1. Introduction
Hearing in mammals relies on the ability of the sensory hair cells to convert sound-evoked mechanical stimuli into electrochemical signals. The hair-bundle deflection induces rapid opening of sensory transduction channels, leading to the generation of an influx of cations into the IHC. This results in a depolarization potential, allowing an influx of calcium through voltage-dependent calcium channels. The coupling of Ca
2+ channels at the presynaptic site of the ribbon synapse triggers high-rate synaptic vesicle fusion and the release of neurotransmitter glutamate from the synaptic cleft. The release of glutamate in the synapse activates Ca
2+-sensitive AMPA receptors. This initiates the generation of neural spikes in spiral ganglion neuron (SGN) fibers, which encodes information about sound stimuli that is sent to the central nervous system. A dysfunction at any level of this complex transduction machinery may disturb the coding of acoustic features, particularly of temporal cues. The potential sites of damage are diverse, including the IHCs, IHC ribbon synapses, or synaptopathy, (e.g., pre-synaptic release of glutamate or postsynaptic terminals dendrites of the spiral ganglion neurons), or can be due to demyelination and axonal loss of the auditory nerve fibers and their targets in the cochlear nucleus (i.e., neuropathy). These auditory pathologies are named auditory neuropathy spectrum disorder (ANSD), in which the activity of outer hair cells (OHCs) is maintained
[1,2,3,4][1][2][3][4].
The clinical profiles of ANSD are quite heterogeneous, depending on the variety of etiologies. ANSD can result from syndromic and non-syndromic genetic abnormalities, as well as environmental causes (e.g., hypoxia, noise-exposure, cytotoxic oncologic drugs) and aging. ANSD is one of the common causes of hearing loss, affecting between 1.2% and 10% of those with hearing loss
[5]. Audiologically, ANSD is characterized by mild to profound sensory neural hearing loss, with impaired or absent compound action potentials (CAP) and auditory brainstem responses (ABRs) and deteriorated speech audiometry in quiet
[6]; these are associated with normal otoacoustic emissions (OAE) or cochlear microphonics (CM), indicating normal OHC function. Additionally, the absence of the middle-ear stapedial reflex and of the contralateral suppression of otoacoustic emissions are usually observed
[1,5,7,8][1][5][7][8].
The clinical profiles of ANSD are quite heterogeneous, depending on the variety of etiologies. ANSD can result from syndromic and non-syndromic genetic abnormalities, as well as environmental causes (e.g., hypoxia, noise-exposure, cytotoxic oncologic drugs) and aging. ANSD is one of the common causes of hearing loss, affecting between 1.2% and 10% of those with hearing loss
[5]. Audiologically, ANSD is characterized by mild to profound sensory neural hearing loss, with impaired or absent compound action potentials (CAP) and auditory brainstem responses (ABRs) and deteriorated speech audiometry in quiet
[6]; these are associated with normal otoacoustic emissions (OAE) or cochlear microphonics (CM), indicating normal OHC function. Additionally, the absence of the middle-ear stapedial reflex and of the contralateral suppression of otoacoustic emissions are usually observed
[1,5,7,8][1][5][7][8].
2. Non-Syndromic Auditory Synaptopathies
Genetic auditory synaptopathies generally only cause deafness, such as the mutations in the CACNA1D gene encoding the Cav1.3L-type Ca2+ channel, the OTOF gene encoding Otoferlin, the SLC17A8 gene encoding Vglut3, or the DIAPH3 gene encoding the diaphanous formin 3.
2.1. Otoferlin-DFNB9
The OTOF gene encodes otoferlin, which is a critical calcium sensor for synaptic exocytosis in cochlear IHCs
[19,20][9][10]. Otoferlin is also involved in vesicular reformation, re-supply, and tethering at the active zone, making otoferlin a multi-tasking protein
[20,21][10][11]. Mutations in the gene encoding otoferlin are responsible for autosomal recessive profound prelingual deafness, DFNB9
[22][12]. To date, about 220 pathogenic variants in
OTOF have been identified
[23][13]. The majority of these mutations are assumed to be nonsense or truncation mutations that provoke the inactivation of otoferlin
[24][14]. Patients with variants in
OTOF displayed milder hearing loss, as well as progressive and temperature-sensitive hearing loss, while OAEs were preserved
[22,25,26,27][12][15][16][17]. Children harboring biallelic mutations of the
OTOF gene displayed profound hearing loss, absence of ABRs and CAP, but preservation of DPOAEs and the amplitude of CM
[28][18]. Otoferlin knock-out mice, which are profoundly deaf due to a failure of sound-evoked neurotransmitter release at the IHC synapse, are likely to be an appropriate animal model for DFNB9
[29,30][19][20]. In these mice, Ca
2+-triggered exocytosis in IHCs is almost abolished
[29,30][19][20]; synaptic vesicles were found near the membrane at the active zone, suggesting that an absence of vesicles did not limit signal transduction, but that a late step of exocytosis was disrupted
[30][20].
2.2. VGLUT3-DFNA25
Vesicular glutamate transporters (VGLUTs) are responsible for glutamate loading into synaptic vesicles, which is essential in order to achieve synaptic transmission
[31][21]. VGLUT3 is expressed in small subsets of neurons in the central nervous system
[31,32][21][22]. In mice, VGLUT3 is expressed in the IHCs
[33,34][23][24] and the OHCs
[35][25]. The genetic ablation of
Slc17a8 in mice results in the absence of CAP or ABRs to acoustic stimuli, while ABRs could be elicited by electrical stimuli, and robust otoacoustic emissions were recorded in these mice
[33,34][23][24]. This thus reflects a failure in activation of the ascending auditory pathway, while the activity in OHCs is unaffected
[33,34,36,37][23][24][26][27]. Patients with a 12q22-q24 deletion in the
SLC17A8 gene at the DFNA25 locus display congenital and non-syndromic autosomal dominant deafness
[33,38,39][23][28][29]. The deafness in patients was characterized as high-frequency, progressive sensorineural hearing loss, with good hearing rescue through cochlear implantation, thus reinforcing the hypothesis of synaptopathy
[33,39][23][29].
VGLUT3A224V/A224V mice harboring the p.A221V mutation (p.A221V in humans corresponds to p.A224V in mice) in the
Slc17a8 gene displayed progressive hearing loss with intact OHC function
[40][30]. The summating potential was, however, reduced, indicating the alteration of the IHC receptor potential. Scanning electron microscopy examinations revealed the collapse of IHC stereocilia bundles, leaving those from OHCs unaffected. In addition, IHC ribbon synapses underwent structural and functional modifications at later stages. These results suggest that DFNA25 stems from a failure in mechano-transduction followed by a change in synaptic transmission
[40][30].
2.3. Cav1.3-SANDD
Calcium influx at the base of the IHCs near the ribbon synapse is mediated via the L-type calcium (Ca
2+) channel Cav1.3, which is the main voltage-gated Ca
2+ channel in IHCs and essential for hearing. Cav1.3 translates sound-induced depolarization into neurotransmitter glutamate release at the synaptic site, resulting in signal transmission to the auditory nerve
[41][31]. Cav1.3-encoding by the
CACNA1D gene is widely distributed across different cells such as OHCs, IHCs, cardiomyocytes, neuroendocrine cells, and neurons. A Cav1.3. mutation in
CACNA1D may cause both sinoatrial node dysfunction and deafness (termed SANDD syndrome) in mice and in humans, in humans closely resembling that of
Cacna1d−/− mice
[41,42][31][32]. Ca
v1.3 is required for normal hearing and cardiac pace making in humans, and loss of function in only a subset of channels is sufficient to cause SANDD syndrome
[42][32]. Loss-of-function mutations in the
CACNA1D gene causes impaired synaptic neurotransmission at the IHC ribbon synapse in KO mice
[41,43][31][33]. Cav1.3 protects the sensory hair cells during cochlear aging through reducing calcium-mediated oxidative stress in C57BL/6J male mice
[44][34] and plays important roles in inner ear differentiation
[45][35].
2.4. CABP2-DFNB93
Calcium-binding protein 2 (CABP2) is a potent modulator of IHC voltage-gated calcium channels CaV1.3. CABP2 regulates Ca
2+ influx at the presynaptic site
[46,47][36][37] and thus also the vesicular release of glutamate. Pathologic mutations in
CABP2 lead to autosomal-recessive, moderate-to-severe non-syndromic hearing impairment DFNB93
[48,49,50,51][38][39][40][41]. DFNB93 patients displayed an auditory synaptopathy phenotype with normal OAEs
[52][42]. Using a knock-out mouse model, Picher et al.
[52][42] demonstrated that DFNB93 hearing impairment may result from an enhanced steady-state inactivation of CaV1.3 channels at the IHC synapse, thus limiting their availability to trigger synaptic transmission, resulting in elevated auditory thresholds
[52][42]. This, however, does not seem to interfere with cochlear development and does not cause the early degeneration of hair cells or their synaptic complex
[52,53][42][43]. These results suggested an extended window for gene therapy.
2.5. DIAPH3-AUNA1
Auditory neuropathy, non-syndromic, autosomal dominant 1 (AUNA1) is a form of delayed-onset, progressive human deafness resulting from a point mutation in the 5′ untranslated region of the Diaphanous homolog 3 (
DIAPH3) gene. The
DIAPH3 mutation leads to the overexpression of the DIAPH3 protein, a formin family member involved in cytoskeleton dynamics
[54][44]. Patients with AUNA1 displayed absent or altered ABR, while OHC functions are still maintained
[1[1][45],
55], thus indicating auditory neuropathy. Transgenic mice overexpressing Diap3 exhibit a progressive threshold shift but maintained a distortion product of otoacoustic emissions (DPOAEs)
[54,56][44][46]. Morphological assessments revealed a selective and early onset alteration of the IHC cuticular plate and fused stereocilia with the eventual loss of the capacity of IHC to transmit incoming sensory stimuli
[54,56][44][46]. Furthermore, a significant reduction in the number of IHC ribbon synapses was observed over 24 weeks in mutant mice, although this reduction did not correlate temporally with the onset and progression of hearing loss or of stereocilia bundle anomalies
[54][44]. Together, these results suggest an important function of Diap3 in regulating the assembly and/or maintenance of actin filaments in IHC stereocilia, as well as a potential role at the IHC ribbon synapse.
3. Syndromic Auditory Neuropathy
Genetic neuropathies frequently affect other neurons, thus leading to syndromic phenotypes such as Charcot–Marie–Tooth disease, autosomal dominant optic atrophy, Leber’s hereditary optic neuropathy, Friedreich’s ataxia, Mohr–Tranebjaerg syndrome, Refsum disease, or Wolfram syndrome
[57,58,59,60][47][48][49][50].
3.1. Charcot–Marie–Tooth
Autosomal-dominant Charcot–Marie–Tooth (CMT) is the most common hereditary peripheral polyneuropathy characterized by the degeneration of peripheral nerves. CMT can be classified into two major categories: TMC type 1 (demyelinating neuropathies) and type 2 (axonal form of neuropathies)
[61,62][51][52]. CMT patients carry mutations in the MPZ genes for myelin protein zero or PMP22 coding for proteins essential for the formation and adhesion of myelin
[2,63,64][2][53][54]. CMT type 1 A (CMT1A) is the predominant subtype, which is a demyelinating peripheral neuropathy characterized by distal muscle weakness, sensory loss, areflexia, and slow motor- and sensory-nerve conduction velocities
[1,62,63][1][52][53]. Hearing impairment is also a relatively common symptom of CMT1A. Compared to controls, CMT1A patients had a significantly decreased speech perception capacity in a noisy environment, as well as decreased temporal and spectral resolution, thus suggesting that demyelination of auditory-nerve fibers in CMT1A causes defective cochlear neurotransmission
[65][55]. Patients with CMT type 1 and 2 showed a delayed or reduced amplitude ABR, as well as an impaired speech intelligibility, which are electrophysiological evidence of auditory neuropathy
[62][52].
3.2. Autosomal-Dominant Optic Atrophy
Autosomal-dominant optic atrophy (DOA) is the most frequent form of hereditary optic neuropathy
[66][56], with a reported frequency of 1:10,000, and is caused by heterozygous variants in the
OPA1 gene encoding a mitochondrial-dynamin-related large GTPase
[67,68,69][57][58][59]. OPA1 is involved in many mitochondrial functions, notably in the maintenance of the respiratory chain and cell membrane potential
[70[60][61][62],
71,72], cristae organization, control of apoptosis
[72[62][63],
73], and mitochondrial DNA maintenance
[74,75,76][64][65][66]. DOA was initially described as a non-syndromic moderate-to-severe loss of visual acuity, with an insidious onset in early childhood caused by a progressive loss of retinal ganglion cells
[77][67]. In the last decade, the clinical spectrum of DOA has been extended to a wide variety of symptoms, including deafness, ataxia, neuropathy, and myopathy, and is now called dominant optic atrophy plus (DOA
plus)
[74,78,79][64][68][69]. Deafness is the second-most prevalent clinical feature in DOA
plus, affecting about 20% of all DOA patients
[17,74,78,79,80][64][68][69][70][71].
The association of DOA and deafness is classically related to the R445H mutation in exon 14, but other
OPA1 mis-sense variants have already been reported in the literature
[79,81][69][72]. Here, hearing loss starts in childhood or early adulthood
[79,82][69][73]. Although the majority of studies broadly qualify the hearing disorder as ‘sensorineural hearing loss’, some authors have proposed auditory neuropathy as the pathophysiological mechanism underlying the hearing impairment in OPA1-DOA
[17,70,83,84][60][70][74][75]. Audiolological examination of OPA1 hearing impaired patients harboring missense mutations showed impaired speech perception and absence or profound alteration of ABRs but preservation of OAE and even enhanced CM potentials reflecting normal OHC function
[28][18].
3.3. Leber Hereditary Optic Neuropathy
Leber hereditary optic neuropathy (LHON) is the most common mitochondrial genetic disease. It is characterized by bilateral, subacute, painless loss of vision, and over 95% of LHON cases are caused by one of three mitochondrial DNA (mtDNA) point mutations: 3460G>A, 11778G>A, and 14484T>C or mutation in the
TMEM126A gene coding a mitochondrial protein. Severe axonal degeneration with demyelination of the optic nerve had been indicated by histological necropsy studies
[85][76]. Patients with Leber hereditary optic neuropathy also show signs of auditory neuropathy
[86,87][77][78].
3.4. Friedreich’s Ataxia
Friedreich’s ataxia (FRDA) is the most frequent autosomal-recessive inherited ataxia caused by mutations in the FXN gene coding for the mitochondrial protein Frataxin involved in regulating iron accumulation in the mitochondria. FRDA is due to an abnormal repetition of the GAA triplet (100 to 2000 GAA triplets) in the
FXN gene
[88][79]. In addition to impaired balance and coordination of voluntary movements, Friedreich’s ataxia is associated with hearing impairment, including difficulty understanding speech in background noise; auditory thresholds were, however, unchanged
[88,89[79][80][81][82],
90,91], nor was OHC function
[92][83]. Most affected individuals show abnormalities in auditory neural and brainstem responses as a result of auditory neuropathy
[92,93,94][83][84][85]. Of FRDA patients, 8 to 13% show sensorineural hearing loss, as revealed in a pure-tone audiogram
[95][86].
3.5. Mohr–Tranebjaerg Syndrome
Mohr–Tranebjaerg syndrome, in which deafness with progressive dystonia and visual impairment are associated, can be classified as a non-isolated auditory neuropathy. Indeed, observation of post-mortem samples shows neuronal loss with preservation of OHCs
[96][87]. Here, again, mutations (DDP1 for deafness-dystonia) of TIMM8A/DDP1, which codes for a polypeptide of 97 amino acids located in the mitochondria, are at the origin of this syndrome.