(A–C) Representative electron micrographs of IHC afferent synapses from P21 RBEWT/WT, RBEWT/KO and RBEKO/KO mice. Ribbonless RBEKO/KO synapses display one or more presynaptic densities (PD) clustering SVs. Scale bars = 200 nm. (D) RBEKO/KO IHC AZ cut tangentially, revealing multiple PDs (here six) per AZ. (E, F) Representative electron micrographs of RBEKO/KO IHC synapses from mice at 6 weeks (E) and 8 months (F) of age: the presence of at least 2 AZs per contact in mature IHCs argues against a developmental delay. (G) Consecutive serial sections of a typical RBEKO/KO P21 IHC synapse showing multiple AZs. Scale bar = 200 nm. (G’) Corresponding serial 3D reconstruction of the synapse in (G) showing four PDs (magenta) surrounded by a total of 48 SVs (gray). Scale bar = 200 nm. (H) Quantification of the number of ribbon/PD per random section. AZs with a single PD are less frequently observed in RBEKO/KO IHCs (n = 108 AZs, N = 2 for RBEKO/KO and n = 106 AZs, N = 2 for RBEWT/WT; p<0.05, NPMC test). (I) Number of PDs and SVs per AZ in P21 RBE KO/KO mice in serial 3D reconstructions of RBEKO/KO afferent synapses. Box plots show 10, 25, 50, 75 and 90th percentiles with individual data points overlaid, as for (M, N, O, P & Q). (J) Schematic drawing illustrating the quantitative analysis of random sections. SV diameter: average of vertical and horizontal measurements from outer rim to outer rim. The ribbon height, width and area were measured as indicated by the gray lines. The length of the PD was determined along the AZ. For ribbon-occupied AZs: Membrane-proximal (MP) SVs (yellow) were counted in a distance of ≤25 nm from the AZ membrane (blue) and ≤80 nm from the PD. Ribbon-associated (RA) SVs were found in the first layer around the ribbon (red) with a maximum distance of 80 nm to the ribbon, quantified as indicated by the gray lines. For ribbonless AZs: Instead of RA-SVs we defined PD-associated SVs (PDA-SVs: all SV at PD with a maximum distance of 80 nm to the PD not matching the MP-SV criteria, defined as above). (K, L) Electron micrographs illustrating the quantification of the MP-SVs (yellow crosses) and the RA/PDA-SVs (green crosses). (M–Q) Quantification of random IHC synapse (P21) sections revealed no significant differences between RBE KO/KO and RBEWT/WT for the PD and PSD length (PD length: n = 108 PDs, N = 2 for RBE KO/KO and n = 108 PDs, N = 2 for RBEWT/WT; p=0.92, NPMC test and PSD length: n = 98 PSDs, N = 2 for RBE KO/KO and n = 113 PSDs, N = 2 for RBEWT/WT; p=0.11, Tukey’s test). However, in the RBEWT/KO IHCs, the PDs were bigger than in the WT IHCs (n = 101 PDs, N = 2; p=0.0004, NPMC test), and the PSDs were smaller than in the knock-out IHCs (n = 100 PSDs, N = 2 for RBEWT/KO; p=0.01, Tukey’s test). MP-SVs (n = 108 AZs, N = 2 for RBEKO/KO, n = 106 AZs, N = 2 for RBEWT/WT; p<0.0001, NPMC test) and RA/PDA-SVs (n = 108 AZs, N = 2 for RBEKO/KO, n = 106 AZs, N = 2 for RBEWT/WT; p<0.0001, Tukey’s test) per AZ, as well as the fraction of RA/PDA-SVs in RBEKO/KO, were significantly reduced (n = 108 AZs, N = 2 for RBEKO/KO, n = 106 AZs, N = 2 for RBEWT/WT; p<0.0001, NPMC test).