Gating of hair cell Ca2+ channels governs the activity of cochlear neurons

Abstract

Our sense of hearing processes sound intensities spanning six orders of magnitude. In the ear, the receptor potential of presynaptic inner hair cells (IHCs) covers the entire intensity range, while postsynaptic spiral ganglion neurons (SGNs) tile the range with their firing rate codes. IHCs vary the voltage dependence of Ca²⁺ channel activation among their active zones (AZs), potentially diversifying SGN firing. Here, we tested this hypothesis in mice modeling the human Ca_V1.3^A749G mutation that causes low-voltage Ca²⁺ channel activation. We demonstrate activation of Ca²⁺ influx and glutamate release of IHC AZs at lower voltages, increased spontaneous firing in SGNs, and lower sound threshold of Ca_V1.3^A749G/A749G mice. Loss of synaptic ribbons in IHCs at ambient sound levels of mouse husbandry indicates that low-voltage Ca²⁺ channel activation poses a risk for noise-induced synaptic damage. We propose that the heterogeneous voltage dependence of Ca_V1.3 activation among presynaptic IHC AZs contributes to the diversity of firing among the postsynaptic SGNs.

Fig. 1.. Shift to lower voltages and altered voltage sensitivity of Ca_V1.3 activation in IHCs of Ca_V1.3^AG/WT and Ca_V1.3^AG/AG mice.

(A) Representative Ca²⁺ current traces from Ca_V1.3^WT/WT (bottom left) and Ca_V1.3^AG/AG (bottom right) IHCs evoked by step depolarizations (top). (B) Whole-cell Ca²⁺ current-voltage relationships (I-V curves) show comparable maximal Ca²⁺ current amplitude in Ca_V1.3^WT/WT, Ca_V1.3^AG/WT, and Ca_V1.3^AG/AG IHCs. Error bars show ± SEM. (C) Ca²⁺ channel activation–voltage relationships calculated from I-V curves show a hyperpolarized shift in Ca_V1.3^AG/WT and Ca_V1.3^AG/AG IHCs. Error bars show ± SEM. (Ca) The voltage of half-maximal activation (V_half) is hyperpolarized in Ca_V1.3^AG/WT and Ca_V1.3^AG/AG IHCs. (Cb) The voltage sensitivity (k) is decreased in Ca_V1.3^AG/WT and increased in Ca_V1.3^AG/AG IHCs compared to the controls. (D) Mean exocytic change in membrane capacitance (ΔC_m) and Ca²⁺ current integrals (Q_Ca) evoked by 100-ms pulses of different depolarizations. (E) Mean exocytic ΔC_m in response to different depolarization durations. Data in (B) to (D) are presented as mean ± SEM. Box-whisker plots with individual data points overlaid show the median, 25th, and 75th percentiles (box) and 10th and 90th percentiles (whiskers). Statistical significances were determined using one-way analysis of variance (ANOVA), followed by Tukey’s post hoc test for (Ca) and (Cb). Significances are reported as **P < 0.01 and ***P < 0.001.

Fig. 2.. Reduced amplitude, hyperpolarized activation, and altered voltage sensitivity of synaptic Ca²⁺ influx in Ca_V1.3^AG/AG IHCs.

(A) Single confocal plane of a representative Ca_V1.3^WT/WT IHC showing TAMRA-conjugated dimeric Ribeye/Ctbp2 peptide and Fluo4-FF fluorescence. Scale bar, 2 μm. Black and gray colors of voltage ramp stimuli and corresponding whole-cell Ca²⁺ currents show IHC response to two voltage ramp depolarizations. Intensity-time profiles of single Ca²⁺ hotspots from one IHC are shown with different colors. bp, binding peptide; a.u., arbitrary units. (B) Average fluorescence-voltage (F-V) relationships of Ca²⁺ influx at single AZs of Ca_V1.3^WT/WT and Ca_V1.3^AG/AG IHCs. Shaded areas show ± SEM. (Ba) The maximal Ca²⁺ influx amplitude (ΔF/F_{0 max}) at single AZs is reduced in IHCs of Ca_V1.3^AG/AG mice. (C) Fractional activation curves of Ca²⁺ channels at individual AZs. Thick, dark lines show the averages, and lighter colors represent individual curves. (Ca and Cb) Box plots showing k (Ca) and V_half (Cb) of Ca²⁺ channels calculated from the Boltzmann fits in (C). (D) The spatial gradient of maximal Ca²⁺ influx is collapsed in Ca_V1.3^AG/AG IHCs. n.s., not significant. (E) The spatial gradient of V_half is maintained in Ca_V1.3^AG/AG IHCs. Polar plots in (D) and (E) show the positions of individual AZs in Ca_V1.3^WT/WT (left) and Ca_V1.3^AG/AG (right) IHCs. Pseudocolor scales represent maximal Ca²⁺ influx amplitude (D) and V_half (E). Box plots compare ΔF/F_{0 max} (D) and V_half (E) at the pillar and modiolar AZs in IHCs of Ca_V1.3^WT/WT and Ca_V1.3^AG/AG mice. Data were acquired from N = 8 (Ca_V1.3^WT/WT) and 6 (Ca_V1.3^AG/AG) mice. Box-whisker plots with individual data points overlaid show the median, 25th, and 75th percentiles (box) and 10th and 90th percentiles (whiskers). Statistical significances were determined using two-tailed Wilcoxon rank sum test for data in (Ba), (Ca), (Cb), (D), and (E). Significances are reported as *P < 0.05, **P < 0.01, and ***P < 0.001.

Fig. 3.. Activation of glutamate release at IHC synapses occurs at lower voltages in Ca_V1.3^AG mice.

(A) Average projection of ΔF images of Rhod-FF fluorescence from multiple planes of a representative Ca_V1.3^WT/WT IHC (top). Average projection of ΔF images of iGluSnFR fluorescence from 50-ms depolarizations at a single IHC plane (middle). Scale bar, 5 μm. (B) Average ΔF/F₀ traces of iGluSnFR fluorescence. Individual traces are in lighter colors. (C) Box plot showing ΔF/F_{0 max} of iGluSnFR signal. (D) Voltage dependence of normalized iGluSnFR signal. Data from individual boutons are shown for Ca_V1.3^WT/WT and Ca_V1.3^AG/AG animals. (E to G) Box plots showing V_half (E), V₁₀ (F), and dynamic range (G) of glutamate release. (H) Voltage dependence of normalized Ca²⁺ influx at single AZs. Individual traces are shown with lighter colors. Dotted lines show the average modified Boltzmann function fit. Thick lines show fits matching the voltage dependency range of iGluSnFR recordings. (I to K) Box plots showing ΔF/F_{0 max} (I), V_half (J), and dynamic range (K) of Ca²⁺ influx. (L and M) Synaptic transfer function at single synapses in Ca_V1.3^WT/WT (L) and Ca_V1.3^AG/AG (M) IHCs. A thick, solid line shows the average. Dotted line represents the power function fitted to the first 25% of the glutamate release. (N) Box plot showing power (m) of Ca²⁺ influx–glutamate release. Data were acquired from N = 12 (Ca_V1.3^WT/WT), 5 (Ca_V1.3^AG/WT), and 9 (Ca_V1.3^AG/AG) mice. Box-whisker plots with individual data points overlaid show the median, 25th, and 75th percentiles (box) and 10th and 90th percentiles (whiskers). Statistical significances were determined using the Kruskal-Wallis test for (C), one-way ANOVA followed by Tukey’s post hoc test for (E) and (F), Kruskal-Wallis followed by Dunn’s test for (G), two-tailed Wilcoxon rank sum test for (I) and (N), and two-tailed t test for (J) and (K). Significances are reported as *P < 0.05, **P < 0.01, and ***P < 0.001.

Fig. 4.. Increased spontaneous rates in SGNs of Ca_V1.3^AG/WT and Ca_V1.3^AG/AG mice.

(A) Average ABR waveforms in response to 80-dB clicks recorded in mice under urethane/xylazine anesthesia. Shaded areas show ± SEM. (B) ABR thresholds in response to click stimuli are lower in Ca_V1.3^AG/AG mice compared to Ca_V1.3^WT/WT littermates. SPL, sound pressure level. (C) ABR P₁-N₁ amplitude was significantly bigger in Ca_V1.3^AG/AG at near-threshold stimulation [40 dB (peak equivalent)] but similar to Ca_V1.3^WT/WT across other sound pressure levels. (D) SRs of SGNs recorded from mice under isoflurane anesthesia show a relative increase in SRs in Ca_V1.3^AG/WT and Ca_V1.3^AG/AG mice compared to Ca_V1.3^WT/WT. (E) Frequency tuning curves of all recorded putative SGNs with the characteristic frequency (CF)/best threshold marked by stars. (F and G) Thresholds at CF between 10 and 20 kHz (G) are comparable in Ca_V1.3^WT/WT, Ca_V1.3^AG/WT, and Ca_V1.3^AG/AG mice. (H) Average peristimulus time histogram (PSTH) in response to 50-ms stimulation at the CF, 30 dB above the threshold level, and stimulation rate of 5 Hz. Shaded areas show ± SEM. (I to K) Onset firing rates (calculated from PSTH as the bin with the highest rate at the sound onset) are not changed (J), but the adapted firing rates (averaged firing rates at 35 to 40 ms after the sound onset) are decreased in SGNs of Ca_V1.3^AG/AG mice (K). Single-unit recordings were obtained from N = 6 (Ca_V1.3^WT/WT), 3 (Ca_V1.3^AG/WT), and 6 (Ca_V1.3^AG/AG) mice. Box-whisker plots with individual data points overlaid show the median, 25th, and 75th percentiles (box) and the range (whiskers). Statistical significances were determined using two-tailed Wilcoxon rank sum test for (B), two-tailed Wilcoxon rank sum test for each sound level for (C), Kruskal-Wallis test followed by Tukey-Kramer multiple comparison test for (D) and (K), and Kruskal-Wallis test for (G) and (J). Significances are reported as *P < 0.05 and **P < 0.01.

Fig. 5.. Loss of synaptic ribbons at a subset of IHC AZs in Ca_V1.3^AG/AG mice.

(A) Schematic illustration of SBEM imaging at apical, middle, and basal segments of the mouse cochlea. (B) Three-dimensional (3D) rendering of reconstructed Ca_V1.3^WT/WT (left) and Ca_V1.3^AG/AG (right) IHCs contacted by SGN terminals with (green) and without (gray, “ribbonless”) associated ribbons (magenta). Scale bars, 5 μm. (C) Representative electron micrographs of ribbon-associated (green; left) and ribbonless (gray; right) SGN terminals. Scale bars, 1 μm. (D) Average numbers of SGN terminals contacting an IHC are comparable between Ca_V1.3^AG/AG and Ca_V1.3^WT/WT mice. (E) Average ribbon counts are significantly smaller in Ca_V1.3^AG/AG IHCs compared to Ca_V1.3^WT/WT at mid- and basal cochlear segments but not in the apex. (F) Mean volumes of ribbons are larger in Ca_V1.3^AG/AG IHCs compared to Ca_V1.3^WT/WT IHCs at mid- and basal cochlear segments, whereas apical Ca_V1.3^AG/AG IHCs appear to have exclusively small ribbons. Each tonotopic location of each genotype represents data from N = 2 mice. Box-whisker plots with individual data points overlaid show median, 25th, and 75th percentiles (box) and the range (whiskers). Statistical significances were determined using two-tailed t test for (D) to (F). Significances are reported as **P < 0.01, ***P < 0.001, and ****P < 0.0001.

Fig. 6.. Ca_V1.3 gating contributes to SGN firing diversity and synaptic vulnerability.

(A) SGN spontaneous rates and thresholds vary according to the voltage dependence of presynaptic Ca²⁺ channel activation and glutamate release in a gene dose–dependent manner. SGN rate-level functions were modeled according to Meddis et al. (58) (see Supplementary Text, fig. S12, and table S1). Top panels summarize the path from sound to SGN code. (B) Hyperpolarized activation of Ca_V1.3 channels in Ca_V1.3^AG/AG mice leads to Ca²⁺ influx at IHC AZs, exceeding that for both pillar and modiolar AZs in WT IHCs at physiological voltages. This results in a homeostatic reduction of Ca²⁺ channels and a general AZ remodeling but does not reach excitotoxic levels of Ca²⁺ influx and glutamate release at the ambient sound levels of the animal facility. Excitotoxic damage is likely to occur at large receptor potentials during intense sound stimulation and may preferentially affect modiolar synapses that harbor a greater Ca²⁺ channel complement in WT IHCs at nonphysiological stimulations, such as moderate noise exposure (shown with the dashed ellipsoid).