Differential modulation of spontaneous and evoked neurotransmitter release from hair cells: some novel hypotheses

Pharmacological characterization and differential expression of NMDA receptor subunits in the chicken vestibular system during development

Previous studies demonstrated that in vitro preparations of the isolated vestibular system of diverse animal species still exhibit stable resting electrical activity and mechanically evoked synaptic transmission between hair cells and primary afferent endings. However, there are no reports related to their neurodevelopment. Therefore, this research aimed to examine whether NMDA receptors mediate these electrical signals in an isolated preparation of the chicken vestibular system at three developmental stages, E15, E18, and E21. We found that the spontaneous and mechanically evoked discharges from primary afferents of the posterior semicircular canal were modulated by agonists NMDA and glycine, but not by the agonist d-serine applied near the synapses. Moreover, the individually applied by bath perfusion of three NMDA receptor antagonists (MK-801, ifenprodil, and 2-naphthoic acid) or high Mg²⁺ decreased the resting discharge rate, the NMDA response, and the discharge rate of mechanically evoked activity from these primary afferents. Furthermore, we found that the vestibular ganglion shows a stage-dependent increase in the expression of NMDA receptor subunits GluN1, GluN2 (A-C), and GluN3 (A-B), being greater at E21, except for GluN2D, which was inversely related to the developmental stage. However, in the crista ampullaris, the expression pattern remained constant throughout development. This could suggest the possible existence of presynaptic NMDA receptors. Our results highlight that although the NMDA receptors are functionally active at the early embryonic stages of the vestibular system, NMDA and glycine reach their mature functionality to increase NMDA responses close to hatching (E21).

Calcium-dependent synaptic vesicle trafficking underlies indefatigable release at the hair cell afferent fiber synapse

Sensory hair cell ribbon synapses respond to graded stimulation in a linear, indefatigable manner, requiring that vesicle trafficking to synapses be rapid and nonrate-limiting. Real-time monitoring of vesicle fusion identified two release components. The first was saturable with both release rate and magnitude varying linearly with Ca(2+), however the magnitude was too small to account for sustained afferent firing rates. A second superlinear release component required recruitment, in a Ca(2+)-dependent manner, of vesicles not in the immediate vicinity of the synapse. The superlinear component had a constant rate with its onset varying with Ca(2+) load. High-speed Ca(2+) imaging revealed a nonlinear increase in internal Ca(2+) correlating with the superlinear capacitance change, implicating release of stored Ca(2+) in driving vesicle recruitment. These data, supported by a mass action model, suggest sustained release at hair cell afferent fiber synapse is dictated by Ca(2+)-dependent vesicle recruitment from a reserve pool.

Trimetazidine modulates AMPA/kainate receptors in rat vestibular ganglion neurons

Trimetazidine (1[2,3,4-trimethoxy-benzyl] piperazine, 2 HCl) is an anti-ischemic agent frequently administered as a prophylactic treatment for episodes of angina pectoris and chorioretinal disturbances. It is also employed as a symptomatic treatment of vertigo but its mechanism of action is yet to be defined. Using Fura-2 fluorescence photometry and whole-cell patch-clamp recordings we investigated the effect of trimetazidine on the [Ca(2+)](i) and current responses induced by the application of non-N-methyl-D-aspartate (NMDA) receptor agonists on low density vestibular ganglion neuronal cultures explanted from 3 day s postnatal rats. Trimetazidine blocked the [Ca(2+)](i) and current responses induced by 100 microM applications of both kainate and alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionate (AMPA). These responses were dependent on external Ca(2+) and were blocked by the voltage-dependent Ca(2+) channel blockers Ni(2+) and Cd(2+) . Trimetazidine only acts on the AMPA/kainate receptors and had no effect on K(+)-induced depolarizations. Dose-dependent curves were obtained for the inhibition by 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) and trimetazidine (IC(50) 7 microM and 0.7 microM) of kainate stimulations. After AMPA stimulation, dose-response inhibition curves showed an IC(50) of 3 microM for CNQX and 25 microM for trimetazidine. These results indicate that trimetazidine could be a potent antagonist of AMPA/kainate receptors in vestibular ganglion neurons. This may explain the protective role of trimetazidine in the inner ear suggesting an anti-excitotoxic activity.

Time and intensity coding at the hair cell's ribbon synapse

The activity of individual afferent neurones in the mammalian cochlea can be driven by neurotransmitter released from a single synaptic ribbon in a single inner hair cell. Thus, a ribbon synapse must be able to transmit all the information on sound frequency, intensity and timing carried centrally. This task is made still more demanding by the process of binaural sound localization that utilizes separate computations of time and intensity, with temporal resolution as fine as 10 micros in central nuclei. These computations may rely in part on the fact that the response phase (at the characteristic frequency) of individual afferent neurones is invariant with intensity. Somehow, the ribbon synapse can provide stronger synaptic drive to signal varying intensity, without accompanying changes in transmission time that ordinarily occur during chemical neurotransmission. Recent ultrastructural and functional studies suggest features of the ribbon that may underlie these capabilities.

Gamma-aminobutyric acid is present in a spatially discrete subpopulation of hair cells in the crista ampullaris of the toadfish Opsanus tau

Although gamma-aminobutyric acid (GABA) and glutamate are known to be present in the vestibular sensory epithelia of a variety of species, the functional relationship between these two transmitters is not clear. The present study addresses the three-dimensional spatial distribution of GABA and glutamate immunoreactivity in the vestibular labyrinth of the oyster toadfish by using whole end organs labeled by immunofluorescence with monoclonal anti-GABA and/or antiglutamate antibodies and visualized as whole mounts by multiphoton confocal microscopy. We find glutamate-immunoreactive hair cells present throughout the sensory epithelium. In contrast, prominent GABA immunoreactivity is restricted to a small population of hair cells located in the central region of the crista. Double immunofluorescence reveals two distinct staining patterns in GABA-labeled hair cells. Most ( approximately 80%) GABA-labeled cells show trace levels of glutamate, appropriate for the metabolic/synthetic role of cytoplasmic glutamate. The remainder of the GABA-stained cells contain substantial levels of both GABA and glutamate, suggesting transmitter colocalization. In the toadfish utricle, glutamatergic hair cells are present throughout the macula. GABA-immunoreactive hair cells follow the arc of the striola, and most GABA-labeled receptor cells coexpress glutamate. The localization of GABA was explored in other species as well. In the pigeon, GABAergic hair cells are present throughout the crista ampullaris. Our findings demonstrate that multiple, neurochemically distinct types of hair cells are present in vestibular sensory epithelia. These observations, together with the excitatory activity generally associated with 8th nerve afferent fibers, strongly suggest that GABA serves an important, specific, and complex role in determining primary afferent response dynamics.

Glutamate induced currents in isolated inner hair cells from guinea-pig cochlea

We investigated the direct action of glutamate (Glu) on the membrane current of isolated inner hair cells of guinea-pig cochlea. Glu elicited inward currents at a holding potential of -70 mV. Eight of 13 cells showed a steady inward current, while five of 13 cells showed a fast and rapidly desensitized current. I-V relationships demonstrated that the reversal potential of Glu-induced current was near 0 mV. Glu-induced currents were dose-dependent, where the half maximum concentration (K(d)) was 41 microM and Hill coefficient (n) was 1.75.

Hyperpolarization-activated, cyclic AMP-gated, HCN1-like cation channel: the primary, full-length HCN isoform expressed in a saccular hair-cell layer

The expression of transcript for hyperpolarization-activated, cyclic nucleotide-sensitive cation channel (HCN) isoforms underlying hyperpolarization-activated, inward current (I(h)) has been determined for a model hair-cell preparation from the saccule of the rainbow trout, Oncorhynchus mykiss. Based upon identification from homology to known vertebrate HCN cDNA sequence, cloning of PCR products amplified with degenerate primers indicated an expression frequency of 7:2:1 (HCN1:HCN2:HCN4) for the hair-cell sheet compared with 1:1:7 for brain. Full-length sequence has been obtained for the HCN1-like isoform representing the primary HCN transcript expressed in the hair-cell preparation. The channel protein is 938 amino acids in length with 93% amino acid identity for the region extending from the S1-S6 membrane spanning domains through the voltage-pore and cyclic nucleotide-binding domains, compared with HCN1 for rabbit, rat, mouse and human. The N- and C-terminal regions are less homologous, with 39-51% and 43-44% amino acid identities, respectively. Compared with other vertebrate HCN1, the hair-cell HCN1 contains additional consensus phosphorylation sites associated with unique repeats in the carboxy terminus. The HCN1-like transcript has been localized to hair cells of the saccular sensory epithelia by in situ hybridization. Previous electrophysiological studies have identified I(h) as the sole inwardly rectifying ion channel in a specific population of hair cells of the saccule of frogs [J Neurophysiol (1995) 73:1484] and fish [J Physiol (1996) 495:665]. I(h) is an important determinant of the resting membrane potential, and for this population of hair cells, is predicted to maintain the membrane potential within a voltage range allowing the voltage-gated calcium channels to open, permitting "spontaneous" release of transmitter. The molecular properties of the HCN1-like isoform underlying I(h) expressed in the saccular hair cells of the teleost, trout, may consequently impact spontaneous release of transmitter from hair cells of the saccule.

Signal discrimination in the semicircular canals: a role for group I metabotropic glutamate receptors

Fluorescence immunocytochemistry indicates that enzymatically isolated semi-circular canal (SCC) hair cells express metabotropic glutamate receptors (mGluRs) 1a and 5. Antibody-antigen preadsorption controls proved entirely negative. Applied while mechanically stimulating the posterior SCC with a piezo-electric bimorph, the non-competitive, mGluR5-selective antagonist MPEP-HCl (1 microM-3 mM) dose-dependently reduces mechanically evoked facilitation of afferent discharge rate (IC50 136 microM; n = 4), while having no effect on tonic, unstimulated afferent discharge. It thus appears that group I mGluRs on SCC hair cells are activated during mechanical stimulation, but are not activated under tonic transmitter release conditions. We conclude that group I mGluRs expressed by SCC hair cells may serve as a mechanism for the selective amplification of mechanically evoked transmitter release, thereby enhancing signal discrimination at the VHC-vestibular afferent synapse.(50)

Transmitter release from Rana pipiens vestibular hair cells via mGluRs: a role for intracellular Ca(++) release

The response of the semicircular canal (SCC) to the group I mGluR-selective agonist dihydroxyphenylglycine (DHPG; 300 microM) - facilitation of afferent discharge rate - was dose-dependently reduced by the phospholipase C inhibitor U-73122 (1-100 microM; IC(50): 22 microM), the smooth endoplasmic reticulum Ca(++) ATPase inhibitor thapsigargin (100 nM-3 microM; IC(50): 500 nM), and xestospongin C (100 pM-1 microM; IC(50): 11 nM), an inositol trisphosphate receptor (IP(3)R) antagonist. Ryanodine, a modulator of Ca(++)-induced Ca(++) release, biphasically facilitated, then suppressed this response (1 nM-1 mM; approximate IC(50): 50 microM). 5 mM caffeine increased the amplitude (34.6+/-13.4%) and duration (453+/-169.8%; n=4) of the response of the SCC to DHPG, while 50 mM caffeine eliminated this response (n=2). The protein kinase C inhibitor bisindolylmaleimide I-HCl (10-100 microM; n=3) and the cyclic-ADP ribose antagonist 8-Br-cyclic-ADP ribose (1-10 microM; n=3) had no effect on the response of the SCC to DHPG. These data suggest that the increase in transmitter release following activation of group I mGluRs on vestibular hair cells is associated with intracellular Ca(++) release from both IP(3)-sensitive and ryanodine/caffeine-sensitive intracellular Ca(++) stores. Such positive feedback on transmitter release may serve to enhance the contrast between the spontaneous and stimulus-evoked modes of hair cell transmitter release, thereby optimizing signal discrimination at the synapse between hair cells and vestibular afferent fibers.

Subcellular immunolocalization of NMDA receptor subunit NR1, 2A, 2B in the rat vestibular periphery

The immunohistochemical localization of the NMDA glutamate receptor subunits NR1, NR2A, and NR2B was investigated in the rat vestibular periphery at the light and electron microscopy level using specific antipeptide antibodies. The afferent calyceal terminals and nerve fibers innervating type I vestibular hair cells were strongly NR1, NR2A, and NR2B immunoreactive. Under electron microscopy, the basolateral type I hair cell membrane was NR1 immunoreactive. The type II hair cell and its afferent boutons were NR1, NR2A, and NR2B non-immunoreactive. Nearly all of Scarpa's ganglion neurons were NR1 immunoreactive, but there was a subset of NR2A non-immunoreactive neurons. Additionally, the larger sized Scarpa's ganglia neurons were NR2B immunoreactive, while the smaller neurons were non-immunoreactive. These findings are strong evidence for functional NMDA receptor mediation or modulation of afferent excitatory neurotransmission from type I but not type II vestibular hair cells to the primary afferent nerve. The receptor subtype(s) may be a combination of NR1/NR2A, NR1/NR2B, and/or NR1/NR2A/NR2B.

Nitric oxide in the afferent synaptic transmission of the axolotl vestibular system

This study was performed using intracellular and multiunit extracellular recording techniques in order to characterize the role of nitric oxide in the afferent synaptic transmission of the vestibular system of the axolotl (Ambystoma tigrinum). Bath application of nitric oxide synthase inhibitors N(G)-nitro-L-arginine (0.01microM to 10microM) and N-nitro-L-arginine methyl ester hydrochloride (0.1microM to 1000microM) elicited a dose-dependent decrease in the basal discharge of the semicircular canal afferent fibers. N(G)-Nitro-L-arginine also diminished the response to mechanical stimuli. Moreover, N(G)-nitro-L-arginine (1microM) produced a hyperpolarization associated with a decrease in the spike discharge and diminished the frequency of the excitatory postsynaptic potentials on afferent fibers recorded intracellularly. Nitric oxide donors were also tested: (i) S-nitroso-N-acetyl-DL-penicillamine (0.1microM to 100microM) increased the basal discharge and the response to mechanical stimuli. At the maximum effective concentration (100microM) this drug affected neither the amplitude nor the frequency of the excitatory postsynaptic potentials. However, it slightly depolarized the afferent neurons and decreased their input resistance. (ii) 3-Morpholino-sydnonimine hydrochloride did not significantly affect the basal discharge or the mechanically evoked peak response of afferent neurons at any of the concentrations used (1microM to 1000microM). However, after 10min of perfusion in the bath, 1microM and 10microM 3-morpholino-sydnonimine hydrochloride significantly modified the baseline of the mechanically evoked response, producing an increase in the mean spike discharge of the afferent fibers. These results indicate that nitric oxide may have a facilitatory role on the basal discharge and on the response to mechanical stimuli of the vestibular afferent fibers. Thus, nitric oxide probably participates in the sensory coding and adaptative changes of vestibular input in normal and pathological conditions.

Evaluation of the vestibular aqueduct in vestibulocochlear disorders by magnetic resonance imaging

The visibility of the vestibular aqueduct (VA) was examined using magnetic resonance imaging (MRI) in 95 patients: 15 patients with Meniere's disease, 4 with vestibular Meniere's disease, 4 with cochlear Meniere's disease and 72 patients with other vestibular and/or cochlear disorders. In order to visualize the VA, the T2-weighted image (T2WI) and the proton-density weighted image (PDWI) were obtained in the sagittal plane with a head coil. The visibility of the VA was classified into 4 grades, i.e. grade 0 (not visible), grade 1 (partially visible with PDWI), grade 2 (partially visible with T2WI) and grade 3 (clearly visible with T2WI). The visibility of the VA was significantly lower bilaterally in Meniere's disease, vestibular Meniere's disease, cochlear Meniere's disease and idiopathic bilateral sensorineural hearing loss (IBSNHL) than in the other diseases. The differences among Meniere's disease, vestibular Meniere's disease and cochlear Meniere's disease were not significant. The significance of decreased visibility in IBSNHL is unknown so far. The VA studies using MRI strongly suggest that the pathogenesis of Meniere's disease, vestibular Meniere's disease, cochlear Meniere's disease and IBSNHL is, at least in part, related to the findings of the VA.

Contribution of glutamate receptors to spontaneous and stimulus-evoked discharge in afferent fibers innervating hair cells of the Xenopus lateral line organ

The relative contributions of NMDA (N-methyl-D-aspartate) and non-NMDA glutamate receptors to spontaneous and stimulus-evoked transmission at the hair cell/afferent fiber synapse were determined in the Xenopus laevis lateral line organ. The non-NMDA receptor antagonist, CNQX (6-cyano-7-nitroquinoxaline-2,3-dione), reversibly reduced both spontaneous and stimulus-evoked discharge rate with an EC(50) of 0.5 microM. NMDA receptor antagonism with the combination of chlorokynurenic acid (100 microM) and elevated magnesium (1.1 mM), or elevated magnesium alone, blocked responses to NMDA without significantly altering spontaneous or stimulus-evoked discharge rate or the responses to kainate. All non-NMDA receptor agonists tested increased discharge rate at low concentrations and, at higher concentrations, increased, then suppressed discharge rate. The EC(50)s were: domoic acid (2.4 mcM)<quisqualic acid (6 mcM)<kainic acid (18 mcM)<AMPA (82 mcM)<<glutamate (1150 mcM). NMDA and ibotenic acid also produced an increase in discharge followed by a suppression, but the suppressive phase of the response predominated and maximum increases in discharge rates were low compared to effects of the non-NMDA agonists. The EC(50)s were: NMDA (148 mcM)<ibotenic acid (463 mcM). The EC(50) for the suppression of afferent discharge that followed the initial excitatory effect was similar to the EC(50) for excitation. Perfusion with active concentrations of kainate, AMPA, or NMDA did not alter the threshold for electrical stimulation of these nerve fibers. We conclude that most of the postsynaptic signal normally seen in afferent fibers is mediated by non-NMDA receptors.

Glycine induced calcium concentration changes in vestibular type I sensory cells

Glutamate is the neurotransmitter of the synapse between vestibular type I hair cells and the afferent nerve calyx. This calyx may also be involved in local feedback, which may modify sensory cell activity via N-methyl-D-aspartate (NMDA) receptors. Glycine is the co-agonist of glutamate in NMDA receptor activation. Both agents have been detected by immunocytochemistry in the nerve calyx. Glutamate and NMDA stimulations cause changes in the intracellular calcium concentration ([Ca(2+)](i)) of isolated type I sensory cells. We investigated the effect of glycine stimulation on [Ca(2+)](i) in guinea pig type I sensory cells by spectrofluorimetry with fura-2. Glycine application to isolated type I sensory cells induced a rapid and transient increase in [Ca(2+)](i). The fluorescence ratio increased by 55% above the resting level. The peak was reached in 9 s and the return to basal level took about 20 s. A specific antagonist of the glycine site on NMDA receptors, 7-chlorokynurenate (10 microM), decreased the calcium response to glycine by 60%. Glycine may activate NMDA receptors. Glycine may also activate the strychnine-sensitive glycine receptor-gated channel. Strychnine (50 microM) decreased the calcium response to glycine by 60%. Thus, glycine probably induces calcium concentration changes in type I vestibular sensory cells via NMDA receptors and/or glycine receptors.

Effects of caloric stimuli on frog ampullar receptors

The observation that caloric nystagmus can be evoked even in microgravity conditions argues against Barany's convective theory. To justify this result, gravity-independent mechanisms (mainly endolymphatic volume changes and direct action of the temperature on vestibular sensors) are believed to contribute to caloric-induced activation of vestibular receptors. To define the importance of both gravity-dependent and gravity-independent mechanisms, the posterior semicircular canal of the frog was thermally stimulated by a microthermistor positioned close to the sensory organ. The stimulus produced a gravity-dependent transcupular pressure difference that, depending on the position of the heater, could result in either excitation or inhibition of ampullar receptor sensory discharge. When the heater was positioned on the ampulla, or when the canal rested on the horizontal plane, no responses could be evoked by thermal stimuli. These results suggest that, in our experimental conditions (DeltaT up to 1.5 degrees C), neither a thermally induced expansion of the endolymph nor a direct action of the temperature on vestibular sensors play any major role.

Stochastic resonance in cochlear signal transduction

The use of a stochastic resonance model contributes crucially to our comprehension of the intensity resolution characteristics of the mammalian cochlea. In guinea pigs, as demonstrated by different statistical methods, the temporal distribution of the interspike intervals of the spontaneous activity reflects an intrinsic cochlear white noise process, demanded as basic requirement for manifest stochastic resonance phenomena. Brownian motion of cochlear fluids is discussed as the underlying white noise motor. Following our model, the amount of white noise, adjusted at the level of the stereocilia of the inner hair cells, determines the threshold, dynamic range and intensity discrimination limen of an individual afferent neuron of the mammalian cochlea.

The metabotropic glutamate receptors of the vestibular organs

This research sought to test the presence and function of metabotropic excitatory amino acid receptors (mGluR) in the frog semicircular canal (SCC). The mGluR agonist +/- 1-aminocyclopentane-trans-1,3-dicarboxylate (ACPD) produced an increase in afferent firing rates of the ampullar nerve of the intact posterior canal. This increase was not due to a stimulation of cholinergic efferent terminals or the acetylcholine (ACh) receptor, since atropine, in concentrations which blocked the response to exogenous acetylcholine, did not affect the response to ACPD. Likewise, ACPD effects were not due to stimulation of postsynaptic NMDA receptors, since the NMDA antagonist D(-)-2-amino-5-phosphonopentanoate (AP-5) did not affect the response to ACPD, reinforcing the reported selectivity of ACPD for mGluRs. When the SCC was superfused with artificial perilymph known to inhibit hair cell transmitter release (i.e. low Ca-high Mg), ACPD failed to increase afferent firing. This suggests that the receptor activated by ACPD is located on the hair cell. Pharmacological evidence suggested that the mGluRs involved in afferent facilitation belong to Group I (i.e. subtypes 1 and 5). In fact, the Group III agonist AP-4 had no effect, and the ACPD facilitatory effect was blocked by the Group I mGluR antagonists (S)-4-carboxyphenylglycine (CPG) and (RS)-1-aminoindan-1,5-dicarboxylic acid (AIDA). Additional pharmacological evidence supported the presence of Group I mGluRs. Interestingly, the mGluR antagonists, AIDA and 4CPG, by themselves did not affect the resting firing rates of ampullar afferents. This may suggest that the mGluRs are not involved in resting activity but perhaps only in evoked activity (as suggested in Guth et al. (1991) Hear. Res. 56, 69-78). In addition, the mRNA for the mGluR1 has been detected in hair cells of both SCC, utricle, and saccule. In summary, the evidence points to an mGluR localized to the hair cell (i.e. an autoreceptor) which may be activated to produce a positive feedback augmentation of evoked but not resting transmitter release and thus affect afferent activity.

Mechanisms and effects of transepithelial polarization in the isolated semicircular canal

Previous studies have shown that galvanic stimulation of semicircular canal organs can modulate their afferent discharge. However, it has not been resolved whether this modulation derived from direct stimulation of hair cells, afferent nerve fibers, some combination of the two, or some as yet unknown path. This problem is addressed in the present study. Experiments were designed first to determine the gross current path necessary for the DC current to modulate afferent firing. These led to the conclusion that the current path had to flow between endolymph and perilymph across the neuroepithelium. Next, the various components in this established path were considered: the afferents, the hair cells, between the hair cells, or some combination of the three. These experiments led to the conclusion that the current pathway was across the hair cells causing transmitter release and thus affecting afferent activity.

Effects of nimodipine on noise-induced hearing loss

The effects of nimodipine, a calcium channel blocker, on noise-induced hearing loss were examined in gerbils. Animals were implanted subcutaneously with a timed-release pellet containing either nimodipine (approximately 10 mg/kg/day) or placebo and exposed to either 102 or 107 dBA noise. Serum levels were tested in two subjects and were in the range known to protect humans from cerebral artery vasospasm and ischemia-related neurologic deficits. Nimodipine and control groups had similar amounts of noise-induced (a) permanent threshold shift; (b) reductions in distortion product otoacoustic emissions; (c) reductions in tuning and suppression of the compound action potential; and (d) loss of outer hair cells. The results suggest that nimodipine, at a dose which results in clinically relevant serum levels, does not provide protection from the effects of moderately intense noise exposures.

Effects of betahistine on vestibular receptors of the frog

Betahistine is widely used in the symptomatic treatment of peripheral and central vestibular disorders. However, its remains unknown whether the drug can act directly on inner ear sensory organs. To this end, the effects of betahistine (10(-7)-10(-2) M) were examined on isolated preparations of frog semicircular canal mounted in a double-celled bath which allowed drug administration both in the endolymphatic and in the perilymphatic fluid. The effects of betahistine were evaluated by recording ampullar receptor potentials and nerve firing rate both at rest and during mechanical stimulation of the isolated preparation. The results demonstrated that endolymphatic administration of betahistine had no effect, whereas its perilymphatic administration could reduce greatly ampullar receptor resting discharge but had little effect on mechanically evoked responses. This observation may explain the anti-vertigo effects of betahistine. Vertigo is normally due to uncontrolled changes in vestibular receptor resting discharge. It is therefore probable that any factor able to reduce the resting firing rate of vestibular receptors and, in consequence, its variations, may have an anti-vertigo action.

A role for chloride in the suppressive effect of acetylcholine on afferent vestibular activity

Afferents of the frog semicircular canal (SCC) respond to acetylcholine (ACh) application (0.3-1.0 mM) with a facilitation of their activity while frog saccular afferents respond with suppression (Guth et al., 1994). All recordings are of resting (i.e., non-stimulated) multiunit activity as previously reported (Guth et al., 1994). Substitution of 80% of external chloride (Cl-) by large, poorly permeant anions of different structures (isethionate, methanesulfonate, methylsulfate, and gluconate) reduced the suppressive effect of ACh in the frog saccular afferents. This substitution did not affect the facilitatory response of SCC afferents to ACh. Chloride channel blockers were also used to test further whether Cl- is involved in the ACh suppressive effect. These included: niflumic and flufenamic acids, picrotoxin, 5-nitro-2-(-3-phenylpropylamino)benzoic acid (NPPB), and 4,4'-dinitrostilbene-2,2'-disulfonic acid (DNDS). As with the Cl- substitutions, all of these agents reduced the suppressive response to ACh in the saccule, but not the facilitatory response seen in the SCC. The suppressive effect of ACh on saccular afferents is considered to be due to activation of a nicotinic-like receptor (Guth et al., 1994; Guth and Norris, 1996). Taking into account the effects of both Cl- substitutions and Cl- channel blockers, we conclude that changes in Cl- availability influence the suppressive effect of ACh and that therefore Cl- may be involved in this effect.

Hyperpolarization-activated current (Ih) in primary auditory neurons

A hyperpolarization-activated current (termed I[h]) is believed to provide a pacemaker depolarization in sinoatrial node cells and in some central and peripheral neurons. In the present study, we examined if such an inward cation current exists in primary auditory neurons using the whole-cell patch-clamp technique. A large inward, non-inactivating current was seen during hyperpolarizing steps negative to the resting potential. A depolarizing sag occurred during hyperpolarizing current injection, and upon termination of the current injection there was an overshoot, or a rebound firing. A low concentration of Cs+, but not Ba2+, reversibly blocked the inward current and depolarizing sag. The activation of the current showed voltage dependence with half-activation occurring at -101 +/- 1 mV. The time course of I(h) activation was fitted by double exponential function and was voltage-dependent (time constants: tau1 and tau2 = 480 and 3125 ms at -100 mV, and 66 and 404 ms at -160 mV). The reversal potential of the current was -36 mV measured from tail currents. The conductance of the current was decreased in Na+-free solution, and increased in high K+ solution. Increases in the levels of intracellular cAMP or cGMP enhanced the current. The results suggest that there exists a hyperpolarization-activated inward cation current in mammalian primary auditory neurons. This current may provide a depolarizing current during the membrane hyperpolarization following each firing of the primary auditory nerve.

AMPA receptors in cultured vestibular ganglion neurons: detection and activation

The presence and the activity of alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) glutamate receptors were investigated in mouse cultured vestibular ganglion neurons using immunocytochemistry and measurement of intracellular calcium concentration ([Ca2+]i) by spectrofluorimetry. Cultures of dissociated vestibular ganglia from 18 gestation day mouse embryos were grown in vitro for 3-4 days. Immunocytochemical labelling of AMPA receptor subunits GluR2/R3 and GluR4 was detected in neuron cell bodies and proximal neurites and more lightly in glial cells. There was no clear selective subcellular localization of the different subunits. For the GluR1 subunit a signal was observed only in some neurons and neurites and was weak. Vestibular ganglion neurons responded to fast application of 1 mM glutamate and 10 mM aspartate through unknown receptors by a transient increase in [Ca2+]i. The mean amplitude of this rapid increase was about nine times the resting level and recovery was complete within 30-45 s after the application. If separated by an interval of at least 10 min, consecutive applications produced similar calcium responses. AMPA (1 mM) application induced the same type of responses. Five minutes prior to the AMPA exposure, the application of a specific AMPA antagonist, 6,7-dinitroquinoxaline-2,3-dione (DNQX, 1.5 mM), in the external medium inhibited the response to AMPA. Chelation of external calcium by EGTA (1.5 mM) abolished the responses to drug applications, indicating that an influx of external calcium is involved in the [Ca2+]i increase. These observations suggest that heteromeric AMPA receptors are expressed in vestibular ganglion neurons in culture and play a functional role in their glutamate-induced depolarization. Experiments are in progress using specific AMPA and NMDA antagonists to characterize the participation of the two types of ionotropic glutamate receptors in the glutamate/aspartate-induced intracellular calcium response.

Immunocytochemical localization of the GTP-binding protein G0 alpha in the vestibular epithelium and ganglion of the guinea-pig

The guanine nucleotide binding protein G0 alpha was immunolocalized in the guinea-pig vestibular system by confocal and electron microscopy. The vestibular sensory epithelia consist of the macula utriculi, macula sacculi and cristae ampullaris of the semicircular canals. Two types of hair cells are present in these epithelia. Type I hair cells are surrounded by an afferent nerve calyx that receives efferent innervation and type II hair cells are innervated directly by the afferent and efferent nerves. G0 alpha protein was observed on the inner face of the afferent calyceal membrane surrounding type I hair cells and in nerve endings in contact with type II hair cells. No labelling was found in the stereocilia and cuticular plate of type I and type II hair cells whereas the cytoplasmic matrix displayed a diffuse labelling. The plasma membrane of the supporting cells showed discreet labelling in the confocal microscope that are still confirmed by electron microscopy. A positive reaction was also observed along the plasma membrane of the vestibular ganglion neurons. Immunoblotting with affinity-purified polyclonal rabbit antibodies selective for the 39 kDa alpha subunit of G0 indicated that G0 alpha protein was present in both the vestibular ganglion. That G0 alpha labelling was observed in the cytoplasm of vestibular hair cells and in nerve endings contacting hair cells suggests that G0 may be involved in the modulation of vestibular neurotransmission.

Immunocytochemical localization of aspartate and glutamate in the peripheral vestibular system

Controversy exists concerning the identity of the neurotransmitter in the mammalian peripheral vestibular system. Several candidates have been proposed, including the excitatory amino acids glutamate and aspartate and the inhibitory amino acid gamma-aminobutyric acid (GABA). Previous studies have demonstrated vestibuloneural electrophysiological activity associated with glutamate and aspartate. Paraffin sections of rat vestibular ganglia and end-organs were examined for the presence of glutamate-like and aspartate-like immunoreactivity. Our results demonstrate the presence of both aspartate-like and glutamate-like immunoreactivity in vestibular hair cells, peripheral vestibular nerve fibers, and vestibular ganglion cells. Minimal immunoreactivity was noted in the tissues surrounding these cells. These data add support to the hypothesis that the excitatory amino acids glutamate and aspartate are involved in vestibular neurotransmission.

Cellular and subcellular localization of AMPA-selective glutamate receptors in the mammalian peripheral vestibular system

The cellular and subcellular distribution of AMPA-selective glutamate receptors in the mammalian peripheral vestibular system was examined using antibodies against peptides corresponding to the C-terminal portions of AMPA receptor subunits: GluR1, GluR2/R3 and GluR4. The light and electron microscopic immunocytochemical studies were carried out on Vibratome sections of rat and guinea pig vestibular sensory epithelial and ganglia. In the epithelium, GluR1 subunit immunoreactivity appeared as accumulations of patches outlining the baso-lateral periphery of the type I sensory cells. The GluR1-immunoreactive microareas were postsynaptically distributed on the membranes of calyceal afferent fibers. GluR2/R3 immunoreactivity was present in the sensory cells. GluR4 was not detected. In the vestibular ganglion, the neurons were densely stained with antibodies to GluR2/R3 and GluR4. The fibroblasts and the Schwann cells were also intensely stained with antibodies to GluR2/R3 and GluR4. In the sensory cells, the AMPA receptors, GluR2/R3, may function as (1) autoreceptors controlling afferent neurotransmitter release or (2) 'postsynaptic' receptors activated by the neurotransmitter release of the afferent calyx. The detection of GluR1 at postsynaptic sites in the afferent fibers provides anatomical evidence for the role of glutamate as a neurotransmitter of sensory cells. In the ganglion neurons, GluR2/R3 and GluR4 may represent reserve intracytoplasmic pools of receptor subunits in transit to the postsynaptic sites. In the Schwann cells, GluR2/R3 and GluR4 may be involved in neuronal-glial signalling at the nodes of Ranvier.

Salicylate-induced abnormal activity in the inferior colliculus of rats

The evaluation of the spontaneous activity of 471 units from the external nucleus of the IC revealed that salicylate induces an increase of the spontaneous activity and the emergence of a bursting type of activity longer than 4 spikes. For sharply tuned units, the affected cells were from the frequency range of 10-16 kHz, which corresponds to the behaviorally measured pitch of salicylate-induced tinnitus in rats. An exogenous calcium supplement, provided under the conditions shown to attenuate the behavioral manifestation of salicylate-induced tinnitus, abolished the modification of the spontaneous activity induced by salicylate. Finally, profound changes of activity were observed for cells not responding to contralateral sound. We propose that the observed long bursts of discharges represent tinnitus-related neuronal activity. The results are consistent with the hypothesis that GABA-mediated disinhibition is involved in the processing of tinnitus-related neuronal activity.

Cholinergic agonists increase intracellular calcium concentration in frog vestibular hair cells

Acetylcholine (ACh) is usually considered to be the neurotransmitter of the efferent vestibular system. The nature and the localization of cholinergic receptors have been investigated on frog isolated vestibular hair cells (VHCs), by measuring variations of intracellular calcium concentration ([Ca2+]i), using calcium sensitive dye fura-2. Focal iontophoretic ACh (1 M, 300 nA.40 ms) application induced a rapid increase in [Ca2+]i, reaching a peak in 20 s and representing about 5-fold the resting level (from 61 +/- 6 to 320 +/- 26 nM). Applications of muscarinic agonists as methacholine and carbachol induced weaker calcium responses (from 78 +/- 25 to 238 +/- 53 nM) than the one obtained with ACh applications. These muscarinic agonists were efficient only in precise zones. Desensitization of muscarinic receptors to successive stimulations was significant. Perfusion of nicotine or 1,1-dimethyl-4-phenyl-piperazinium (DMPP), a nicotinic agonist, induced an increase in [Ca2+]i only in some cells (4/28 with DMPP). These results indicated the presence of cholinergic receptors on frog VHCs: muscarinic receptors were more responsive than nicotinic receptors. Presence of muscarinic and nicotinic receptors in the membrane of VHCs could indicate different modulations of VHCs activity mediated by [Ca2+]i and involving an efferent control which represents a central regulation of the vestibular afferent message.

The cholinergic pharmacology of the frog saccule

Stimulation of the efferent nerves to the vestibular organs of the frog's inner ear produces either facilitation or inhibition of afferent firing. Similarly, application of acetylcholine (ACH), the major transmitter of the efferents, can produce both facilitation and/or inhibition as previously reported [Guth et al. (1986) Acta Otolaryngol. 102, 194-204; Norris et al. (1988) Hear. Res. 32, 197-206]. The firing rates of afferent neurons of the semicircular canal (SCC) using multiunit recordings are generally facilitated by ACH. Conversely, the firing rates of afferent units innervating the saccule are generally inhibited by ACH. This latter inhibition is antagonized by strychnine more potently than by curare, which is more potent than atropine. When inhibition is antagonized by strychnine or curare an underlying facilitation is revealed. The inhibition of saccular afferents by ACH shows desensitization requiring about 20 min to recover. The ACH-induced inhibition is mimicked by nicotine at very high concentrations but not by dimethyl phenylpiperazinium or cytisine. The fact that multiunit afferent firing from the SCC is generally facilitated while that from the saccule is generally inhibited by ACH suggests a different distribution of ACH receptors and receptor types (i.e. muscarinic or nicotinic and their subtypes) in the two organs and demonstrates the usefulness of recording from multiple units simultaneously. The difference in distribution of ACH receptors may be important for understanding the physiology of vestibular efferents.

The effect of L-glutamate on the afferent resting activity in the cephalopod statocyst

The effects of bath application of L-glutamate and of excitatory amino acid agonists and antagonists on the resting activity of afferent crista fibers were studied in isolated preparations of the statocyst of the cuttlefish, Sepia officinalis. L-Glutamate (threshold 10(-5) M) and its agonists quisqualate and kainate (thresholds 10(-6) M) increased the resting activity in a dose-dependent manner. Glutamine (threshold 10(-5) M) was also excitatory, while D-glutamate had no effect. Also, no obvious excitatory effects were seen for NMDA and L-aspartate, nor was any antagonistic effect seen for the selective NMDA-receptor antagonist D-2-amino-5-phosphonovaleric acid (D-AP-5). The spider toxin Argiotoxin636 (threshold 10(-11) M), 2-amino-4-phosphonobutyric acid (AP-4), glutamic acid diethyl ester (GDEE), gamma-D-glutamylaminomethyl-sulfonic acid (GAMS), and kynurenic acid decreased the resting activity and effectively blocked or reversed the effect of L-glutamate and its non-NMDA agonists. Preliminary experiments with statocysts from the squid Sepioteuthis lessoniana and the octopod Octopus bimaculoides gave comparable results. All data show that in cephalopod statocysts L-glutamate, via non-NMDA receptors, has an excitatory effect on the activity of afferent fibers, an effect consistent with its possible function as a hair cell transmitter.

Hair cells of different shapes and their placement along the frog crista ampullaris

The list of distinguishing morphological features of hair cells includes: Type I and Type II afferent innervation, and length, shapes and arrangements of stereo- and kinocilia. We now add to this list the shapes of the hair cells themselves and their placement within the mechanosensory organ, in this case the semicircular canal. Although hair cells of the crista ampullaris of the frog are only of Type II they may now be further classified into three sub-groups according to shape: club-, cigar- and pear-shaped. The cigar- and club-shaped hair cells are each about 40% while the pear-shaped cells are about 20% of the total numbers of hair cells in the crista. The differently-shaped hair cells also distribute differently along the crista. The cigar- and club-shaped are more-or-less uniformly distributed with somewhat higher concentrations at the ends of the crista than in the center. The pear-shaped hair cells, on the other hand, are mostly concentrated toward the center of the crista. This distribution of the pear-shaped hair cells, and their shape is reminiscent of the distribution of calyceal endings (Type I hair cell) in the cristae of amniotes [Goldberg et al., Hear. Res. 49, 89-102 (1990) in Chinchilla; Fernandez et al., Soc. Neurosci. Abstr. 17, 312 (1991) in Monkey]. There are some quantitative differences between hair cells of the same shape but from different portions of the crista. For instance, pear-shaped hair cells of the center are generally of greater cross-sectional area than those of the ends.(ABSTRACT TRUNCATED AT 250 WORDS)

Evidence for NMDA receptor in the afferent synaptic transmission of the vestibular system

This study aimed to define the pharmacology and physiological role of the N-methyl-D-aspartate (NMDA) receptor in the synapse between the hair cells and primary afferent neurons in the vestibular system. The spontaneous and mechanically evoked spike discharges of vestibular nerve fibers were extracellularly recorded in isolated inner ear from the axolotl (Ambystoma tigrinum). Pressure ejection of NMDA (10(-6) to 10(-3) M) elicited a dose-dependent increase of the basal spike discharge from the vestibular nerve fibers. Extracellular magnesium antagonized the NMDA effect in a dose-dependent manner. D(-)-2-amino-5-phosphonovaleric acid (AP5, 10(-5) to 10(-3) M) and 7-chloro-kynurenic acid (7ClKyn, 10(-6) to 10(-3) M) inhibited the basal activity of the vestibular nerve fibers. 7ClKyn also diminished the responses elicited by the mechanical stimulation of the preparation. Glycine (10(-9) to 10(-6) M) applied by bath substitution enhanced the NMDA response, and the glycine agonist D-serine partially reversed the 7ClKyn inhibitory action. These results suggest that NMDA receptors participate in the generation of the basal spike discharge of vestibular system primary afferent neurons, but its activation is not critical for the response to brief mechanical stimuli.

The vestibular type I hair cells: a self-regulated system?

In this study the functional role of afferent nerve calyx surrounding the type I vestibular hair cells was investigated. Synaptic microvesicles were present at the apex of the calyx in the vestibular epithelium of human foetuses at 9 weeks from gestation. Whole cell clamped type I hair cells isolated from guinea pig epithelium presented active movements as shortening of the neck and tilting of the cuticular plate at the cessation of the depolarising step. These movements were calcium dependent. With the aim of establishing the kinetics of calcium influx during the cell depolarisation, intracellular free calcium rate variations were investigated by coupling cytofluorimetry technique with whole cell patch clamp. An increase of intracellular calcium was only observed at the repolarisation of type I hair cells. Thus, a regulatory short-loop is thought to exist to control adaptation phenomena at the upper part of the type I hair cell. It is suggested that this occurs through the release of a neurotransmitter from the apex of the afferent calyx.

Glutamate receptors on type I vestibular hair cells of guinea-pig

Afferent nerve calyces which surround type I vestibular hair cells (VHCI) have recently been shown to contain synaptic-like vesicles and to be immunoreactive to glutamate antibodies. In order to understand the physiological significance of these observations, the presence of glutamate receptors on type I vestibular sensory cells has been investigated. The effect of excitatory amino acids applied by iontophoresis was examined by spectrofluorimetry using fura-2 sensitive dye. Glutamate application caused a rapid and transient increase in intracellular calcium concentration ([Ca2+]i), in a dose-dependent manner. The ionotropic glutamate receptors agonists N-methyl-D-aspartic acid (NMDA), alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) and quisqualic acid (QA) induced an increase of [Ca2+]i. The NMDA receptor antagonist 2-amino-5-phosphonovaleric acid and the AMPA receptor antagonist 6,7-dinitroquinoxaline-2,3-dione partially blocked the glutamate response, by 39 +/- 10 and 53 +/- 11% respectively. Metabotropic receptors were also revealed by the specific agonist trans-1-amino-cyclopentyl-1,3-dicarboxylate. The presence of different glutamate receptors on the VHCI membrane suggests two kinds of feedback. (i) At the base of the sensory cell, autoreceptors may locally control the synaptic transmission. (ii) At the apex, postsynaptic receptors may modulate sensory transduction from glutamate release at the upper part of the afferent nerve calyx. These feedbacks suggest presynaptic modulation of the vestibular hair cell response which could affect its sensitivity.

The correlated blanching of synaptic bodies and reduction in afferent firing rates caused by transmitter-depleting agents in the frog semicircular canal

Synaptic bodies (SBs) associated with rings of synaptic vesicles and well-defined, pre- and post-synaptic membrane structures are indicators of maturity in most hair cell-afferent nerve junctions. The role of the SBs remains elusive despite several experiments showing that they may be involved in storage of neurotransmitter. Our results demonstrate that SBs of the adult posterior semicircular canal (SCC) cristae hair cells become less electron dense following incubation of the SCC with the transmitter-depleting drug tetrabenazine (TBZ). Objective quantification and comparison of the densities of the SBs in untreated and TBZ-treated frog SCC demonstrated that TBZ significantly decreased the electron density of SBs. This reduction in electron density was accompanied by a reduction in firing rates of afferent fibers innervating the posterior SCC. A second transmitter-depleting drug, guanethidine, previously shown to reduce the electron density of hair cell SBs, also reduced the firing rates of afferent fibers innervating the posterior SCC. In contrast, the electron density of dense granules (DG), similar in size and shape to synaptic bodies (SB) in hair cells, did not change after incubation in TBZ, thus indicating that granules and SBs are not similar in regard to their electron density. The role of SBs in synaptic transmission and the transmitter, if any, stored in the SBs remain unknown. Nonetheless, the association of the lessening of electron density with a reduction in afferent firing rate provides impetus for the further investigation of the SB's role in neurotransmission.

Effects of cromakalim (BRL 34915) on resting and evoked activity in frog semicircular canals

The effects of endolymphatic administration of cromakalim and glibenclamide were tested in isolated semicircular canals of the frog. The actions of the drugs were evaluated by recording: (1) the transepithelial potential between the endolymphatic and the perilymphatic sides of the crista ampullaris; (2) the slow nerve potential from the ampullar nerve; (3) the action potential discharge in afferent ampullar nerve fibres; and (4) the perilympathatic potassium concentration in the fluid bathing the outer surface of the crista ampullaris. The above mentioned parameters were recorded both at rest and during mechanical stimulation of the sensory organ. The results demonstrated that the endolymphatic administration of cromakalim (10(-4) M) produced an increase in both ampullar receptor resting activity and in perilymphatic K+ resting levels. By contrast all the parameters related to the mechanically evoked responses were practically unaffected. Glibenclamide (10(-4) M) proved able to cancel or to prevent cromakalim effects. These data suggest that the membrane of the hair cells is endowed with K+ channels regulated by internal ATP whose activation is mainly involved in the processes sustaining ampullar receptors' resting firing rate.