Abstract Quantal synaptic transmission in vestibular endorgans is glutamatergic. Here, we investigated the vestibular phenotype of deaf Vglut3 −/− (Slc17a8 −/− ) mice from the cellular to behavioral levels. In Vglut3 −/− mice, quantal synaptic transmission in utricular calyces was reduced in rate and amplitude by > 95%. In vivo recordings of spontaneous activity in the vestibular nerve revealed no significant effect of VGLUT3 deletion on afferent rate and regularity, suggesting a divergent underlying mechanism compared to the silent Vglut3 −/− auditory nerve. In behavioral studies, Vglut3 −/− mice did not exhibit considerable sensorimotor or balance deficits. Type-II vestibular hair cells (VHCs) in Vglut3 +/+ mice were strongly immunoreactive for VGLUT3, while type-I VHCs showed weak immunoreactivity. Collectively, these data support the view that non-quantal transmission is the predominant mode of neurotransmission between type-I VHCs and vestibular calyceal afferent neurons. We propose that non-quantal transmission is sufficient to support vestibular nerve physiology and behavioral function in Vglut3 −/− mice.

Noise-induced hearing loss (NIHL) is a major public health problem caused by damage to cochlear hair cells, synapses, and spiral ganglion neurons (SGNs). Since effective treatments are lacking, we investigated cellular stress responses induced by moderate and loud noise in a mouse model of cochlear synaptopathy. RNA sequencing and spatial transcriptomics revealed that noise exposure elicited a robust but transient upregulation of endoplasmic reticulum chaperones and proteasome subunits in SGNs and their supporting cells. To target this response, we administered TRC051384, a small-molecule activator of the heat shock transcription factor Hsf1, prior to noise exposure. TRC051384 crossed the blood-labyrinth barrier and reached the cochlea, induced heat shock protein gene expression, and restored ubiquitin-proteasome function in SGNs. Notably, TRC051384 treatment enhanced auditory brainstem response threshold recovery, preserved Wave I amplitudes, and maintained ribbon synapse density. Together with the existing literature, these findings identify proteotoxic stress in spiral ganglion neurons as a contributor to noise-induced hearing loss and support pharmacological activation of HSF1 as a promising therapeutic strategy.

Quantal synaptic transmission in vestibular end-organs is glutamatergic. Although genetic deletion of Slc17a8 (termed Vglut3 ) leads to deafness in mice, the dependence of vestibular function on VGLUT3-mediated quantal transmission is unknown. Here, we investigated the vestibular phenotype of Vglut3-/- mice at the cellular, systems, and behavioral levels. The type-II vestibular hair cells (VHCs) in Vglut3+/+ mice were strongly immunoreactive for VGLUT3, while type-I VHCs showed poor immunoreactivity. In Vglut3-/- mice quantal synaptic transmission in utricular calyces was reduced in rate and amplitude by > 95%. In vivo recordings of spontaneous activity in the vestibular nerve revealed similar action potential rates and regularity in Vglut3+/+ and Vglut3-/- mice, suggesting a divergent underlying mechanism compared to the silent Vglut3-/- auditory nerve. In behavioral studies, Vglut3-/- mice did not exhibit considerable sensorimotor or balance deficits. Collectively, these data support the view that non-quantal transmission is the predominant mode of neurotransmission between type I VHCs and vestibular calyceal afferent neurons. We propose that non-quantal transmission alone underlies the apparently normal vestibular nerve physiology and behavioral function in Vglut3-/- mice.

AMPA-type glutamate receptors (AMPAR) mediate excitatory cochlear transmission. However, the unique roles of AMPAR subunits are unresolved. Lack of subunit GluA3 ( Gria3 KO ) in male mice reduced cochlear output by 8-weeks of age. Since Gria3 is X-linked and considering sex differences in hearing vulnerability, we hypothesized accelerated presbycusis in Gria3 KO females. Here, auditory brainstem responses (ABR) were similar in 3-week-old female Gria3 WT and Gria3 KO mice. However, when raised in ambient sound, ABR thresholds were elevated and wave-1 amplitudes were diminished at 5-weeks and older in Gria3 KO . In contrast, these metrics were similar between genotypes when raised in quiet. Paired synapses were similar in number, but lone ribbons and ribbonless synapses were increased in female Gria3 KO mice in ambient sound compared to Gria3 WT or to either genotype raised in quiet. Synaptic GluA4:GluA2 ratios increased relative to Gria3 WT , particularly in ambient sound, suggesting an activity-dependent increase in calcium-permeable AMPARs in Gria3 KO . Swollen afferent terminals were observed by 5-weeks only in Gria3 KO females reared in ambient sound. We propose that lack of GluA3 induces sex-dependent vulnerability to AMPAR-mediated excitotoxicity.

Our sense of hearing is mediated by cochlear hair cells, localized within the sensory epithelium called the organ of Corti. There are two types of hair cells in the cochlea, which are organized in one row of inner hair cells and three rows of outer hair cells. Each cochlea contains a few thousands of hair cells, and their survival is essential for our perception of sound because they are terminally differentiated and do not regenerate after insult. It is often desirable in hearing research to quantify the number of hair cells within cochlear samples, in both pathological conditions, and in response to treatment. However, the sheer number of cells along the cochlea makes manual quantification impractical. Machine learning can be used to overcome this challenge by automating the quantification process but requires a vast and diverse dataset for effective training. In this study, we present a large collection of annotated cochlear hair-cell datasets, labeled with commonly used hair-cell markers and imaged using various fluorescence microscopy techniques. The collection includes samples from mouse, human, pig and guinea pig cochlear tissue, from normal conditions and following in-vivo and in-vitro ototoxic drug application. The dataset includes over 90'000 hair cells, all of which have been manually identified and annotated as one of two cell types: inner hair cells and outer hair cells. This dataset is the result of a collaborative effort from multiple laboratories and has been carefully curated to represent a variety of imaging techniques. With suggested usage parameters and a well-described annotation procedure, this collection can facilitate the development of generalizable cochlear hair cell detection models or serve as a starting point for fine-tuning models for other analysis tasks. By providing this dataset, we aim to supply other groups within the hearing research community with the opportunity to develop their own tools with which to analyze cochlear imaging data more fully, accurately, and with greater ease.

Objective. To describe the collection methods for perilymph fluid biopsy during cochlear implantation, detect levels of amyloid beta 42 and 40 (A beta(42) and A beta(40)), and total tau (tTau) analytes with a high-precision assay, to compare these levels with patient age and Montreal Cognitive Assessment (MoCA) scores, and explore potential mechanisms and relationships with otic pathology. Study design. Prospective study. Setting. Tertiary referral center. Methods. Perilymph was collected from 25 patients using polyimide tubing to avoid amyloid adherence to glass, and analyzed with a single-molecule array advanced digital enzyme-linked immunosorbent assay platform for A beta(40), A beta(42), and tTau. Cognition was assessed by MoCA. Results. Perilymph volumes ranged from similar to 1 to 13 mu L, with analyte concentrations spanning 2.67 to 1088.26 pg/mL. All samples had detectable levels of tTau, A beta(40), and A beta(42), with a significant positive correlation between A beta(42) and A beta(40) levels. Levels of A beta(42), A beta(40), and tTau were positively correlated with age, while MoCA scores were inversely correlated with age. tTau and A beta(42)/A beta(40)-ratios were significantly correlated with MoCA scores. Conclusion. Alzheimer's disease-associated peptides A beta(42), A beta(40), and tau analytes are detectable in human perilymph at levels approximately 10-fold lower than those found in cerebrospinal fluid (CSF). These species increase with age and correlate with cognitive impairment indicators, suggesting their potential utility as biomarkers for cognitive impairment in patients undergoing cochlear implantation. Future research should investigate the origin of these analytes in the perilymph and their potential links to inner ear pathologies and hearing loss, as well as their relationships to CSF and plasma levels in individuals.

Glutamate excitotoxicity in the cochlea Mark Rutherford, Associate Professor at Washington University School of Medicine, discusses the burden and causes of hearing dysfunction, as well as the possible solutions for mitigating glutamate excitotoxicity. The cochlea, the snail-shaped organ of the inner ear located deep in the skull, contains approximately 15,000 sound receptors called sensory ‘hair cells’ and approximately 50,000 nerve fibers sending sound information to the brain. Although that may seem like a lot, it is not compared to the approximately 100 billion neurons in the brain. These hair cells and nerve fibers are all we get for life, and they are gradually lost over time due to aging. Unfortunately, environmental factors accelerate cochlear decline over time. One of those environmental factors is loud sounds.Today, we live in an environment of sound at levels that far exceed those under which our species evolved, especially those of us with noisy occupations or recreational activities. For example, military personnel are particularly vulnerable to hearing loss, making hearing disorders (including tinnitus) the number one disability claim of the U.S. Veterans Administration.

Cochlear sound encoding depends on α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid receptors (AMPARs), but reliance on specific pore-forming subunits is unknown. With 5-week-old male C57BL/6J Gria3-knockout mice (i.e., subunit GluA3KO) we determined cochlear function, synapse ultrastructure, and AMPAR molecular anatomy at ribbon synapses between inner hair cells (IHCs) and spiral ganglion neurons. GluA3KO and wild-type (GluA3WT) mice reared in ambient sound pressure level (SPL) of 55–75 dB had similar auditory brainstem response (ABR) thresholds, wave-1 amplitudes, and latencies. Postsynaptic densities (PSDs), presynaptic ribbons, and synaptic vesicle sizes were all larger on the modiolar side of the IHCs from GluA3WT, but not GluA3KO, demonstrating GluA3 is required for modiolar–pillar synapse differentiation. Presynaptic ribbons juxtaposed with postsynaptic GluA2/4 subunits were similar in quantity, however, lone ribbons were more frequent in GluA3KO and GluA2-lacking synapses were observed only in GluA3KO. GluA2 and GluA4 immunofluorescence volumes were smaller on the pillar side than the modiolar side in GluA3KO, despite increased pillar-side PSD size. Overall, the fluorescent puncta volumes of GluA2 and GluA4 were smaller in GluA3KO than GluA3WT. However, GluA3KO contained less GluA2 and greater GluA4 immunofluorescence intensity relative to GluA3WT (threefold greater mean GluA4:GluA2 ratio). Thus, GluA3 is essential in development, as germline disruption of Gria3 caused anatomical synapse pathology before cochlear output became symptomatic by ABR. We propose the hearing loss in older male GluA3KO mice results from progressive synaptopathy evident in 5-week-old mice as decreased abundance of GluA2 subunits and an increase in GluA2-lacking, GluA4-monomeric Ca2+-permeable AMPARs.

The encoding of acoustic signals in the cochlea depends on α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid receptors (AMPARs), but relatively little is known about their reliance on specific pore-forming subunits. With 5-week-old male GluA3 KO mice, we determined cochlear function, synapse ultrastructure, and AMPAR subunit molecular anatomy at ribbon synapses between inner hair cells (IHCs) and spiral ganglion neurons (SGNs). GluA3 KO and wild-type ( GluA3 WT ) mice reared in ambient sound pressure level (SPL) of 55-75 dB had similar ABR thresholds, wave-1 amplitudes, and latencies. Ultrastructurally, the IHC modiolar-pillar differences in presynaptic ribbon size and shape, and synaptic vesicle size seen in GluA3 WT were diminished or reversed in GluA3 KO . The quantity of paired synapses (presynaptic ribbons juxtaposed with postsynaptic GluA2 and GluA4) was similar, however, GluA2-lacking synapses (ribbons paired with GluA4 but not GluA2) were observed only in GluA3 KO . SGNs of GluA3 KO mice had AMPAR arrays of smaller overall volume, containing less GluA2 and greater GluA4 immunofluorescence intensity relative to GluA3 WT (3-fold difference in mean GluA4:GluA2 ratio). The expected modiolar-pillar gradient in ribbon volume was observed in IHCs of GluA3 WT but not GluA3 KO . Unexpected modiolar-pillar gradients in GluA2 and GluA4 volume were present in GluA3 KO . GluA3 is essential to the morphology and molecular composition of IHC-ribbon synapses. We propose the hearing loss seen in older male GluA3 KO mice results from progressive synaptopathy evident in 5-week-old mice as increased abundance of GluA2-lacking, GluA4 monomeric, Ca 2+ -permeable AMPARs.

Hearing depends on precise synaptic transmission between cochlear inner hair cells and spiral ganglion neurons through afferent ribbon synapses.Neuroligins (Nlgns) facilitate synapse maturation in the brain, but they have gone unstudied in the cochlea. We report Nlgn3 and Nlgn1 knockout (KO) cochleae have fewer ribbon synapses and have impaired hearing. Nlgn3 KO is more vulnerable to noise trauma with limited activity at high frequencies one day after noise. Furthermore, Nlgn3 KO cochleae have a 5-fold reduction in synapse number compared to wild type after two weeks of recovery. Double KO cochlear phenotypes are more prominent than the KOs, for example, 5-fold smaller synapses, 25% reduction in synapse density, and 30% less synaptic output. These observations indicate Nlgn3 and Nlgn1 are essential to cochlear ribbon synapse maturation and function.

Ribbon-class synapses in the ear achieve analog to digital transformation of a continuously graded membrane potential to all-or-none spikes. In mammals, several auditory nerve fibres (ANFs) carry information from each inner hair cell (IHC) to the brain in parallel. Heterogeneity of transmission among synapses contributes to the diversity of ANF sound-response properties. In addition to the place code for sound frequency and the rate code for sound level, there is also a temporal code. In series with cochlear amplification and frequency tuning, neural representation of temporal cues over a broad range of sound levels enables auditory comprehension in noisy multi-speaker settings. The IHC membrane time constant introduces a low-pass filter that attenuates fluctuations of the receptor potential above 1-2 kHz. The ANF spike generator adds a high-pass filter via its depolarization-rate threshold that rejects slow changes in the postsynaptic potential and its phasic response property that ensures one spike per depolarization. Synaptic transmission involves several stochastic subcellular processes between IHC depolarization and ANF spike generation, introducing delay and jitter that limits the speed and precision of spike timing. ANFs spike at a preferred phase of periodic sounds in a process called phase-locking that is limited to frequencies below a few kilohertz by both the IHC receptor potential and the jitter in synaptic transmission. During phase-locking to periodic sounds of increasing intensity, faster and facilitated activation of synaptic transmission and spike generation may be offset by presynaptic depletion of synaptic vesicles, resulting in relatively small changes in response phase. Here we review encoding of spike-timing at cochlear ribbon synapses.

Auditory nerve fibers (ANFs) innervating the same inner hair cell (IHC) may have identical frequency tuning but different sound response properties. In cat and guinea pig, ANF response properties correlate with afferent synapse morphology and position on the IHC, suggesting a causal structure-function relationship. In mice, this relationship has not been fully characterized. Here we measured the emergence of synaptic morphological heterogeneities during maturation of the C57BL/6J mouse cochlea by comparing postnatal day 17 (p17, ∼3 days after hearing onset) with p34, when the mouse cochlea is mature. Using serial block face scanning electron microscopy and three-dimensional reconstruction we measured the size, shape, vesicle content, and position of 70 ribbon synapses from the mid-cochlea. Several features matured over late postnatal development. From p17 to p34, presynaptic densities (PDs) and post-synaptic densities (PSDs) became smaller on average (PDs: 0.75 to 0.33; PSDs: 0.58 to 0.31 μm2) and less round as their short axes shortened predominantly on the modiolar side, from 770 to 360 nm. Membrane-associated synaptic vesicles decreased in number from 53 to 30 per synapse from p17 to p34. Anatomical coupling, measured as PSD to ribbon distance, tightened predominantly on the pillar side. Ribbons became less spherical as long-axes lengthened only on the modiolar side of the IHC, from 372 to 541 nm. A decreasing gradient of synaptic ribbon size along the modiolar-pillar axis was detected only at p34 after aligning synapses of adjacent IHCs to a common reference frame (median volumes in nm3 × 106: modiolar 4.87; pillar 2.38). The number of ribbon-associated synaptic vesicles scaled with ribbon size (range 67 to 346 per synapse at p34), thus acquiring a modiolar-pillar gradient at p34, but overall medians were similar at p17 (120) and p34 (127), like ribbon surface area (0.36 vs. 0.34 μm2). PD and PSD morphologies were tightly correlated to each other at individual synapses, more so at p34 than p17, but not to ribbon morphology. These observations suggest that PDs and PSDs mature according to different cues than ribbons, and that ribbon size may be more influenced by cues from the IHC than the surrounding tissue.

Noise-induced hearing loss has gained relevance as one of the most common forms of hearing impairment. The anatomical correlates of hearing loss, principally cell damage and/or death, are relatively well-understood histologically. However, much less is known about the physiological aspects of damaged, surviving cells. Here we addressed the functional consequences of noise exposure on the capacity of inner hair cells (IHCs) to release synaptic vesicles at synapses with spiral ganglion neurons (SGNs). Mice of either sex at postnatal day (P) 15–16 were exposed to 1–12 kHz noise at 120 dB sound pressure level (SPL), for 1 h. Exocytosis was measured by tracking changes in membrane capacitance (ΔCm) from IHCs of the apical cochlea. Upon IHC depolarization to different membrane potentials, ΔCm showed the typical bell-shaped curve that mirrors the voltage dependence of Ca2+ influx, in both exposed and unexposed cells. Surprisingly, from IHCs at 1-day after exposure (d.a.e.), we found potentiation of exocytosis at the peak of the bell-shaped curve. The increase in exocytosis was not accompanied by changes in whole-cell Ca2+ influx, suggesting a modification in coupling between Ca2+ channels and synaptic vesicles. Consistent with this notion, noise exposure also changed the Ca2+-dependence of exocytosis from linear to supralinear. Noise exposure did not cause loss of IHCs, but did result in a small reduction in the number of IHC-SGN synapses at 1-d.a.e. which recovered by 14-d.a.e. In contrast, a strong reduction in auditory brainstem response wave-I amplitude (representing synchronous firing of SGNs) and distortion product otoacoustic emissions (reflecting outer hair cell function) indicated a profound hearing loss at 1- and 14-d.a.e. To determine the role of glutamate release in the noise-induced potentiation of exocytosis, we evaluated vesicular glutamate transporter-3 (Vglut3) knock-out (KO) mice. Unlike WT, IHCs from Vglut3KO mice showed a noise-induced reduction in ΔCm and Ca2+ influx with no change in the Ca2+-dependence of exocytosis. Together, these results indicate that traumatic noise exposure triggers changes of IHC synaptic function including a Vglut3-dependent potentiation of exocytosis.

Hearing depends on glutamatergic synaptic transmission mediated by α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptors (AMPARs). AMPARs are tetramers, where inclusion of the GluA2 subunit reduces overall channel conductance and Ca2+ permeability. Cochlear afferent synapses between inner hair cells (IHCs) and auditory nerve fibers (ANFs) contain the AMPAR subunits GluA2, 3, and 4. However, the tetrameric complement of cochlear AMPAR subunits is not known. It was recently shown in mice that chronic intracochlear delivery of IEM-1460, an antagonist selective for GluA2-lacking AMPARs [also known as Ca2+-permeable AMPARs (CP-AMPARs)], before, during, and after acoustic overexposure prevented both the trauma to ANF synapses and the ensuing reduction of cochlear nerve activity in response to sound. Surprisingly, baseline measurements of cochlear function before exposure were unaffected by chronic intracochlear delivery of IEM-1460. This suggested that cochlear afferent synapses contain GluA2-lacking CP-AMPARs alongside GluA2-containing Ca2+-impermeable AMPA receptors (CI-AMPARs), and that the former can be antagonized for protection while the latter remain conductive. Here, we investigated hearing function in the guinea pig during acute local or systemic delivery of CP-AMPAR antagonists. Acute intracochlear delivery of IEM-1460 or systemic delivery of IEM-1460 or IEM-1925 reduced the amplitude of the ANF compound action potential (CAP) significantly, for all tone levels and frequencies, by > 50% without affecting CAP thresholds or distortion product otoacoustic emissions (DPOAE). Following systemic dosing, IEM-1460 levels in cochlear perilymph were ~ 30% of blood levels, on average, consistent with pharmacokinetic properties predicting permeation of the compounds into the brain and ear. Both compounds were metabolically stable with half-lives >5 h in vitro, and elimination half-lives in vivo of 118 min (IEM-1460) and 68 min (IEM-1925). Heart rate monitoring and off-target binding assays suggest an enhanced safety profile for IEM-1925 over IEM-1460. Compound potency on CAP reduction (IC50 ~ 73 μM IEM-1460) was consistent with a mixture of GluA2-lacking and GluA2-containing AMPARs. These data strongly imply that cochlear afferent synapses of the guinea pig contain GluA2-lacking CP-AMPARs. We propose these CP-AMPARs may be acutely antagonized with systemic dosing, to protect from glutamate excitotoxicity, while transmission at GluA2-containing AMPARs persists to mediate hearing during the protection.

Article Figures and data Abstract Editor's evaluation Introduction Results Discussion Materials and methods Data availability References Decision letter Author response Article and author information Metrics Abstract Acoustic overexposure and aging can damage auditory synapses in the inner ear by a process known as synaptopathy. These insults may also damage hair bundles and the sensory transduction apparatus in auditory hair cells. However, a connection between sensory transduction and synaptopathy has not been established. To evaluate potential contributions of sensory transduction to synapse formation and development, we assessed inner hair cell synapses in several genetic models of dysfunctional sensory transduction, including mice lacking transmembrane channel-like (Tmc) 1, Tmc2, or both, in Beethoven mice which carry a dominant Tmc1 mutation and in Spinner mice which carry a recessive mutation in transmembrane inner ear (Tmie). Our analyses reveal loss of synapses in the absence of sensory transduction and preservation of synapses in Tmc1-null mice following restoration of sensory transduction via Tmc1 gene therapy. These results provide insight into the requirement of sensory transduction for hair cell synapse development and maturation. Editor's evaluation Deafness is often caused by a defect in mechanotransduction. Lately it has become clear that synaptopathy, a defect at the first synapse in the auditory pathway, also causes hearing loss. Here the authors show that synapses can be lost following a loss of hair cell mechanotransduction, but that restoration of mechanotransduction can prevent the synaptic loss. These results are important for understanding hearing loss and restoration. https://doi.org/10.7554/eLife.69433.sa0 Decision letter eLife's review process Introduction Acoustic overexposure and aging can cause significant loss of synaptic connections between inner hair cells (IHCs) and the afferent fibers of spiral ganglion neurons (SGNs), which function to relay information to the brain via the eighth cranial nerve (Kujawa and Liberman, 2009; Kujawa and Liberman, 2006; Sergeyenko et al., 2013). Loss of IHC-SGN synapses, known as cochlear synaptopathy, is evident following noise exposure producing both temporary threshold shifts (TTS) and permanent threshold shifts (PTS), with rapid loss of up to 40–50% of synapses and gradual degeneration of SGNs even in the absence of hair cell death (Kujawa and Liberman, 2009). Aging mice show similar synaptopathy, with a steady loss of synapses and a delayed but similar decrease in the number of SGNs that precedes age-related hair cell loss (Sergeyenko et al., 2013). Damage to stereocilia, the mechanosensory microvilli, is also observed in aging and noise-exposed cochleas. Following PTS-inducing noise exposure, irreversible disarray, fusion, and loss of IHC stereocilia are visible even in the absence of hair cell death (Wang et al., 2002). Similarly, IHC stereocilia in aging mice undergo a number of changes including fusion, elongation, and internalization without obvious hair cell loss (Bullen et al., 2019). Stereocilia defects compromise the integrity of the mechanosensory transduction apparatus, resulting in absent or impaired sensory transduction and thus contribute to diminished hearing sensitivity (Pickles et al., 1987; Assad et al., 1991). Despite the concurrence of mechanosensory insult and cochlear synaptopathy in noise exposure and aging, impaired sensory transduction has not been implicated as a mechanism contributing to the loss of hair cell synapses. To begin to address the contributions of sensory transduction to the development and maturation of IHC-SGN synapses, we evaluated IHC ribbon synapses across multiple developmental timepoints in five genetic models with disrupted sensory transduction, including mice lacking Tmc1, Tmc2, or both (Tmc1Δ/Δ, Tmc2Δ/Δ, Tmc1Δ/Δ;Tmc2Δ/Δ), Beethoven mice which carry a dominant mutation in Tmc1 (Bth) and Spinner mice (Tmiesr), which carry a mutation in Tmie. TMC proteins form the pore of hair cell transduction channels (Pan et al., 2018) and TMIE is a necessary component of the hair cell mechanosensory transduction complex (Zhao et al., 2014). Mutations in these proteins cause transduction dysfunction and hearing loss in mice and humans (Kurima et al., 2003; Vreugde et al., 2002; Kawashima et al., 2011; Naz et al., 2002; Mitchem et al., 2002; Zhao et al., 2014). Our analyses reveal an unanticipated and complex relationship between sensory transduction, synaptogenesis, and synaptopathy. In mouse models with genetic disruption of sensory transduction, we find that IHC synapses undergo exuberant synaptogenesis during the first postnatal week and that the number of synapses declines drastically over the following few weeks. We also investigated whether restoration of sensory transduction, using an established Tmc1 gene therapy strategy capable of targeting nearly 100% of IHCs (Lee et al., 2020; Wu et al., 2021), preserved normal synaptic development and maturation. We report that Tmc1 gene therapy in Tmc1Δ/Δ mice preserves synapses in mature mice, and that synapse counts were correlated with recovery of ABR thresholds. Together, these data provide insight into synaptic changes associated with sensory transduction and suggest that dysfunction of sensory transduction may contribute to cochlear synaptopathy. Results Tmc deletion alters synapse development and maturation To determine whether initial ribbon formation was influenced by the presence of functional mechanosensory transduction channels, the number of CtBP2-positive puncta per IHC was quantified in five wild-type (WT) and six Tmc1Δ/Δ;Tmc2Δ/Δ mice at postnatal day 2 (P2) from cochlear regions corresponding to 8, 11.3, 16, 22.6, and 32 kHz. Counts of CtBP2-positive puncta were similar in both groups of mice and to those previously reported for WT C57/BL6 mice at P2 (Huang et al., 2012). No significant differences in counts were observed between WT and Tmc1Δ/Δ;Tmc2Δ/Δ groups across all frequency regions (Figure 1, Supplementary file 1), suggesting the initial formation of presynaptic ribbon precursors is independent of the acquisition of sensory transduction, which begins at the base at P0 and progresses tonotopically toward the apex by P4 (Lelli et al., 2009). Figure 1 Download asset Open asset CtBP2+ puncta counts in Tmc1Δ/Δ;Tmc2Δ/Δ mice lacking sensory transduction are similar to those in wild-type (WT) mice at postnatal day 2 (P2). (A) Representative 3D projections of confocal z-stacks of P2 WT and Tmc1Δ/Δ;Tmc2Δ/Δ (B) inner hair cells (IHCs) from the 16 kHz region. Scale bar: 7 µm. The tissue was immunostained with anti-Myosin7a (blue) and anti-CtBP2 (red). (C) To determine a mean number of ribbon precursors/IHC, the total number of CtBP2+ puncta from 8 to 10 IHCs was counted for each frequency region. Individual points represent counts from one mouse. Data from WT (black; n = 5) and Tmc1Δ/Δ; Tmc2Δ/Δ (red; n = 5–6) groups are illustrated. Bold lines indicate mean ± SD. Figure 1—source data 1 CtBP2+ puncta counts in Tmc1Δ/Δ;Tmc2Δ/Δ mice lacking sensory transduction are similar to those in wild-type (WT) mice at postnatal day 2 (P2). https://cdn.elifesciences.org/articles/69433/elife-69433-fig1-data1-v2.xlsx Download elife-69433-fig1-data1-v2.xlsx To assess the consequences of Tmc1 and Tmc2 deletion on synaptic development, ribbon synapses were further evaluated at P7, P14, and P28 in WT and Tmc1Δ/Δ;Tmc2Δ/Δ mice. These three timepoints were selected to encompass events crucial to the maturation of the inner ear with P7 following acquisition and maturation of IHC sensory transduction, P14 following hearing onset in mice, and P28 reflecting the nearly mature organ of Corti and auditory function. For each timepoint and cochlear region, the average number of synapses per IHC was estimated by counting the total number of presynaptic CtBP2 juxtaposed to postsynaptic GluA2 puncta, divided by the number of IHCs sampled in each region (Figure 2). Figure 2 Download asset Open asset Synapse counts are elevated at postnatal day 7 (P7) and reduced at P28 in Tmc1Δ/Δ;Tmc2Δ/Δ mice relative to wild-type (WT) mice. (A–B) Representative image of P7 WT and Tmc1Δ/Δ;Tmc2Δ/Δ inner hair cells (IHCs) from 16 kHz region immunostained for anti-Myosin7a (gray), anti-CtBP2 (red), and anti-GluA2 (green). Higher magnification images are shown below. (C–D) P28 WT and Tmc1Δ/Δ;Tmc2Δ/Δ IHCs from 16 kHz region. Scale bars: 10 µm (upper) and 5 µm (lower). While ribbon counts were unaltered in P2 Tmc1Δ/Δ;Tmc2Δ/Δ mice, the lack of Tmc1 and Tmc2 resulted in elevated synapse counts at P7 across all frequency regions relative to those in WT mice (Figure 3A, Supplementary file 2A). There were ~43% more synapses in Tmc1Δ/Δ;Tmc2Δ/Δ mice, suggesting there may be a correlation between the developmental acquisition of sensory transduction in IHCs during the first postnatal week and the developmental increase in the number of ribbons and postsynaptic densities. In WT mice, ~50% of ribbon synapses are lost between the end of the first and second postnatal weeks (Sundaresan et al., 2016). This decrease is thought to result from pruning, refinement, and fusion of ribbons and postsynaptic densities (Michanski et al., 2019; Sundaresan et al., 2016; Wong et al., 2014; Huang et al., 2012; Sendin et al., 2007). Consistent with these findings, we found a 46–57% reduction in synapses counts in WT mice between P7 and P14 between 8 and 32 kHz. In Tmc1Δ/Δ;Tmc2Δ/Δ mice, a more drastic decrease in synapse counts was evident during the second postnatal week. The number of synapses decreased by 67 to 73% at P14 relative to P7 between 8 and 32 kHz (Figure 3A and B). As a result, synapse numbers at P14 did not differ significantly between WT and Tmc1Δ/Δ;Tmc2Δ/Δ groups except at the 32 kHz region where Tmc1Δ/Δ;Tmc2Δ/Δ mice exhibited ~27% fewer synapse counts (Figure 3B, Supplementary file 2B). Figure 3 Download asset Open asset Synapse counts were elevated at postnatal day 7 (P7) in the absence of both Tmc1 and Tmc2 and diminished at P28 in the absence of Tmc1 and Tmc2 or Tmc1 alone. (A–C) The mean number of synapses/inner hair cell (IHC) was calculated for each frequency region. Data from wild-type (WT) (black), Tmc1Δ/Δ;Tmc2Δ/Δ (red), Tmc1Δ/Δ (blue), and Tmc2Δ/Δ (dark yellow) groups are shown. Individual points represent counts from one mouse. Temporal changes in synapse counts differed by genotype (two-way ANOVA; p < 0.001 based on two-way interaction between genotype and timepoints for all frequencies; Supplementary file 3A). However, genotype-specific trajectories of synaptic development did not vary by frequency (three-way ANOVA; p = 0.73 based on three-way interaction between group, time, frequency; Supplementary file 3A). Bolded lines depict mean ± SD. Black horizontal bars and asterisks represent statistically significant differences between group means (multiple pairwise comparisons, *p < 0.05, **p < 0.01, ***p < 0.001; p values are listed in Supplementary file 2A-C). Number of cochleas: 4–9 WT, 5–12 Tmc1Δ/Δ;Tmc2Δ/Δ, 4 Tmc1Δ/Δ, 4–5 Tmc2Δ/Δ. Figure 3—source data 1 Synapse counts were elevated at postnatal day 7 (P7) in the absence of both Tmc1 and Tmc2 and diminished at P28 in the absence of Tmc1 and Tmc2 or Tmc1 alone. https://cdn.elifesciences.org/articles/69433/elife-69433-fig3-data1-v2.xlsx Download elife-69433-fig3-data1-v2.xlsx Following hearing onset in WT mice, ribbon synapse counts remain stable into adulthood (Michanski et al., 2019; Huang et al., 2012). As expected, our WT synapse counts did not differ significantly between P14 and P28 in any of the frequency regions examined (Figure 3B and C). However, synapse counts were markedly decreased in P28 Tmc1Δ/Δ;Tmc2Δ/Δ mice relative to WT mice and relative to P14 Tmc1Δ/Δ;Tmc2Δ/Δ numbers at all frequencies (Figure 3C, Supplementary file 2C). Though synapse counts remain stable in WT mice post-hearing onset, changes in the distribution of ribbon sizes and continued reduction in the sizes of presynaptic/postsynaptic densities are observed until their full development around P34 (Payne et al., 2021). As with the abnormally elevated synapse counts at P7 in Tmc1Δ/Δ;Tmc2Δ/Δ mice, the rapid loss of more mature ribbon synapses further suggests a role for sensory transduction in the development of synapses. Synapse counts in Tmc single knockouts differ from those of double knockouts Since Tmc2 expression coincides with the developmental onset of sensory transduction in OHCs (Lelli et al., 2009; Kawashima et al., 2011) and IHCs (Pan et al., 2013) at P0 in the base and P2-P4 in the apex and is followed several days later by expression of Tmc1, we wondered whether genetic deletion of Tmc2 would cause a delay in the developmental pattern of synapse development. In addition, to determine whether the changes observed in Tmc1Δ/Δ;Tmc2Δ/Δ synapses were due to deletion of Tmc1, we examined synapses in single knockouts of Tmc1 or Tmc2. At P7 and P14, synapse counts in neither Tmc1Δ/Δ nor Tmc2Δ/Δ mice differed significantly from those of WT mice across all frequencies (Figure 3A and B, Supplementary file 2A and B), suggesting that expression of either Tmc1 or Tmc2 was sufficient for acquisition of normal synapse counts and juxtaposition of presynaptic ribbons and their postsynaptic densities. However, it remains possible that synaptic function and development are altered in mice lacking Tmc1, Tmc2, or both in ways that cannot be captured by immunohistochemistry and synapse counts alone. At P28, there was a significant deviation in the number of synapses observed between Tmc1Δ/Δ and Tmc2Δ/Δ groups. Synapse counts in Tmc1Δ/Δ mice were significantly reduced compared to WT mice and similar to those of Tmc1Δ/Δ;Tmc2Δ/Δ mice at all frequencies (Figure 3C, Supplementary file 2C). Synapse counts in Tmc2Δ/Δ mice, on the other hand, were not different from those in WT mice at any of the frequencies (Figure 3C, Supplementary file 2C). Thus, it appears the absence of Tmc1 specifically, in Tmc1Δ/Δ and Tmc1Δ/Δ;Tmc2Δ/Δ mice, accounts for the loss of synapses after hearing onset. Tmie deletion leads to loss of synapses While the results obtained from Tmc1Δ/Δ, Tmc2Δ/Δ, and Tmc1Δ/Δ;Tmc2Δ/Δ mice implicated a role for sensory transduction in the development and maturation of ribbon synapses, the possibility of ribbon synapses being affected by targeted deletion of Tmc1, Tmc2, or both via a mechanism unrelated to inhibition of sensory transduction remained a possibility. To determine whether the change in synapse counts was a specific consequence of Tmc deletion or a general consequence of the loss of sensory transduction, we investigated deletion of a different gene known to cause loss of hair cell sensory transduction. Synapses were counted in Spinner mice, which carry a spontaneous nonsense mutation at the Tmie locus (Tmiesr) (Mitchem et al., 2002). TMIE encodes transmembrane inner ear protein, which is an essential component of the transduction channel complex (Gleason et al., 2009; Shen et al., 2008; Zhao et al., 2014). While loss of functional TMIE in homozygous Spinner mice causes complete absence of sensory transduction (Zhao et al., 2014) and mislocalization of TMC proteins (Pacentine and Nicolson, 2019; Cunningham et al., 2020), it does not affect TMC expression (Cunningham et al., 2020). As shown in Figure 4A, Tmiesr synapse counts at P7 were significantly elevated compared to WT mice. There were no significant differences between Tmiesr and Tmc1Δ/Δ;Tmc2Δ/Δ synapse counts in any frequency region (Figure 4A, Supplementary file 4A). At P28, Tmiesr mice showed decreased synapse counts at all frequencies, with values that did not differ significantly from those in Tmc1Δ/Δ;Tmc2Δ/Δ mice (Figure 4B, Supplementary file 4B). The similarities in synapse counts in Tmc1Δ/Δ;Tmc2Δ/Δ and Tmiesr mice at P7 and P28 provide compelling evidence that the loss of normal synapse development and maturation is a consequence of the absence of sensory transduction and not disruption of a specific gene. Figure 4 Download asset Open asset Loss of sensory transduction, not of Tmc1 and Tmc2 specifically, results in the synaptic differences observed at postnatal day 7 (P7) and P28. (A–C) The mean number of synapses/inner hair cell (IHC) at each frequency region. Data from wild-type (WT) (black), Tmc1Δ/Δ;Tmc2Δ/Δ (red), Tmc1Δ/Δ (blue), Tmiesr (purple), and Tmc1Bth (gold) groups are shown. WT, Tmc1Δ/Δ, and Tmc1Δ/Δ;Tmc2Δ/Δ data are the same as those depicted in Figure 3, reprinted here to facilitate comparison. Individual points represent counts from one mouse. In A and B two-way interactions between genotype and timepoints were statistically significant for all frequencies based on two-way ANOVA (p < 0.001; Supplementary file 3B), suggesting the trajectory of synaptic development varies by genotype. The genotype-specific trajectory of synaptic development did not vary by frequency (three-way ANOVA; p = 0.68 based on three-way interaction between genotype, timepoints, and frequency; Supplementary file 3B). Frequency-specific synapse counts differed by genotype in C (two-way ANOVA; p = 0.002 based on two-way interaction between genotype and frequency; Supplementary file 3B). Bolded lines depict mean ± SD. Black horizontal bars and asterisks represent statistically significant differences in group means (multiple pairwise comparisons, *p < 0.05, **p < 0.01, ***p < 0.001; exact p values listed in Supplementary file 4A-C). Number of cochleas: 4 WT, 5–12 Tmc1Δ/Δ;Tmc2Δ/Δ, 4 Tmc1Δ/Δ, 5–6 Tmiesr, 4–6 Tmc1Bth. Figure 4—source data 1 Loss of sensory transduction, not of Tmc1 and Tmc2 specifically, results in the synaptic differences observed at postnatal day 7 (P7) and P28. https://cdn.elifesciences.org/articles/69433/elife-69433-fig4-data1-v2.xlsx Download elife-69433-fig4-data1-v2.xlsx Altered sensory transduction permeability is inconsequential for synapses To explore whether the changes in synapses were the result of calcium entry through sensory transduction channels, we performed a similar analysis in Beethoven mice with a dominant mutation in Tmc1. IHCs in Tmc1Bth mice (Vreugde et al., 2002) have normal sensory transduction current amplitudes, but reduced calcium permeability relative to control IHCs (Pan et al., 2013; Beurg et al., 2015; Corns et al., 2016). Marcotti et al., 2006 reported Tmc1Bth hair cells have normal resting potentials but altered basolateral currents, leading Pan et al., 2013 to speculate that altered calcium permeability in Tmc1Bth hair cell may have consequences for basolateral hair cell functions, perhaps including synapse development and maturation. No significant differences in P28 synapse counts were evident between Tmc1Bth and WT groups except in the 32 kHz region (Figure 4C, Supplementary file 4C), suggesting reduced calcium entry via sensory transduction channels does not cause a significant loss of ribbon synapses. However, the difference observed in the 32 kHz region suggests a tonotopic role for calcium signaling cannot be discounted. Restoration of sensory transduction preserves auditory synapses Next, we wondered whether gene therapy restoration of sensory transduction could prevent synapse loss. Recent inner ear gene therapy studies have demonstrated the efficacy of a novel utricle injection technique and a synthetic AAV9-PHP.B capsid for transducing nearly 100% of cochlear hair cells with high specificity (Lee et al., 2020) and restoring ABR thresholds with WT Tmc1 injected into Tmc1Δ/Δ mouse inner ears (Wu et al., 2021). Although AAV-mediated gene therapy for otoferlin, a synaptic protein, has demonstrated recovery of auditory synapses (Akil et al., 2019; Al-Moyed et al., 2019), the consequences of Tmc1 gene therapy on auditory synapses has not been assessed. To characterize the consequences of Tmc1 gene therapy on IHC synapses, 12 Tmc1Δ/Δ mice were injected at P1 with AAV9-PHP.B-Tmc1 via the utricle and ABRs/synapses were assessed at P28. At P28, Tmc1Δ/Δ mice are profoundly deaf with no measurable ABR thresholds up to 110 dB sound pressure level (SPL) (Wu et al., 2021). All 12 Tmc1Δ/Δ mice that were injected showed some degree of improvement in auditory function relative to uninjected Tmc1Δ/Δ mice (Figure 5A and B, Supplementary file 5). Recovery was variable across mice, but consistently greater at lower frequencies. On average, the injected mice showed a recovery of 59 dB SPL from 8 to 16 kHz (8 kHz: 61.7 ± 18.3 dB SPL; 11 kHz: 57.9 ± 18.4 dB SPL; 16 kHz: 56.3 ± 25.3 dB SPL). The four best performers out of the 12 injected mice demonstrated mean thresholds of 39 dB SPL across 8–16 kHz (Figure 5B, Supplementary file 5, 8 kHz: 46.3 ± 6.3 dB SPL; 11 kHz: 40.0 ± 7.1 dB SPL; 16 kHz: 31.3 ± 7.5 dB SPL). Figure 5 Download asset Open asset AAV9-PHP.B-Tmc1 restores ABR thresholds in Tmc1Δ/Δ mice. (A) Representative ABR waveforms recorded at postnatal day 28 (P28) using 16 kHz tone bursts at sound pressure levels of increasing 5 dB increments. Waveforms from uninjected Tmc1Δ/Δ mouse (left), three Tmc1Δ/Δ mice injected with AAV9-PHP.B-Tmc1, representing best (green), average (blue), and worst (red) recovery and one wild-type (WT) control (black) are shown. Thresholds determined by the presence of Peak 1 and indicated by bolded, colored traces. (B) ABR thresholds plotted as a function of stimulus frequency for 12 Tmc1Δ/Δ mice injected with AAV9-PHP.B-Tmc1 tested at P28 (gray traces). Mice with the best (green), median (blue), and worst (red) recovery are indicated and correspond to the best, average, and worst traces in (A). Black dotted lines show mean ± SD thresholds from eight previously tested WT mice. Green dotted line shows mean ± SD thresholds from the 12 injected mice. (C) Peak 1 amplitudes measured from 16 kHz ABR waveforms (A) for 12 Tmc1Δ/Δ mice injected with AAV9-PHP.B-Tmc1. Colors correspond to conditions indicated in B. (D) Peak 1 latencies measured from 16 kHz ABR waveforms (A) for 12 Tmc1Δ/Δ mice injected with AAV9-PHP.B-Tmc1. Colors correspond to conditions indicated in B. Figure 5—source data 1 AAV9-PHP. B-Tmc1 restores ABR thresholds in Tmc1Δ/Δ mice. https://cdn.elifesciences.org/articles/69433/elife-69433-fig5-data1-v2.xlsx Download elife-69433-fig5-data1-v2.xlsx Following ABR recording, injected cochleas were immediately harvested and stained for synapses. Average synapse counts were greater in injected Tmc1Δ/Δ mice at all frequency regions (11.3 kHz p = 0.008, 16 kHz p = 0.001, 22.6 kHz p = 0.008, 32 kHz p = 0.012) except at 8 kHz (p = 0.377), than in uninjected Tmc1Δ/Δ mice (Figure 6A, Supplementary file 6A). Synapse counts were variable across injected mice, similar to ABR threshold recoveries. However, we did note a correlation between the level of ABR threshold recovery and the number of synapses/IHC. We calculated pure tone average (PTA) thresholds from 8 to 22.6 kHz and plotted the values as a function of average number of IHC synapses across the corresponding cochlear region for 12 mice (Figure 6B). The data were fit with a linear regression (r = 0.86) which indicated a statistically significant correlation (p = 0.00032) with an improvement of PTA threshold by 13 dB/synapse. The increased synapse counts in Tmc1Δ/Δ mice injected with AAV9-PHP.B-Tmc1 provide a novel line of evidence supporting the potential for Tmc1 gene therapy in preserving ribbon synapses, a requirement for normal auditory function. Figure 6 Download asset Open asset AAV9-PHP.B-Tmc1 preserves synapse counts and ribbon volume distributions in Tmc1Δ/Δ mice. (A) The mean number of synapses/inner hair cell (IHC) was counted at each frequency region from postnatal day 28 (P28) wild-type (WT) (black), Tmc1Δ/Δ (blue), and injected Tmc1Δ/Δ (green) mice. WT and Tmc1Δ/Δ ribbon counts are the same as those depicted in Figure 3. Individual points represent counts from one mouse. Frequency-specific synapse counts did not differ by group (two-way ANOVA; p = 0.49 based on two-way interaction between genotype and frequency; Supplementary file 3C). Bolded lines indicate mean ± SD. Black horizontal bars and asterisks represent statistically significant differences in group means (multiple pairwise comparisons, *p < 0.05, **p < 0.01, ***p < 0.001). Number of cochleas: 4 WT, 4 Tmc1Δ/Δ cochleas, 11–12 injected Tmc1Δ/Δ cochleas. (B) Pure tone average (PTA) thresholds were calculated for frequencies between 8 and 22.6 kHz for 12 Tmc1Δ/Δ mice injected with 1 µL AAV9-PHP.B-Tmc1 at P1. PTA thresholds were plotted as a function of the mean synapses/IHC (circles) based on synapse counts from corresponding cochlear regions. The data were fitted with a linear regression (red line) that had a slope of 13 dB/synapse, a correlation coefficient of 0.86; p = 0.00032. (C) Histograms showing distributions of ribbon volumes from confocal z-stacks plotted for each frequency region. Scale bars indicate volume counts on the Y-axis and ribbon volumes in µm3 on the X-axis. Data were obtained from four P28 WT (gray) cochleas, four Tmc1Δ/Δ (blue) cochleas, and four injected Tmc1Δ/Δ (green) cochleas for each of the 8–32 kHz regions. Figure 6—source data 1 AAV9-PHP. B-Tmc1 preserves synapse counts and ribbon volume distributions in Tmc1Δ/Δ mice. https://cdn.elifesciences.org/articles/69433/elife-69433-fig6-data1-v2.xlsx Download elife-69433-fig6-data1-v2.xlsx Lastly, we wondered whether the distribution of ribbon volumes was altered by loss and recovery of sensory transduction. Confocal z-stacks of cochleas stained with GluA2 and CtBP2 were acquired using Zeiss Airyscan from four WT mice, four uninjected Tmc1Δ/Δ mice, and four injected Tmc1Δ/Δ mice with the best ABR threshold recovery and synapse count preservation. For each z-stack, ribbon volumes were calculated using the 'Surfaces' module in Imaris. The distributions of ribbon volumes are illustrated in Figure 6C. Standard deviations of ribbon volumes were compared to evaluate differences in variation between the three groups. P28 uninjected Tmc1Δ/Δ mice showed larger distributions of ribbon volumes than WT mice at the same age (Supplementary file 6B). These differences were statistically significant at all five frequency regions (Supplementary file 6C). Injected Tmc1Δ/Δ mice showed narrower distributions more similar to those in WT mice, suggesting Tmc1 gene therapy preserves both synapse counts and the distribution of ribbon volumes. However, the standard deviations of injected mice were not significantly different from those of uninjected mice (Supplementary file 6C). Discussion IHC synapses undergo significant developmental changes during the first few postnatal weeks (Sobkowicz et al., 1982; Huang et al., 2012; Yu and Goodrich, 2014; Michanski et al., 2019; Voorn and Vogl, 2020). Sensory transduction is also maturing during this time frame and is required for maturation and maintenance of IHC morphology and a number of biophysical properties (Corns et al., 2018). Here, we report that sensory transduction also contributes to the development and maturation of IHC synapses during the first postnatal month. We began with Tmc1Δ/Δ;Tmc2Δ/Δ mice, which fail to acquire sensory transduction (Kawashima et al., 2011), and found they had a similar number of CtBP2-positive puntca to WT mice at P2. This was consistent with prior work suggesting that formation of presynaptic ribbon complexes is an intrinsic property of immature hair cells (Sobkowicz et al., 1986). Despite the similarities at P2, synapse counts in Tmc1Δ/Δ;Tmc2Δ/Δ mice were significantly elevated at P7 relative to WT mice (Figure 7). These unexpected results are the first to show elevated synapse counts in neonatal mice lacking sensory transduction and suggest that acquisition of sensory transduction may help regulate synaptic pruning and proper synaptic numbers. Figure 7 Download asset Open asset Developmental changes in inner hair cell (IHC) synapses in genetic models of sensory transduction dysfunction. (A) Number of synapses/IHC from the 32 kHz region of wild-type (WT), Tmc1Δ/Δ;Tmc2Δ/Δ, Tmc1Δ/Δ, and Tmc2Δ/Δ cochleas as a function of age. (B) Schematic diagram summarizing changes in IHC synapses between postnatal day 7 (P7) and P28 in five genetic models. In WT mice, the number of synapses progressively increases and then undergoes a reduction during the second postnatal week until the onset of hearing (Sundaresan et al., 2016; Wong et al., 2014; Huang et al., 2012; Sendin et al., 2007). While the reduction of synapses still occurred in our Tmc1Δ/Δ;Tmc2Δ/Δ mice, the lack of sensory transduction led to a larger decline in the number of synapses by P14. Tmc1Δ/Δ;Tmc2Δ/Δ mice also displayed reduced synapse counts at P28, in a manner similar to that reported in previous studies assessing synaptic consequences of genetic disruption of synaptic transmission. For example, mice lacking Cav1.3 Ca2+ channels, which are critical for calcium-dependent neurotransmission at IHC synapses (Platzer et al., 2000; Brandt et al., 2005), lost IHC synapses 4 weeks after birth (Nemzou N. et al., 2006; Brandt et al., 2003). Another study assessed the synaptic consequences of genetic deletion of vesicular glutamate transporter type 3 (Kim et al., 2019) and found roughly half of IHC synapses were lost by 9 weeks of age. Although the genetic targets are different, disruption of synaptic proteins or transduction proteins lead to similar loss of IHC-SGN synapses suggesting they may converge upon similar molecular pathways that regulate synaptic development and maturation. Genetic deletion of Tmc1 or Tmc2 alone did not alter synapse counts prior to hearing onset; Tmc1Δ/Δ and Tmc2

Recent studies have identified sex-differences in auditory physiology and in the susceptibility to noise-induced hearing loss (NIHL). We hypothesize that 17β-estradiol (E2), a known modulator of auditory physiology, may underpin sex-differences in the response to noise trauma. Here, we gonadectomized B6CBAF1/J mice and used a combination of electrophysiological and histological techniques to study the effects of estrogen replacement on peripheral auditory physiology in the absence of noise exposure and on protection from NIHL. Functional analysis of auditory physiology in gonadectomized female mice revealed that E2-treatment modulated the peripheral response to sound in the absence of changes to the endocochlear potential compared to vehicle-treatment. E2-replacement in gonadectomized female mice protected against hearing loss following permanent threshold shift (PTS)- and temporary threshold shift (TTS)-inducing noise exposures. Histological analysis of the cochlear tissue revealed that E2-replacement mitigated outer hair cell loss and cochlear synaptopathy following noise exposure compared to vehicle-treatment. Lastly, using fluorescent in situ hybridization, we demonstrate co-localization of estrogen receptor-2 with type-1C, high threshold spiral ganglion neurons, suggesting that the observed protection from cochlear synaptopathy may occur through E2-mediated preservation of these neurons. Taken together, these data indicate the estrogen signaling pathways may be harnessed for the prevention and treatment of NIHL.

There is a strong association between endolymphatic hydrops and low-frequency hearing loss, but the origin of the hearing loss remains unknown. A reduction in the number of cochlear afferent synapses between inner hair cells and auditory nerve fibres may be the origin of the low-frequency hearing loss, but this hypothesis has not been directly tested in humans or animals. In humans, measurements of hearing loss and postmortem temporal-bone based measurements of endolymphatic hydrops are generally separated by large amounts of time. In animals, there has not been a good objective, physiologic, and minimally invasive measurement of low-frequency hearing. We overcame this obstacle with the combined use of a reliable surgical approach to ablate the endolymphatic sac in guinea pigs and create endolymphatic hydrops, the Auditory Nerve Overlapped Waveform to measure low-frequency hearing loss (≤ 1 kHz), and immunohistofluorescence-based confocal microscopy to count cochlear synapses. Results showed low- and mid-(1–4 kHz) frequency hearing loss at all postoperative days, 1, 4, and 30. There was no statistically significant loss of cochlear synapses, and there was no correlation between synapse loss and hearing function. We conclude that cochlear afferent synaptic loss is not the origin of the low-frequency hearing loss in the early days following endolymphatic sac ablation. Understanding what is, and is not, the origin of a hearing loss can help guide preventative and therapeutic development.

Exposure to loud sound damages the postsynaptic terminals of spiral ganglion neurons (SGNs) on cochlear inner hair cells (IHCs), resulting in loss of synapses, a process termed synaptopathy. Glutamatergic neurotransmission via α-amino-3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA)-type receptors is required for synaptopathy, and here we identify a possible involvement of GluA2-lacking Ca2+-permeable AMPA receptors (CP-AMPARs) using IEM-1460, which has been shown to block GluA2-lacking AMPARs. In CBA/CaJ mice, a 2-h exposure to 100-dB sound pressure level octave band (8 to 16 kHz) noise results in no permanent threshold shift but does cause significant synaptopathy and a reduction in auditory brainstem response (ABR) wave-I amplitude. Chronic intracochlear perfusion of IEM-1460 in artificial perilymph (AP) into adult CBA/CaJ mice prevented the decrease in ABR wave-I amplitude and the synaptopathy relative to intracochlear perfusion of AP alone. Interestingly, IEM-1460 itself did not affect the ABR threshold, presumably because GluA2-containing AMPARs can sustain sufficient synaptic transmission to evoke low-threshold responses during blockade of GluA2-lacking AMPARs. On individual postsynaptic densities, we observed GluA2-lacking nanodomains alongside regions with robust GluA2 expression, consistent with the idea that individual synapses have both CP-AMPARs and Ca2+-impermeable AMPARs. SGNs innervating the same IHC differ in their relative vulnerability to noise. We found local heterogeneity among synapses in the relative abundance of GluA2 subunits that may underlie such differences in vulnerability. We propose a role for GluA2-lacking CP-AMPARs in noise-induced cochlear synaptopathy whereby differences among synapses account for differences in excitotoxic susceptibility. These data suggest a means of maintaining normal hearing thresholds while protecting against noise-induced synaptopathy, via selective blockade of CP-AMPARs.

The perception of sound relies on sensory hair cells in the cochlea that convert the mechanical energy of sound into release of glutamate onto postsynaptic auditory nerve fibers. The hair cell receptor potential regulates the strength of synaptic transmission and is shaped by a variety of voltage-dependent conductances. Among these conductances, the Ca2+- and voltage-activated large conductance Ca2+-activated K+ channel (BK) current is prominent, and in mammalian inner hair cells (IHCs) displays unusual properties. First, BK currents activate at unprecedentedly negative membrane potentials (-60 mV) even in the absence of intracellular Ca2+ elevations. Second, BK channels are positioned in clusters away from the voltage-dependent Ca2+ channels that mediate glutamate release from IHCs. Here, we test the contributions of two recently identified leucine-rich-repeat-containing (LRRC) regulatory γ subunits, LRRC26 and LRRC52, to BK channel function and localization in mouse IHCs. Whereas BK currents and channel localization were unaltered in IHCs from Lrrc26 knockout (KO) mice, BK current activation was shifted more than +200 mV in IHCs from Lrrc52 KO mice. Furthermore, the absence of LRRC52 disrupted BK channel localization in the IHCs. Given that heterologous coexpression of LRRC52 with BK α subunits shifts BK current gating about -90 mV, to account for the profound change in BK activation range caused by removal of LRRC52, we suggest that additional factors may help define the IHC BK gating range. LRRC52, through stabilization of a macromolecular complex, may help retain some other components essential both for activation of BK currents at negative membrane potentials and for appropriate BK channel positioning.

Little is known about the spatial origins of auditory nerve (AN) compound action potentials (CAPs) evoked by moderate to intense sounds. We studied the spatial origins of AN CAPs evoked by 2- to 16-kHz tone bursts at several sound levels by slowly injecting kainic acid solution into the cochlear apex of anesthetized guinea pigs. As the solution flowed from apex to base, it sequentially reduced CAP responses from low- to high-frequency cochlear regions. The times at which CAPs were reduced, combined with the cochlear location traversed by the solution at that time, showed the cochlear origin of the removed CAP component. For low-level tone bursts, the CAP origin along the cochlea was centered at the characteristic frequency (CF). As sound level increased, the CAP center shifted basally for low-frequency tone bursts but apically for high-frequency tone bursts. The apical shift was surprising because it is opposite the shift expected from AN tuning curve and basilar membrane motion asymmetries. For almost all high-level tone bursts, CAP spatial origins extended over 2 octaves along the cochlea. Surprisingly, CAPs evoked by high-level low-frequency (including 2 kHz) tone bursts showed little CAP contribution from CF regions ≤ 2 kHz. Our results can be mostly explained by spectral splatter from the tone-burst rise times, excitation in AN tuning-curve "tails," and asynchronous AN responses to high-level energy ≤ 2 kHz. This is the first time CAP origins have been identified by a spatially specific technique. Our results show the need for revising the interpretation of the cochlear origins of high-level CAPs-ABR wave 1. NEW & NOTEWORTHY Cochlear compound action potentials (CAPs) and auditory brain stem responses (ABRs) are routinely used in laboratories and clinics. They are typically interpreted as arising from the cochlear region tuned to the stimulus frequency. However, as sound level is increased, the cochlear origins of CAPs from tone bursts of all frequencies become very wide and their centers shift toward the most sensitive cochlear region. The standard interpretation of CAPs and ABRs from moderate to intense stimuli needs revision.