Brain Research 786 Ž1998.

47–54

Research report

Selective activation of carotid nerve fibers by acetylcholine applied to the cat

petrosal ganglion in vitro
a, )
Julio Alcayaga , Rodrigo Iturriaga b , Rodrigo Varas a , Jorge Arroyo a , Patricio Zapata b

a
Laboratory of Neurobiology, Faculty of Sciences, UniÕersity of Chile, P.O. Box 653, Santiago 1, Chile
b
Laboratory of Neurobiology, Catholic UniÕersity of Chile, P.O. Box 114-D, Santiago 1, Chile

Accepted 18 November 1997

Abstract

The petrosal ganglion innervates carotid body chemoreceptors through the carotid Žsinus. nerve. These primary sensory neurons are
activated by transmitters released from receptor Žglomus. cells, acetylcholine ŽACh. having been proposed as one of the transmitters
involved in this process. Since the perikarya of primary sensory neurons share several properties with peripheral sensory endings, we
studied the electrical responses of the carotid nerve and glossopharyngeal branch to ACh locally applied to the cat petrosal ganglion
superfused in vitro. Ganglionar applications of AChCl Ž1 m g y 1 mg. generated bursts of action potentials conducted along the carotid
nerve, while only a few spikes were exceptionally recorded from the glossopharyngeal branch in response to the largest doses. Carotid
nerve responses to ACh were dose-dependent, the higher doses inducing transient desensitization. Application of nicotine to the petrosal
ganglion also evoked dose-dependent excitatory responses in the carotid nerve. Responses to ACh were reversibly antagonized by adding
hexamethonium to the superfusate, more intense and prolonged block of ACh responses being produced by mecamylamine. Ganglionar
applications of g-amino butyric acid and serotonin, in doses of up to 5 mg, did not induce firing of action potentials in any of the
branches of the glossopharyngeal nerve. Our results indicate that petrosal ganglion neurons projecting through the carotid nerve are
selectively activated by ACh acting on nicotinic ACh receptors located in the somata of these neurons. Thus, cholinosensitivity would be
shared by the membranes of peripheral endings and perikarya of primary sensory neurons involved in arterial chemoreception. q 1998
Elsevier Science B.V.

Keywords: Acetylcholine; Arterial chemoreceptors; Carotid body; Carotid Žsinus. nerve; Petrosal ganglion; Sensory ganglion

1. Introduction elicits discharges in chemosensory nerve fibers Žsee Ref.

Since the pioneering work of de Castro w6x, it is known
that the petrosal ganglion contains the perikarya of the
sensory neurons innervating the carotid body chemorecep-
tors. The transduction of physiological chemical stimuli
Žlowering PaO 2 , increasing PaCO 2 , lowering pH a . within
arterial chemoreceptor organs is commonly attributed to
glomus cells, which in turn release junctional transmitters
to activate the apposed chemosensory nerve endings Žsee
Refs. w9,15x.. Acetylcholine ŽACh. was the first substance
advocated as transmitter in the carotid body w33x and its
application to the carotid bodies of all mammalian species
) ganglion—where the perikarya of carotid body sensory
´ Facultad de Cien-
Corresponding author. Departamento de Biologıa,
cias, Universidad de Chile, Casilla 653, Santiago, Chile. Fax: q56-2-
271-2983; E-mail: jalcayag@codon.ciencias.uchile.cl
w9x..
It has been recently proposed that numerous properties
are shared by the cell membranes of peripheral nerve
endings and the respective somata of primary afferent
neurons w13,31x. Since the protein components of transduc-
ing elements andror chemical receptors anchored on the
membrane of the peripheral endings of primary sensory
neurons are conveyed to these sites by fast axoplasmic
transport from the respective somata w19x, these transducer
and receptor elements may also be found anchored on the
perikaryal membrane on sensory ganglia.
Based on the above considerations, we decided to test
the hypothesis that ACh applied to the petrosal
neurons are located—would induce excitation of the carotid
nerve chemosensory fibers.

0006-8993r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved.

PII S 0 0 0 6 - 8 9 9 3 Ž 9 7 . 0 1 4 2 4 - 8
48 J. Alcayaga et al.r Brain Research 786 (1998) 47–54
2. Materials and methods
Petrosal ganglia Ž n s 25. were obtained from 22 adult
cats, weighing 2.0 " 0.5 kg Žmean " S.D.., of either sex,
anaesthetized with sodium pentobarbitone Ž40 mgrkg. i.p.,
supplemented when necessary with additional doses Ž12
mg. given i.v. The animals were placed in the supine
position, the neck opened through a midline incision, and
the full trajectory of one glossopharyngeal nerve was
exposed. Its carotid nerve branch was dissected up to its
entrance to the carotid body, where it was cut. The so-called
glossopharyngeal branch w26x, going to the pharynx and
tongue, was also cleaned from surrounding tissue and
sectioned distally. After removing the tympanic bulla and
the ventral wall of the jugular foramen to expose the
petrosal ganglion, its central process was cut at about 2
mm from the apparent central border of the ganglion. The
petrosal ganglion with its central and peripheral processes
was placed in a Petri dish with ice-chilled Hank’s balanced
solution ŽSigma.. Under a stereomicroscope, the capsule
was removed from the ganglion and the epineurium along
the full extension of the glossopharyngeal nerve and its
branches. After excision of one or both petrosal ganglia,
the cat was sacrificed with an overdose of sodium barbi-
tonerurethane given i.v.
The petrosal ganglion with its peripheral nerve branches
was transferred to a thermostated chamber kept at 38.0 "
0.58C, superfused with Hank’s balanced solution, supple-
mented with 5 mM HEPES buffer, pH 7.35 Žadjusted with
NaOH to desired value., equilibrated with room air Ž21%
O 2 ., and flowing at 1.5 mlrmin ŽFig. 1.. The ganglion
Fig. 1. Diagram of experimental set-up. Lucite chamber for superfusion
of petrosal ganglion, to which square electrical stimulating pulses may be
applied. The chamber is thermoregulated. Inflow of air bubbled saline
provided from thermoregulated bottle Žleft side.; outflow Žright side.
controlled by the same roller pumping device. The peripheral process of
the glossopharyngeal nerve is lifted to the mineral oil phase, where
electrodes pairs record action potentials from glossopharyngeal branch
and carotid Žsinus. nerve, after stages of preamplification, displays on
oscilloscope and storage of digital coded signals on video-cassette recorder
Župper right part of the figure..
was placed in the 0.5 ml capacity lower chamber, over a
pair of stimulating Pt–Ir electrodes, and pinned to the
bottom of the chamber. The electrodes were connected
through a stimulus isolation unit to a stimulator. The tip of
a thermistor, immersed in the superfusion channel, was
located at 4 mm from the ganglion surface. The carotid
Žsinus. nerve ŽCN. and the glossopharyngeal branch ŽGPB.
were placed on recording paired Pt–Ir electrodes Žinter-
electrodes distances: 2.3 and 4.1 mm, respectively., and
lifted into the upper compartment of the superfusion cham-
ber filled with mineral oil. The recording electrodes were
connected to AC-preamplifiers, and the resulting elec-
troneurograms were amplified, displayed on a multiple
beam oscilloscope, and recorded in a video cassette
recorder after digital pulse code modulation ŽPCM.. The
electroneurograms were also fed to a spike amplitude
discriminator, whose standardized pulses were counted and
integrated at 1 s intervals to assess the frequencies Ž f . of
discharges of both nerves, which were digitized through an
AD interface board ŽLabMaster DMA. and displayed on a
computer ŽAT 386..
Agonists were applied in 10 m l boluses by means of a
microdispenser, whose tip was placed about 1 mm distant
from the exposed surface of the petrosal ganglion. Acetyl-
choline chloride ŽSigma., nicotine tartrate ŽICN-K& K
Labs.., g-amino butyric acid ŽGABA. ŽSigma. and sero-
tonin Ž5-HT. hydrochloride ŽRBI. were applied in doses of
from 0.1 m g to 5 mg, at ca. 15 min intervals. Neostigmine
methyl sulfate ŽHermann Strack, Germany., hexametho-
nium chloride ŽMatheson, Coleman and Bell. and mecamy-
lamine hydrochloride ŽMerck. were applied at constant
concentrations through the superfusion medium. Doses
correspond to the salts.
The changes in discharge frequency Ž D f . induced by
the drugs were calculated as the differences between the
maximal frequency Žmax f . achieved during a single
response and the mean basal activity Žbas f ., computed
along the 30 s period prior to drug administration. The
relation between D f ’s and the doses of the tested drugs
Žw x x. was assessed using the best fit of the standardized
responses Ž D frmax D f . to a sigmoid curve Ž R s 1rŽ1 q
 ED50 rw x x4 S .., using a simplex algorithm. Differences be-
tween dose-response curves were assessed using Friedman
and Quade analyses of variance.
3. Results
3.1. Electrical stimulation
The functional integrity of the preparation was evalu-
ated at the beginning and end of each experiment by
applying brief electrical pulses Ž50 m s. to the ganglion and
recording the compound action potentials in both periph-
eral nerve branches. Several components of different con-
 J. Alcayaga et al.r Brain Research 786 (1998) 47–54 49

duction velocities ŽCVs. and relative amplitudes were
recorded. Conduction times were estimated from beginning
of stimulation artifacts to those peaks of compound action
potentials clearly discernible and present in all successive
recordings taken at the beginning of experiments. For the
fibers following the route of the GPB, fast components
were recorded in four experiments, with CVs of 52.3, 45.7,
31.0 and 20.1 mrs; intermediate components were recorded
in eight experiments, with a CV of 10.6 " 1.1 mrs Žmean
" S.E.M..; and slowest components were clearly dis-
cernible in only two experiments, with CVs of 0.85 and
1.0 mrs. For the fibers comprised within the CN, the
fastest component was observed only in one experiment, at
a CV of 25.3 mrs; intermediate components had a bi-
modal distribution in four experiments, with CVs of 9.7 "
1.1 and 4.6 " 0.2 mrs; and slowest components were
recorded in five experiments, with a CV of 0.94 " 0.05
mrs.
detected from CN recordings in all preparations ŽFig. 2,
left side.. The basal activity recorded from CN’s consisted
of low frequency, randomly distributed discharges, al-
though clusters of higher frequency spikes were seldomly
observed. Fig. 3 corresponds to a high speed display of
recordings obtained in one experiment, where three units
were recognizable, firing spontaneously at 15, 10 and 3.3
Hz. This shows that the gross spontaneous electrical activ-
ity recorded from the entire CN corresponds to basal
discharges from its units.
At the end of some experiments, novocaine hydrochlo-
ride was injected in boluses of 200 m g each at the entrance
of the superfusion channel, until obtaining abolition of the
compound action potentials recorded in both peripheral
nerve branches in response to electrical stimulation of the
petrosal ganglion, along with the disappearance of sponta-
neous activity, thus discriminated from electronic noise.

3.2. Basal actiÕity 3.3. Effects of ACh and nicotine applied to the ganglion

While scarce spontaneous activity was recorded from In all experiments, topical applications of ACh to the
the GPB in only two preparations, such activity was petrosal ganglion produced bursts of discharges recorded

Fig. 2. Basal and ACh-evoked activities in peripheral branches. Electrical activities recorded from carotid nerve ŽCN. and glossopharyngeal branch ŽGPB.,
at same amplifications, before Žy1.6 s, left. and after Ž1.6 s, right. the application of 100 m g AChCl to the petrosal ganglion. Simultaneous displays at two
different rates, as indicated by time signals. The response to ACh is restricted to the CN.
50 J. Alcayaga et al.r Brain Research 786 (1998) 47–54
Fig. 3. Carotid nerve units discharging in basal conditions. Display at
high speed of the electrical activity recorded from one carotid nerve.
Time, 1 msrdivision; Amplitude, q140 mV to y140 mV; interrupted
line, 0 mV. Three spontaneously discharging units are recognized, marked
Ž1., Ž2. and Ž3., which discharged at 10, 15 and 3.3 Hz, respectively,
along 1.2 s of continuous recording, comprising the 10-ms samples here
illustrated.
from the CN ŽFig. 2, right side.. At the end of these
experiments and after crushing the glossopharyngeal nerve
trunk at the distal border of the petrosal ganglion, ACh
applications to the ganglion failed to evoke electrical
activity in the CN. After displacing the tip of the microdis-
penser to the nerve segment immediately distal to the
crush, ACh applications failed to evoke electrical activity
in the CN.
The threshold dose of ACh necessary to evoke CN
responses varied from 1 to 5 m g and maximal responses
were obtained with doses of between 100 and 500 m g, as
illustrated in Fig. 4A,B. Carotid nerve responses evoked by
ganglionar applications of ACh were dose-dependent, in
terms of the maximal frequencies Žmax f . of evoked
discharges. However, CN responses were slightly pro-
longed at low doses, but were shortened at larger doses,
becoming more abrupt in onset but also more rapid in
decay ŽFig. 4A.. Occasionally, responses to large doses of
ACh were biphasic Žsee responses to 100, 200 and 500 m g
in Fig. 4B.. Furthermore, brief silencing was sometimes
observed immediately following responses to high doses of
ACh ŽFig. 4C,D., revealing that a low frequency basal
spontaneous activity was present before and after applica-
tion of the drug.
In two experiments Žnot included in the present series.
in which the petrosal ganglion was bathed at room temper-
ature Ž18–208C., no responses from the CN could be
evoked by ganglionar applications of ACh in doses up to 5
mg, but the doses of 1 and 5 mg reduced by 25–50% the
amplitude of CN compound action potentials evoked by
electrical stimulation of the ganglion Žnot illustrated.. This
observation, together with those reported in the previous
paragraph, suggest that high doses of ACh may cause
depolarization block.
When testing CN responses to repeated ganglionar ap-
plications of the same doses of ACh, transient desensitiza-
tion was observed, especially for doses above 10–20 m g.
An interval of 5 min between applications of the same
dose was required to achieve similar neural responses.
Thus, 15 min intervals between applications of increasing
doses were used when obtaining data for dose-response
curves.
Fig. 5 summarizes the maximal frequencies of re-
sponses recorded from carotid nerves to applications of
increasing doses of ACh to the petrosal ganglia in ten
different preparations. The dose-response curve reveals a
mean effective dose of 29 m g, a slope of 0.9, and a
maximal response for doses of 500 m g.
In two experiments, neostigmine 20 m M was added to
the superfusate of the petrosal ganglia, provoking a shift to
the left of dose-response curves to AChCl, whose ED50
was displaced from 63.8 to 3.2 m g Žnot illustrated..
Just a few spikes were recorded from the GPB in
response to the largest doses of AChCl Ž) 100 m g. in only
3 out of 25 experiments. These doses are well above the
minimal effective doses of ACh eliciting bursts of dis-
charges in the CN.
In three experiments, nicotine tartrate 100–500 m g was
applied to the petrosal ganglion, evoking increases in the
frequency of CN discharges similar to those evoked by
equivalent doses of ACh, but of slower time course Žrising
and decay times.. When given at short intervals, the
responses evoked by nicotine presented more prolonged
desensitization than that induced by ACh, as well as
crossed desensitization with responses induced by gan-
glionar applications of ACh. A dose-response curve Žnot
illustrated., constructed from one experiment in which
nicotine injections were given at 30 min intervals, revealed
discernible response with 1 m g and maximal responses
within 0.1 to 1 mg, of similar amplitude to those evoked
by two test injections of 200 m g of AChCl. Nicotine failed
to evoke responses from the GPB.
3.4. Effects of cholinergic blockers
Fig. 6A summarizes the dose-response curve obtained
from three different experiments in which increasing doses
 J. Alcayaga et al.r Brain Research 786 (1998) 47–54 51

Fig. 4. Changes in frequency of discharges Ž f . recorded from carotid nerves in response to applications of ACh to petrosal ganglia Žarrowheads.. ŽA,B.
Responses to increasing doses of AChCl applied Žin two different preparations. in boluses of 500 ng, 1, 2, 5, 10, 20, 50, 200, 500 m g and 1 mg ŽA., and 5,
10, 20, 50, 100, 200 and 500 m g ŽB.. Recordings interrupted after evoked responses subsided. ŽC,D. Pattern of discharge evoked by single AChCl 100 m g
bolus, followed by a brief rebound Žsilencing. of electrical activity. Continuous line, mean basal frequency of discharge. In recording displayed at higher
amplification ŽD., the interrupted lines indicate upper and lower limits of the 95% normal distribution of mean basal frequency.

of ACh were applied to the petrosal ganglion, first in
control conditions, then during superfusion with 10 m M
hexamethonium, and finally after washing hexamethonium
Fig. 5. Dose-response curve of ACh-evoked responses. Dose-response
curve for changes in frequency Ž D f . of discharge of carotid nerves
Žexpressed as percentages of maximal responses of each preparation.
evoked by AChCl applied to petrosal ganglia. Means"S.E.M.’s from
data collected from 10 preparations. Curve adjusted only to the range
1–500 m g. Inset: correlation, mean effective dose, slope and significance
level.
from the superfusing solution. Hexamethonium blocked
partially but consistently the excitatory effects of ACh and
this blockade was reversible after washing. The CN re-
sponses evoked by electrical stimulation of the petrosal
ganglia immediately before and at the end of the superfu-
sion with hexamethonium 10 m M ŽFig. 6A, inset. reveal
that the block of ACh responses was achieved without
affecting the conduction properties of CN fibers.
The effects of the nicotinic cholinergic blocker
mecamylamine were also tested in three preparations. In
one experiment, 100 m M mecamylamine produced com-
plete and irreversible Žwithin 3 h. block of CN responses
to full series of ganglionic applications of ACh, but this
effect was associated with conduction block of the faster
conducting fibers of the CN, and partial block of slower
conducting fibers Žnot illustrated.. In another experiment,
mecamylamine 1 m M produced full block of ACh re-
sponses, not associated with conduction block of CN fibers,
and followed by partial recovery of ACh responses after 3
h of washing the antagonist ŽFig. 6B.. In a third experi-
ment, mecamylamine 100 nM produced partial block of
ACh responses, with partial recovery after 3 h of washing
the antagonist.
52 J. Alcayaga et al.r Brain Research 786 (1998) 47–54
Fig. 6. Block by cholinergic antagonists of ACh-evoked responses.
Dose-response curves for changes in frequency Ž D f . of discharge recorded
from carotid nerves evoked by increasing doses of AChCl applied to
petrosal ganglia. ŽA. Mean values of data from 3 preparations Ž D f
expressed as percentage of maximal evoked discharge in each prepara-
tion.. Evoked responses obtained in control conditions Žfilled squares.,
during superfusion with 10 m M hexamethonium for 110 min Žtriangles.
and after at least 50 min of washing Žopen squares.. ŽB. Data obtained
from one preparation in control conditions Žfilled squares., during super-
fusion with 1 m M mecamylamine for 110 min Žtriangles. and after at
least 50 min of washing Žopen squares.. Insets, recordings of compound
action potentials from carotid nerve evoked by electrical stimulation of
petrosal ganglion in control conditions Župper records. and at the end of
antagonists superfusions Žlower records..
3.5. Effects of other putatiÕe transmitters applied to the
ganglion
Since g-amino butyric acid ŽGABA. has been reported
to induce depolarization of cat primary afferent neurons in
vitro w12x, we decided to test whether petrosal ganglion
neurons were excited by this agent. Although the hypothe-
sis that serotonin Ž5-HT. should serve as chemical trans-
mitter between taste receptor cells and glossopharyngeal
nerve endings has been mostly dismissed w27x, the possibil-
ity of testing the effects of 5-HT upon the perikarya of
their sensory neurons was also attractive.
GABA in doses of from 1 m g to 10 mg was assayed in
three different experiments. Serotonin in doses of from 1
m g to 1 mg was assayed in two different experiments.
Neither GABA nor 5-HT applied to the petrosal ganglion
had any noticeable effect on the electrical activities
recorded from both peripheral glossopharyngeal nerve
branches.
4. Discussion
The electrical activity recorded from the CN may be
carried by chemosensory nerve fibers innervating the
carotid body and barosensory nerve fibers innervating the
carotid sinus. Otherwise, the electrical activity recorded
from the GPB may be carried by chemosensory nerve
fibers innervating the taste buds located in the posterior
third of the tongue and nasopharynx, and by mechanosen-
sory fibers innervating the lingual mucosa and pharyngeal
structures. A few efferent fibers are contained in the CN
Žsee Ref. w9x., whereas abundant motor fibers are present in
the GPB. However, efferent fibers cannot contribute to the
spontaneous electrical activity reported in this paper, since
their perikarya are located within the medulla oblongata
and therefore not included in our in vitro preparation.
Leaving sensory fibers as the only source of the sponta-
neous and drug-induced activities here recorded, such ac-
tivities cannot originate from peripheral sensory nerve
endings which were also not included in the preparation
used here. Thus, the intraganglionar segment of sensory
neurons remains as the only possible origin of the electri-
cal activity present in this preparation.
Spontaneous electrical activity may be recorded from
primary afferent fibers within a few hours of peripheral
nerve section w25x. Furthermore, the recording of impulses
from the CN central to a section is not new. These
impulses may still be recorded after intracranial section of
glossopharyngeal and vagal nerve roots w39x, an in situ
preparation with the same remaining structures as those
included in the present in vitro preparation. The observa-
tion reported here that electrical activity was evoked in the
CN by application of ACh or nicotine to the petrosal
ganglion, but not to applications distal to the ganglion,
suggests that the impulses observed in both in situ and in
vitro preparations may be generated from the perikarya of
primary sensory neurons. It has also been recently shown
that petrosal ganglion neurons from rat pups co-cultured
with carotid body glomus cells may exhibit spontaneous
action potentials and subthreshold depolarizing potentials
w41x. It is known that the petrosal ganglia only contain the
perikarya of sensory neurons, the presence there of auto-
nomic efferent somata having been discarded w36x.
The petrosal ganglion contains the perikarya of
mechanosensory neurons innervating the carotid sinus, lin-
gual and pharyngeal mucosae, in addition to the chemosen-
sory neurons innervating the carotid body and circumval-
late and foliate papillae of the tongue w9,17,38x. The above
mentioned mechanoreceptors are considered simple recep-
 J. Alcayaga et al.r Brain Research 786 (1998) 47–54
tors, in which transduction is performed by sensory nerve
endings themselves w3x. In contrast, arterial and taste
chemoreceptors are considered composite receptors, in
which transduction is performed by glomus and taste cells,
respectively, which in turn release transmitters for the
excitation of the closely apposed sensory nerve endings
Žsee Refs. w9,15,32x.. The exact nature of the transmitter
substances released by receptor cells of the carotid body
and taste cells of gustatory papillae is still controversial,
but ACh has been considered a putative transmitter in both
preparations.
The carotid body parenchyma contains ACh, its synthe-
sizing enzyme Žcholine acetyltransferase., its degrading
enzyme Žacetylcholinesterase. and a high-affinity incorpo-
ration system for the precursor choline Žsee Ref. w9x.. Our
finding that neostigmine produced a displacement to the
left of dose-response curves to ACh is consistent with an
anticholinesterase action exerted at the level of the petrosal
ganglion. Acetylcholine release has been demonstrated
from the carotid body in situ w24x and in vitro w8,10x, as
well as from cultured glomus cells w34x. ACh produces
dose-dependent chemosensory excitation of carotid bodies
in situ and in vitro, an effect that is blocked by nicotinic
blockers mecamylamine, hexamethonium and curare, in
this order of potency Žw8x; see also Ref. w22x.. It has been
shown recently that selective perfusion of the cat carotid
body with mecamylamine produces a dose-dependent re-
duction of chemosensory responses to hypoxia w11x. In
response to ACh applied to the carotid body in vitro, slow
depolarizing potentials have been recorded intracellularly
from chemosensory fibers, presumably not far from their
endings w16x. Although ACh receptors were expected to be
located in the membrane of chemosensory nerve endings,
presumed post-junctional site, nearly two-thirds of a-
bungarotoxin binding sites were located in the membrane
of glomus cells, presumed prejunctional site w5x. Another
study showed that specific a-bungarotoxin binding in sen-
sory-denervated carotid bodies was similar to that ob-
served in intact carotid bodies w7x.
The presence of nicotinic ACh receptors in the mem-
brane of sensory ganglion cells has been inferred from
electrical responses of cultured neurons from rat nodose
ganglia w1x, as well as from neonatal rabbit nodose gan-
glion cells co-cultured with carotid glomus cells w14x. The
presence of nerve growth factor in the culture medium is
required for nodose ganglion neurons to express nicotinic
ACh receptors w21x. There is a recent brief report w41x on
recordings from neonatal rat petrosal ganglion neurons,
cultured alone or in co-culture with carotid body glomus
cells, showing slow depolarizing and action potentials in
response to ACh applied in the vicinity. There is also a
recent immunocytochemical demonstration of the expres-
sion of nAChR a 4 subunits in cultured neurons obtained
from cat’s petrosal ganglia as well as in the somata of
frozen tissue sections of the same ganglia w18x. However,
we do not know of any reported physiological or phar-
53
macological demonstration of ACh receptors operating in
the perikarya of primary sensory neurons located in pet-
rosal ganglia of preparations in situ or recently excised and
superfused in vitro.
The excitatory response of sensory nerve endings to
ACh is a prominent feature of the chemosensory fibers
innervating the carotid bodies w8,22x, while ACh has no
direct effect on the carotid discharges initiated from
barosensory endings w23x. It has been shown that the
perikarya of chemosensory fibers have different electrical
membrane properties than those of barosensory neurons
whose axons course through the CN w2x. However, the
present work does not allow to conclude that CN fibers
activated by ACh applied to the petrosal ganglion corre-
spond exclusively to chemosensory fibers contained within
the CN.
The initial enhancement followed by depression of dis-
charges recorded from gustatory nerve fibers in frogs after
applying ACh to the tongue surface led to the proposal that
ACh was the transmitter between taste cells and gustatory
nerve endings w20,30x; however, intra-arterial perfusion of
frog’s tongue with ACh depressed glossopharyngeal re-
sponses to salts but not those to HCl, quinine and water
Žcited by Ref. w28x.. Nevertheless, the nerve endings of
primary sensory neurons innervating the skin w35x and
cornea w37x are also activated by ACh. It is also known that
nicotinic ACh receptors, traced by a-bungarotoxin binding
sites, are manufactured in the somata of primary afferent
neurons located in dorsal root ganglia, and are transported
by their peripheral and central axonal processes to their
respective endings w29x. Furthermore, acetylcholinesterase
is widely distributed among sensory neurons of dorsal root
ganglia w4x.
The presence of nicotinic ACh receptors in both
perikarya and peripheral endings of chemosensory neurons
does not imply that both regions are equally sensitive to
ACh. Respiratory and circulatory reflexes induced by nico-
tine 5–50 m grkg i.v. in cats disappear after sectioning the
carotid and vagus nerves w40x, in spite of the anatomical
continuity between the somata of petrosal and nodose
ganglia with the brain stem, an indication of the higher
sensitivity to nicotine of peripheral chemosensory endings.
Acknowledgements
Work supported by grant 96r06PF from the Research
and Postgraduate Division of the Catholic University of
Chile ŽDIPUC., and grant 1971013 from the National Fund
for Scientific and Technological Development ŽFONDE-
CYT.. J.A. also received support from a Data Foundation
Award. Thanks are due to Mrs Carolina Larraın ´ for her
help in the preparation of the illustrations.
54 J. Alcayaga et al.r Brain Research 786 (1998) 47–54
References
w1x P.I. Baccaglini, E. Cooper, Influences on the expression of acetyl-
choline receptors on rat nodose neurones in cell culture, J. Physiol.
Lond. 324 Ž1982. 441–451.
w2x C. Belmonte, R. Gallego, Membrane properties of cat sensory
neurones with chemoreceptor and baroreceptor endings, J. Physiol.
Lond. 342 Ž1983. 603–614.
w3x A.M. Brown, Receptors under pressure. An update on baroreceptors,
Circ. Res. 46 Ž1980. 1–10.
w4x F. Castrignano, M.E. De Stefano, F. Leone, B. Mulatero, A.M. Tata,
A. Fasolo, G. Augusti-Tocco, Ontogeny of acetylcholinesterase,
substance P and calcitonin gene-related peptide-like immunoreactiv-
ity in chick dorsal root ganglia, Neuroscience 34 Ž1990. 499–510.
w5x I.-L. Chen, J.A. Mascorro, R.D. Yates, Autoradiographic localiza-
tion of a-bungarotoxin binding sites in the carotid body of the rat,
Cell Tissue Res. 219 Ž1981. 609–618.
w6x F. De Castro, Sur la structure et l’innervation de la glande intercaro-
tidienne Ž glomus caroticum. de l’homme et des mammiferes, ` et sur
` d’innervation autonome du nerf glossopharyn-
un nouveau systeme
gien, Trab. Lab. Invest. Biol. Univ. Madrid 24 Ž1926. 365–432.
w7x B. Dinger, C. Gonzalez,
´ K. Yoshizaki, S. Fidone, Alpha-bungaroto-
xin binding in cat carotid body, Brain Res. 205 Ž1981. 187–193.
w8x C. Eyzaguirre, P. Zapata, The release of acetylcholine from carotid
body tissues. Further study on the effects of acetylcholine and
cholinergic blocking agents on the chemosensory discharge, J. Phys-
iol. Lond. 195 Ž1968. 589–607.
w9x C. Eyzaguirre, P. Zapata, Perspectives in carotid body research, J.
Appl. Physiol. 57 Ž1984. 931–957.
w10x R.S. Fitzgerald, M. Shirahata, Release of acetylcholine from the in
vitro cat carotid body, Adv. Exp. Med. Biol. 410 Ž1996. 227–232.
w11x R.S. Fitzgerald, M. Shirahata, T. Ide, Further cholinergic aspects of
carotid body chemotransduction of hypoxia in cats, J. Appl. Physiol.
82 Ž1997. 819–827.
w12x J.P. Gallagher, H. Higashi, S. Nishi, Characterization and ionic basis
of GABA-induced depolarizations recorded in vitro from cat pri-
mary afferent neurones, J. Physiol. Lond. 275 Ž1978. 263–282.
w13x M.S. Gold, S. Dastmalchi, J.D. Levine, Co-expression of nociceptor
properties in dorsal root ganglion neurons from the adult rat in vitro,
Neuroscience 71 Ž1996. 265–275.
w14x W.F. Goldman, M. Sato, L.J. Stensaas, C. Eyzaguirre, Acetyl-
choline- and dopamine-induced excitation of cultured newborn rab-
bit nodose ganglion neurons: effects of co-culture with carotid body
fragments, in: J.A. Ribeiro, D.J. Pallot ŽEds.., Chemoreceptors in
Respiratory Control, Croom Helm, London, 1987, pp. 284–295.
w15x C. Gonzalez,
´ L. Almaraz, A. Obeso, R. Rigual, Carotid body
chemoreceptors: from natural stimuli to sensory discharges, Physiol.
Rev. 74 Ž1994. 829–898.
w16x Y. Hayashida, H. Koyano, C. Eyzaguirre, An intracellular study of
chemosensory fibers and endings, J. Neurophysiol. 44 Ž1980. 1077–
1088.
w17x H. Ichikawa, A. Rabchevsky, C.J. Helke, Presence and coexistence
of putative transmitters in carotid sinus baro- and chemoreceptor
afferent neurons, Brain Res. 611 Ž1993. 67–74.
w18x Y. Ishizawa, R.S. Fitzgerald, M. Shirahata, B. Schofield, Localiza-
tion of nicotinic acetylcholine receptors in cat carotid body and
petrosal ganglion, Adv. Exp. Med. Biol. 410 Ž1996. 253–256.
w19x G.M. Koschorke, R.A. Meyer, J.N. Campbell, Cellular components
necessary for mechanoelectrical transduction are conveyed to pri-
mary afferent terminals by fast axonal transport, Brain Res. 641
Ž1994. 99–104.
w20x S. Landgren, G. Liljestrand, Y. Zotterman, Chemical transmission in
taste fibre endings, Acta Physiol. Scand. 30 Ž1954. 105–114.
w21x A. Mandelzys, E. Cooper, V.M.K. Verge, P.M. Richardson, Nerve
growth factor induces functional nicotinic acetylcholine receptors on
rat sensory neurons in culture, Neuroscience 37 Ž1990. 523–530.
w22x D.S. McQueen, A quantitative study of the effects of cholinergic
drugs on carotid chemoreceptors in the cat, J. Physiol. Lond. 273
Ž1977. 515–532.
w23x D.S. McQueen, Effects of acetylcholine and sodium cyanide on cat
carotid baroreceptors, Br. J. Pharmacol. 69 Ž1980. 433–440.
w24x B. Metz, Release of acetylcholine from the carotid body by hypoxia
and hypoxia plus hypercapnia, Respir. Physiol. 6 Ž1969. 386–394.
w25x M. Michaelis, K.H. Blenk, W. Janig, ¨ C. Vogel, Development of
spontaneous activity and mechano-sensitivity in axotomized afferent
nerve fibers during the first hours after nerve transection in rats, J.
Neurophysiol. 74 Ž1995. 1020–1027.
w26x A. Morales, I. Ivorra, R. Gallego, Membrane properties of glos-
sopharyngeal sensory neurons in the petrosal ganglion of the cat,
Brain Res. 401 Ž1987. 340–346.
w27x K. Morimoto, M. Sato, Is serotonin a chemical transmitter in the
frog taste organ?, Life Sci. 21 Ž1977. 1685–1696.
w28x K. Morimoto, M. Sato, Role of monoamines in afferent synaptic
transmission in frog taste organ, Jpn. J. Physiol. 32 Ž1982. 855–871.
w29x M. Ninkovic, S.P. Hunt, a-Bungarotoxin binding sites on sensory
neurons and their axonal transport in sensory afferents, Brain Res.
272 Ž1983. 57–69.
w30x G. Rapuzzi, A. Pedrini, Importanza dell’acetilcolina sull’attivazione
dei recettori linguali di rana, Boll. Soc. Ital. Biol. Sper. 39 Ž1963.
1851–1853.
w31x D.B. Reichling, L. Barratt, J.D. Levine, Heat-induced cobalt entry:
an assay for heat transduction in cultured rat dorsal root ganglion
neurons, Neuroscience 77 Ž1997. 291–294.
w32x S.D. Roper, The cell biology of vertebrate taste receptors, Annu.
Rev. Neurosci. 12 Ž1989. 329–353.
w33x A. Schweitzer, S. Wright, Action of prostigmine and acetylcholine
on respiration, Q. J. Exp. Physiol. 28 Ž1938. 33–47.
w34x M. Shirahata, Y. Ishizawa, A. Garashi, R.S. Fitzgerald, Release of
acetylcholine from cultured cat and pig glomus cells, Adv. Exp.
Med. Biol. 410 Ž1996. 233–237.
w35x K. Steen, P. Reeh, Actions of cholinergic agonists and antagonists
on sensory nerve endings in rat skin in vitro, J. Neurophysiol. 70
Ž1993. 397–405.
w36x L.J. Stensaas, S.J. Fidone, An ultrastructural study of cat petrosal
ganglia: a search for autonomic ganglion cells, Brain Res. 124
Ž1977. 29–39.
w37x D.L. Tanelian, Cholinergic activation of a population of corneal
afferent nerves, Exp. Brain Res. 86 Ž1991. 414–420.
w38x E.H. Vidruk, J.A. Dempsey, Carotid body chemoreceptor activity as
recorded from the petrosal ganglion in cats, Brain Res. 181 Ž1980.
455–459.
w39x P. Willshaw, R.M. McAllen, Sinus nerve efferent activity originates
outside the brain, in: C. Belmonte, D. Pallot, H. Acker, S. Fidone
ŽEds.., Arterial Chemoreceptors, Leicester Univ. Press, Leicester,
UK, 1981, pp. 440–447.
w40x P. Zapata, A. Zuazo, F. Llados, Respiratory and circulatory reflexes
induced by nicotine injections: role of carotid body chemoreceptors,
Arch. Intl. Pharmacodyn. Ther.´ 219 Ž1976. 128–139.
w41x H. Zhong, C. Nurse, Co-cultures of rat petrosal neurons and carotid
body type I cells. A model for studying chemosensory mechanisms,
Adv. Exp. Med. Biol. 410 Ž1996. 189–193.