At the first auditory synapse in mammals, one ribbon-type AZ of the IHC drives one postsynaptic spiral ganglion neuron (SGN) to spike at rates exceeding 100 Hz in silence and 1 kHz upon sound onset 1, 2 . Moreover, SGNs sustain firing rates of several hundred Hz during ongoing acoustic stimulation. In such a steady state, vesicle replenishment has to balance vesicle fusion at the IHC AZ. Accordingly, high rates of initial and sustained exocytosis have been found in hair cells [3][4][5][6][7][8] . Ribbon-type AZs of IHCs replenish readily releasable vesicles at hundreds of Hz over several seconds of stimulation, faster than ribbon synapses in the eye 9-15 and most non-ribbon-type AZs
16(but see ref. 17). This efficient vesicle re-supply maintains a large standing pool of fusion competent synaptic vesicles, which appears to be critical for reliable and temporally precise sound encoding [18][19][20] . and Otof +/+ mice ( Fig. 2a-b The observations of normal RRP size after resting the synapse for more than 30 seconds and of reduced vesicle re-supply during ongoing stimulation prompted us to explore RRP recovery from depletion in paired-pulse experiments (Fig. 3a). RRP recovery, assessed as the paired-pulse ratio for different inter-pulse intervals, was impaired in Otof Pga/Pga mice (Fig. 3b). This indicated a deficit in vesicle replenishment also in the rest period between stimuli. The ΔC m pattern elicited by trains of ten short (10 ms) depolarizations demonstrates the exocytosis phenotype found after a period of rest (30 s voltage-clamp at -84 mV): normal RRP exocytosis but subsequent failure (Fig. 3c). Studying the C m decline after exocytosis we observed normal endocytic membrane retrieval (Fig. 3d).
Normal synaptic ultrastructure in Otof Pga/Pga IHCsIn order to explore whether a docking or a priming defect underlies the impairment of vesicle replenishment at Otof Pga/Pga IHC synapses, we studied their ultrastructure using electron microscopy (EM). Both, EM of single ultrathin sections (perpendicular to the 7 plasma membrane and the long axis of the ribbon; Supplementary Fig. 2) as well as EM tomography (Fig. 3e-f High-resolution EM tomography ( Fig. 3e-f) was used to measure the distance of membrane-proximal synaptic vesicles (labeled orange in Fig. 3f) from the plasma membrane under both conditions. The average membrane-membrane distance was approximately 6 nm regardless of condition and genotype (Supplementary
Reduced rates but maintained size variability of EPSCsThe stark contrast between the absence of auditory neuron population responses in vivo (Fig. 1) (Fig. 4a-b). We pooled the data from recordings in 5. Fig. 4h-i; Kolmogorov-Smirnov test, p = 0.14). Moreover, we detected action potential generation by recording action currents in the loose-patch configuration (Fig. 4g). Together these results suggest that Otof Pga/Pga synapses should be capable of encoding sound into spiking activity in auditory nerve fibers, albeit at lower rates.In addition, we recorded from SGNs of Otof -/-mice of the same age a...