Auditory Research publications

Hearing is an extremely delicate sense that is particularly vulnerable to insults from environment, including drugs and noise. Unsurprisingly, mice of different genetic backgrounds show different susceptibility to hearing loss. In particular, CBA/CaJ (CBA) mice maintain relatively stable hearing over age while C57BL/6J (B6) mice show a steady decline of hearing, making them a popular model for early onset hearing loss. To reveal possible underlying mechanisms, we examined cellular differences in the cochlea of these two mouse strains. Although the ABR threshold and Wave I latency are comparable between them, B6 mice have a smaller Wave I amplitude. This difference is probably due to fewer spiral ganglion neurons found in B6 mice, as the number of ribbon synapses per inner hair cell (IHC) is comparable between the two mouse strains. Next, we compared the outer hair cell (OHC) function and we found OHCs from B6 mice are larger in size but the prestin density is similar among them, consistent with the finding that they share similar hearing thresholds. Lastly, we examined the IHC function and we found IHCs from B6 mice have a larger Ca2+ current, release more synaptic vesicles and recycle synaptic vesicles more quickly. Taken together, our results suggest that excessive exocytosis from IHCs in B6 mice may raise the probability of glutamate toxicity in ribbon synapses, which could accumulate over time and eventually lead to early onset hearing loss.

With the advent of recent genetic technologies for mice, it is now feasible to investigate the circuit mechanisms of brain functions in an unprecedented manner. Although transgenic mice are commonly used on C57BL/6J (C57) background, hearing research has typically relied on different genetic backgrounds, such as CBA/Ca or CBA due to the genetic defect of C57 mice for early age-related hearing loss. This limits the utilization of available genetic resources for hearing research. Here we report congenic (>F10) Cre-dependent channelrhodopsin2 (ChR2) mice on CBA/Ca background. By crossing this line with Cre-driver mice on C57 background, F1 hybrids restored the hearing deficit of C57 mice. We also found a linear relationship between aging and hearing loss, with progression rates varied depending on genetic backgrounds (3.39 dB/month for C57; 0.82 dB/month for F1 hybrid). We further demonstrate that this approach allows to express ChR2 in a specific type of inhibitory neurons in the auditory cortex and that they can be identified within a simultaneously recorded population of neurons in awake mice. Thus, our Cre-dependent optogenetic transgenic mice on CBA/Ca background are a valuable tool to investigate the circuit mechanisms of hearing across lifespan.

Significance In humans, inner ear development is completed in utero, with hearing onset at ∼20 weeks of gestation. However, genetic forms of congenital deafness are typically diagnosed during the neonatal period. Gene therapy approaches in animal models should therefore be tested after the period of hearing onset, to determine whether they can reverse an existing deafness phenotype. Here, we used a mouse model of DFNB9, a human deafness form accounting for 2–8% of all cases of congenital genetic deafness. We show that local gene therapy in the mutant mice not only prevents deafness when administered to immature hearing organs, but also durably restores hearing when administered at a mature stage, raising hopes for future gene therapy trials in DFNB9 patients.

Skill learning is fundamental to the acquisition of many complex behaviors that emerge during development. For example, years of practice give rise to perceptual improvements that contribute to mature speech and language skills. While fully honed learning skills might be thought to offer an advantage during the juvenile period, the ability to learn actually continues to develop through childhood and adolescence, suggesting that the neural mechanisms that support skill learning are slow to mature. To address this issue, we asked whether the rate and magnitude of perceptual learning varies as a function of age as male and female gerbils trained on an auditory task. Adolescents displayed a slower rate of perceptual learning compared with their young and mature counterparts. We recorded auditory cortical neuron activity from a subset of adolescent and adult gerbils as they underwent perceptual training. While training enhanced the sensitivity of most adult units, the sensitivity of many adolescent units remained unchanged, or even declined across training days. Therefore, the average rate of cortical improvement was significantly slower in adolescents compared with adults. Both smaller differences between sound-evoked response magnitudes and greater trial-to-trial response fluctuations contributed to the poorer sensitivity of individual adolescent neurons. Together, these findings suggest that elevated sensory neural variability limits adolescent skill learning. SIGNIFICANCE STATEMENT The ability to learn new skills emerges gradually as children age. This prolonged development, often lasting well into adolescence, suggests that children, teens, and adults may rely on distinct neural strategies to improve their sensory and motor capabilities. Here, we found that practice-based improvement on a sound detection task is slower in adolescent gerbils than in younger or older animals. Neural recordings made during training revealed that practice enhanced the sound sensitivity of adult cortical neurons, but had a weaker effect in adolescents. This latter finding was partially explained by the fact that adolescent neural responses were more variable than in adults. Our results suggest that one mechanistic basis of adult-like skill learning is a reduction in neural response variability.

Betaine‐homocysteine S‐methyltransferases (BHMTs) are methionine cycle enzymes that remethylate homocysteine; hence, their malfunction leads to hyperhomocysteinemia. Epidemiologic and experimental studies have revealed a correlation between hyperhomocysteinemia and hearing loss. Here, we have studied the expression of methionine cycle genes in the mouse cochlea and the impact of knocking out the Bhmt gene in the auditory receptor. We evaluated age‐related changes in mouse hearing by recording auditory brainstem responses before and following exposure to noise. Also, we measured cochlear cytoarchitecture, gene expression by RNA‐arrays and quantitative RT‐PCR, and metabolite levels in liver and plasma by HPLC. Our results indicate that there is an age‐dependent strain‐specific expression of methionine cycle genes in the mouse cochlea and a further regulation during the response to noise damage. Loss of Bhmt did not cause an evident impact in the hearing acuity of young mice, but it produced higher threshold shifts and poorer recovery following noise challenge. Hearing loss was associated with increased cochlear injury, outer hair cell loss, altered expression of cochlear methionine cycle genes, and hyperhomocysteinemia. Our results suggest that BHMT plays a central role in the homeostasis of cochlear methionine metabolism and that Bhmt2 up‐regulation could carry out a compensatory role in cochlear protection against noise injury in the absence of BHMT.—Partearroyo, T., Murillo‐Cuesta, S., Vallecillo, N., Bermúdez‐Muñoz, J. M., Rodríguez‐de la Rosa, L., Mandruzzato, G., Celaya, A. M., Zeisel, S. H., Pajares, M. A., Varela‐Moreiras, G., Varela‐Nieto, I. Betaine‐homocysteine S ‐methyltransferase deficiency causes increased susceptibility to noise‐induced hearing loss associated with plasma hyperhomocysteinemia. FASEB J. 33, 5942–5956 (2019). www.fasebj.org

Objectives: The goal of this study was to identify the effects of auditory deprivation (age-related hearing loss) and auditory stimulation (history of hearing aid use) on the neural registration of sound across two stimulus presentation conditions: (1) equal sound pressure level and (2) equal sensation level. Design: We used a between-groups design, involving three groups of 14 older adults (n = 42; 62 to 84 years): (1) clinically defined normal hearing (≤25 dB from 250 to 8000 Hz, bilaterally), (2) bilateral mild–moderate/moderately severe sensorineural hearing loss who have never used hearing aids, and (3) bilateral mild–moderate/moderately severe sensorineural hearing loss who have worn bilateral hearing aids for at least the past 2 years. Results: There were significant delays in the auditory P1-N1-P2 complex in older adults with hearing loss compared with their normal hearing peers when using equal sound pressure levels for all participants. However, when the degree and configuration of hearing loss were accounted for through the presentation of equal sensation level stimuli, no latency delays were observed. These results suggest that stimulus audibility modulates P1-N1-P2 morphology and should be controlled for when defining deprivation and stimulus-related neuroplasticity in people with hearing loss. Moreover, a history of auditory stimulation, in the form of hearing aid use, does not appreciably alter the neural registration of unaided auditory evoked brain activity when quantified by the P1-N1-P2. Conclusions: When comparing auditory cortical responses in older adults with and without hearing loss, stimulus audibility, and not hearing loss–related neurophysiological changes, results in delayed response latency for those with age-related hearing loss. Future studies should carefully consider stimulus presentation levels when drawing conclusions about deprivation- and stimulation-related neuroplasticity. Additionally, auditory stimulation, in the form of a history of hearing aid use, does not significantly affect the neural registration of sound when quantified using the P1-N1-P2–evoked response.

Cortical layers (L) 5 and 6 are populated by intermingled cell-types with distinct inputs and downstream targets. Here, we made optogenetically guided recordings from L5 corticofugal (CF) and L6 corticothalamic (CT) neurons in the auditory cortex of awake mice to discern differences in sensory processing and underlying patterns of functional connectivity. Whereas L5 CF neurons showed broad stimulus selectivity with sluggish response latencies and extended temporal non-linearities, L6 CTs exhibited sparse selectivity and rapid temporal processing. L5 CF spikes lagged behind neighboring units and imposed weak feedforward excitation within the local column. By contrast, L6 CT spikes drove robust and sustained activity, particularly in local fast-spiking interneurons. Our findings underscore a duality among sub-cortical projection neurons, where L5 CF units are canonical broadcast neurons that integrate sensory inputs for transmission to distributed downstream targets, while L6 CT neurons are positioned to regulate thalamocortical response gain and selectivity.

The brain estimates the two-dimensional direction of sounds from the pressure-induced displacements of the eardrums. Accurate localization along the horizontal plane (azimuth angle) is enabled by binaural difference cues in timing and intensity. Localization along the vertical plane (elevation angle), including frontal and rear directions, relies on spectral cues made possible by the elevation dependent filtering in the idiosyncratic pinna cavities. However, the problem of extracting elevation from the sensory input is ill-posed, since the spectrum results from a convolution between source spectrum and the particular head-related transfer function (HRTF) associated with the source elevation, which are both unknown to the system. It is not clear how the auditory system deals with this problem, or which implicit assumptions it makes about source spectra. By varying the spectral contrast of broadband sounds around the 6–9 kHz band, which falls within the human pinna’s most prominent elevation-related spectral notch, we here suggest that the auditory system performs a weighted spectral analysis across different frequency bands to estimate source elevation. We explain our results by a model, in which the auditory system weighs the different spectral bands, and compares the convolved weighted sensory spectrum with stored information about its own HRTFs, and spatial prior assumptions.

Ribbon synapses of cochlear inner hair cells (IHCs) operate with high rates of neurotransmission; yet, the molecular regulation of synaptic vesicle (SV) recycling at these synapses remains poorly understood. Here, we studied the role of endophilins-A1-3, endocytic adaptors with curvature-sensing and curvature-generating properties, in mouse IHCs. Single-cell RT-PCR indicated the expression of endophilins-A1-3 in IHCs, and immunoblotting confirmed the presence of endophilin-A1 and endophilin-A2 in the cochlea. Patch-clamp recordings from endophilin-A-deficient IHCs revealed a reduction of Ca2+ influx and exocytosis, which we attribute to a decreased abundance of presynaptic Ca2+ channels and impaired SV replenishment. Slow endocytic membrane retrieval, thought to reflect clathrin-mediated endocytosis, was impaired. Otoferlin, essential for IHC exocytosis, co-immunoprecipitated with purified endophilin-A1 protein, suggestive of a molecular interaction that might aid exocytosis-endocytosis coupling. Electron microscopy revealed lower SV numbers, but an increased occurrence of coated structures and endosome-like vacuoles at IHC active zones. In summary, endophilins regulate Ca2+ influx and promote SV recycling in IHCs, likely via coupling exocytosis to endocytosis, and contributing to membrane retrieval and SV reformation.

Gentamicin is a widely used antibiotic for the treatment of gram-negative bacterial infections; however, its use often results in significant and permanent hearing loss. Hearing loss resulting from hair cell (HC) degeneration affects millions of people worldwide, and one major cause is the loss of sensory HCs in the inner ear due to aminoglycoside exposure. Strategies to overcome the apparently irreversible loss of HCs in mammals are crucial for hearing protection. Here, we report that the somatostatin analog pasireotide protects mouse cochlear HCs from gentamicin damage using a well-established in vitro gentamicin-induced HC loss model and that the otoprotective effects of pasireotide are due to Akt up-regulation via the PI3K–Akt signal pathway activation. We demonstrate active caspase signal in organ of Corti (OC) explants exposed to gentamicin and show that pasireotide treatment activates survival genes, reduces caspase signal, and increases HC survival. The neuropeptide somatostatin and its selective analogs have provided neuroprotection by activating five somatostatin receptor (SSTR1–SSTR5) subtypes. Pasireotide has a high affinity for SSTR2 and SSTR5, and the addition of SSTR2- and SSTR5-specific antagonists leads to a loss of protection. The otoprotective effects of pasireotide were also observed in a gentamicin-injured animal model. In vivo studies have shown that 13 days of subcutaneous pasireotide application prevents gentamicin-induced HC death and permanent hearing loss in mice. Auditory brainstem response analysis confirmed the protective effect of pasireotide, and we found a significant threshold shift at all measured frequencies (4, 8, 16, 24, and 32 kHz). Together, these findings indicate that pasireotide is a novel otoprotective peptide acting via the PI3K–Akt pathway and may be of therapeutic value for HC protection from ototoxic insults.

Hearing loss has been implicated as a risk factor for schizophrenia, but it is not known whether this association arises from common etiology, top-down influences (e.g., social isolation), bottom-up neurobiological mechanisms, or combinations of these factors. Patients with 22q11.2 Deletion Syndrome (22q11.2DS) have a 25-30% risk of developing schizophrenia, and also suffer frequent hearing loss. Here, we used the Df1 /+ mouse model of 22q11.2DS to investigate the relationship between hearing loss and susceptibility to schizophrenia-relevant brain and behavioral abnormalities. Df1 /+ mice have a multi-gene deletion analogous to the chromosomal microdeletion that causes human 22q11.2DS, and like human 22q11.2DS patients exhibit high rates of hearing loss arising primarily from susceptibility to middle ear inflammation. We found that hearing loss in Df1 /+ mice affected schizophrenia-relevant endophenotypes, including electrophysiological measures of central auditory gain and behavioral measures of auditory sensorimotor gating. Moreover, PV+ inhibitory interneurons, another marker for schizophrenia pathology, were significantly reduced in density in auditory cortex but not secondary motor cortex of Df1 /+ mice with hearing loss. These results reveal bottom-up neurobiological mechanisms through which peripheral hearing loss arising from the 22q11.2 deletion may promote the emergence of schizophrenia-relevant auditory brain and behavioral abnormalities, and also suggest a link between conductive hearing loss and reduced PV+ interneuron density in the auditory cortex. SIGNIFICANCE STATEMENT Hearing loss is a known risk factor for schizophrenia. Deletion of chromosomal locus 22q11.2 is associated with both schizophrenia and hearing loss in humans. In the Df1 /+ mouse model of human 22q11.2 Deletion Syndrome, we find that hearing loss shapes measures that are considered schizophrenia-relevant endophenotypes, such as central auditory gain and auditory sensorimotor gating. Moreover, we report a reduction in density of PV+ inhibitory interneurons in the auditory cortex, but not secondary motor cortex, of Df1 /+ mice with hearing loss. These results suggest mechanisms through which hearing loss associated with the 22q11.2 deletion may promote emergence of schizophrenia-relevant auditory brain and behavioral abnormalities and indicate that conductive hearing loss may influence PV+ interneuron density in the auditory cortex.

Objectives: To investigate how mouse cochleae are affected by the striking noise of titanium head golf driver. Methods: Thirty-two BALB/c mice (20-22 g) with normal hearing were used. The impact acoustic stimulus generated by the striking of titanium golf driver head was centered around 4.5 kHz with 120.5 dB sound pressure level. The recorded impact noise was provided to mice in two ways with the same exposure time of 288 seconds: 1,440 repetitions and an impact duration of 0.2 seconds for 2 hours (repetitive noise) or serially connected impact noise for 288 seconds (continuous noise). Auditory brainstem responses were measured at baseline, day 7, and day 14 after exposure. The mice were then sacrificed for histology. Results: Both groups showed statistically significant threshold shifts immediately after noise exposure. Mice in the continuous exposure group, except for those exposed to 32 kHz noise, recovered from threshold shifts 1-2 weeks after noise exposure. However, in the repetitive exposure group, threshold shifts remained for 2 weeks after exposure. The repetitive exposure group had greater hair cell damage than did the continuous exposure group. Structural changes in the stria vascularis were observed in the repetitive exposure group. Conclusion: Overexposure to impact noise caused by hitting of titanium head golf driver may be hazardous to the cochlea, and repetitive exposure may induce greater damage than continuous exposure.

Aminoglycoside antibiotics-induced hearing loss is a common sensorineural impairment. Spiral ganglion neurons (SGNs) are first-order neurons of the auditory pathway and are critical for the maintenance of normal hearing. In the present study, we investigated the time-course of morphological changes and the degeneration process of spiral ganglion cells (SGCs) following chronic kanamycin-induced deafness and determined whether the endoplasmic reticulum (ER) stress was involved in the degeneration of SGNs. We detected density changes in SGCs and the expressions of Bip, inositol requirement 1 (IRE1)α, activating transcription factor-6α, p-PERK, p-eIF2α, CHOP, and caspase-12 at each time point after kanamycin treatment. Terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) staining was also performed. The number of SGC deletions reached ∼50% at the 70th day after kanamycin administration and the ER of most SGCs were dilated. The expression of p-PERK, p-eIF2α, p-IRE1α, Bip, caspase-12, and Chop was significantly unregulated after kanamycin treatment. The number of SGCs that were positive for both TUNEL and caspase-12 increased from day 7 to 28. Taken together, these data demonstrate that ER stress was involved in kanamycin-induced apoptosis of SGNs. Kanamycin-induced SGN apoptosis is mediated, at least in part, by ER stress-induced upregulation of CHOP and caspase-12.

Based on anatomical connectivity and basic response characteristics, primate auditory cortex is divided into a central core surrounded by belt and parabelt regions. The encoding of pitch, a prototypical element of sound identity, has been studied in primary auditory cortex (A1) but little is known about how it is encoded and represented beyond A1. The caudal auditory belt and parabelt cortical fields process spatial information but also contain information on non-spatial aspects of sounds. In this study, we examined neuronal responses in these areas to pitch-varied marmoset vocalizations, to derive the consequent representation of pitch in these regions and the potential underlying mechanisms, to compare to the encoding and representation of pitch of the same sounds in A1. With respect to response patterns to the vocalizations, neurons in caudal medial belt (CM) showed similar short-latency and short-duration response patterns to A1, but caudal lateral belt (CL) neurons at the same hierarchical level and caudal parabelt (CPB) neurons at a higher hierarchical level showed delayed or much delayed response onset and prolonged response durations. With respect to encoding of pitch, neurons in all cortical fields showed sensitivity to variations in the vocalization pitch either through modulation of spike-count or of first spike-latency. The utility of the encoding mechanism differed between fields: pitch sensitivity was reliably represented by spike-count variations in A1 and CM, while first spike-latency variation was better for encoding pitch in CL and CPB. In summary, our data show that (a) the traditionally-defined belt area CM is functionally very similar to A1 with respect to the representation and encoding of complex naturalistic sounds, (b) the CL belt area, at the same hierarchical level as CM, and the CPB area, at a higher hierarchical level, have very different response patterns and appear to use different pitch-encoding mechanisms, and (c) caudal auditory fields, proposed to be specialized for encoding spatial location, can also contain robust representations of sound identity.

Fragile X Syndrome (FXS) is a leading genetic cause of autism and intellectual disabilities. Sensory-processing deficits are common in humans with FXS and an animal model, the Fmr1 knockout (KO) mouse, manifesting in the auditory system as debilitating hypersensitivity and abnormal electroencephalographic (EEG) and event-related potential (ERP) phenotypes. FXS is a neurodevelopmental disorder, but how EEG/ERP phenotypes change during development is unclear. Therefore, we characterized baseline and stimulus-evoked EEG in auditory and frontal cortex of developing (postnatal day (P) 21 and P30) and adult (P60) wildtype (WT) and Fmr1 KO mice with the FVB genetic background. We found that baseline gamma-band power and N1 amplitude of auditory ERP were increased in frontal cortex of Fmr1 KO mice during development and in adults. Baseline gamma power was increased in auditory cortex at P30. Genotype differences in stimulus-evoked gamma power were present in both cortical regions, but the direction and strength of the changes were age-dependent. These findings suggest that cortical deficits are present during early development and may contribute to sensory-processing deficits in FXS, which in turn may lead to anxiety and delayed language. Developmental changes in EEG measures indicate that observations at a single time-point during development are not reflective of FXS disease progression and highlight the need to identify developmental trajectories and optimal windows for treatment.

Auditory stimulus reconstruction is a technique that finds the best approximation of the acoustic stimulus from the population of evoked neural activity. Reconstructing speech from the human auditory cortex creates the possibility of a speech neuroprosthetic to establish a direct communication with the brain and has been shown to be possible in both overt and covert conditions. However, the low quality of the reconstructed speech has severely limited the utility of this method for brain-computer interface (BCI) applications. To advance the state-of-the-art in speech neuroprosthesis, we combined the recent advances in deep learning with the latest innovations in speech synthesis technologies to reconstruct closed-set intelligible speech from the human auditory cortex. We investigated the dependence of reconstruction accuracy on linear and nonlinear (deep neural network) regression methods and the acoustic representation that is used as the target of reconstruction, including auditory spectrogram and speech synthesis parameters. In addition, we compared the reconstruction accuracy from low and high neural frequency ranges. Our results show that a deep neural network model that directly estimates the parameters of a speech synthesizer from all neural frequencies achieves the highest subjective and objective scores on a digit recognition task, improving the intelligibility by 65% over the baseline method which used linear regression to reconstruct the auditory spectrogram. These results demonstrate the efficacy of deep learning and speech synthesis algorithms for designing the next generation of speech BCI systems, which not only can restore communications for paralyzed patients but also have the potential to transform human-computer interaction technologies.

Adeno-associated virus (AAV) has been successfully used to deliver gene therapy to improve auditory function in mouse models of hereditary hearing loss. Many forms of hereditary hearing loss have mutations which affect the cochlear hair cells, the mechanosensory cells which allow for sound detection and processing. While most conventional AAVs infect inner hair cells (IHCs) with various efficiencies, they infect outer hair cells (OHCs) and supporting cells at lower levels in the cochlea. Here we examine the infection patterns of two synthetic AAVs (AAV2.7m8 and AAV8BP2) in the mouse inner ear. AAV2.7m8 infects both IHCs and OHCs with high efficiency. In addition, AAV2.7m8 infects inner pillar cells and inner phalangeal cells with high efficiency. Our results suggest that AAV2.7m8 is an excellent viral vector for inner ear gene therapy targeting cochlear hair cells and supporting cells, and it will likely greatly expand the potential applications for inner ear gene therapy.

Active listening involves dynamically switching attention between competing talkers and is essential to following conversations in everyday environments. Previous investigations in human listeners have examined the neural mechanisms that support switching auditory attention within the acoustic featural cues of pitch and auditory space. Here, we explored the cortical circuitry underlying endogenous switching of auditory attention between pitch and spatial cues necessary to discern target from masker words. Because these tasks are of unequal difficulty, we expected an asymmetry in behavioral switch costs for hard-to-easy versus easy-to-hard switches, mirroring prior evidence from vision-based cognitive task-switching paradigms. We investigated the neural correlates of this behavioral switch asymmetry and associated cognitive control operations in the present auditory paradigm. Behaviorally, we observed no switch-cost asymmetry, i.e., no performance difference for switching from the more difficult attend-pitch to the easier attend-space condition (P→S) versus switching from easy-to-hard (S→P). However, left lateral prefrontal cortex activity, correlated with improved performance, was observed during a silent gap period when listeners switched attention from P→S, relative to switching within pitch cues. No such differential activity was seen for the analogous easy-to-hard switch. We hypothesize that this neural switch asymmetry reflects proactive cognitive control mechanisms that successfully reconfigured neurally-specified task parameters and resolved competition from other such "task sets," thereby obviating the expected behavioral switch-cost asymmetry. The neural switch activity observed was generally consistent with that seen in cognitive paradigms, suggesting that established cognitive models of attention switching may be productively applied to better understand similar processes in audition.

Fifty percent of inner ear disorders are caused by genetic mutations. To develop treatments for genetic inner ear disorders, we designed gene replacement therapies using synthetic adeno-associated viral vectors to deliver the coding sequence for Transmembrane Channel-Like ( Tmc ) 1 or 2 into sensory hair cells of mice with hearing and balance deficits due to mutations in Tmc1 and closely related Tmc2 . Here we report restoration of function in inner and outer hair cells, enhanced hair cell survival, restoration of cochlear and vestibular function, restoration of neural responses in auditory cortex and recovery of behavioral responses to auditory and vestibular stimulation. Secondarily, we find that inner ear Tmc gene therapy restores breeding efficiency, litter survival and normal growth rates in mouse models of genetic inner ear dysfunction. Although challenges remain, the data suggest that Tmc gene therapy may be well suited for further development and perhaps translation to clinical application.

This paper describes a technique to assess the correspondence between patterns of similarity in the brain's response to speech sounds and the patterns of similarity encoded in phonological feature systems, by quantifying the recoverability of phonological features from the neural data using supervised learning. The technique is applied to EEG recordings collected during passive listening to consonant-vowel syllables. Three published phonological feature systems are compared, and are shown to differ in their ability to recover certain speech sound contrasts from the neural data. For the phonological feature system that best reflects patterns of similarity in the neural data, a leave-one-out analysis indicates some consistency across subjects in which features have greatest impact on the fit, but considerable across-subject heterogeneity remains in the rank ordering of features in this regard.