Strong inactivation of IHC CaV1.3 channels upon deletion of Cabp1 and Cabp2.

(A) Schematic of the IHC and one of its several ribbon synapses. CaBPs and CaM modulate voltage-gated calcium channels at the IHC ribbon synapses to shape the presynaptic Ca2+ signal. (B) Peak normalized Ca2+-currents after a 500-ms depolarization step to the maximum current potential in the apical IHCs of WT and Cabp1/2-DKO animals. Note a pronounced inactivation of the Ca2+ current in the absence of CaBP1 and 2 and partial recovery upon Cabp2 gene replacement. (C) The fraction of the remaining Ca2+-current after a 500-ms depolarization step is depicted for different mouse models acquired at comparable conditions (room temperature, 2 mM extracellular [Ca2+]). The histogram combines the results of this study (Cabp1/2-DKO) with previously published data on CaBP1 12 and CaBP2 13. The WT data combines pooled data from the current study and controls as obtained previously 12. Note that the recordings from Cabp2-KO animals were acquired at the lower apex to mid-cochlear tonotopic positions as the phenotype in the apex (low-frequency positions investigated otherwise) is mild 13. (D) Ca2+-currents upon a train of 10-ms long depolarization steps normalized to the amplitude of the first pulse. (E) Ca2+-current-voltage relationships show slightly reduced amplitudes in the IHCs of the Cabp1/2-DKO animals transduced with the PHP.eB-Cabp2. (F) Activation curves were calculated with Boltzmann fits. Note a significant difference in the half-voltage activation (V0.5) and slope in the CaBP1/2-deficient IHCs and the persistence of the voltage shift upon re-expression of CaBP2 (inserts). See main text for the p values. Data in B and D-F was acquired in 1.3 mM, data in C in 2 mM [Ca]2+.

Voltage- and Ca2+-dependent inactivation are enhanced after deletion of Cabp1 and Cabp2.

(A-B) Peak normalized Ca2+- and Ba2+ currents in WT and CaBP1/2-deficient IHCs recorded in the extracellular solution containing 2 mM of the respective divalent cation. The two types of inactivation were calculated as depicted. (C) VDI and CDI were both significantly enhanced in the IHCs of Cabp1/2-DKO animals (asterisks; Student’s t test, p < 0.0001).

Reduced sustained component of exocytosis and calcium charge transfer (QCa) can be partially rescued by Cabp2-transgene-expression.

(A) Capacitance increments with the corresponding QCa were probed by different depolarization durations from a holding potential of - 85 mV (unpaired t tests with Welch correction). CaBP1/2-deficient IHCs showed significant impairment of the sustained exocytosis (p < 0.05 for 20-ms, and < 0.001 for 100- and 200-ms step, respectively) and reduced cumulative Ca2+ influx (p < 0.005 for the longest two pulses) as compared to WT controls (black asterisks). AAV-mediated delivery of Cabp2 improved IHC synaptic function as compared to non-injected Cabp1/2-DKO controls (p < 0.01 for ΔCm upon 100- and 200-ms step, and p < 0.05 for QCa upon 200-ms step; green asterisks). (B) Representative Ca2+-current traces and corresponding membrane capacitance changes upon 100-ms long depolarization steps to the peak Ca2+-current potential. (C) Note a reduced efficiency of Ca2+-dependent exocytosis (p < 0.005 for the two longest test pulses; black asterisks), which can be efficiently rescued by intracochlear delivery of PHP.eB-Cabp2 (p < 0.01 and 0.05 for 100- and 200-ms pulse, respectively; green asterisks). (D) Capacitance measurements at a holding potential of -55 mV reveal an aggravation of the phenotype by additional activation of calcium channels between the test pulses (p < 0.01 for 50-ms, and p < 0.005 for 20-, 100- and 200-ms pulse) and a further discrepancy between the QCa of WT and CaBP1/2-deficient IHCs (2-10 ms: p < 0.05; 20-200 ms: p < 0.001). Also in these recording conditions, the IHCs from Cabp2-injected Cabp1/2-DKO animals showed increased QCa (20-200 ms: p < 0.05) and exocytosis (100-200 ms: p < 0.005) as compared to non-injected controls. (E) Capacitance increments and the corresponding QCa upon 100-ms depolarization steps as recorded in different conditions. Note worsening of the phenotype with increasing IHC activation between the test pulses.

The speed of recovery of Ca2+ currents is affected in CaBP1/2-deficient IHCs.

(A) Ca2+ currents were probed by 5-ms long depolarization steps to the maximum current potential. An initial test pulse was followed by a conditioning stimulus (a 500-ms depolarization step or a 120-s long sine wave). Subsequently, Ca2+ currents were tested at various time points to assess the recovery of Ca2+-current amplitudes. (A’) Example Ca2+-current recording using a 500-ms long depolarization step as a conditioning stimulus. (A’’) Representative Ca2+-current traces in WT and CaBP1/2-deficient IHCs at different time points after a 500-ms long conditioning stimulus. (B-C) Ca2+-current amplitudes normalized to the amplitude of the pre-conditioning test pulse up to 150 s after a 500-ms depolarization step (B) or a 2-min long sine wave stimulus (C). The inserts show the initial 20 s after the conditioning stimulus. The data were fitted by a single or double exponential function (fit parameters in the Table 1). Note stronger inactivation in the Cabp1/2-DKO IHCs and delayed recovery of Ca2+ currents after the presentation of a long sine wave and a partial recovery upon viral gene transfer of Cabp2. Recovery upon a 500-ms pulse compared to the 2-min long sine wave stimulus was much faster. Surprisingly, in these conditions recovery of calcium currents was slowest in hair cells with virus-mediated re-expression of CaBP2.

Ca2+-current recovery fit parameters.

The data was fitted with the single or the double exponential with the following equations:

y(t) = y0 - A1 × exp(-(t/τ1)) and y(t) = y0 - A1 × exp(-(t/τ1)) - A2 × exp(-(t/τ2)), where y(t) represents the peak current amplitude at the time t after the end of conditioning pulse (either 500-ms depolarization or 2-min sine wave) given in % of the initial pre-conditioning peak current amplitude (100 × Ipost/Ipre). While current recovery upon 500-ms depolarization in the IHCs of WT and DKO animals could be well fitted by a single exponential, the data obtained in the injected DKO animals required a double exponential fitting. Upon 2-min sine wave, the current recovery in DKO IHCs (both, injected and non-injected group) also required fitting with a double exponential function.

Comparable synapse density with changed morphology in Cabp1/2-DKO IHCs.

(A) CaV1.3-immunofluorescence appeared normal in Cabp1/2-DKO-mice. Scale bars: 5 µm. (B) Ribbon synapse immunostainings using RibeyeA as a presynaptic and Homer1 as a postsynaptic marker. Ribbon synapses were identified as juxtaposed pre- (RibeyeA/CtBP2) and postsynaptic (Homer1/GluA3) immunofluorescent spots. (C) SK2-immunofluorescence was absent from the majority of IHCs in 3-week-old animals, but prominent in immature cochleae of 1-week-old animals and together with synapsin 1 labeled direct synaptic contacts from the efferent fibers to the IHCs. Occasionally, however rarely, isolated SK2 immunofluorescent spots were still detected in close proximity to the synapsin marker in the IHCs of 3-week-old DKO animals (arrows). (D) Expression of SK2 beyond the second postnatal week was also observed in some WT IHCs (arrows). (E-F) BK clusters at 3 weeks of age appeared smaller in the non-treated animals, and less so in the ears of PHP.eB-Cabp2 injected Cabp1/2-DKO animals. This difference became even stronger at 14 weeks of age. Scale bars are 10 µm if not stated otherwise. (G) A significantly higher immunofluorescence intensity was observed for pre- (RibeyeA or CtBP2) and postsynaptic (Homer1) markers in Cabp1/2-DKO mice (analysis of 614 and 441 immunospots from WT and Cabp1/2-DKO organs of Corti, respectively; N=5 for both genotypes; Wilcoxon rank-sum test, p < 0.005). (H) The volume of BK-channel immunofluorescent spots around the neck of the IHCs was significantly smaller in DKO IHCs (Student’s t-test p < 0.0001, N=6 organs per genotype).

AAV-mediated transgene-expression of Cabp2 ameliorates hearing loss and recovers ABR amplitude responses in Cabp1/2-DKO mice.

(A) Significant improvement of ABR thresholds in the injected ears of 3-4-week-old Cabp1/2-DKO animals for all tone burst frequencies tested as well as 20-Hz-click stimulus (see Table S1 for the p values). (B) Average click-evoked responses to an 80-dB 20-Hz-click stimulus. (C) A strong recovery of the amplitude of the ABR wave I (Tukey’s multicomparison test; p ≤ 0.0001 for DKO vs WT or vs injected DKO) upon Cabp2-transgene expression.

Impaired sound encoding in Cabp1/2-DKO mice.

Single neuron responses to sound stimulation in vivo in the region of the auditory nerve of Cabp1/2-DKO mice were extremely scarce. (A) WT poststimulus time histograms of SGNs to 50-ms suprathreshold tone or noise burst stimuli presented at a rate of 5 Hz showed a very high sound onset rate gradually adapting to a sustained rate persisting throughout the stimulus duration. Spike rates were much lower in the mutants. (B) Applying a lower stimulus rate of 0.5 Hz partially restored sound onset responses in Cabp1/2-DKOs, highlighting the enhanced strength of adaptation. WT and mutant neurons sustained their adapted spiking response throughout the longer 500 ms stimulus used in these recordings. For direct comparison, the average 5-Hz poststimulus time histograms from the same neurons are shown. (C) Correlations of peak and adapted rates for different stimulus rates. (D) Peak rates for different stimulus rates / interstimulus intervals in WT SGNs strongly increase between 10 and 5 Hz, while in DKOs a longer period of silence is required to observe such difference. (E) Cabp1/2-DKO neurons had lower spontaneous spike rates than WT SGNs.