AMPA-preferring glutamate receptors in cochlear physiology of adult guinea-pig

Abstract

1. The present study was designed to determine which glutamate (Glu) receptors are involved in excitatory neurotransmission at the first auditory synapse between the inner hair cells and the spiral ganglion neurons. 2. The Glu receptors present at the membrane level were investigated on isolated spiral ganglion neuron somata from guinea-pigs by whole-cell voltage-clamp measurements. Glu and AMPA induced a fast onset inward current that was rapidly desensitized, while kainate induced only a non-desensitizing, steady-state current. NMDA induced no detectable current. 3. To further discriminate between the AMPA and kainate receptors present, we used the receptor-specific desensitization blockers, cyclothiazide and concanavalin A. While no effect was observed with concanavalin A, cyclothiazide greatly enhanced the Glu-, AMPA- and kainate-induced steady-state currents and potentiated Glu-induced membrane depolarization. 4. To extrapolate the results obtained from the somata to the events occurring in situ at the dendrites, the effects of these drugs were evaluated in vivo. Cyclothiazide reversibly increased spontaneous activity of single auditory nerve fibres, while concanavalin A had no effect, suggesting that the functional Glu receptors on the somata may be the same as those at the dendrites. 5. The combination of a moderate-level sound together with cyclothiazide increased and subsequently abolished the spontaneous and the sound-evoked activity of the auditory nerve fibres. Histological examination revealed destruction of the dendrites, suggesting that cyclothiazide potentiates sound-induced Glu excitotoxicity via AMPA receptors. 6. Our results reveal that fast synaptic transmission in the cochlea is mainly mediated by desensitizing AMPA receptors.

Figure 1. Glutamate receptor agonist-induced current responses in spiral ganglion neurons at a holding potential of -70 mV

A, glutamate (Glu) induced an inward current response. B, AMPA-induced current response. C, kainic acid (KA)-induced current response. Inset, NMDA induced no current response at a concentration of 1 mM delivered in a Mg²⁺-free solution containing 10 μM glycine.

Figure 2. Effect of concanavalin A (Con A) and cyclothiazide (CT) on the Glu-induced current response in spiral ganglion neurons

A, Glu (100 μM)-induced response. B, co-application of Glu (100 μM) and concanavalin A (0.5 mg ml⁻¹) following pretreatment with concanavalin A (0.5 mg ml⁻¹) for 3 min. C, Glu-induced response after 3 min of the co-application of Glu and concanavalin A. D, co-application of Glu and cyclothiazide (100 μM) after pretreatment with cyclothiazide for 1 min. E, Glu-induced response after 20 min of washout. Data in A-E are from the same cell. F, Glu (1 mM)-induced response. G, co-application of Glu (1 mM) and concanavalin A (0.5 mg ml⁻¹) following pretreatment with concanavalin A (0.5 mg ml⁻¹) for 3 min. H, co-application of Glu (1 mM) and cyclothiazide (100 μM) after pretreatment with cyclothiazide for 1 min. In A-H the membrane potential was held at -70 mV.

Figure 3. Dose-response curves of the Glu-induced current in the absence (control) and presence of concanavalin A (con A; 0.5 mg ml⁻¹) and cyclothiazide (CT; 100 μM) in isolated spiral ganglion neurons

The membrane potential was held at -70 mV. All currents were normalized to the current elicited by 100 μM Glu. Continuous lines are non-linear least-square fits to a logistic equation with a Hill coefficient of 1.5 for the control and concanavalin A groups (K_d = 95 μM) and 1.8 for the cyclothiazide group (K_d = 175 μM). Each point is the mean ±s.e.m. of 3-93 neurons. Inset, the dose-response curve of the control (Glu, alone).

Figure 4. Effect of concanavalin A (Con A) and cyclothiazide (CT) on the Glu-induced membrane depolarization under current-clamp mode

A, Glu (100 μM)-induced membrane depolarization. B, co-application of Glu (100 μM) and concanavalin A (0.5 mg ml⁻¹) following pretreatment with concanavalin A (0.5 mg ml⁻¹) for 3 min. C, co-application of Glu and cyclothiazide (100 μM) after pre-treatment with cyclothiazide (100 μM) for 1 min. D, summary of the effects of concanavalin A (0.5 mg ml⁻¹) and cyclothiazide (100 μM) on the Glu (100 μM)-induced membrane depolarization in spiral ganglion neurons. Each bar represents the mean ±s.e.m. for the indicated number of cells. The asterisk indicates a statistically (P < 0.05) significant difference from Glu alone (control, ANOVA and Student-Newmann-Keuls test).

Figure 5. Effect of concanavalin A (Con A) and cyclothiazide (CT) on the AMPA-induced current response in spiral ganglion neurons at a holding potential of -70 mV

A, AMPA (100 μM)-induced inward current. B, co-application of AMPA (100 μM) and concanavalin A (0.5 mg ml⁻¹) following pre-treatment with concanavalin A (0.5 mg ml⁻¹) for 3 min. C, co-application of AMPA and cyclothiazide (100 μM) after pre-treatment with cyclothiazide (100 μM) for 1 min. D, summary of the effects of concanavalin A (0.5 mg ml⁻¹) and cyclothiazide (100 μM) on the AMPA (100 μM)-induced current response in spiral ganglion neurons. Each bar represents the mean ±s.e.m. for the indicated number of cells. The asterisk indicates a statistically (P < 0.05) significant difference from AMPA alone (control).

Figure 6. Effect of concanavalin A (Con A) and cyclothiazide (CT) on the kainate (KA)-induced current response in spiral ganglion neurons at a holding potential of -70 mV

A, KA (100 μM)-induced inward current. B, co-application of KA (100 μM) and concanavalin A (0.5 mg ml⁻¹) following pre-treatment with concanavalin A (0.5 mg ml⁻¹) for 3 min. C, co-application of KA (100 μM) and cyclothiazide (100 μM) after pre-treatment with cyclothiazide (100 μM) for 1 min. D, summary of the effect of concanavalin A (0.5 mg ml⁻¹) and cyclothiazide (100 μM) on the KA (100 μM)-induced current response in spiral ganglion neurons. Each bar represents the mean ±s.e.m. for the indicated number of cells. The asterisk indicates a statistically (P < 0.05) significant difference from KA alone (control).

Figure 7. Effects of intracochlear perfusions of increasing concentrations of cyclothiazide

The data shown represent the effects of cyclothiazide on compound action potentials (CAP) and its N1 latency evoked by 90 dB SPL and 50 dB SPL tone bursts. The recordings were done before any perfusion, after perfusion of artificial perilymph alone, after perfusion with the control artificial perilymph containing 0.1% DMSO and after increasing concentrations of cyclothiazide (in μM: 1, 10, 100, 200, 300, 400 and 500). A last recording was done after a final perfusion with control artificial perilymph containing 0.1% DMSO. The asterisks indicate a statistically (ANOVA and Student-Newmann-Keuls test; P < 0.05) significant difference from the values obtained after the first perfusion with the control artificial perilymph. Each bar represents the mean ±s.e.m. calculated from 5 animals.

Figure 8. Effect of concanavalin A (Con A) and cyclothiazide (CT) on spontaneous rate of the auditory nerve fibres

A, the number of spikes per second during perfusion with artificial perilymph (AP) and 20 mg ml⁻¹ concanavalin A. The characteristic frequency of this fibre was 14 kHz with a spontaneous rate of 62 spikes s⁻¹. B, the number of spikes per second during perfusion with artificial perilymph (AP) and 200 μM cyclothiazide (CT). The characteristic frequency of this fibre was 8.5 kHz with a spontaneous rate of 75 spikes s⁻¹. Note that cyclothiazide drastically increases the spontaneous rate activity of the fibre to about 150 spikes s⁻¹ and that this effect was totally reversible after washing the cyclothiazide out of the cochlea with artificial perilymph. C, blocking effect of GYKI 52466 on cyclothiazide-increased activity. Here, the characteristic frequency of this fibre was 19 kHz and its spontaneous rate was 82 spikes s⁻¹. Note that the cyclothiazide effect was not only blocked, but reversed, by 200 μM GYKI 52466, and a full recovery was observed after washing the drugs out of the cochlea with artificial perilymph.

Figure 9. Effect of cyclothiazide (CT) on sound-evoked activity of a single auditory nerve fibre

Post-stimulus time histograms (PSTHs) were constructed from 40 tone bursts of 500 ms presented 60 dB above the threshold of the fibre at a rate of 1 s⁻¹. In this figure, t = 0 coincides with the onset of the tone burst. The spontaneous rate of this fibre was 3 spikes s⁻¹ and its characteristic frequency was 8 kHz. The histograms were obtained during artificial perilymph perfusion and 2, 5 and 10 min after the beginning of a second perfusion of artificial perilymph containing 200 μM cyclothiazide. Application of 200 μM cyclothiazide transiently increased single unit sound-evoked activity within the 2 first min and subsequently abolished it within 5-10 min. As the responses were totally abolished, no PSTHs could be constructed after 20 min perfusion.

Figure 10. Micrographs from the cochleal basal turn of the animal presented in Fig. 9

a, micrograph from the non-treated contralateral cochlea. b-e, micrographs from the cochlea which received cyclothiazide in the presence of moderate sound stimulation. b, all spiral ganglion dendrites below the inner hair cells (IHC) are disrupted (stars). c, an enlargement of the frame presented in b. The presynaptic body is facing an empty space. Note the presence of some remaining postsynaptic membrane. d and e, micrographs from the third turn. Note that the dendrites are not disrupted, but a clear mitochondrial swelling (arrows) is observed within the spiral ganglion dendrites and in the lateral efferents (e). Scale bars = 5 μm (a); 1 μm (b); 0.5 μm (c); 1 μm (d); 1 μm (e).