Download Biophysical Properties and Responses to Neurotransmitters of

Biophysical Properties and Responses to Neurotransmitters of
Petrosal and Geniculate Ganglion Neurons Innervating the Tongue
TOMOSHIGE KOGA1 AND ROBERT M. BRADLEY1,2
Department of Biologic and Materials Sciences, School of Dentistry, University of Michigan, Ann Arbor 48109-1078; and
2
Department of Physiology, Medical School, University of Michigan, Ann Arbor, Michigan 48109-0622
1
Received 3 February 2000; accepted in final form 30 May 2000
Investigators of the peripheral gustatory system have examined the properties of afferent taste fibers by extracellular
recording from dissected single fibers. This classical approach
has revealed a wealth of information on the response characteristics of afferent taste fibers (for a review see Smith and
Frank 1993). In contrast, the current knowledge of the biophysical and neurochemical properties of afferent taste fibers is
less complete because of the practical difficulty of examining
these fibers in vivo. Researchers of nociceptors facing a similar
problem have made profitable use of in vitro techniques to
define the biophysical and chemical sensitivity of afferent
sensory neurons supplying cutaneous receptors because the
properties of the ganglion cells of these fibers reflects the
properties of the afferent fibers and endings (Dray 1996). The
goal of the present investigation was to apply similar techniques to afferent sensory neurons supplying taste buds to
characterize both the biophysical properties and responses of
these neurons to neurotransmitters that have been shown using
anatomical techniques to be associated with taste receptors
(Nagai et al. 1996).
Afferent fibers transmitting gustatory information from the
anterior tongue and soft palate travel in the chorda tympani and
greater superficial petrosal nerves with cell bodies in the geniculate ganglion (GG), while taste buds on the posterior tongue
are supplied by the glossopharyngeal nerves with cell bodies in
the petrosal ganglion (PG). Electrophysiological studies of the
chorda tympani, greater superficial petrosal, and glossopharyngeal nerves have revealed that these nerves have heterogeneous
response properties (Frank 1991; Frank et al. 1988; Nejad
1986; Ogawa et al. 1968), but it is not known if these differences are reflected in the biophysical properties of the ganglion
cells. In addition to having heterogeneous response properties,
afferent taste neurons also express multiple putative neurotransmitters and neuromodulators. For example, the PG has
been shown to immunostain for substance P (SP), tyrosine
hydroxylase, vasoactive intestinal polypeptide, calcitonin
gene-related peptide, galanin, glutamate, and asparate
(Czyzyk-Kreska et al. 1991; Finley et al. 1992; Helke and Hill
1988; Helke and Niederer 1990; Helke and Rabchevsky 1991;
Ichikawa et al. 1991; Okada and Miura 1992).
Not only have these neurochemicals been shown to be
present in the PG and GG, they have been implicated in taste
function at both the peripheral and central processes of the
peripheral taste system. Serotonin (5HT) is believed to modulate chemosensory responses of taste receptor cells (Delay and
Roper 1988; Nagai et al. 1998) and acetylcholine (ACh), ␥
aminobutyric acid (GABA), and SP have been suggested to
function at the synapse between taste buds and primary afferent
neurons (Jain and Roper 1991; Nagai et al. 1986; Nagy 1982;
Paran and Mattern 1975). Moreover, PG neurons respond to
ACh at concentrations in the physiological range (Zhong and
Nurse 1997). At the first relay in the central taste pathway,
GABA and SP have been shown to function as inhibitory and
Address for reprint requests: R. M. Bradley, Dept. of Biologic and Materials
Sciences, Room 6228, School of Dentistry, University of Michigan, Ann
Arbor, MI 48109-1078 (E-mail: [email protected]).
The costs of publication of this article were defrayed in part by the payment
of page charges. The article must therefore be hereby marked ‘‘advertisement’’
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
INTRODUCTION
1404
0022-3077/00 $5.00 Copyright © 2000 The American Physiological Society
www.jn.physiology.org
Downloaded from http://jn.physiology.org/ by 10.220.32.246 on April 28, 2017
Koga, Tomoshige and Robert M. Bradley. Biophysical properties
and responses to neurotransmitters of petrosal and geniculate ganglion
neurons innervating the tongue. J Neurophysiol 84: 1404 –1413, 2000.
The properties of afferent sensory neurons supplying taste receptors
on the tongue were examined in vitro. Neurons in the geniculate (GG)
and petrosal ganglia (PG) supplying the tongue were fluorescently
labeled, acutely dissociated, and then analyzed using patch-clamp
recording. Measurement of the dissociated neurons revealed that PG
neurons were significantly larger than GG neurons. The active and
passive membrane properties of these ganglion neurons were examined and compared with each other. There were significant differences
between the properties of neurons in the PG and GG ganglia. The
mean membrane time constant, spike threshold, action potential halfwidth, and action potential decay time of GG neurons was significantly less than those of PG neurons. Neurons in the PG had action
potentials that had a fast rise and fall time (sharp action potentials) as
well as action potentials with a deflection or hump on the falling phase
(humped action potentials), whereas action potentials of GG neurons
were all sharp. There were also significant differences in the response
of PG and GG neurons to the application of acetylcholine (ACh),
serotonin (5HT), substance P (SP), and GABA. Whereas PG neurons
responded to ACh, 5HT, SP, and GABA, GG neurons only responded
to SP and GABA. In addition, the properties of GG neurons were
more homogeneous than those of the PG because all the GG neurons
had sharp spikes and when responses to neurotransmitters occurred,
either all or most of the neurons responded. These differences between
neurons of the GG and PG may relate to the type of receptor innervated. PG ganglion neurons innervate a number of receptor types on
the posterior tongue and have more heterogeneous properties, while
GG neurons predominantly innervate taste buds and have more homogeneous properties.
PROPERTIES OF SENSORY NEURONS INNERVATING THE TONGUE
excitatory neurotransmitters at the synapse of the primary
afferent fibers in the nucleus of solitary tract (Davis and Smith
1997; Du and Bradley 1998; Grabauskas and Bradley 1996,
1998a; King et al. 1993; Wang and Bradley 1993).
To gain insight into the biophysical properties of afferent
neurons supplying taste receptors, we have used acutely isolated GG and PG neurons innervating the tongue that were
retrogradely labeled by Fluorogold injected into the area of the
fungiform, foliate, and circumvallate papillae. We have also
characterized the responses of PG and GG neurons to application of 5HT, SP, ACh, and GABA, which have been demonstrated to be involved in the afferent taste system.
METHODS
Cell labeling and isolation
clamp recording procedures with an Axoclamp 2A amplifier (Axon
Instruments). Current- and voltage-clamp protocols, data acquisition,
and analysis were performed using pCLAMP software (Axon Instruments). Bridge balance was carefully monitored throughout the experiments and adjusted when necessary. The junction potential due to
potassium gluconate (10 mV) was subtracted from the recorded membrane voltages (Standen and Stanfield 1992). Both Fluorogold and
nonlabeled neurons were investigated to determine any possible effects of the label on the recordings. Criteria for a successful recording
included a minimum of 10 min recording time with a stable resting
membrane potential of ⬎ ⫺40 mV and an action potential amplitude
⬎70 mV. The statistics are expressed as mean ⫾ SD and differences
between groups measured with the Students t-test were considered
significant when P ⱕ 0.05.
Drug application
Neurotransmitters or neuromodulators were prepared daily for experiments from stock solution in HEPES buffer stored at ⫺80°C and
diluted to the desired final concentration just before use. A three- or
seven-barrel pipette filled with a different concentration of drug was
positioned ⬃40 ␮m from the neuronal cell body. SP (0.1 ␮M-1 mM),
ACh (1 ␮M-1 mM), 5HT (0.01 ␮M-1 mM), GABA (1 ␮M-1 mM),
the gamma aminobutyric acid-A (GABAA) agonist muscimol (1
␮M-1 mM), or the gamma aminobutyric acid-B (GABAB) agonist
baclofen (0.1–2 mM) was ejected from the pipette using a Picospritzer
at a low pressure (2 psi). Concentration-response curves were fitted
using the Hill equation. All experiments were conducted at room
temperature.
RESULTS
Identification of neurons labeled with Fluorogold
The PG and GG were surgically removed 3–12 days after
injection of Fluorogold into their receptive fields. Dissociated
PG and GG neurons were round or ovoid in shape and some of
them had short processes. The lengths of the long axis of the
isolated PG and GG neurons were 32.2 ⫾ 4.4 and 26.0 ⫾ 2.6
␮m, respectively (mean ⫾ SD), while their short axes were
28.6 ⫾ 4.2 and 23.0 ⫾ 2.6 ␮m. The average diameter of the PG
neurons was significantly larger than those of the GG (P ⬍
0.001) (Fig. 3A). Figure 1 is a photomicrograph of PG neurons
viewed under normal (A) and fluorescent illumination (B). Two
of these neurons were strongly fluorescent. Once a labeled
Electrophysiological recording
Patch electrodes, pulled in two stages from 1.5-mm OD borosilicate
filament glass, were filled with a solution containing (in mM) 130
potassium gluconate, 10 HEPES, 10 EGTA, 1 MgCl2, 1 CaCl2, and 2
ATP. The pipette solution was adjusted to a pH of 7.2 with KOH.
Electrode resistance was between 5 and 8 M⍀. Recordings were
obtained between 30 min and 3 h after plating. The petri dish containing the neurons was mounted on the stage of an inverted microscope equipped with epifluorescence and Hoffman modulation optics.
Fluorogold-labeled neurons were identified by fluorescence with an
exposure of sufficient length to identify labeled neurons.
Electrodes were manipulated under visual control and membrane
potential and currents were recorded using standard whole-cell patch-
FIG. 1. Dissociated petrosal ganglion neurons viewed under tungsten (A)
and fluorescence (B) illumination. Two of these cells were identified as
Fluorogold labeled neurons.
Downloaded from http://jn.physiology.org/ by 10.220.32.246 on April 28, 2017
The cell labeling technique is based on methods developed in an
earlier study (Bradley et al. 1985). Female Sprague-Dawley rats aged
20 – 40 days were anesthetized with an intramuscular injection of
mixture of Rompun (10 mg/kg) and ketamine hydrochloride (90
mg/kg). The lower jaw was retracted and the tongue depressed. Using
a dissecting microscope, the single circumvallate papilla was visualized and 10 –15 ␮l of 5% solution of Fluorogold (Fluorochrome,
Englewood, CO) mixed in distilled water was injected just beneath the
papilla from a Hamilton syringe. To study GG neurons innervating the
anterior tongue, 5–10 ␮l of 5% solution of Fluorogold was injected
bilaterally just beneath both the foliate and fungiform papillae.
After a postoperative survival time of 3–12 days, the rats were
reanesthetized with sodium pentobarbital (50 mg/kg, ip) and decapitated. The skull was opened and the forebrain removed just rostral to
the brainstem. Using the exit of the facial nerve from the brainstem as
a landmark, the petrous portion of the temporal bone was removed to
expose the GG which was then excised. The PG was exposed in the
neck by following the glossopharyngeal nerve centrally and then
removed. The GG or PG were placed in HEPES buffer containing (in
mM) 124 NaCl, 5 KCl, 5 MgCl2, 10 sodium succinate, 15 dextrose, 15
HEPES, and 2 CaCl2, and gassed with O2. The ganglia were transferred to HEPES buffer containing 0.5 mg/ml trypsin (type III) and
0.5 mg/ml collagenase (type IVA) and incubated for 1 h at 37°C. For
incubation of the GG, the enzyme concentrations were reduced by one
half. The ganglia were then triturated with a series of progressively
smaller diameter, fire-polished Pasteur pipettes to produce a suspension of dissociated neurons which were placed in a 35-mm-diameter
plastic petri dish. The cell suspension was continuously perfused at
about 2 ml/min with oxygenated HEPES buffer by gravity flow and
the fluid level in the recording chamber was maintained relatively
constant by suction of excess fluid. After isolation, the majority of
cells were spherical, devoid of processes, and became loosely attached
to the substrate after 10 min. The enzymes were obtained from Sigma
and prepared before each experiment.
1405
1406
T. KOGA AND R. M. BRADLEY
neuron was identified, whole-cell recordings were made under
normal illumination.
Biophysical properties of labeled dissociated cells
FIG. 2. Recordings from Fluorogold labeled neurons dissociated from the petrosal (A) and geniculate (B) ganglia. Aa: an
example of a hump neuron with a deflection or hump on the descending limb of the action potential. Ab: a nonhump neuron with
a narrow action potential. (i): neuronal responses to a depolarizing current injection applied through the patch electrode. (ii):
membrane responses to a series of depolarizing and hyperpolarizing current pulses. (iii): membrane responses to a ⫺500 pA of
hyperpolarizing current pulse. Regardless of having humped or narrow spikes, all neurons had a pronounced inward rectifier
response to current injection of ⫺500 pA. DT, time between 90% and 10% of the action potential amplitude; HD, duration
measured at half height of action potential.
Downloaded from http://jn.physiology.org/ by 10.220.32.246 on April 28, 2017
Recordings were made from 130 labeled PG and 103 labeled
GG neurons. Recordings were also made from nonlabeled PG
(n ⫽ 45) and GG (n ⫽ 26) neurons to determine if they were
different from the labeled neurons. No neurons were spontaneously active. Depolarizing and hyperpolarizing currents
were injected to investigate action potential and passive membrane properties, respectively.
Since the ganglion cells were obtained from animals aged
20 – 40 days, it is possible that developmental changes could be
still occurring to the peripheral taste system in these animals.
The full complement of rat fungiform taste buds is present by
20 days (Mistretta 1972), but taste buds in the circumvallate
papilla reach adult numbers at 90 days (Hosley and Oakley
1987). Thus, at the ages studied, the receptive field of the GG
neurons was probably not changing, whereas significant
changes were ongoing in the receptive field of PG neurons.
How this impacts on the biophysical properties of the ganglion
cells is not known, but no systematic differences due to animal
age were measured.
Resting membrane potentials ranged
from ⫺40 to ⫺78 mV, with a mean of ⫺59 ⫾ 8 mV. Input
resistance ranged from 300 to 951 M⍀, with a mean of 542 ⫾
147 M⍀. Membrane time constants, measured after a ⫺50 pA
current was injected, averaged 34.5 ⫾ 10.2 ms. All the neurons
recorded exhibited varying degrees of membrane rectification
evident from their “sagging” voltage responses to intracellular
injections of hyperpolarizing current pulses (Fig. 2).
Mean threshold for generation of action potential was 122
pA, which was measured in each neuron by increasing 10 pA
steps of depolarizing current. Action potential amplitudes were
between 72 and 128 mV with mean of 101 ⫾ 13 mV. As
previously described in other ganglion neurons (Gallego and
Eyzaguirre 1978; Jaffe and Sampson 1976), the action potentials recorded from many neurons had a deflection or “hump”
on the repolarization phase of the spike evoked by intracellular
depolarizing current injection (Fig. 2Aa). Fifty-one of 130
labeled neurons were classified as hump neurons. A hump
neuron generally had a longer spike half-width when compared
with a sharp or nonhump neuron. Thus, duration measured at
half-amplitude of hump neurons (half duration: HD) was significantly greater than that of nonhump neurons (Table 1).
Additionally, the repolarizing time of hump neurons from 90 to
LABELED PG NEURONS.
PROPERTIES OF SENSORY NEURONS INNERVATING THE TONGUE
TABLE
1407
1. Intrinsic membrane properties of hump and nonhump neurons in Fluorogold labeled petrosal ganglion neurons
Hump neurons
(n ⫽ 51)
Nonhump neurons
(n ⫽ 79)
Cell
Diameter
(␮m)
RMP
(mV)
Time
Constant
(ms)
Input
Resistance
(M⍀)
Spike
Threshold
(pA)
Spike
Amplitude
(mV)
Half-Width
(ms)
Decay
Time
(ms)
Spike Area
(mV ⫻ ms)
32.8 ⫾ 4.2
⫺56 ⫾ 7
37.4 ⫾ 10.5
520 ⫾ 120
110 ⫾ 71
102.3 ⫾ 13.4
5.1 ⫾ 1.6
5.9 ⫾ 2.1
769 ⫾ 320
31.8 ⫾ 4.6
⫺61 ⫾ 8
32.7 ⫾ 9.6
556 ⫾ 161
100 ⫾ 92
100.7 ⫾ 12
2.5 ⫾ 1*
2.5 ⫾ 1*
591 ⫾ 317*
Values are expressed as mean ⫾ SD. * Significantly different at P ⬍ 0.01.
pA and the mean action potential amplitude was 90.5 ⫾ 12.7
mV. In contrast to PG neurons, no hump action potentials were
observed in GG neurons. Most (88/103) GG neurons responded with multiple spikes to an above threshold depolarizing current pulse (Fig. 2B).
Comparison of GG and PG neuron biophysical properties
Membrane time constant (Fig. 3B) and input resistance (Fig.
3C) of labeled PG and GG neurons were measured from the
response to a ⫺50 pA current injection. The characteristics of
the action potentials were measured in the first spike at the
threshold depolarizing current injection. Values measured were
action potential threshold (Fig. 3D), action potential decay time
(Fig. 3E), action potential half-width (Fig. 3F), and action
potential area (Fig. 3G). The mean time constant of GG neurons was significantly less than those of PG neurons. However,
there was no difference in input resistance measured at the
steady phase of membrane potential between the two groups.
The mean spike threshold of GG neurons was significantly
(P ⬍ 0.05) lower than that of PG neurons.
The most remarkable difference between the biophysical
properties of PG and GG neurons was the shape of the action
potential. As described above, no hump neurons were detected
FIG. 3. Comparison of several anatomical and electrophysiological characteristics of petrosal (filled columns) and geniculate (open columns) ganglia neurons. PG,
petrosal ganglion neurons; GG, geniculate
ganglion neurons. Significant levels are
indicated by *: P ⬍ 0.0001 (Student’s
t-test).
Downloaded from http://jn.physiology.org/ by 10.220.32.246 on April 28, 2017
10% of spike amplitude (decrease time: DT) was longer than
that of nonhump neurons (Table 1). However, there was no
significant difference in the other basic membrane properties
between the neurons with different types of action potential. In
general, the numbers of spikes increased with increasing magnitude depolarizating current pulses. Less than half (53/130) of
the cells fired only a single action potential at the onset of a
depolarizing step, even when the depolarizing current was
adjusted to about twice threshold. The remaining 77 neurons
responded with more than two spikes in response to an above
threshold depolarizing current pulse. Mean threshold current
for neurons with multiple spikes (70 pA) was significantly
lower than for singly discharging neurons (169 pA). Multiple
spike discharge was elicited from neurons that exhibited either
hump (23/77) or nonhump spikes (54/77). Thus, 45% of the
hump neurons and 68% of the nonhump neurons responded
with multiple spikes, but the difference was not significant.
LABELED GG NEURONS. Resting membrane potentials ranged
from ⫺43 to ⫺84 mV, with a mean of ⫺64 ⫾ 9 mV. Input
resistance ranged from 322 to 975 M⍀, with a mean of 574 ⫾
137 M⍀. Membrane time constants averaged 25.0 ⫾ 6.0 ms.
Like PG neurons, all of the GG neurons had a sagging phase in
response to hyperpolarizing currents (Fig. 2B).
Mean threshold for generation of action potentials was 66
1408
T. KOGA AND R. M. BRADLEY
in GG labeled neurons. All GG neurons had narrow spikes and
the decay time was much shorter (Fig. 3E). The action potential
amplitude of GG neurons was significantly (P ⬍ 0.05) smaller
than that of PG neurons (Fig. 3G).
Comparison between biophysical properties of labeled and
nonlabeled neurons
Effects of putative neurotransmitters
Sensitivity to ACh was investigated in
48 PG and 32 GG neurons. The sensitivity to ACh was examined during ⫺100 pA, 100 ms, hyperpolarizing current injection at 1 Hz to determine changes in membrane potential and
input resistance. Neurons were considered responsive if they
depolarized with a decrease in input resistance (current clamp)
or produced an inward current (voltage clamp at ⫺60 mV
holding potential) during application of ACh concentrations up
to 0.5 mM. Thus, if neurons did not respond to 0.5 mM ACh,
they were considered to be insensitive. Concentrations of ACh
were selected between 1 ␮M and 1 mM, because 0.3 mM ACh
is reported to evoke almost maximal current measured under
voltage clamp in PG neurons (Zhong and Nurse 1997). The
duration of the ACh application was 5 s.
ACh-induced depolarization was accompanied by a decrease
in input resistance in 50% of the PG neurons (Fig. 4, A and C).
The response was dose-dependent and returned to control
levels within 30 s after termination of the application (Fig. 4,
A and E). However, the level of sensitivity to ACh, i.e.,
membrane potential and input resistance changes, was different
from cell to cell. An ACh dose-response relationship was
measured under voltage-clamp condition. The mean peak current at each ACh concentration was normalized to that elicited
by 100 ␮M ACh (Fig. 3, D and E). The ACh elicited response
saturated at ⬃0.5 mM. The interval between each application
was at least 3 min to avoid receptor desensitization and/or
ACh-induced channel block. The complete dose-relation curve
was fitted by the Hill equation
ACETYLCHOLINE (ACH).
I/I max ⫽ 1/关1 ⫹ 共EC 50 /关ACh兴 n 兲兴
where I is the measured peak current, Imax is the maximal
response, n is the Hill coefficient, and EC50 is the concentration
of ACh required for half-maximal activation. The EC50 for
FIG. 4. Responses to ACh application. A and B: membrane responses of
petrosal (A) and geniculate (B) neurons to constant ⫺100 pA current hyperporalizing pulses during ACh applications at indicated concentrations. C: ratio
of neurons classified by their response to 0.5 mM ACh in labeled petrosal (PG)
and geniculate (GG) ganglion neurons. D: concentration-dependent ACh induced inward current responses recorded from a petrosal ganglion neuron
when the voltage was clamped at ⫺60 mV. E: concentration-response relationships for ACh response. Peak amplitudes of currents induced at various
concentrations were normalized to current evoked by 100 ␮M ACh (*). Each
point is the average of 4 – 6 cells.
receptor activation was approximately 40.5 ␮M and the Hill
coefficient was 1.54.
In contrast to PG neurons, the majority of GG neurons were
insensitive to 0.5 mM ACh application (Fig. 4C). Only 3 of 32
GG neurons indicated some slight response as shown in Fig.
4B, and this particular neuron was the most ACh sensitive of
all tested GG neurons.
SEROTONIN (5HT). Sensitivity to 5HT was examined in 30 PG
and 21 GG neurons. 5HT rapidly depolarized over 60% of the
PG neurons and initiated action potentials (Fig. 5A). The amplitude of the response to a second application of 5HT decreased when compared with the first application, especially
with high concentrations of 5HT. Thus, a 5-min interval between the first and second applications was required to obtain
consistent responses. The membrane potential and input resistance changes in response to 5HT were different in each cell.
However, 11 of 30 PG neurons were insensitive to application
of relatively high concentrations of 5HT (0.5–1 mM). Figure
5D shows the 5HT dose-response relation recorded under
voltage clamp. The mean peak current at each 5HT concentration was normalized to that elicited by 100 ␮M 5HT. This
concentration of 5HT elicited a saturated response (Fig. 5, C
Downloaded from http://jn.physiology.org/ by 10.220.32.246 on April 28, 2017
Differences in the biophysical properties of labeled and
nonlabeled neurons were examined by recording from an additional 45 PG and 26 GG nonlabeled neurons. Forty-two
percent of the nonlabeled PG neurons were classified as hump
neurons, and this ratio was similar to that of the PG labeled
neurons. In addition, no significant differences were found in
any biophysical properties between PG labeled and nonlabeled
neurons, indicating that the label did not alter the biophysical
properties of the neurons. In nonlabeled GG neurons, no differences were detected in passive membrane properties. However, 3 of 26 GG nonlabeled neurons had a hump in their action
potential. As a result, the decay time and action potential area
of nonlabeled neurons were significantly larger than labeled
GG neurons. Similarly, the mean half-width of nonlabeled
neurons was larger than the labeled neurons, but the difference
was not significant.
PROPERTIES OF SENSORY NEURONS INNERVATING THE TONGUE
1409
and D). The complete dose-relation curve was also described
by the Hill equation. The EC50 for receptor activation was
approximately 4.6 ␮M and the Hill coefficient was 0.93, suggesting one 5HT agonist binding site per receptor. The 5HTinduced currents were activated more rapidly at higher concentrations (Fig. 5C). A typical example of the activation and
short-term desensitization phases of an inward current is indicated in 100 and 1000 ␮M 5HT.
In contrast to PG neurons, the majority of the GG neurons
(20 of 21) were insensitive to 1 mM 5HT application (Fig. 5B).
However, one neuron responded to 5HT with a rapid depolarization and a high-frequency of spikes (not shown).
SUBSTANCE P (SP). The effect of SP was first examined at
concentration of 100 nM–1 ␮M in PG neurons. A 10-s application of 1 ␮M SP did not produce an observable response.
However, a longer application of the same concentration of SP
(30 – 40 s) was an effective stimulus (Fig. 6A), producing either
a membrane depolarization with an input resistance decrease
(Fig. 6, Aa and c), or a membrane hyperpolarization with an
increase in input resistance (Fig. 6Ab). More than 80% of both
PG and GG neurons were sensitive to SP.
The effect of a single concentration of SP (0.5 mM) was
investigated in 52 PG and 40 GG neurons. Figure 6B shows a
typical response to SP on GG neurons. The hyperpolarizing
and depolarizing response patterns that were observed in PG
neurons were also seen in GG neurons. Most PG neurons have
a depolarizing response as shown in Fig. 6, Aa and c, whereas
a hyperpolaring response accompanied by an increase in input
resistance was most frequently observed in GG neurons (Fig.
6Bb). Depolarizing and hyperpolarizing responses to SP have
been reported in other neurons (Dray and Pinnock 1982; King
et al. 1993; Plata-Salaman et al. 1989).
GAMMA AMINOBUTYRIC ACID (GABA). The sensitivity of PG
and GG neurons to GABA as well as the GABAA and GABAB
receptor agonists muscimol and baclofen was investigated to
determine differences in distribution of each GABA receptor
subtype.
Sensitivity to GABA was examined in 17 PG and 11 GG
neurons. Six of the 17 PG neurons were hyperpolarized with a
decrease in input resistance (Fig. 7Ab). In eight PG neurons,
however, GABA produced a small depolarization associated
with a decrease in input resistance (Fig. 7Aa). In contrast, most
GG neurons were hyperpolarized with a marked decrease in
input resistance (Fig. 7B). In summary, 82% of PG neurons and
all GG neurons were sensitive to GABA accompanied with a
decrease in input resistance. The ratio of response patterns is
summarized in Fig. 7C.
FIG. 6. Responses to substance P application in petrosal (A) and geniculate
(B) ganglion neurons. Aa: a depolarizing response accompanied by a decrease
in input resistance. Ab: a hyperpolarizing response with an increase in input
resistance. Ac: a depolarizing responses accompanied by a decrease in input
resistance. B: both depolarizing (Ba) and hyperpolarizing responses to application of 0.5 mM SP were observed in geniculate ganglion neurons.
Downloaded from http://jn.physiology.org/ by 10.220.32.246 on April 28, 2017
FIG. 5. Responses to 5HT application.
A: an example of a depolarizing response
to 100 ␮M 5HT in a petrosal ganglion
neuron. B: ratio of neurons classified by
their responses to 0.5–1 mM 5HT. C: concentration-dependent 5HT induced inward
current responses recorded from a petrosal
ganglion neuron. Holding potential was
⫺60 mV. D: concentration-response relationships for 5HT response. Peak amplitudes of currents induced at various concentrations were normalized to current
evoked by 100 ␮M 5HT (*). Each point is
the average of 4 –9 cells.
1410
T. KOGA AND R. M. BRADLEY
DISCUSSION
The GABAA agonist muscimol was applied to 13
PG and 18 GG neurons. Most of PG neurons (9/13) and all of
GG neurons were hyperpolarized by muscimol (Fig. 8B). Some
PG neurons were depolarized by muscimol; however, the
membrane resistance of all neurons was decreased in a dosedependent manner (Fig. 8, A, C, and D). The concentrationresponse curves were fitted using the Hill equation (Fig. 8D).
The EC50 of PG and GG neurons were approximately 36.8 and
27.1 ␮M, respectively. The Hill coefficient of PG and GG
neurons were 1.02 and 0.98, respectively, suggesting that one
molecule of muscimol binds to one ion channel. In summary,
the membrane resistance of all PG and GG neurons was decreased by muscimol in a dose-dependent manner.
MUSCIMOL.
The GABAB receptor agonist baclofen was applied to seven PG and 10 GG neurons. A slight hyperpolarization with membrane conductance increase was observed only
at high concentration (1 mM) in two PG and six GG neurons.
An example of response is indicated in Fig. 8E. However, no
obvious effect was observed in all neurons at concentrations of
⬍100 ␮M for 10-s application. Therefore, GABAB receptors
do not play a major role in the GABA response of these
neurons.
BACLOFEN.
Downloaded from http://jn.physiology.org/ by 10.220.32.246 on April 28, 2017
FIG. 7. Responses to GABA application. A: responses of petrosal ganglion
neurons to 100 ␮M GABA application. B: response of a geniculate ganglion
neuron to 100 ␮M GABA. C: ratio of neurons classified by their responses to
100 ␮M GABA.
Compared with the dorsal root, nodose, and trigeminal ganglia, the biophysical and neuropharmacological properties of
the PG and GG ganglia have received little attention. Moreover, when these ganglia have been studied, investigators have
not always recorded from a selected population of ganglion
cells even though there is evidence from immunochemical and
anatomical studies that these ganglia are heterogeneous and
innervate very different populations of receptors (Hall et al.
1997). The present study is thus the first in which an attempt
has been made to record from a defined population of ganglion
cells similar to recent experiments on tooth pulp neurons in the
trigeminal ganglion (Chiego et al. 1997; Cook et al. 1998).
The results indicate that there are significant differences
between the biophysical and neuropharmacological properties
of neurons in the PG and GG ganglia. For example, neurons in
the PG have both sharp and humped action potentials and PG
neurons respond to ACh, 5HT, SP, and GABA, whereas GG
neurons have sharp action potentials and only respond to SP
and GABA. In addition, the properties of GG neurons are more
homogeneous than those of the PG because all the neurons
have sharp spikes and when responses to neurotransmitters
occurred, either all or most of the neurons responded.
In other sensory ganglia, a relationship has been established
between axon diameter, action potential duration, and the type
of receptor innervated (Djouhri et al. 1998; Harper and Lawson
1985; Koerber et al. 1988; Rose et al. 1986; Villière and
McLachlan 1996). Even when a single receptor type is innervated, the ganglion cells can have different membrane properties (Belmonte and Gallego 1983). Afferent fibers in the rat and
hamster glossopharyngeal and chorda tympani nerves range in
diameter from 0.2 to 4.1 ␮m and innervate a number of
different receptors (Farbman and Hellekant 1978; Jang and
Davis 1987; Miller et al. 1978). By labeling a selective receptive field, the types of receptor innervated are reduced. Thus, a
label injected into the posterior tongue will label PG cells that
innervate taste and somatosensory receptors and a label injected into the anterior tongue will label GG cells that innervate
taste buds (Krimm and Hill 1998). The more heterogeneous
properties of the PG neurons probably reflects its role in
innervating taste and somatosensory receptors, while the homogeneous response characteristics of the GG neurons reflects
its main role in innervating taste buds on the anterior tongue.
However, even though GG neurons innervating the anterior
tongue have homogeneous biophysical and neuropharmacological properties, they differ in other characteristics. For example, fibers that innervate taste buds on the anterior tongue are
of different diameters, the receptive field properties of chorda
tympani fibers differ (Nagai et al. 1988), and the response
properties of chorda tympani fibers to taste stimuli applied to
the tongue are not homogeneous (Frank et al. 1988). Chorda
tympani fibers with small receptive fields respond best to NaCl,
whereas fibers with large receptive fields respond to a larger
number of salts (Nagai et al. 1988). Additional experiments in
which intracellular recordings are made from the whole ganglion while stimulating the afferent input are needed to determine if GG neurons with fibers of different diameters have
different membrane properties.
While little is known about the biophysical and neuropharmacological properties of the GG, there is information on the
PROPERTIES OF SENSORY NEURONS INNERVATING THE TONGUE
1411
PG. In early pioneering work, Gallego and his co-workers
performed the first intracellular analysis of PG ganglion neurons (Belmonte and Gallego 1983; Gallego 1983). These investigators were particularly interested in neurons innervating
chemoreceptors and baroreceptors involved in the cardiovascular system. Intracellular recordings with sharp electrodes
were made using an in vitro preparation of the whole ganglion
allowing electrical stimulation of the carotid sinus nerve to
isolate neurons innervating baroreceptors or carotid body chemoreceptors. As in the present study, neurons with both sharp
and humped action potentials were described. All the neurons
innervating the carotid body had humped action potentials and
even with prolonged depolarization only produced a single
spike. In contrast, the neurons innervating baroreceptors were
of two types based on the conduction velocity of their axons
and these neurons were capable of firing multiple action potentials when depolarized. It is possible, therefore, that the
ganglion cells innervating the posterior tongue with humped
and sharp action potentials innervate different populations of
receptors on the posterior tongue. In other ganglia, the humped
neurons have small diameter, slow conducting axons, and
innervate nociceptors (Koerber et al. 1988), so it is possible
that the PG neurons with humped spikes also innervate posterior tongue nociceptors, and the neurons with sharp action
potentials innervate taste receptors. This possibility is supported by the results from the labeled anterior tongue neurons
in the GG which supply taste receptors and all have sharp
spikes.
More recently, Nurse and co-workers have examined the
sensitivity of cultured PG cells to application of ACh and 5HT
(Zhong et al. 1999; Zong and Nurse 1997). Application of ACh
resulted in a rapid depolarization accompanied by an increase
in membrane conductance in 68% of the neurons which is
similar to the results obtained in the present study. In addition,
the EC50 for receptor activation in the present study was 40.5
␮M and the Hill coefficient was 1.54 ␮M, which is similar to
those reported by Zhong and Nurse (1997), who suggested that
more than two ACh agonist binding sites exist per receptor
from the value of the Hill coefficient. 5HT depolarized 43% of
the PG cells with an increase in membrane conductance, but in
a small percentage of neurons (6%) the depolarization was
accompanied with a decrease in membrane conductance. The
EC50 for 5HT receptor activation in the present study was 4.6
␮M and the Hill coefficient 0.93 ␮m. Zhong et al. (1999)
report similar results and suggested that more than two binding
sites are required for maximal activation of the receptor.
A number of investigators using immunohistochemical
methods have demonstrated SP-like immunoreactivity in PG
neurons and there is one report of intense SP immunoreactive
neurons in the GG (Ichiyama et al. 1997). However, there are
no reports of electrophysiological studies with SP. Guinea-pig
trigeminal ganglion neurons and bullfrog dorsal root ganglion
neurons are depolarized by exogeneous application of SP
(Dray and Pinnock 1982; Ishimatsu 1994; Spiegelman and Puil
1990). However, in a recent study in ferrets, SP in nanomolar
concentrations were found to hyperpolarize the membrane
potential of nodose ganglion neurons by activation of an outward calcium-dependent potassium current (Jafri and Weinreich 1998).
To our knowledge, this is the first study of the sensitivity of
PG and GG to GABA. Electrophysiological studies have been
performed to study the sensitivity of chick dorsal root ganglion
cells (Choi and Fishbach 1981; Choi et al. 1981), rat sympathetic ganglion cells (Adams and Brown 1998; Bowery and
Brown 1974), and rabbit nodose ganglion cells (Higashi et al.
1982) to GABA. In all these ganglia, GABA depolarizes the
cell membrane accompanied by an increase in membrane conductance, whereas in the present study GABA hyperpolarizes
Downloaded from http://jn.physiology.org/ by 10.220.32.246 on April 28, 2017
FIG. 8. Responses to muscimol and
baclofen application. A: concentrationdependent muscimol induced decrease in
input resistance in 2 petrosal ganglion
neurons. Aa: an example of depolarizing
type of response. Ab: an example of hyperpolaring response type. B: ratio of
neurons classified by their responses to
100 ␮M of muscimol. C: concentrationdependent muscimol induced hyperpolarization of membrane potential and decrease in input resistance in a geniculate
ganglion neuron. D: concentration response curve of % change of geniculate
(filled circles and a solid line) and petrosal (open squares and dashed line) ganglion neurons in input resistance to increasing concentration of muscimol.
Each point is the average of 4 –15 cells.
E: an example of the effect of 1 mM
baclofen on a geniculate ganglion neuron.
1412
T. KOGA AND R. M. BRADLEY
the majority of the ganglion cells in both the PG and GG with
an increase in membrane conductance. However, the earlier
studies were on different ganglia and used both extracellular
and intracellular recording techniques and were accomplished
before the use of whole-cell recording. Not all the neurons
were hyperpolarized by GABA and neurons with lower resting
membrane potentials were in fact hyperpolarized by GABA
(Adams and Brown 1998).
Functional significance
This work was supported by National Institute on Deafness and Other
Communication Disorders Grant DC-00288 to R. M. Bradley. T. Koga was the
recipient of a subsidy from the Nakayama Foundation for Human Science.
Present address of T. Koga: Dept. of Restorative Science, Kawasaki University of Medical Welfare, Kurashiki 701-0193, Japan.
REFERENCES
ADAMS PR AND BROWN DA. Actions of ␥-aminobutyric acid on sympathetic
ganglion cells. J Physiol (Lond) 250: 85–120, 1998.
BELMONTE C AND GALLEGO R. Membrane properties of cat sensory neurones
with chemoreceptor and baroreceptor endings. J Physiol (Lond) 342: 603–
614, 1983.
BOWERY NG AND BROWN DA. Depolarizing action of ␥-aminobutyric acid and
related compounds on rat superior cervical ganglion in vitro. Br J Pharmacol 50: 205–218, 1974.
BRADLEY RM, MISTRETTA CM, BATES CA, AND KILLACKEY HP. Transganglionic transport of HRP from the circumvallate papilla of the rat. Brain Res
361: 154 –161, 1985.
Downloaded from http://jn.physiology.org/ by 10.220.32.246 on April 28, 2017
SP has been shown to be present in centrally projecting
fibers of the glossopharyngeal nerve (Helke and Hill 1988) and
in the terminal fields of the glossopharyngeal nerve in the
nucleus of the solitary tract (Cuello and Kanazawa 1978;
Hokfelt et al. 1975). Moreover, King et al. (1993) have shown
that SP excites neurons of the rostral gustatory portion of the
nucleus of the solitary tract (rNST), suggesting that SP may be
involved with afferent transmission of taste information at the
first central synapse in the taste pathway. Similarly, GABA has
been shown to have a major influence on neurons processing
taste information in the rNTS (Grabauskas and Bradley 1996,
1998b; Wang and Bradley 1993, 1995) and this influence may
be partially acounted for via presynaptic inhibition at the
primary afferent excitatory synapse.
At present, the neurochemical characteristics of afferent
taste fibers is not known and mechanisms of synaptic transmission between the taste bud and the primary afferent fiber are
also unknown. Investigators of peripheral cutaneous afferent
fibers have determined their chemosensitive properties using
isolated dorsal root ganglion neurons (Dray 1996). It is possible, therefore, that the receptors on the distal afferent fibers
innervating taste buds are similar to those characterized in the
present study on isolated PG and GG ganglion neurons. Thus,
neural transmission between glossopharyngeal nerves and taste
buds could be mediated by ACh, SP, 5HT, and GABA,
whereas transmission in taste buds innervated by the facial
nerve could be mediated by 5HT and GABA but not by ACh
and SP. Of course, other untested neurotransmitters could also
be involved. The approach of using isolated ganglion cells to
examine neurotransmission in taste buds provides a practical
approach to what so far has proved to be a difficult technical
problem.
CHIEGO DJ, BRADLEY RM, AND MISTRETTA CM. Biophysical characterization
of trigeminal ganglion neurons innervating the rat molar dental pulp. Soc
Neurosci Abstr 23: 1257, 1997.
CHOI DW, FARB DH, AND FISCHBACH GD. GABA-mediated synaptic potentials
in chick spinal cord and sensory neurons. J Neurophysiol 45: 632– 643,
1981.
CHOI DW AND FISCHBACH GD. GABA conductance of chick spinal cord and
dorsal root ganglion neurons in cell culture. J Neurophysiol 45: 605– 620,
1981.
COOK SP, VULCHANOVA L, HARGREAVES KM, ELDE R, AND MCCLESKEY EW.
Distinct ATP receptors on pain-sensing and stretch-sensing neurons. Nature
387: 505–508, 1998.
CUELLO AC AND KANAZAWA I. The distribution of substance P immunoreactive
fibers in the rat central nervous system. J Comp Neurol 178: 129 –156, 1978.
CZYZYK-KRZESKA MF, BAYLISS DA, SEROOGY KB, AND MILLHORN DE. Gene
expression for pepetides in neurons of the petrosal and nodose ganglia in rat.
Exp Brain Res 83: 411– 418, 1991.
DAVIS BJ AND SMITH DV. Substance P modulates taste responses in the nucleus
of the solitary tract of the hamster. Neuroreport 8: 1723–1727, 1997.
DELAY RJ AND ROPER SD. Ultrastructure of taste cells and synapses in the
mudpuppy, Necturus maculosus. J Comp Neurol 277: 268 –280, 1988.
DJOUHRI L, BLEAZARD L, AND LAWSON SN. Association of somatic action
potential shape with sensory receptive properties in guinea-pig dorsal root
ganglion neurones. J Physiol (Lond) 513: 857– 872, 1998.
DRAY A. Chemosensitivity of cultured dorsal root ganglion cells. News Physiol
Sci 11: 288 –292, 1996.
DRAY A AND PINNOCK RD. Effects of substance P on adult rat sensory ganglion
neurones in vitro. Neurosci Lett 33: 61– 66, 1982.
DU J AND BRADLEY RM. Effect of GABA on acutely isolated neurons from the
gustatory zone of the rat nucleus of the solitary tract. Chem Senses 23:
683– 688, 1998.
FARBMAN AI AND HELLEKANT G. Quantitative analyses of the fiber population
in rat chorda tympani nerves and fungiform papillae. Am J Anat 153:
509 –522, 1978.
FINLEY JCW, POLAK J, AND KATZ DM. Transmitter diversity in carotid body
afferent neurons: dopaminergic and peptidergic phenotypes. Neuroscience
51: 973–987, 1992.
FRANK ME. Taste-responsive neurons of the glossopharyngeal nerve of the rat.
J Neurophysiol 65: 1452–1463, 1991.
FRANK ME, BIEBER SL, AND SMITH DV. The organization of taste sensibilities
in hamster chorda tympani nerve fibers. J Gen Physiol 91: 861– 896, 1988.
GALLEGO R. The ionic basis of action potentials in petrosal ganglion cells of
the cat. J Physiol (Lond) 342: 591– 602, 1983.
GALLEGO R AND EYZAGUIRRE C. Membrane and action potential characteristics
of A and C nodose ganglion cells studied in whole ganglia and in tissue
slices. J Neurophysiol 41: 1217–1232, 1978.
GRABAUSKAS G AND BRADLEY RM. Synaptic interactions due to convergent
input from gustatory afferent fibers in the rostral nucleus of the solitary tract.
J Neurophysiol 76: 2919 –2927, 1996.
GRABAUSKAS G AND BRADLEY RM. Ionic mechanism of GABAA biphasic
synaptic potentials in gustatory nucleus of the solitary tract. Ann NY Acad
Sci 855: 486 – 487, 1998a.
GRABAUSKAS G AND BRADLEY RM. Tetanic stimulation induces short-term
potentiation of inhibitory synaptic activity in the rostral nucleus of the
solitary tract. J Neurophysiol 79: 595– 604, 1998b.
HALL AK, AI X, HICKMAN GE, MACPHEDRAN SE, NDUAGUBA CO, AND
ROBERTSON CP. The generation of neuronal heterogeneity in a rat sensory
ganglion. J Neurosci 17: 2775–2784, 1997.
HARPER AA AND LAWSON SN. Electrical properties of rat dorsal root ganglion
neurones with different peripheral nerve conduction velocities. J Physiol
(Lond) 359: 47– 63, 1985.
HELKE CJ AND HILL KM. Immunohistochemical study of neuropeptides in
vagal and glossopharyngeal afferent neurons in the rat. Neuroscience 26:
539 –551, 1988.
HELKE CJ AND NIEDERER AJ. Studies on the coexistance of substance P with
other putative neurotransmitters in the nodose and petrosal ganglia. Synapse
5: 144 –151, 1990.
HELKE CJ AND RABCHEVSKY A. Axotomy alters putative neurotransmitters in
visceral sensory neurons of the nodose and petrosal ganglia. Brain Res 551:
44 –51, 1991.
HIGASHI H, UEDA N, NISHI S, GALLAGHER JP, AND SHINNICK-GALLAGHER P.
Chemoreceptors for seratonon (5-HT), actylcholine (ACh), bradykinin
(BK), histamine (H) and ␥-aminobutyric acid (GABA) on rabbit visceral
afferent neurons. Brain Res Bull 8: 23–32, 1982.
PROPERTIES OF SENSORY NEURONS INNERVATING THE TONGUE
NAGAI T, MISTRETTA CM, AND BRADLEY RM. Developmental decrease in size
of peripheral receptive fields of single chorda tympani nerve fibers and
relation to increasing NaCl taste sensitivity. J Neurosci Meth 8: 64 –72,
1988.
NAGY JI, GOEDERT M, HUNT SP, AND BOND A. The nature of the substance
P-containing nerve fibers in taste papillae of the rat tongue. Neuroscience 7:
3137–3151, 1982.
NEJAD MS. The neural activities of the greater superficial petrosal nerve of the
rat in response to chemical stimualtion of the palate. Chem Senses 11:
283–293, 1986.
OGAWA H, SATO M, AND YAMASHITA S. Multiple sensitivity of chorda tympani
fibers of the rat and hamster to gustatory and thermal stimuli. J Physiol
(Lond) 199: 223–240, 1968.
OKADA J AND MIURA M. Transmitter substances contained in the petrosal
ganglion cells determined by a double-labeling method in the rat. Neurosci
Lett 146: 33–36, 1992.
PARAN N AND MATTERN CFT. The distribution of actylcholinesterase in buds
of the rat vallate papilla as determined by electron microscope histochemistry. J Comp Neurol 159: 29 – 43, 1975.
PLATA-SALAMAN CR, FUKUDA A, MINAMI T, AND OOMURA Y. Substance P
effects on the dorsal motor nucleus of the vagus. Brain Res Bull 23:
149 –153, 1989.
ROSE RD, KOERBER HR, SEDIVEC MJ, AND MENDELL LM. Somal action
potential duration differs in identified primary afferents. Neurosci Lett 63:
259 –264, 1986.
SMITH DV AND FRANK ME. Sensory coding in peripheral taste fibers. In:
Mechanisms of Taste Transduction, edited by Simon SA and Roper SD.
Boca Raton, FL: CRC, 1993, p. 295–338.
SPIGELMAN I AND PUIL E. Ionic mechanisms of substance P actions on neurons
in trigeminal root ganglia. J Neurophysiol 64: 273–281, 1990.
STANDEN NB AND STANFIELD PR. Patch clamp methods for single channel and
whole cell recordings. In: Monitoring Neural Activity: A Practical Approach, edited by Stamford JA. Oxford, UK: IRL Press, 1992, p. 59 – 83.
VILLIÈRE V AND MCLACHLAN EM. Electrophysiological properties of neurons
in intact rat dorsal root ganglia classified by conduction velocity and action
potential duration. J Neurophysiol 76: 1924 –1941, 1996.
WANG L AND BRADLEY RM. Influence of GABA on neurons of the gustatory
zone of the rat nucleus of the solitary tract. Brain Res 616: 144 –153, 1993.
WANG L AND BRADLEY RM. In vitro study of afferent synaptic transmission in
the rostral gustatory zone of the rat nucleus of the solitary tract. Brain Res
702: 188 –198, 1995.
ZHONG HJ AND NURSE CA. Nicotinic acetylcholine sensitivity of rat petrosal
sensory neurons in dissociated cell culture. Brain Res 766: 153–161, 1997.
ZHONG HJ, ZHANG M, AND NURSE CA. Electrophysiological characterization of
5-HT receptors on rat petrosal neurons in dissociated cell culture. Brain Res
816: 544 –553, 1999.
Downloaded from http://jn.physiology.org/ by 10.220.32.246 on April 28, 2017
HOKFELT T, KELLERTH J-O, NILSSON G, AND PERNOW B. Experimental immunohistochemical studies on the localization and distribution of substance P
in cat primary sensory neurons. Brain Res 100: 235–252, 1975.
HOSLEY MA AND OAKLEY B. Postnatal development of the vallate papilla and
taste buds in rats. Anat Rec 218: 216 –222, 1987.
ICHIKAWA H, JACOBOWITZ DM, WINSKY L, AND HELKE CJ. Calretinin-immunoreactivity in vagal and glossopharyngeal sensory neurons of the rat:
distribution and coexistence with putative transmitter agents. Brain Res 557:
316 –321, 1991.
ICHIYAMA M, ITOH M, MIKI T, XIE Q, KANETO T, AND TAKEUCHI Y. Central
distribution of sensory fibers in the facial nerve: an anatomical and immunohistochemical study. Okajimas Folia Anat Jpn 74: 53– 64, 1997.
ISHIMATSU M. Substance P produces an inward current by supressing voltagedependent and -independent K⫹ currents in bullfrog primary afferent neurons. Neurosci Res 19: 9 –20, 1994.
JAFFE RA AND SAMPSON SR. Analysis of passive and active electrophysiologic
properties of neurons in mammalian nodose ganglion maintained in vitro.
J Neurophysiol 39: 802– 815, 1976.
JAFRI MS AND WEINREICH, D. Substance P regulates Ih via a NK-1 receptor in
vagal sensory neurons of the ferret. J Neurophysiol 79: 769 –777, 1998.
JAIN S AND ROPER SD. Immunocytochemistry of gamma-aminobutyric acid,
glutamate, serotonin, and histamine in Necturus taste buds. J Comp Neurol
307: 675– 682, 1991.
JANG T AND DAVIS BJ. The chorda tympani and glossopharyngeal nerves in the
adult hamster. Chem Senses 12: 381–395, 1987.
KING MS, WANG L, AND BRADLEY RM. Substance P excites neurons in the
gustatory zone of the rat nucleus tractus solitarius. Brain Res 619: 120 –130,
1993.
KOERBER HR, DRUZINSKY RE, AND MENDELL LM. Properties of somata of
spinal dorsal root ganglion cells differ according to peripheral receptor
innervated. J Neurophysiol 60: 1585–1596, 1988.
KRIMM RF AND HILL DL. Innervation of single fungiform taste buds during
development in rat. J Comp Neurol 398: 13–24, 1998.
MILLER IJ JR, GOMEZ MM, AND LUBARSKY EH. Distribution of the facial nerve
to taste receptors in the rat. Chem Senses 3: 397– 411, 1978.
MISTRETTA CM. Topographical and histological study of the developing rat
tongue palate and taste buds. In: Third Symposium on Oral Sensation and
Perception: The Mouth of the Infant, edited by Bosma JF. Springfield, IL:
Thomas, 1972, p. 163–187.
NAGAI T, DELAY RJ, WELTON J, AND ROPER SD. Uptake and release of
neurotransmitter candidates, [3H]serotonin, [3H]glutamate, and [3H]gammaaminobutyric acid, in taste buds of the mudpuppy, Necturus maculosus.
J Comp Neurol 392: 199 –208, 1998.
NAGAI T, KIM DJ, DELAY RJ, AND ROPER SD. Neuromodulation of transduction
and signal processing in the end organs of taste. Chem Senses 21: 353–365,
1996.
1413