Abstract
Neural diversity can expand the encoding capacity of a circuitry. A striking example of diverse structure and function is presented by the afferent synapses between inner hair cells (IHCs) and spiral ganglion neurons (SGNs) in the cochlea. Presynaptic active zones at the pillar IHC side activate at lower IHC potentials than those of the modiolar side that have more presynaptic Ca2+- channels. The postsynaptic SGNs differ in their spontaneous firing rates, sound thresholds and operating ranges. While a causal relationship between synaptic heterogeneity and neural response diversity seems likely, experimental evidence linking synaptic and SGN physiology has remained difficult to obtain. Here, we aimed at bridging this gap by ex vivo paired recordings of IHCs and postsynaptic SGN boutons with stimuli and conditions aimed to mimic those of in vivo SGN-characterization. Synapses with high spontaneous rate of release (SR) were found predominantly on the pillar side of the IHC. These high SR synapses had larger and more compact spontaneous EPSCs, lower voltage-thresholds, tighter coupling of Ca2+ channels and vesicular release sites, shorter response latencies and higher initial release rates. This study indicates that synaptic heterogeneity in IHCs directly contributes to the diversity of spontaneous and sound-evoked firing of SGNs.
Significance Statement
Sound encoding relies on spiral ganglion neurons (SGNs) with diverse spontaneous firing, sound thresholds of firing and sound-intensity range over which SGN firing rate changes. Such functional SGN diversity might originate from different input from afferent synapses with inner hair cells (IHCs). The present study addresses this hypothesis by using recordings from individual IHC-SGN synapses of hearing mice under ex vivo conditions aimed to mimic cochlear physiology. The results provide evidence that synaptic heterogeneity in IHCs contributes to SGN firing diversity. Thus, the cochlea employs heterogeneous synapses to decompose sound information into different neural pathways that collectively inform the brain about sound intensity.
Introduction
Chemical synapses represent diverse and plastic neural contacts that are adapted to the specific needs of neural computation. Synaptic diversity is expressed across the nervous system, within a given circuit and even within the same neuron (recent reviews in Ref. Nusser, 2018; Wichmann and Kuner, 2022). Synaptic diversity occurs at various levels: from synapse shape and size, to the ultrastructure of pre-and postsynaptic specializations, to their molecular composition. The auditory system harbors striking examples of synaptic diversity. Glutamatergic ribbon synapses in the cochlea, calyceal synapses in the brainstem, and bouton synapses throughout the central auditory system differ greatly from each other (Moser et al., 2006; Wichmann and Kuner, 2022). Beyond the diversity across synapses formed by different neurons, heterogeneity exists even among auditory synapses of an individual presynaptic inner hair cell (IHC) with its 5-30 postsynaptic spiral ganglion neurons (SGNs, (Gómez-Casati and Goutman, 2021; reviews in: Meyer and Moser, 2010; Moser, 2020; Reijntjes and Pyott, 2016)). Such synaptic heterogeneity is a candidate mechanism for how the cochlea decomposes acoustic information (Grant et al., 2010; Meyer et al., 2009; Ohn et al., 2016; Özçete and Moser, 2021). For example, the cochlea might use heterogeneous afferent synapses to break down sound intensity information into complementary spike rate codes of SGNs that have been reported across species (Huet et al., 2016; Kiang et al., 1965; Sachs and Abbas, 1974; Taberner and Liberman, 2005; Winter et al., 1990).
Decades of in vivo recordings from single SGNs have demonstrated functional diversity of SGNs with comparable frequency tuning, i.e., receiving input from IHCs at a given tonotopic place or potentially even the same IHC. Such functional diversity is present in both spontaneous and sound-evoked firing. The spontaneous firing rate (SR) in the absence of sound varies from less than 1 spikes/s to more than 100 spikes/s (Barbary, 1991; Evans, 1972; Kiang et al., 1965; Schmiedt, 1989; Taberner and Liberman, 2005). In response to increasing sound pressure, SGNs with high SR show a low sound threshold and a steep rise in the spike rate to increasing sound intensities until the rate saturates. SGNs with low SR have a higher sound thresholds, shallower spike rate rise and saturate at higher sound intensities (Ohlemiller et al., 1991; Winter et al., 1990). Additionally, SGNs show differences in their excitability (Crozier and Davis, 2014; Markowitz and Kalluri, 2020; Smith et al., 2015), morphological features (Liberman, 1980; Merchan-Perez and Liberman, 1996; Tsuji and Liberman, 1997) and heterogeneous molecular profiles (Li et al., 2020; Petitpré et al., 2020, 2018; Shrestha et al., 2018; Sun et al., 2018). Yet, it has remained challenging to demonstrate a causal link of a candidate mechanism to the physiological SGN diversity.
One common approach has been to capitalize on a pioneering study that employed in vivo labelling of physiologically characterized SGNs in cats and proposed that synapses formed by low and high SR SGNs segregate on the basal IHC pole (Liberman, 1982). High SR SGNs preferentially contacted the pillar side of the IHC (facing pillar cells), while low SR SGNs synapsed on the opposite, modiolar side of the IHC (facing the cochlear modiolus). Interestingly, a segregation has also been found for afferent and efferent synaptic properties, as well as molecular and biophysical SGN properties (Frank et al., 2009; Grant et al., 2010; Hua et al., 2021; Kantardzhieva et al., 2013; Liberman et al., 2011; Markowitz and Kalluri, 2020; Merchan-Perez and Liberman, 1996; Meyer et al., 2009; Neef et al., 2018; Ohn et al., 2016; Özçete and Moser, 2021; Shrestha et al., 2018; Sun et al., 2018; Yin et al., 2014). For instance, type Ib and Ic SGNs preferentially synapse on the modiolar side (Sherrill et al., 2019; Shrestha et al., 2018; Sun et al., 2018) and show low SR (Siebald et al., 2023). Pillar synapses are preferentially formed by type Ia SGNs (Shrestha et al., 2018; Siebald et al., 2023), have smaller IHC active zones (AZs) (Liberman et al., 2011; Ohn et al., 2016; Reijntjes et al., 2020), and activate at voltages as low as the IHC resting potential (Ohn et al., 2016; Özçete and Moser, 2021). The low voltage of activation of release at pillar synapses shown ex vivo could underly the high SR and low sound threshold of firing of SGNs found in vivo but direct demonstration of such a link has been missing.
Here, we aimed to bridge the gap between ex vivo presynaptic physiology and in vivo SGN neurophysiology. We performed paired IHC and SGN bouton recordings in acutely explanted organs of Corti from hearing mice with stimuli and conditions aimed to mimic in vivo SGN-characterization. This approach tightly controls IHC Ca2+ influx and records the postsynaptic SGN response to glutamate release at a single afferent synapse. Using the rate of spontaneous excitatory postsynaptic currents (rate of sEPSCs, SR) as a surrogate for SGN SR, we demonstrate that high SR synapses have larger and more compact sEPSCs as well as lower voltage thresholds, shorter latencies of evoked EPSCs, tighter Ca2+ channel coupling to vesicle release and higher initial release rates. 90% of these high SR synapses were located on the pillar side of the IHC. Our findings suggest that synaptic heterogeneity accounts for much of the SGN firing diversity.
Results
Simultaneous paired patch-clamp recordings from IHCs and one of the postsynaptic SGN boutons were performed on mice after the onset of hearing (postnatal days (p) 14-23). We performed perforated-patch whole-cell configuration from IHCs, held at their presumed physiological resting potential (−58 mV (Johnson, 2015)), and ruptured-patch whole-cell recordings from one of the postsynaptic SGN boutons (Fig. 1). Due to the technical difficulty of establishing the paired recording, typically only one bouton was recorded per IHC. Recordings were made at body temperature and in artificial perilymph-like solution (Wangemann and Schacht, 1996). To establish the paired recording, we approached boutons facing either the pillar or the modiolar side of the IHC in an effort to elucidate synaptic differences between both sides (Fig. 1A). We nickname the synapse location as “pillar" and “modiolar” based on the DIC-image, but note that efforts to stain and image the recorded boutons by fluorescence microscopy were not routinely successful. In addition, the recorded boutons were classified based on their spontaneous rate of synaptic transmission (SR, Low SR < 1 sEPSC/s vs High SR > 1 sEPSC/s according to (Taberner and Liberman, 2005)).
Spontaneous synaptic transmission
In order to recapitulate synaptic transmission in the absence of sound stimulation, we held the IHC at their presumed physiological resting potential (−58 mV (Johnson, 2015). Since CaV1.3 Ca2+- channels activate at low voltages (−65 to −45 mV (Koschak et al., 2001; Picher et al., 2017; Platzer et al., 2000; Xu and Lipscombe, 2001)), their open probability at the IHC resting potential is thought to be sufficient to trigger spontaneous release (Glowatzki and Fuchs, 2002; Özçete and Moser, 2021; Robertson and Paki, 2002). Under our experimental conditions, spontaneous, i.e., excitatory postsynaptic currents in the absence of IHC stimulation (sEPSCs) were observed in 23 of 33 pairs (Fig. 2). We used the rate of sEPSCs (SR) as a surrogate of SGN SR as sEPSCs trigger SGN firing with >90% probability (Rutherford et al., 2012). Regardless of all efforts to maintain the physiological integrity in the ex vivo experiments, we expect our estimated rates of sEPSC to underestimate the SGN SR for the same age group (Wong et al., 2013). For each paired recording, we quantified the SR during 5 or 10 s of a continuous recording, or during the segment before and after the step depolarization protocols (Supplementary Table 1). Amplitudes of sEPSCs typically varied from around −10 pA to −400 pA across all recorded IHC-SGN synapses (Fig. 2B). The amplitude histogram for all pairs was slightly skewed towards larger amplitudes (skewness of 1.06) with a coefficient of variation (CV) of 0.68. The charge distribution for all pairs displayed a prominent peak at 40 fC, with a skewness of 2.00 and a CV of 0.77. SR ranged from 0 to about 18 sEPSC/s (Fig. 2H). The SR distribution was highly skewed, with a median of 0.2 sEPSC/s. Following the spontaneous firing rate classification of SGNs by Taberner and Liberman (2005), we classified the synapses into low (< 1 sEPSC/s, ∼70%) or high (≥ 1 sEPSC/s, ∼30%) SR synapses. Next, we analyzed the recordings as a function of i) synapse position and ii) rate of spontaneous synaptic transmission.
i) Position dependence of synaptic transmission
Of the 33 obtained paired recordings, 20 were classified as pillar synapses. SR was significantly higher for the pillar synapses (mean SR of 2.30 ± 0.96 sEPSC/s; median 0.59; n = 20) compared to modiolar ones (mean SR of 0.22 ± 0.09 sEPSC/s; median 0.05; n = 13; p = 0.0311, Mann-Whitney U test). The mean amplitude (Fig. 2C) and charge (Fig. 2D) of the sEPSCs were comparable between modiolar and pillar SGN boutons (AmpsEPSC −69.37 ± 13.91 pA (n = 8 modiolar) vs −87.21 ± 9.49 pA (n = 13 pillar); p = 0.2913, unpaired t-test; QsEPSC 63.74 ± 7.77 fC (modiolar) vs 72.54 ± 9.57 fC (pillar); p = 0.5463, unpaired t-test). sEPSC of pillar SGN boutons showed significantly shorter 10-90% rise times than the modiolar ones (Figure 2–figure supplement 1, panel A and C; 0.38 ± 0.04 ms vs 0.57 ± 0.06 ms; p = 0.0111, unpaired t-test), yet similar decay times and full-width half-maximum (FWHM, Figure 2–figure supplement 1, panels A, D-E; p = 0.7997 and p = 0.9198, respectively, unpaired t-test). As a second approach to sEPSC properties, we quantified the percentage of monophasic (or compact) sEPSCs (Chapochnikov et al., 2014; Glowatzki and Fuchs, 2002) and found a non-significant trend towards higher percentages of monophasic sEPSCs for pillar synapses (Fig. 2G; 25.57 ± 6.9% for modiolar vs 46.65 ± 7.06% for pillar; p = 0.0681, unpaired t-test).
ii) Relation of synaptic properties to the rate of spontaneous synaptic transmission
High SR synapses had significantly larger sEPSCs (Fig. 2J and Figure 2–figure supplement 1, panel Bi; average sEPSC amplitude of −105.2 ± 8.47 pA for high SR (n = 10) vs. −62.39 ± 9.67 pA for low SR (n = 13); p = 0.0042, unpaired t-test). Likewise, sEPSC charge tended to be larger in high SR SGN synapses (Fig. 2K and Figure 2–figure supplement 1, panel Bii; QsEPSC 84.23 ± 10.38 fC for high SR vs. 58.14 ± 7.79 fC for low SR; p = 0.0527, unpaired t-test). The fraction of monophasic sEPSCs was significantly higher in high SR synapses (Fig. 2L; 52.08 ± 6.65%; median 43.71%) compared to low SR synapses (29.49 ± 7.43%; median 22.22%; p = 0.0185, Mann-Whitney U test). High SR synapses also showed a significantly faster 10-90% rise times (Figure 2– figure supplement 1, panel F; 0.36 ± 0.04 ms) than low SR synapses (0.51 ± 0.05 ms; p = 0.0420, unpaired t-test).
Other sEPSCs kinetics, such as decay time constant and FWHM, were not different between low and high SR pairs (Figure 2–figure supplement 1, panel G,H; p = 0.7969 and p = 0.9948, respectively, unpaired t-test). Taken together, these results indicate that high SR synapses are characterized by sEPSCs with larger amplitudes and charges, faster rising times and a more compact waveform. Interestingly, 9 out of 10 high SR synapses were located on the pillar side of the IHC.
Evoked synaptic transmission differs between afferent synapses with high and low SR
Next, we compared the physiology of afferent synapses with high and low SR by adapting stimulation protocols routinely employed for in vivo characterization of sound encoding by SGNs. We used step depolarizations to emulate physiological receptor potentials given that mature IHCs of the “high frequency” mouse cochlea have graded receptor potentials that primarily represent the rectified envelope of an acoustic stimulus (i.e. the DC component (Russell and Sellick, 1978).
i) Stimulus intensity encoding at IHC synapses
Sound intensity encoding by SGNs primarily relies on a spike rate code: the average discharge rate increases with the strength of the acoustic stimuli from threshold to saturation of the response. These so-called “rate level functions” are typically analyzed by fitting a sigmoidal function, of which the range of sound pressure level between 10 and 90% of the maximal discharge rate represents the operational or dynamic range (Sachs and Abbas, 1974; Taberner and Liberman, 2005; Winter et al., 1990). To understand stimulus intensity coding at mouse IHC synapses, we measured whole-cell IHC Ca2+ currents and the evoked EPSCs (eEPSCs) of SGNs in 31 paired recordings. We stimulated the IHC with 10 ms depolarizations to different potentials ranging from −70 to 0 mV in 5 mV steps (IV protocol; Fig. 3A). We deemed it incompatible with a reasonable productivity of the technically challenging, low-throughput paired recordings to combine them with imaging of Ca2+ at single AZs. Therefore, this study relies on analysis of the presynaptic Ca2+ influx at the level of the whole IHC (i.e. summing over all synapses and a low density of extrasynaptic Ca2+ channels, (Frank et al., 2009; Wong et al., 2014)). IHCs with synapses classified as high (n = 10) or low SR (n = 21) had similar Ca2+ current-voltage (IV) relationships: comparable maximal Ca2+ currents (Fig. 3Bi; p = 0.6939, Mann-Whitney U test) elicited at similar potentials (Fig. 3Bii; p = 0.1795, unpaired t-test) and comparable reversal potentials (Figure 3–figure supplement 1, panel A; p = 0.4034, unpaired t-test). The fractional activation of Ca2+ channels was determined from the normalized chord conductance of the IHC. Fitting a Boltzmann function to these activation curves (Fig. 3C, upper panel), we obtained the voltages of half-maximal activation (Vhalf Ca) and the voltage-sensitivity of activation (slope) of the IHC Ca2+ channels.
“Release-stimulus intensity” curves, akin of an ex vivo representation of the SGN rate-level function, were constructed from the normalized QEPSC (Figure 3–figure supplement 1, panel C). The voltage dependence of synaptic vesicle release per active zone was approximated by the fit of a sigmoidal function to the individual release-intensity curves (Fig. 3C, lower panel). From these sigmoidal fits, we obtained voltage of 10%-maximal release (Q10, EPSC), voltage of half-maximal release (Q50, EPSC), voltage of 90%-maximal release (Q90, EPSC) and the voltage sensitivity of release (slope). For two low SR paired recordings (Fig. 3Dii, grey dotted lines), a sigmoidal function did not properly fit the release-intensity curve (assessed by visual inspection) which led us to exclude them from the statistical analysis.
The whole-cell voltage dependence of Ca2+ influx was similar between IHCs contacted by high and low SR boutons: threshold of Ca2+ influx (Fig. 3Ei; p = 0.2393, unpaired t-test), Vhalf Ca (Fig 3Fi; p = 0.3479, unpaired t-test) and voltage sensitivity of Ca2+ influx (Figure 3–figure supplement 1, panel B; p = 0.3470, unpaired t-test) did not differ significantly between IHCs contacted by low or high SR synapses. This seems to rule out a potential scenario in which the diverse SGN firing properties would be caused by varying average properties of Ca2+ channels among different presynaptic IHCs. Q50, EPSC (Fig. 3Fii) of high SR synapses (−47.76 ± 1.4 mV; n = 10) was 6.76 ± 2.0 mV more negative compared to low SR synapses (−41.00 ± 1.2 mV; n = 19; p = 0.0021, unpaired t-test). Accordingly, high SR synapses had lower thresholds of release than low SR synapses (Fig. 3Eii, Q10, EPSC of - 56.91 ± 1.5 mV (median −56.64 mV) vs. −47.39 ± 1.4 mV (median −48.89 mV); p = 0.0001, Mann-Whitney U test). The hyperpolarized shift was not significant for Q90 EPSC (Figure 3–figure supplement 1, panel D; p = 0.1706, unpaired t-test). The voltage sensitivity of release, determined by a slope factor, tended to be higher in high SR (4.16 ± 0.75 mV) than in low SR (2.9 ± 0.35 mV) synapses without reaching statistical significance (Fig. 3 – figure supplement 1, panel E; p = 0.0940, unpaired t-test;). The dynamic range, defined as the voltage range for which the exocytosis changes from 10-90% (Q90 EPSC – Q10 EPSC) tended to be larger for high SR synapses without reaching significance (Fig. 3G; 18.30 ± 3.3 mV for high SR synapses vs 12.78 ± 1.5 mV for low SR synapses; p = 0.0940, unpaired t-test). The voltage dependence of Ca2+ influx and of synaptic vesicle release did not differ significantly when the synapses where grouped based on their topographical location at the IHC (n = 12 modiolar synapses vs n = 17 pillar synapses; Figure 3– figure supplement 1, panels F-K). However, pillar SR synapses had a tendency to lower thresholds of release than modiolar synapses (Figure 3–figure supplement 1, panel Gii; Q10, EPSC of −52.78 ± 1.8 mV (median −53.73 mV) vs. −47.68 ± 1.8 mV (median −49.30 mV); p = 0.0725, Mann-Whitney U test). Altogether, these results demonstrate that high SR synapses have a more hyperpolarized voltage-dependence of release.
Finally, we studied the apparent Ca2+ dependence of SV release during the aforementioned IV protocol, i.e., in the range of IHC receptor potentials. This protocol varies Ca2+ influx mainly via changing the channel open probability and to a lesser extent by changing the single channel current. We note that a supralinear intrinsic Ca2+ dependence of exocytosis in IHCs (m ∼3-4) has been observed for IHCs of the cochlear apex in mice after hearing onset (Brandt, 2005; Özçete and Moser, 2021; Wong et al., 2014). We related changes of release at individual synapses (ΔQEPSC) to the change of the integrated IHC Ca2+ influx (ΔQCa). We fitted power functions (QEPSC = a(QCa)m) to the relationships for individual synapses (Figure 3–figure supplement 2) and found Ca2+ cooperativities of m < 2 for all but 2 synapses. This result suggests a tight, Ca2+ nanodomain-like control of release sites by one or few Ca2+ channel(s). Interestingly, however, high SR synapses, on average, had significantly lower Ca2+ cooperativities than low SR synapses (Fig. 3H; mhighSR of 0.8 ± 0.1 (median 0.75; n = 10) vs mlowSR of 1.4 ± 0.1 (median 1.37; n = 21); p = 0.0016, Mann-Whitney U test). The fit to pooled normalized data of high and low SR synapses yielded the same Ca2+ cooperativities of mhighSR of 0.8 and mlowSR of 1.4 (Fig. 3I). When grouped based on their modiolar or pillar location, pillar synapses showed significantly lower Ca2+ cooperativities than modiolar synapses (Figure 3–figure supplement 1, panel L; mpillar of 1.0 ± 0.08 (median 0.88; n = 19) vs mmodiolar of 1.6 ± 0.2 (median 1.3; n = 12); p = 0.0202, Mann-Whitney U test). Our findings indicate that most afferent IHC synapses of hearing mice employ a tight, Ca2+ nanodomain-like control of release sites by one or few Ca2+ channel(s) for physiological sound encoding. Yet, quantitative differences in coupling seem to exist between high SR/pillar synapses and low SR/modiolar synapses, whereby a control of SV release by ∼1 Ca2+ channel prevails at high SR/pillar synapses.
ii) Synaptic vesicle pool dynamics at individual IHC active zones
In 13 of the 31 aforementioned paired recordings (6 classified as low SR and 7 as high SR; 2 belonging to modiolar and 11 to pillar synapses), we employed a forward masking paradigm to study synaptic vesicle (SV) pool dynamics of single afferent synapses. The forward masking paradigm (Harris and Dallos, 1979) is commonly used for in vivo analysis of SGN spike rate adaptation and recovery from adaptation, which has been attributed to the depletion of readily releasable pool of SVs (RRP) and the recovery from depletion (Avissar et al., 2013; Frank et al., 2010; Furukawa and Matsuura, 1978; Goutman, 2017; Goutman and Glowatzki, 2007; Li et al., 2009; Moser and Beutner, 2000; Schroeder and Hall, 1974). Typically, the in vivo protocol is applied at saturating sound pressure levels, which we aimed to mimic using strong step IHC depolarizations (to −18 mV from −58 mV) separated by different interstimulus intervals (ISI: 4, 16, 64 and 256 ms) (Fig. 4A). In analogy to the in vivo forward masking paradigm, the first stimulus - called masker, as it depresses the response to a subsequent stimulus when applied in rapid succession-had a duration of 100 ms. The second stimulus-denominated probe- lasted for 15 ms. The recordings included a time frame of 400 ms preceding the “masker” and 400 ms following the “probe”, and the interval between masker and masker was 20 s. Applied to recordings of eEPSCs, the forward masking protocol provides experimental access to the initial RRP release rates, kinetics and extent of RRP depletion, sustained exocytosis, as well as recovery from RRP depletion. To accommodate the stochasticity of SV release from the RRP of IHC active zones, we run each protocol several times (≥3 to ≤ 20), which is routinely done for in vivo SGN physiology, but challenging ex vivo given the fragile and typically short-lived paired pre-and postsynaptic recordings (e.g. Goutman, 2017; Goutman and Glowatzki, 2007). Note that we did not employ cyclothiazide to inhibit AMPA receptor desensitization and reduce its contribution to postsynaptic eEPSC depression (Goutman, 2017), given the potential presynaptic effects of cyclothiazide in synaptic release (Diamond and Jahr, 1995; Dittman and Regehr, 1998).
For the analysis of evoked release dynamics, we focused on the response evoked by the masker. At the presynaptic level, there was no difference in the peak, initial and final IHC Ca2+ currents (ICa, Figure 4–figure supplement 1, panel A-C) and Ca2+ current charge (QCa, Fig. 4D) between the recordings from high and low SR synapses. We calculated the synaptic delay from the onset of the masker to the onset of the eEPSC. High SR synapses had significantly shorter synaptic delays (Fig. 4B), with mean latencies of the first eEPSC after stimulus onset of 1.21 ± 0.09 ms (median 1.18) compared to 3.34 ± 1.03 ms (median 2.09) in low SR pairs (p = 0.0221, Mann-Whitney U test). This result was corroborated in a bigger sample size, when we compared the synaptic latencies of the 31 pairs to 10 ms pulses to −18 mV (Figure 4–figure supplement 1, panel D; latency of 1.19 ± 0.14 ms (median 1.14) in high SR compared to 2.57 ± 0.48 ms (median 1.79) in low SR synapses; p = 0.0101, Mann-Whitney U test). The latency jitter (measured as the standard deviation of the eEPSC latencies; Fig. 4C) was also significantly smaller in high SR synapses compared to low SR synapses (0.38 ± 0.07 ms (median 0.41) vs 1.67 ± 0.72 ms (median 0.98), respectively; p = 0.0012, Mann-Whitney U test). Additionally, high SR synapses had significantly larger peak amplitudes of the masker-evoked EPSCs which reports the initial release from the RRP (Figure 4–figure supplement 1, panel E; −421.4 ± 54.22 pA for high SR synapses vs −240.2 ± 23.25 pA for low SR synapses; p = 0.0147, unpaired t-test). In contrast, the charge of masker-evoked EPSCs (QEPSC MASKER) was more comparable (Figure 4–figure supplement 1, panel F; 4.818 ± 0.88 pC for high SR vs 3.223 ± 0.47 pC for low SR pairs; p = 0.1574, unpaired t-test). We fitted the first 50 ms of the QEPSC MASKER with the sum of a single exponential and a line function in order to analyze RRP depletion and sustained release (Figure 4–figure supplement 1, panel G; Fig. 4G and dashed lines in Fig. 4D). The amplitude of the exponential component (A1), thought to reflect RRP exocytosis, was not different between high and low SR synapses (Figure 4–figure supplement 1, panel H; p = 0.4092, unpaired t-test). Likewise, the slope of the linear component, reflecting sustained exocytosis, did not differ significantly between the two groups (Figure 4–figure supplement 1, panel I; p = 0.1807, Mann-Whitney U test).
To quantify synaptic release in terms of SVs, we divided QEPSC MASKER by the mean QsEPSC recorded for each pair. This builds on our assumption that each sEPSC corresponds to a unitary release event (“univesicular mode of release” (Chapochnikov et al., 2014; Grabner and Moser, 2018; Huang and Moser, 2018)). The quantal content (RRP size in #SV) was comparable between high and low SR synapses (Fig. 4F, 13.99 ± 2.239 SVs vs 14.28 ± 4.839 SVs, respectively; p = 0.9553, unpaired t-test). However, the RRP depleted significantly faster in high SR synapses (Fig. 4G): τRRP depletion was 6.347 ± 1.096 ms (median 6.13) for high SR vs. 20.88 ± 5.063 ms (median 19.93) for low SR synapses (p = 0.0140, Mann-Whitney U test). Accordingly, high SR synapses showed significantly higher initial release rates compared to low SR synapses (Fig. 4H; 2651 ± 660.0 SV/s vs. 927.4 ± 307.4 SV/s; p = 0.0472, unpaired t-test), which given the comparable RRP size, indicates a higher release probability of the high SR rare synapses. Moreover, high SR synapses showed stronger depression of the release rate (Fig. 4J; the ratio of initial/sustained release rate was 5.941 ± 0.916 for high SR pairs and 2.023 ± 0.6251 for low SR pairs; p = 0.006, unpaired t-test), despite comparable sustained release rates (Fig. 4I; p = 0.9258, unpaired t-test).
Finally, we determined RRP recovery from depletion using the ratio QEPSC Probe/QEPSC Masker, whereby QEPSC for masker and probe was estimated for the first 10 ms of stimulation. Figure 4K plots the ratio for each interstimulus interval (ISI), including the masker-to-masker interval. Surprisingly, and contrary to a previous report in auditory bullfrog synapses (Cho et al., 2011), approximately half of the synapses showed a lower ratio at 16 ms than at 4 ms regardless of their SR. Therefore, to determine the kinetics of recovery from RRP depletion, we fitted a single exponential to the recovery data over the ISI range from 16 ms to 20 s (Fig. 4L). We did not observe a significant difference between high SR and low SR synapses (Fig. 4M; τrecovery RRP: 451.5 ± 187.0 ms (median 192.7) vs 2416 ± 2064 ms (median 157.9), respectively; p = 0.5338, Mann-Whitney U test). In 4 high SR and 4 low SR synapses, spontaneous activity was resumed shortly after the offset of the probe. The time to this first sEPSC took longer in high SR synapses compared to low ones (Figure 4–figure supplement 1, panel J; 185.2 ± 25.25 ms (median 161.6) vs 104.5 ± 24.10 ms (median 108.4), respectively; p = 0.0286, Mann-Whitney U test), again indicating stronger synaptic depression of hight SR synapses.
Discussion
Much of the information on synaptic sound encoding at afferent IHC-SGN synapses has been obtained from either juxtacellular recordings of SGN firing in vivo or from ex vivo patch-clamp recordings. Yet, it has remained difficult to reconcile those in vivo and ex vivo results and to establish a unified account of sound intensity coding in the auditory nerve given differences in experimental conditions, animal models and protocols employed. Here, we biophysically characterized the heterogeneous function of afferent SGN synapses in hearing mice with reference to their rate of spontaneous transmission (SR) as a surrogate of SGN SR that informs their functional properties. We performed paired pre-and post-synaptic patch-clamp recordings from single IHC synapses of hearing mice under near-physiological conditions using protocols adapted from in vivo characterization of SGN’s response properties. Using this approach, we were able to distinguish synapses with low and high SR, which we propose to provide the input into low and high SR SGNs. We found that about 90% of high SR synapses were located at the pillar side of the IHC. High SR synapses had larger sEPSCs with a more compact (monophasic) waveform, lower voltage-thresholds of release, shorter synaptic delays, tighter coupling of release sites to Ca2+ channels as well as higher initial release rates and shorter RRP depletion time constants. We conclude that high SR synapses exhibit higher release probability which likely reflects the tighter coupling of Ca2+ channels and release sites. RRP size, rate of sustained exocytosis and kinetics of RRP recovery from depletion were comparable between high and low SR synapses.
Diversity of spontaneous release and their topographical segregation
The SR range observed in our paired recordings from mouse afferent synapses (0 to 18 EPSC/s) agrees with previous reports of bouton recordings from rat synapses (p15-p17, Wu et al., 2016). However, the maximum rate is considerably smaller than those recorded in vivo from single SGNs [up to 60 spikes/s in p14-p21 mice (Wong et al., 2013) and up to 120 spikes/s in 8-17 weeks old mice (Taberner and Liberman, 2005)]. Given that less than 35% of murine SGNs have in vivo SRs above 20 spikes/s (Taberner and Liberman, 2005), our lower maximum rate could also originate from a sampling bias. In addition, the larger, faster and more compact sEPSC observed in high SR synapses would increase the speed and timing of the generated spike(s) (Rutherford et al., 2012), perhaps allowing for several spikes to be generated and leading to a further increase in the SGN spike SR. Other important factors that might explain the higher SR reported in vivo include: i) the intrinsic biophysical properties of SGNs which could further expand the firing rate distribution (Markowitz and Kalluri, 2020); ii) the developmental recruitment of high SR SGNs with age (Niwa et al., 2021; Romand, 1984; Walsh and McGee, 1987; Wong et al., 2013; Wu et al., 2016); and iii) compromised physiological conditions as well as a deviation of the chosen IHC holding potential from the IHC resting potential despite our efforts to mimic physiological conditions during the ex vivo recordings.
In analogy to the pioneering finding of a synaptic segregation of cat SGNs according to SR along the pillar–modiolar axis of IHCs (Liberman, 1982; Merchan-Perez and Liberman, 1996), we found that about 90% of high SR were located on the pillar IHC side. Yet, not all the synapses of the pillar IHC side had high SR, which agrees with a recent study of molecularly tagged SGNs (Siebald et al., 2023). These findings suggests that high frequency and large amplitude of sEPSCs occur predominantly in synapses with smaller ribbons and AZs, opposing results from retinal cells in which smaller ribbons resulted in reduced frequency and amplitude of EPSCs (Mehta et al., 2013). It is important to point out that our modiolar/pillar classification is less precise than that of other studies in which the synapse position was quantitatively assigned (Frank et al., 2009; Liberman et al., 2011; Ohn et al., 2016; Özçete and Moser, 2021). Moreover, other studies also support an overall pillar-modiolar gradient with “salt and pepper” intermingling of synaptic properties rather than their strict segregation (Ohn et al., 2016; Özçete and Moser, 2021).
Similar to previous reports (e.g. Chapochnikov et al., 2014; e.g. Glowatzki and Fuchs, 2002; Grant et al., 2010; Huang and Moser, 2018; Rutherford et al., 2012), we found a high variability in the waveform and amplitude of sEPSCs between synapses, which apparently do not strictly depend on the topographical location of the synapse (the present work and (Niwa et al., 2021)). Yet, high SR synapses had larger and more compact (monophasic) sEPSCs. The difference in the percentage of monophasic sEPSCs in low and high SR synapses could arise from variability in the fusion pore dynamics on the way to vesicle fusion and/or on the number of SVs released in timely manner. In the view of the multivesicular hypothesis of spontaneous release (Glowatzki and Fuchs, 2002; Li et al., 2009; Niwa et al., 2021; Schnee et al., 2013), SVs of a modiolar AZ might fuse in an uncoordinated manner, creating a EPSCs with a less compact waveform. Alternatively and our favorite hypothesis, each sEPSC corresponds to a unitary release event (“univesicular mode of release” (Chapochnikov et al., 2014; Grabner and Moser, 2018; Huang and Moser, 2018; Young et al., 2021)) that has the capacity to drive action potential firing (Rutherford et al., 2012). In the framework of the univesicular hypothesis of spontaneous release, the flickering of the fusion pore, prior to or instead of full collapse fusion of the SV, might be favored in low SR synapses, leading to a lower percentage of monophasic sEPSCs. Such heterogeneity of fusion pore dynamics has been reported in chromaffin cells, calyx of Held and hippocampal neurons (Chang et al., 2021; Henkel et al., 2019; Shin et al., 2020, 2018).
Candidate mechanisms distinguishing evoked release at low and high SR synapses
The temporal and quantal resolution offered by paired recordings allowed us to analyze the biophysical properties of evoked synaptic transmission in relation to the SR of the given synapse. In an intriguing resemblance with in vivo evoked firing properties of high SR SGNs (Bourien et al., 2014; Buran et al., 2010; Relkin and Doucet, 1991; Rhode and Smith, 1985; Taberner and Liberman, 2005), high SR synapses showed lower voltage (∼sound pressure in vivo) thresholds of synaptic transmission (∼firing in vivo), shorter and less variable synaptic latencies (∼first spike latencies in vivo), and higher initial release rates (∼onset firing rate in vivo). In addition, we found stronger synaptic depression at high SR synapses, which agrees well with the finding of a greater ratio of peak to adapted firing rate in high SR SGNs recorded in vivo (Taberner and Liberman, 2005). These results support the hypothesis that IHC synaptic heterogeneity (Frank et al., 2009; Hua et al., 2021; Ohn et al., 2016; Özçete and Moser, 2021; Reijntjes et al., 2020) contributes to the diversity of spontaneous and sound-evoked SGN firing.
How do high SR synapses with likely smaller ribbons and lower maximal Ca2+ influx achieve a shorter latency and higher initial release rate? Our hypothesis is that, in high SR synapses, a more hyperpolarized Ca2+ channel activation (Ohn et al., 2016) in combination with tighter coupling between the Ca2+ channels and the Ca2+ sensor of fusion (this work and (Özçete and Moser, 2021)) would enable a faster response with a greater initial SV release probability for a given stimulus. Interestingly, genetic disruptions that shift the voltage-dependence have a greater impact on the in vivo distribution of SR than mutations that change the maximal synaptic Ca2+ influx (Jean et al., 2018; Ohn et al., 2016). The spatial coupling of the Ca2+ channel to the SV release site has also been shown to greatly affect release probability in other synapses (Eggermann et al., 2012; Fekete et al., 2019; Moser et al., 2019; Rebola et al., 2019). Other factors that affect release probability include variations in the number of open Ca2+ channels at the AZ (Gratz et al., 2019; Holderith et al., 2012; Scimemi and Diamond, 2012; Sheng et al., 2012; Wong et al., 2013) and the fusion competence of the SV (Klenchin and Martin, 2000), including the priming and docking state (Lin et al., 2022; Neher and Brose, 2018). The ensuing Ca2+ influx in pillar synapses might facilitate Ca2+ channels and priming of SVs (Cho and von Gersdorff, 2012; Goutman and Glowatzki, 2011, 2007; Michalski et al., 2017; Moser and Beutner, 2000; Pangrsic et al., 2015; Schnee et al., 2011; Spassova et al., 2004) and contribute to the observed results in high SR synapses. Regarding the fusion competence of SVs, it is unknown whether modiolar and pillar synapses exhibit different numbers of docked and primed SVs. To date, ultrastructural studies that resolve docked and tethered SVs have not address the topographical location of the AZ in the murine IHC (Chakrabarti et al., 2022, 2018).
Besides SV release probability, RRP size co-determines neurotransmitter release. Our estimated RRP of about 14 SVs in both high and low SR synapses compares well to prior estimates obtained using ex vivo electrophysiology (10-40 SVs: Refs. (Goutman and Glowatzki, 2007; Jean et al., 2018; Johnson et al., 2005; Khimich et al., 2005; Moser and Beutner, 2000; Pangrsic et al., 2010; Schnee et al., 2005), model-based analysis of SGN firing (4-40 SVs (Frank et al., 2010; Jean et al., 2018; Peterson et al., 2014)) and electron microscopy (10-16 SVs within 50 nm of the presynaptic membrane (Chakrabarti et al., 2018; Frank et al., 2010; Graydon et al., 2011; Kantardzhieva et al., 2013; Khimich et al., 2005; Pangrsic et al., 2010). However, previous reports based on electron microscopy (Kantardzhieva et al., 2013; Merchan-Perez and Liberman, 1996; Michanski et al., 2019) suggested larger pools of SVs at modiolar synapses, while our electrophysiological estimate of RRP size was comparable between low and high SR synapses. This finding argues against a strong contribution of RRP size to the observed differences in neurotransmitter release. However, higher release probability with comparable RRP size explains higher initial release rates, which likely explain the faster and temporally more precise postsynaptic depolarization that is likely to turn into shorter first spike latencies and lower first spike latency jitter (this study and (Buran 2010).
Finally, the heterogeneity in the functional properties of IHC synapses could also arise from molecular heterogeneity of the AZ. In central glutamatergic synapses, molecular heterogeneity of synaptic proteins plays a critical role in the modulation of SV release probability and priming state (Neher and Brose, 2018; Wichmann and Kuner, 2022). For instance, differential isoforms of priming factors and scaffold proteins have been suggested to tune the functional synaptic diversity of central synapses (Fulterer et al., 2018; Rebola et al., 2019; Rosenmund et al., 2002). Cochlear IHCs have an unconventional fusion machinery that appears to work without neuronal SNARES (Nouvian et al., 2011) (but see Calvet et al., 2022) and priming factors such as Munc13 and CAPS (Vogl et al., 2015). Therefore, future studies will need to determine the molecular nanoanatomy underlying the specific AZ nanophysiology and functional synaptic heterogeneity at IHCs. Promising candidates include RBPs (Butola et al., 2021; Grauel et al., 2016; Krinner et al., 2021, 2017; Petzoldt et al., 2020), RIMs (Jung et al., 2015; Maria M. Picher et al., 2017), and Septin (Fekete et al., 2019; Yang et al., 2010).
Challenges for relating synaptic and neural response properties
Next to providing support for the presynaptic hypothesis of functional SGN diversity, the present study also highlights some of the challenges met when aiming to bridge the gap between presynaptic hair cell function and neural sound encoding. Despite major efforts undertaken to match experimental conditions and protocols, it remains difficult to reconcile some findings of ex vivo and in vivo physiology. Parameters such as RRP size (∼# spikes of the rapidly adapting component of firing), sustained exocytosis (∼adapted firing rate in vivo), recovery of spontaneous and evoked release (∼recovery from forward masking in vivo) did not differ among our high and low SR synapses, and contrasts with in vivo data (e.g. Refs. (Bourien et al., 2014; Buran et al., 2010; Relkin and Doucet, 1991; Rhode and Smith, 1985; Taberner and Liberman, 2005)).
Of particular interest is that the dynamic ranges and slope of release-intensity relationship of high and low SR synapses diverge from the expectations if assuming that high SR SGNs are driven by high SR synapses. High SR synapses tended to show broader dynamic ranges with shallower slopes, while, in vivo, high SR SGNs show smaller dynamic ranges and steeper slopes than the low SR ones (Ohlemiller et al., 1991; Winter et al., 1990). Could this reflect the non-linear saturating properties of the basilar membrane (Sachs et al., 1989; Sachs and Abbas, 1974; Yates et al., 1990) (discussed in Ref. (Ohlemiller et al., 1991)) which might widen the rate level function of low SR SGNs? Or is it due to a partial depletion of the “standing” RRP (i.e. the occupancy of the RRP release sites with a fusion-competent SV (Moser, 2020; Pangrsic et al., 2010)) at high SR synapses in vivo? It remains to be determined whether this and the other aforementioned differences between our data and in vivo reports could be attributed to mechanisms downstream of glutamate release and AMPA receptor activation. The possible mechanisms include but are not limited to: i) different spike rates due to diverse EPSC waveforms (Rutherford et al., 2012); ii) differences in SGN excitability (Crozier and Davis, 2014; Markowitz and Kalluri, 2020; Smith et al., 2015) due to heterogenous molecular (Petitpré et al., 2020, 2018; Shrestha et al., 2018; Sun et al., 2018) and morphological profiles (Liberman, 1980; Merchan-Perez and Liberman, 1996; Tsuji and Liberman, 1997); and iii) differences in efferent innervation of SGNs (Hua et al., 2021; Liberman, 1990; Ruel et al., 2001; Wu et al., 2020; Yin et al., 2014). Certainly, caution is to be applied for the comparison of ex vivo and in vivo data due to partial disruption of the physiological milieu despite our efforts to maintain near-physiological conditions. This also extends to the alteration of efferent signaling in our ex vivo preparation.
Clearly more work is needed to elucidate the mechanisms of SGN firing diversity in the cochlea. Ideally, future studies will combine in vivo and ex vivo experiments, such as combining physiological SGN characterization with neural backtracing and synaptic morphology of labelled SGNs using volume imaging of afferent and efferent connectivity (Hua et al., 2021). Moreover, combining optogenetic IHC stimulation with imaging of SGN activity could provide higher throughput and serve posthoc morphology. Finally, paired patch-clamp recordings, as done in the present study, could be combined with SGN subtype-specific molecular labeling, fiber tracing and immunolabeling to further relate synaptic transmission and SGN neurophysiology.
Materials and Methods
Animals and tissue preparation
c57BL/6N mice of either sex between postnatal day 14-23 (p14-23) were used. The animal handling and experiments complied with national animal care guidelines and were approved by the University of Göttingen Board for animal welfare and the Animal Welfare Office of the State of Lower Saxony. Animals were sacrificed by decapitation and the cochleae were extracted in modified Hepes Hank’s solution containing: 5.36 mM KCl, 141.7 mM NaCl, 1 mM MgCl2-6H2O, 0.5 mM MgSO4-7H2O, 10 mM HEPES, 0.5 mg/ml L-glutamine, and 1 mg/ml D-glucose (pH 7.2, osmolarity of ∼300 mOsm). The apical coil of the organ of Corti was dissected and placed under a grid in the recording chamber. Pillar or modiolar supporting cells were removed using soda glass pipettes in order to gain access to the basolateral face of the IHCs and to the postsynaptic boutons of type I SGNs. Dissection of the organ of Corti and cleaning of the supporting cells were performed at room temperature (20-25°C).
Electrophysiological recordings
Pre-and postsynaptic paired patch clamp recordings were performed at near physiological temperature (32-37°C) using an EPC-9 amplifier (HEKA electronics) (Fig. 1). Patch electrodes were positioned using a PatchStar micromanipulator (Scientifica, UK). Whole-cell recordings from IHCs were achieved using the perforated-patch clamp technique (Moser and Beutner, 2000) using Sylgard™–coated 1.5 mm borosilicate pipettes with typical resistances between 3.5 and 6 MΩ. The IHC pipette solution contained: 129 mM Cs-gluconate, 10 mM tetraethylammonium (TEA)-Cl, 10 mM 4-AP, 10 mM HEPES, 1 mM MgCl2 (pH 7.2, osmolarity of ∼290 mOsm), as well as 300 μg/ml amphotericin B added prior to the experiment. Once the series resistance of the IHC reached below 30 MΩ, whole-cell voltage-clamp recordings from a contacting bouton was established as described in previous studies (Glowatzki and Fuchs, 2002; Grant et al., 2010; Huang and Moser, 2018). Sylgard™-coated 1.0 mm borosilicate pipettes with typical resistances between 7 and 12 MΩ were used for the postsynaptic recordings. The bouton pipette solution contained: 137 mM KCl, 5 mM EGTA, 5 mM HEPES, 1 Mm Na2-GTP, 2.5 mM Na2-ATP, 3.5 mM MgCl2·6H2O and 0.1 mM CaCl2 (pH 7.2 and osmolarity of ∼290 mOsm). The organ of Corti was continuously perfused with an extracellular solution containing 4.2 mM KCl, 95-100 mM NaCl, 25 mM NaHCO3, 30 mM TEA-Cl, 1mM Na-Pyruvate, 0.7 mM NH2PO4·H2O, 1mM CsCl, 1 mM MgCl2·H2O, 1.3 mM CaCl2, and 11.1 mM D-glucose (pH 7.3, osmolarity of ∼310 mOsm). 2.5 µM tetrodotoxin (Tocris or Santa Cruz) was added to block voltage-gated Na+ channels in the postsynaptic bouton.
Data were acquired using the Patchmaster software (HEKA electronics). The current signal was filtered at 5-10 kHz and sampled at 20-50 kHz. IHC were voltage-clamped at a holding potential of −58 mV, corresponding to the presumed in vivo resting potential(Johnson, 2015). The bouton was held at a potential of −94 mV. All reported potentials were corrected for the liquid junction potential (19 mV for the IHC and 4 mV for the bouton), measured experimentally. Ca2+ current recordings were corrected for the linear leak current using a P/n protocol. We excluded IHCs and boutons with leak currents exceeding −60 pA and -100 pA at holding potential, respectively. The series resistance of the IHCs was typically below 30 MΩ. The apparent series resistance of the bouton was calculated from the capacitive transient in response to a 10-mV test pulse. The actual series resistance (Rs) was offline calculated as reported in (Huang and Moser, 2018). Recordings with bouton Rs > 80 MΩ were discarded.
The threshold for sEPSC detection was 4 times SD of the baseline. Spontaneous activity was calculated from time windows without stimulation with the IHC held at −58 mV; either from a 5 – 10 s recording or by averaging the number of events from the segments before and after a depolarizing pulse (Fig. 1B, Supplementary Table 1). To study the depletion and recovery of the pool of vesicles, we used a protocol adapted from the forward masking protocol performed during in vivo extracellular recordings of SGNs (Harris and Dallos, 1979; Jean et al., 2018). It consisted of two consecutive depolarizing pulses to the voltage that elicited the highest peak of Ca2+ current (−18 mV; Fig. 1C). The first pulse, called masker, lasted 100 ms and it was followed by a second pulse, called probe, which lasted 15 ms. The two pulses were separated by intervals without depolarization (interstimulus intervals, ISI) that lasted 4, 16, 64 and 256 ms. The waiting time between masker and masker was 20 s and each protocol was repeated between 3 – 20 times. To study the dynamic voltage range of synaptic transmission, we used a current-voltage (IV) protocol with 10 ms pulses of increasing voltage (from −70 mV/-60 mV to 70 mV in 5 mV steps). The interval between two stimuli was 1.5 s.
Data Analysis
Electrophysiological data was analyzed using the IgorPro 6 Software Package (Wavemetrics), GraphPad Prism 9 and Excel. Ca2+ charge (QCa) and EPSC charge (QEPSC) were estimated by taking the integral of the current. Kinetics of sEPSCs, such as amplitude, 10-90% rise time, time constant of decay (τdecay) and full-width half-maximum (FWHM), were calculated with Neuromatic (Rothman and Silver, 2018).
To obtain IV curves, we averaged the evoked Ca2+-currents (ICa) during 5 to 10 ms after the start of each depolarization. Fractional activation of the Ca2+ channels was obtained from the normalized chord conductance, g,
where V is the membrane potential and Vrev is the reversal potential determined by fitting a line function between the voltage of ICa peak + 10 mV and the maximal depolarization. The activation curve was approximated by a first-order Boltzmann equation:
where gmax is the maximum chord conductance, Vhalf Ca is the membrane potential at which the conductance is half activated, and S is the slope factor describing the voltage sensitivity of activation.
Release intensity curves were obtained by calculating QEPSC by the end of each depolarization step and fitted using a sigmoidal function:
where Qmax is the maximal QEPSC (normalized to 1), Vh EPSC corresponds to the voltage of half-maximal release (or Q50, EPSC) and Q is the EPSC charge. The dynamic range was determined as the voltage range between 10% and 90% of the maximal vesicle release. For statistical analysis of dynamic range, we included only pairs for which both the Ca2+ fractional activation and the rate level curves were possible to fit.
The apparent Ca2+ dependence of neurotransmitter release was studied from the 10 ms step-depolarizations of the IV curves. The resulting QEPSC vs IHC QCa plots from each individual pair were fitted with a power function:
Where m corresponds to the Ca2+ cooperativity. Some pairs showed a clear saturation of release at high IHC QCa. In these cases, the fit was restricted to the datapoints before the plateau, which was determined by visual inspection. For the pooled data, the power function was fitted to the normalized QEPSC vs normalized QCa. For the pairs with saturation of release, QCa was normalized to a point before the plateau.
For forward masking experiments, the postsynaptic response was averaged for all the repetitions for each paired recording (between 3 and 20, depending on the stability of the pair). Single active zone pool dynamics were determined by fitting an exponential plus line function to the individual EPSC charge plots,
where A1 is the amplitude of the exponential component, τ is the time constant of the exponential component. RRP size (in vesicles) was estimated from dividing A1 by the charge of the average sEPSC for each pair. Sustained exocytosis rate (in vesicles per s) was calculated from the slope of the linear function divided the charge of the average sEPSC. Individual recovery kinetics were determined from the ratio of probe and masker responses at 10 ms of the depolarization, with the ratio between masker and masker being 1. The recovery traces were fitted with a single exponential function from 16 ms to 20 000 ms to determine the time constant of RRP recovery.
Data was prepared for presentation using Adobe Illustrator. Skewness analysis, PCA and other statistical analysis were performed using GraphPad Prism 9. Statistical significance was assessed with unpaired t-test or non-parametric Mann-Whitney U test depending on the normal distribution and equality of variances of the data (Saphiro-Wilk test and F test). Data is expressed as mean ± sem. The box plots show 25th, 50th and 75th percentiles with the individual data points overlaid.
Acknowledgements
We would like to thank Dr. Chao-Hua Huang and Dr. Jakob Neef for their experimental and analytical input. Dr. Antoine Huet for the discussion regarding in vivo response properties of spiral ganglion neurons. Dres. Erwin Neher, Manfred Lindau and Jakob Neef for critical input into this project. LMJT was a recipient of the Erwin Neher Fellowship and T.M. is a Max-Planck Fellow at the Max Planck Institute for Multidisciplinary Sciences. This work was further supported by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) via the Collaborative Research Center 889 (project A02) and the Leibniz Program (MO896/5 to T.M), the European Research Council through the Advanced Grant ‘DynaHear” to T.M. under the European Union’s Horizon 2020 Research and Innovation program (grant agreement No. 101054467), and by Fondation Pour l’Audition (FPA RD-2020-10). LMJT is a member of the Hertha Sponer College from the Cluster of Excellence Multiscale Bioimaging (MBExC). Open access funding provided by Max Planck Society.
References
- Refractoriness Enhances Temporal Coding by Auditory Nerve FibersJournal of Neuroscience 33:7681–7690https://doi.org/10.1523/JNEUROSCI.3405-12.2013
- Auditory nerve of the normal and jaundiced ratI. Spontaneous discharge rate and cochlear nerve histology. Hearing Research 54:75–90https://doi.org/10.1016/0378-5955(91)90138-Y
- Contribution of auditory nerve fibers to compound action potential of the auditory nerveJournal of Neurophysiology 112:1025–1039https://doi.org/10.1152/jn.00738.2013
- Few CaV1.3 Channels Regulate the Exocytosis of a Synaptic Vesicle at the Hair Cell Ribbon SynapseJournal of Neuroscience 25:11577–11585https://doi.org/10.1523/JNEUROSCI.3411-05.2005
- Onset coding is degraded in auditory nerve fibers from mutant mice lacking synaptic ribbonsJ Neurosci 30:7587–7597https://doi.org/10.1523/JNEUROSCI.0389-10.2010
- RIM-Binding Protein 2 Organizes Ca2+ Channel Topography and Regulates Release Probability and Vesicle Replenishment at a Fast Central SynapseJ Neurosci 41:7742–7767https://doi.org/10.1523/JNEUROSCI.0586-21.2021
- The SNARE protein SNAP-25 is required for normal exocytosis at auditory hair cell ribbon synapsesiScience 25https://doi.org/10.1016/j.isci.2022.105628
- Optogenetics and electron tomography for structure-function analysis of cochlear ribbon synapseseLife 11https://doi.org/10.7554/eLife.79494
- Vesicle sub-pool organization at inner hair cell ribbon synapsesEMBO reports e 44937https://doi.org/10.15252/embr.201744937
- Synaptophysin Regulates Fusion Pores and Exocytosis Mode in Chromaffin CellsJ Neurosci 41:3563–3578https://doi.org/10.1523/JNEUROSCI.2833-20.2021
- Uniquantal Release through a Dynamic Fusion Pore Is a Candidate Mechanism of Hair Cell ExocytosisNeuron 17:1389–1403https://doi.org/10.1016/j.neuron.2014.08.003
- Recovery from Short-Term Depression and Facilitation Is Ultrafast and Ca2+ Dependent at Auditory Hair Cell SynapsesThe Journal of Neuroscience 31:5682–5692https://doi.org/10.1523/JNEUROSCI.5453-10.2011
- Ca2+ influx and neurotransmitter release at ribbon synapsesCell Calcium 52:208–216https://doi.org/10.1016/j.ceca.2012.06.004
- Unmasking of Spiral Ganglion Neuron Firing Dynamics by Membrane Potential and Neurotrophin-3J Neurosci 34:9688–9702https://doi.org/10.1523/JNEUROSCI.4552-13.2014
- Asynchronous release of synaptic vesicles determines the time course of the AMPA receptor-mediated EPSCNeuron 15:1097–1107
- Calcium Dependence and Recovery Kinetics of Presynaptic Depression at the Climbing Fiber to Purkinje Cell SynapseThe Journal of Neuroscience 18:6147–6162
- Nanodomain coupling between Ca2+ channels and sensors of exocytosis at fast mammalian synapsesNat Rev Neurosci 13:7–21https://doi.org/10.1038/nrn3125
- The frequency response and other properties of single fibres in the guinea-pig cochlear nerveThe Journal of Physiology 226:263–287https://doi.org/10.1113/jphysiol.1972.sp009984
- Underpinning heterogeneity in synaptic transmission by presynaptic ensembles of distinct morphological modulesNat Commun 10https://doi.org/10.1038/s41467-019-08452-2
- Mechanisms contributing to synaptic Ca2+ signals and their heterogeneity in hair cellsProceedings of the National Academy of Sciences 106:4483–4488https://doi.org/10.1073/pnas.0813213106
- Bassoon and the synaptic ribbon organize Ca2+ channels and vesicles to add release sites and promote refillingNeuron 68:724–738https://doi.org/10.1016/j.neuron.2010.10.027
- Active Zone Scaffold Protein Ratios Tune Functional Diversity across Brain SynapsesCell Rep 23:1259–1274https://doi.org/10.1016/j.celrep.2018.03.126
- Adaptive rundown of excitatory post-synaptic potentials at synapses between hair cells and eight nerve fibres in the goldfishJ Physiol (Lond 276:193–209
- Transmitter release at the hair cell ribbon synapseNat Neurosci 5:147–154https://doi.org/10.1038/nn796
- Divide and conquer acoustic diversityThe EMBO Journal 40https://doi.org/10.15252/embj.2020107531
- Mechanisms of synaptic depression at the hair cell ribbon synapse that support auditory nerve functionProc Natl Acad Sci USA https://doi.org/10.1073/pnas.1706160114
- Short-Term Facilitation Modulates Size and Timing of the Synaptic Response at the Inner Hair Cell Ribbon SynapseJ Neurosci 31:7974–7981https://doi.org/10.1523/JNEUROSCI.0604-11.2011
- Time course and calcium dependence of transmitter release at a single ribbon synapseProc Natl Acad Sci U S A 104:16341–16346https://doi.org/10.1073/pnas.0705756104
- Individual synaptic vesicles mediate stimulated exocytosis from cochlear inner hair cellsProc Natl Acad Sci USA 115:12811–12816https://doi.org/10.1073/pnas.1811814115
- Two Modes of Release Shape the Postsynaptic Response at the Inner Hair Cell Ribbon SynapseJ Neurosci 30:4210–4220https://doi.org/10.1523/JNEUROSCI.4439-09.2010
- Endogenous Tagging Reveals Differential Regulation of Ca2+ Channels at Single Active Zones during Presynaptic Homeostatic Potentiation and DepressionJ Neurosci 39:2416–2429https://doi.org/10.1523/JNEUROSCI.3068-18.2019
- RIM-binding protein 2 regulates release probability by fine-tuning calcium channel localization at murine hippocampal synapsesPNAS 113:11615–11620https://doi.org/10.1073/pnas.1605256113
- Sharp Ca2+ Nanodomains beneath the Ribbon Promote Highly Synchronous Multivesicular Release at Hair Cell SynapsesJ Neurosci 31:16637–16650https://doi.org/10.1523/JNEUROSCI.1866-11.2011
- Forward masking of auditory nerve fiber responsesJ Neurophysiol 42:1083–1107
- Differential Release of Exocytosis Marker Dyes Indicates Stimulation-Dependent Regulation of Synaptic ActivityFrontiers in Neuroscience 13
- Release probability of hippocampal glutamatergic terminals scales with the size of the active zoneNature Neuroscience 15:988–997https://doi.org/10.1038/nn.3137
- Electron Microscopic Reconstruction of Neural Circuitry in the CochleaCell Reports 34https://doi.org/10.1016/j.celrep.2020.108551
- Ca2+ Regulates the Kinetics of Synaptic Vesicle Fusion at the Afferent Inner Hair Cell SynapseFront Cell Neurosci 12https://doi.org/10.3389/fncel.2018.00364
- Sound coding in the auditory nerve of gerbilsHear Res https://doi.org/10.1016/j.heares.2016.05.006
- The synaptic ribbon is critical for sound encoding at high rates and with temporal precisionElife 7https://doi.org/10.7554/eLife.29275
- Membrane properties specialize mammalian inner hair cells for frequency or intensity encodingeLife 4https://doi.org/10.7554/eLife.08177
- Increase in efficiency and reduction in Ca2+ dependence of exocytosis during development of mouse inner hair cellsJ Physiol 563:177–191https://doi.org/10.1113/jphysiol.2004.074740
- Rab3-interacting molecules 2α and 2β promote the abundance of voltage-gated CaV1.3 Ca2+ channels at hair cell active zonesProc Natl Acad Sci USA 112:E3141–E3149https://doi.org/10.1073/pnas.1417207112
- Quantitative analysis of ribbons, vesicles, and cisterns at the cat inner hair cell synapse: correlations with spontaneous rateJ Comp Neurol 521:3260–3271https://doi.org/10.1002/cne.23345
- Hair cell synaptic ribbons are essential for synchronous auditory signallingNature 434:889–894https://doi.org/10.1038/nature03418
- Discharge Patterns of Single Fibers in the Cat’s Auditory Nerve. CambridgeMassachusetts: MIT Press
- Priming in exocytosis: Attaining fusion-competence after vesicle dockingBiochimie 82:399–407https://doi.org/10.1016/S0300-9084(00)00208-X
- . alpha 1D (Cav1.3) subunits can form l-type Ca2+ channels activating at negative voltagesJ Biol Chem 276:22100–22106https://doi.org/10.1074/jbc.M101469200
- RIM-Binding Protein 2 Promotes a Large Number of CaV1.3 Ca2+-Channels and Contributes to Fast Synaptic Vesicle Replenishment at Hair Cell Active ZonesFront Cell Neurosci 11https://doi.org/10.3389/fncel.2017.00334
- RIM-Binding Proteins Are Required for Normal Sound-Encoding at Afferent Inner Hair Cell SynapsesFront Mol Neurosci 14https://doi.org/10.3389/fnmol.2021.651935
- Comprehensive transcriptome analysis of cochlear spiral ganglion neurons at multiple agesElife 9https://doi.org/10.7554/eLife.50491
- The Unitary Event Underlying Multiquantal EPSCs at a Hair Cell’s Ribbon SynapseJournal of Neuroscience 29:7558–7568https://doi.org/10.1523/JNEUROSCI.0514-09.2009
- Opposing Gradients of Ribbon Size and AMPA Receptor Expression Underlie Sensitivity Differences among Cochlear-Nerve/Hair-Cell SynapsesJ Neurosci 31:801–808https://doi.org/10.1523/JNEUROSCI.3389-10.2011
- Effects of chronic cochlear de-efferentation on auditory-nerve responseHear Res 49:209–223
- Single-Neuron Labeling in the Cat Auditory NerveScience 216:1239–1241https://doi.org/10.1126/science.7079757
- Morphological differences among radial afferent fibers in the cat cochlea: an electron-microscopic study of serial sectionsHear Res 3:45–63https://doi.org/10.1016/0378-5955(80)90007-6
- A sequential two-step priming scheme reproduces diversity in synaptic strength and short-term plasticityProceedings of the National Academy of Sciences 119https://doi.org/10.1073/pnas.2207987119
- Gradients in the biophysical properties of neonatal auditory neurons align with synaptic contact position and the intensity coding map of inner hair cellsElife 9https://doi.org/10.7554/eLife.55378
- Synaptic Ribbons Influence the Size and Frequency of Miniature-like Evoked Postsynaptic CurrentsNeuron 77:516–527https://doi.org/10.1016/j.neuron.2012.11.024
- Ultrastructural differences among afferent synapses on cochlear hair cells: correlations with spontaneous discharge rateJ Comp Neurol 371:208–221https://doi.org/10.1002/(SICI)1096-9861(19960722)371:2<208::AID-CNE2>3.0.CO;2-6
- Tuning of synapse number, structure and function in the cochleaNat Neurosci 12:444–453https://doi.org/10.1038/nn.2293
- Structure and function of cochlear afferent innervationCurr Opin Otolaryngol Head Neck Surg 18:441–446https://doi.org/10.1097/MOO.0b013e32833e0586
- Otoferlin acts as a Ca2+ sensor for vesicle fusion and vesicle pool replenishment at auditory hair cell ribbon synapseseLife 6https://doi.org/10.7554/eLife.31013
- Mapping developmental maturation of inner hair cell ribbon synapses in the apical mouse cochleaPNAS 116:6415–6424https://doi.org/10.1073/pnas.1812029116
- Presynaptic Physiology of Cochlear Inner Hair CellsThe Senses: A Comprehensive ReferenceElsevier :441–467https://doi.org/10.1016/B978-0-12-809324-5.24185-0
- Kinetics of exocytosis and endocytosis at the cochlear inner hair cell afferent synapse of the mouseProc Natl Acad Sci U S A 97:883–888https://doi.org/10.1073/pnas.97.2.883
- Sensory processing at ribbon synapses in the retina and the cochleaPhysiol Rev 100:103–144https://doi.org/10.1152/physrev.00026.2018
- Mechanisms underlying the temporal precision of sound coding at the inner hair cell ribbon synapseJ Physiol 576:55–62https://doi.org/10.1113/jphysiol.2006.114835
- Quantitative optical nanophysiology of Ca2+ signaling at inner hair cell active zonesNat Commun 9https://doi.org/10.1038/s41467-017-02612-y
- Dynamically Primed Synaptic Vesicle States: Key to Understand Synaptic Short-Term PlasticityNeuron 100:1283–1291https://doi.org/10.1016/j.neuron.2018.11.024
- Functional subgroups of cochlear inner hair cell ribbon synapses differently modulate their EPSC properties in response to stimulationJournal of Neurophysiology 125:2461–2479https://doi.org/10.1152/jn.00452.2020
- Exocytosis at the hair cell ribbon synapse apparently operates without neuronal SNARE proteinsNat Neurosci 14:411–413https://doi.org/10.1038/nn.2774
- Creating diverse synapses from the same moleculesCurr Opin Neurobiol 51:8–15https://doi.org/10.1016/j.conb.2018.01.001
- Factors that influence rate-versus-intensity relations in single cochlear nerve fibers of the gerbilThe Journal of the Acoustical Society of America 90:274–287https://doi.org/10.1121/1.401298
- Hair cells use active zones with different voltage dependence of Ca2+ influx to decompose sounds into complementary neural codesPNAS 113:E4716–E4725https://doi.org/10.1073/pnas.1605737113
- A sensory cell diversifies its output by varying Ca2+ influx-release coupling among active zonesEMBO J 40https://doi.org/10.15252/embj.2020106010
- EF-hand protein Ca2+ buffers regulate Ca2+ influx and exocytosis in sensory hair cellsP Natl Acad Sci USA 112:E1028–E1037https://doi.org/10.1073/pnas.1416424112
- Hearing requires otoferlin-dependent efficient replenishment of synaptic vesicles in hair cellsNat Neurosci 13:869–876
- A model of synaptic vesicle-pool depletion and replenishment can account for the interspike interval distributions and nonrenewal properties of spontaneous spike trains of auditory-nerve fibersJ Neurosci 34:15097–15109https://doi.org/10.1523/JNEUROSCI.0903-14.2014
- Genetic and functional diversity of primary auditory afferents. Current Opinion in PhysiologyPhysiology of Mammalian Hearing 18:85–94https://doi.org/10.1016/j.cophys.2020.09.011
- Neuronal heterogeneity and stereotyped connectivity in the auditory afferent systemNat Commun 9https://doi.org/10.1038/s41467-018-06033-3
- RIM-binding protein couples synaptic vesicle recruitment to release sitesJournal of Cell Biology 219https://doi.org/10.1083/jcb.201902059
- Ca2+-binding protein 2 inhibits Ca2+-channel inactivation in mouse inner hair cellsPNAS 114:E1717–E1726https://doi.org/10.1073/pnas.1617533114
- Rab Interacting Molecules 2 and 3 Directly Interact with the Pore-Forming CaV1.3 Ca2+ Channel Subunit and Promote Its Membrane ExpressionFront Cell Neurosci 11https://doi.org/10.3389/fncel.2017.00160
- Congenital deafness and sinoatrial node dysfunction in mice lacking class D L-type Ca2+ channelsCell 102:89–97
- Distinct Nanoscale Calcium Channel and Synaptic Vesicle Topographies Contribute to the Diversity of Synaptic FunctionNeuron 104:693–710https://doi.org/10.1016/j.neuron.2019.08.014
- Volume gradients in inner hair cell-auditory nerve fiber pre-and postsynaptic proteins differ across mouse strainsHearing Research 390https://doi.org/10.1016/j.heares.2020.107933
- The afferent signaling complex: Regulation of type I spiral ganglion neuron responses in the auditory peripheryHear Res 336:1–16https://doi.org/10.1016/j.heares.2016.03.011
- Recovery from prior stimulation. I: Relationship to spontaneous firing rates of primary auditory neuronsHear Res 55:215–222
- Characteristics of tone-pip response patterns in relationship to spontaneous rate in cat auditory nerve fibersHear Res 18:159–168
- Role of L-Type Ca2+ Channels in Transmitter Release From Mammalian Inner Hair CellsII. Single-Neuron Activity. Journal of Neurophysiology 87:2734–2740
- Functional properties of auditory-nerve fibers during postnatal development in the kittenExperimental Brain Research 56:395–402
- Differential control of vesicle priming and short-term plasticity by Munc13 isoformsNeuron 33:411–424https://doi.org/10.1016/S0896-6273(02)00568-8
- NeuroMatic: An Integrated Open-Source Software Toolkit for Acquisition, Analysis and Simulation of Electrophysiological DataFrontiers in Neuroinformatics 12
- Dopamine inhibition of auditory nerve activity in the adult mammalian cochleaEuropean Journal of Neuroscience 14:977–986https://doi.org/10.1046/j.0953-816x.2001.01721.x
- Intracellular studies of hair cells in the mammalian cochleaJ Physiol (Lond 284:261–290
- Spike Encoding of Neurotransmitter Release Timing by Spiral Ganglion Neurons of the CochleaJ Neurosci 32:4773–4789https://doi.org/10.1523/JNEUROSCI.4511-11.2012
- Rate versus level functions for auditory-nerve fibers in cats: tone-burst stimuliJ Acoust Soc Am 56:1835–1847https://doi.org/10.1121/1.1903521
- A computational model for rate-level functions from cat auditory-nerve fibersHear Res 41:61–69https://doi.org/10.1016/0378-5955(89)90179-2
- Spontaneous rates, thresholds and tuning of auditory-nerve fibers in the gerbil: comparisons to cat dataHear Res 42:23–35
- Response properties from turtle auditory hair cell afferent fibers suggest spike generation is driven by synchronized release both between and within synapsesJournal of Neurophysiology 110:204–220https://doi.org/10.1152/jn.00121.2013
- Auditory hair cell-afferent fiber synapses are specialized to operate at their best frequenciesNeuron 47:243–254https://doi.org/10.1016/j.neuron.2005.06.004
- Calcium-Dependent Synaptic Vesicle Trafficking Underlies Indefatigable Release at the Hair Cell Afferent Fiber SynapseNeuron 70:326–338https://doi.org/10.1016/j.neuron.2011.01.031
- Model for mechanical to neural transduction in the auditory receptorJ Acoust Soc Am 55:1055–1060
- The number and organization of Ca2+ channels in the active zone shapes neurotransmitter release from Schaffer collateral synapsesJ Neurosci 32:18157–18176https://doi.org/10.1523/JNEUROSCI.3827-12.2012
- Calcium-channel number critically influences synaptic strength and plasticity at the active zoneNat Neurosci 15:998–1006https://doi.org/10.1038/nn.3129
- Pou4f1 Defines a Subgroup of Type I Spiral Ganglion Neurons and Is Necessary for Normal Inner Hair Cell Presynaptic Ca2+ SignalingJ Neurosci 39:5284–5298https://doi.org/10.1523/JNEUROSCI.2728-18.2019
- Vesicle Shrinking and Enlargement Play Opposing Roles in the Release of Exocytotic ContentsCell Rep 30:421–431https://doi.org/10.1016/j.celrep.2019.12.044
- Visualization of Membrane Pore in Live Cells Reveals a Dynamic-Pore Theory Governing Fusion and EndocytosisCell 173:934–945https://doi.org/10.1016/j.cell.2018.02.062
- Sensory Neuron Diversity in the Inner Ear Is Shaped by ActivityCell 174:1229–1246https://doi.org/10.1016/j.cell.2018.07.007
- Molecular signatures define subtypes of auditory afferents with distinct peripheral projection patterns and physiological propertiesProceedings of the National Academy of Sciences 120https://doi.org/10.1073/pnas.2217033120
- Phosphoinositide Modulation of Heteromeric Kv1 Channels Adjusts Output of Spiral Ganglion Neurons from Hearing MiceJ Neurosci 35:11221–11232https://doi.org/10.1523/JNEUROSCI.0496-15.2015
- Evidence that rapid vesicle replenishment of the synaptic ribbon mediates recovery from short-term adaptation at the hair cell afferent synapseJ Assoc Res Otolaryngol 5:376–390
- Hair Cell Mechanotransduction Regulates Spontaneous Activity and Spiral Ganglion Subtype Specification in the Auditory SystemCell 174:1247–1263https://doi.org/10.1016/j.cell.2018.07.008
- Response Properties of Single Auditory Nerve Fibers in the MouseJ Neurophysiol 93:557–569https://doi.org/10.1152/jn.00574.2004
- Intracellular labeling of auditory nerve fibers in guinea pig: central and peripheral projectionsJ Comp Neurol 381:188–202
- Unconventional molecular regulation of synaptic vesicle replenishment in cochlear inner hair cellsJ Cell Sci 128:638–644https://doi.org/10.1242/jcs.162099
- Postnatal development of auditory nerve and cochlear nucleus neuronal responses in kittensHear Res 28:97–116
- Homeostatic Mechanisms in the Cochlea In: Dallos P, Popper AN, Fay RR, editors. The CochleaSpringer Handbook of Auditory Research. Springer New York :130–185https://doi.org/10.1007/978-1-4612-0757-3_3
- Heterogeneity of glutamatergic synapses: cellular mechanisms and network consequencesPhysiological Reviews 102:269–318https://doi.org/10.1152/physrev.00039.2020
- Diversity of characteristic frequency rate-intensity functions in guinea pig auditory nerve fibresHear Res 45:191–202https://doi.org/10.1016/0378-5955(90)90120-E
- Concurrent Maturation of Inner Hair Cell Synaptic Ca2+ Influx and Auditory Nerve Spontaneous Activity around Hearing Onset in MiceJ Neurosci 33:10661–10666https://doi.org/10.1523/JNEUROSCI.1215-13.2013
- Developmental refinement of hair cell synapses tightens the coupling of Ca2+ influx to exocytosisEMBO J 33:247–264https://doi.org/10.1002/embj.201387110
- Sound exposure dynamically induces dopamine synthesis in cholinergic LOC efferents for feedback to auditory nerve fiberseLife 9https://doi.org/10.7554/eLife.52419
- Maturation of Spontaneous Firing Properties after Hearing Onset in Rat Auditory Nerve Fibers: Spontaneous Rates, Refractoriness, and Interfiber CorrelationsJ Neurosci 36:10584–10597https://doi.org/10.1523/JNEUROSCI.1187-16.2016
- Neuronal CaV1.3α1 L-Type Channels Activate at Relatively Hyperpolarized Membrane Potentials and Are Incompletely Inhibited by DihydropyridinesJ Neurosci 21:5944–5951
- Septins regulate developmental switching from microdomain to nanodomain coupling of Ca2+ influx to neurotransmitter release at a central synapseNeuron 67:100–115https://doi.org/10.1016/j.neuron.2010.06.003
- Basilar membrane nonlinearity determines auditory nerve rate-intensity functions and cochlear dynamic rangeHearing research 45:203–219
- Olivocochlear innervation maintains the normal modiolar-pillar and habenular-cuticular gradients in cochlear synaptic morphologyJ Assoc Res Otolaryngol 15:571–583https://doi.org/10.1007/s10162-014-0462-z
- Resolution of subcomponents of synaptic release from postsynaptic currents in rat hair-cell/auditory-nerve fiber synapsesJ Neurophysiol 125:2444–2460https://doi.org/10.1152/jn.00450.2020
Article and author information
Author information
Version history
- Sent for peer review:
- Preprint posted:
- Reviewed Preprint version 1:
- Reviewed Preprint version 2:
- Reviewed Preprint version 3:
Copyright
© 2023, Lina María Jaime Tobón & Tobias Moser
This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.
Metrics
- views
- 859
- downloads
- 21
- citations
- 0
Views, downloads and citations are aggregated across all versions of this paper published by eLife.