties (data on geniculate ganglion) (Al Hadlaq et al., 2001b). However, our work with the geniculate ganglion demonstrates that when the geniculate is supported in culture with NGF, rather than BDNF, neuronal properties do not become more similar to those of the trigeminal ganglion maintained with NGF. We designed experiments in the present paper to compare embryonic trigeminal and geniculate neurons when sustained with neurotrophins that would support most neurons and that would best mimic the biological receptive field in vivo. Further experiments to compare properties for either the trigeminal or geniculate ganglion with varying neurotrophin support can be informed by data from the current study.

Use of ganglion explant cultures, and neuron morphology

To study neural properties of developing trigeminal and geniculate ganglia, we used explants of entire or half ganglia instead of dissociated cells to maintain a cellular and molecular environment in ganglion cultures that is more similar to that in vivo. The significance of using ganglion explants is clear from studies of developmental changes in neurotrophin dependence in embryonic trigeminal ganglion, demonstrating emergence of different neurotrophin responses in explant versus dissociated neurons (Enokido et al. 1999). In addition, embryonic DRG neurons maintained in explants have more uniform axon morphology than dissociated cells (Ichinose and Snider 2000). Acute isolation procedures for dissociating neurons also may modify properties of membrane proteins, including ion channels (Molitor and Manis 1999) and can eliminate receptors and channels clustered in distal neurite membranes (Du and Bradley 1996). Furthermore, in dissociated E14 spinal motor neurons grown in culture, sodium currents are only detected in association with neurite outgrowth (Allesandri-Haber et al. 1999).

Neurons in explant cultures of embryonic trigeminal and geniculate ganglia had rounded or ovoid soma and neurite morphologies that included pseudounipolar, bipolar, and multipolar extensions. This varied neuron phenotype is similar to reports of embryonic and neonatal rat dorsal root (Matsuda and Uehara 1984; Matsuda et al. 1996) and nodose ganglia (De Koninck et al. 1994) and chick dorsal root ganglia in vivo and in culture (Matsuda et al. 1996; Riederer and Barakat-Walker 1992). In sensory ganglia in vivo, an initial embryonic, bipolar neurite morphology develops to form the pseudounipolar configuration (Lieberman 1976; Matsuda and Uehara 1984), so embryonic ganglia contain both bipolar and pseudounipolar neurons. Multipolar neurons also are sometimes noted (Lieberman 1976), especially in younger rat embryos (E14-15) (Matsuda and Uehara 1984).

Culture conditions can alter morphology of embryonic neurons. For example, dissociated neonatal rat, nodose neurons all acquire a multipolar morphology when cultured without ganglion satellite cells (De Koninck et al. 1994), and whereas dissociated adult leech ganglion neurons cultured on native substrate have a pseudounipolar morphology, experimental substrates result in neurite retraction and altered patterns (De Miguel and Vargas 2000). Because we studied single neurons in whole (geniculate) or half (trigeminal) ganglion explants and used a matrix deposited by living cells (MacCallum et al. 1982), our explant cultures can retain ganglion and matrix properties that are lost in dissociated preparations.

We recorded from neurons that had migrated from the central core of explanted trigeminal or geniculate ganglion. Molecules in the neuronal matrix environment can alter cell cytoskeletal components that in turn affect cell motility (Song and Poo 2001). We know of no studies that directly address effects of migration on basic neurophysiological properties of sensory neurons but cannot exclude the possibility of altered properties with migration.

Passive membrane and action potential properties of trigeminal and geniculate ganglion neurons alter gradually over time in explant culture

In cultured explants of embryonic trigeminal and geniculate ganglia, passive membrane properties altered gradually over several days in culture. Whereas resting membrane potential and time constant did not change, resistance decreased and capacitance increased in association with a small progressive increase in soma size. Several action potential properties also differed with days in culture. Action potential duration decreased and rise and fall slopes increased so that action potentials were shorter and sharper. Amplitude remained constant but threshold current needed to elicit the spike increased slightly in both trigeminal and geniculate neurons.

The membrane and action potential differences over days in culture are similar to the reported increase in soma size, decreased membrane resistance, and increased capacitance and shorter, more rapidly rising action potentials observed in cultured, embryonic hippocampal (Porter et al. 1997), septal (Thinschmidt et al. 1999), and neocortical neurons (Yamada et al. 1999), and in postnatal brain slice preparations of rat nucleus accumbens (Belleau and Warren 2000). However, the differences for each neuron type, trigeminal or geniculate, are often small, always very gradual, and statistical significance was achieved primarily through contribution of data from days 9 and 10 in culture. Explant cultures fed continuously with the appropriate neurotrophin to maximize survival and neurite growth can provide an optimal environment for neuron maintenance, and our data demonstrate the utility of this system for studying and comparing ganglion properties over a period of several days in culture.

Embryonic trigeminal and geniculate ganglion neurons have different passive membrane properties

Passive membrane and action potential properties of cultured E16 trigeminal and geniculate neurons are significantly and substantially different from each other even at this relatively early embryonic age. Although the mature target fields of trigeminal and geniculate ganglion neurons are in direct proximity within the fungiform papillae, the trigeminal and geniculate neurons innervate distinct specialized receptors in the rat tongue and have different embryonic origins (Graham and Begbie 2000), so neurophysiological differences might be expected. However, we have found that differences in basic electrophysiological properties of neurons in these two sensory ganglia are established well in advance of peripheral maturation of the taste system and do not present a homogeneous neurophysiological profile at E16.

Our data indicate that trigeminal ganglion neurons in the rat embryo have larger average soma size, lower resting membrane potential, lower input resistance, and shorter time constant compared with geniculate neurons. There are some data from postnatal and adult rodent to indicate that the direction of these differences is similar to that in mature ganglia. For example, as reported in RESULTS, embryonic soma diameters overlap with ranges reported in postnatal animals (Janigro et al. 1997; Kim and Chung 1999; King and Bradley 2000; Koga and Bradley 2000; Liu and Simon 1996) and indicate smaller soma in adult geniculate than in trigeminal ganglion. The average resting membrane potential of embryonic trigeminal neurons in our experiments, 59 mV, is between those reported from studies of adult rat trigeminal, 62 mV (Janigro et al. 1997) and 52 mV (Liu at al. 2001). Similarly, the average resting membrane potential of 57 mV from embryonic geniculate neurons in our experiments is between values reported from acutely isolated geniculate neurons from postnatal and adult rat, 55 mV (King and Bradley 2000) and 64 mV (Koga and Bradley 2000). The average input resistance in E16 geniculate neurons (525 M) also is between values reported postnatally, 336 M (King and Bradley 2000) and 574 M (Koga and Bradley 2000). Reported membrane time constant averages 25 ms in adult (Koga and Bradley 2000) compared with 30 ms in embryonic geniculate.

Different action potential and discharge properties in embryonic trigeminal and geniculate ganglion neurons

Trigeminal and geniculate ganglion neurons also have very different action potential properties at E16. However, without a direct comparison between postnatal trigeminal and geniculate ganglia, it is not clear whether these particular properties are still as substantially different after birth. Average action potential amplitude for our embryonic trigeminal neurons (105 mV) is lower than averages for various neuron types in adult rat trigeminal (113-128 mV) (Liu et al. 2001) and the threshold in E16 neurons (36 mV) is lower than in adult (12 to 27 mV) (Liu et al. 2001). This suggests that embryonic trigeminal action potential properties will have a considerable postnatal maturation. There are no data for adult geniculate to compare to our embryonic data, but substantial maturation is expected as in trigeminal.

In E16 explant cultures, geniculate action potentials are of lower amplitude, shorter duration, and steeper rise time than trigeminal. Furthermore, whereas all trigeminal neurons have one action potential at threshold, a large proportion of geniculate neurons generates multiple spikes. Embryonic geniculate neurons, in comparison to trigeminal, also have a much lower threshold for action potential generation by depolarizing current and a substantially shorter AHP. Lower threshold and shorter AHP are properties that also characterize repetitively firing neurons in the adult DRG, compared with single-spiking neurons (Villiere and McLachlan 1996).

The contrast in embryonic trigeminal and geniculate ganglia in discharge properties is seen also in adult ganglia. Depolarization of trigeminal neurons from adult rat elicited only one or two spikes (Janigro et al. 1997) similar to the single spikes from embryonic neurons in our experiments. In contrast, large proportions of geniculate neurons from adult rat were multiple spiking (King and Bradley 2000; Koga and Bradley 2000); again this is similar to our data on embryonic geniculate neurons. Furthermore, adult trigeminal neurons also required much greater current (Janigro et al. 1997) to elicit an action potential than geniculate (King and Bradley 2000; Koga and Bradley 2000).

In summary, some of the observed, distinctive differences in action potential properties between postnatal and adult geniculate and trigeminal ganglion cells, including discharge properties and threshold of excitation, are present already at E16 in rat. Maturation of geniculate ganglion cell innervation patterns of taste buds in fungiform papillae is a prolonged process, through 40 days postnatal in rat (Krimm and Hill 1998, 2000). Furthermore, taste responses from the chorda tympani nerve, which derives from geniculate ganglion cells, alter dramatically during the first postnatal month (Hill 2001). The extent to which these developmental changes in ganglion cell innervation of taste buds and papillae and in afferent response properties relate to electrophysiological properties of the ganglion neurons themselves is not known. However, our experiments have shown that several basic electrophysiological properties of ganglion cells are established well in advance of mature patterns of target receptor organ innervation.

Sodium currents and TTX sensitivity in embryonic ganglia

Distinct ionic currents must contribute to the different passive membrane and action potential properties of trigeminal and geniculate ganglion neurons. Differences in the sodium currents participating in action potential generation are revealed by sensitivity to TTX. Although action potentials of mature sensory neurons can exhibit differing sensitivity to TTX, in general the proportion of neurons with TTX-sensitive compared with TTX-resistant action potentials increases during maturation (Koerber and Mendell 1992; Omri and Meiri 1990; Roy and Narahashi 1992).

However, we observed important differences in sensitivity to TTX between embryonic trigeminal and geniculate neurons. High concentrations of TTX increased the threshold of excitation for E16 trigeminal neurons, but ability to spike was always sustained. In contrast, all geniculate neurons lacked ability to spike after treatment with low concentrations of TTX; that is, all neurons had TTX-sensitive action potentials. In adult rat trigeminal neurons, sodium current recordings demonstrated presence of both TTX-sensitive and -resistant channels (Kim and Chung 1999; Liu et al. 2001), similar to our findings for embryonic trigeminal ganglion where action potentials are resistant to TTX but the threshold to spike is increased.

No other data are published on sodium currents in geniculate neurons. In other cultured cranial sensory ganglia of rat (petrosal and nodose), about 90% of neurons possessed both TTX-sensitive and -resistant components of sodium current (Bossu and Feltz 1984). The complete TTX-sensitivity for action potentials in all of the E16 geniculate neurons that we studied seems distinctive, but we cannot exclude a possible developmental addition of TTX-resistant currents in later embryos or postnatally.

In adult DRG neurons, large-diameter cells have only TTX-sensitive currents, whereas TTX-resistant currents are found only in small diameter neurons (Caffrey et al. 1992; Elliot and Elliott 1993; McLean et al. 1988; Strassman and Raymond 1999). However, some small cells can also contain TTX-sensitive currents (Villiere and McLachlan 1996). Recently a direct role for the specifically characterized TTX-resistant sodium channel, Na_v1.8, has been demonstrated in action potential generation in small, C-type, mouse DRG neurons (Renganathan et al. 2001). Furthermore, the Na_v1.8 channel is expressed as early as E15 in rat, reaches adult levels postnatally, and is distributed mainly in DRG neurons that have unmyelinated C fibers (Benn et al. 2001). Because nociceptive neurons comprise a major population of small sensory neurons, the distinctive TTX-resistant sodium currents that are particularly characteristic of small diameter sensory neurons are thought to be important in nociception. Furthermore, polymodal nociceptive, C type neurons in adult mouse cornea have broad, TTX-resistant action potentials with a hump on the repolarization phase (De Armentia et al. 2000). These are characteristics of E16 trigeminal neurons. One of the principal sensory functions of the trigeminal ganglion is transmission of nociceptive information, so the potential functional role of TTX-resistant sodium currents in trigeminal should be further explored.

Calcium currents in embryonic trigeminal and geniculate neurons

Calcium plays an important role in development of neurons. For example, neurite outgrowth, growth cone motility and synaptogenesis all depend on changes in calcium conductances, and voltage-gated calcium channels provide the major pathway for calcium ions into the cell (Mattson and Kater 1987; Turrigiano et al. 1995