Stages of degradation of timing information in the cochlea: a comparison of hair-cell and nerve-fiber responses in the alligator lizard

Models optimized for real-world tasks reveal the task-dependent necessity of precise temporal coding in hearing

Neurons encode information in the timing of their spikes in addition to their firing rates. Spike timing is particularly precise in the auditory nerve, where action potentials phase lock to sound with sub-millisecond precision, but its behavioral relevance remains uncertain. We optimized machine learning models to perform real-world hearing tasks with simulated cochlear input, assessing the precision of auditory nerve spike timing needed to reproduce human behavior. Models with high-fidelity phase locking exhibited more human-like sound localization and speech perception than models without, consistent with an essential role in human hearing. However, the temporal precision needed to reproduce human-like behavior varied across tasks, as did the precision that benefited real-world task performance. These effects suggest that perceptual domains incorporate phase locking to different extents depending on the demands of real-world hearing. The results illustrate how optimizing models for realistic tasks can clarify the role of candidate neural codes in perception.

Models optimized for real-world tasks reveal the task-dependent necessity of precise temporal coding in hearing

Neurons encode information in the timing of their spikes in addition to their firing rates. Spike timing is particularly precise in the auditory nerve, where action potentials phase lock to sound with sub-millisecond precision, but its behavioral relevance remains uncertain. We optimized machine learning models to perform real-world hearing tasks with simulated cochlear input, assessing the precision of auditory nerve spike timing needed to reproduce human behavior. Models with high-fidelity phase locking exhibited more human-like sound localization and speech perception than models without, consistent with an essential role in human hearing. However, the temporal precision needed to reproduce human-like behavior varied across tasks, as did the precision that benefited real-world task performance. These effects suggest that perceptual domains incorporate phase locking to different extents depending on the demands of real-world hearing. The results illustrate how optimizing models for realistic tasks can clarify the role of candidate neural codes in perception.

Characterization of the decline in auditory nerve phase locking at high frequencies

The frequency dependence of phase locking in the auditory nerve influences various auditory coding mechanisms. The decline of phase locking with increasing frequency is commonly described by a low-pass filter. This study compares fitted low-pass filter parameters with the actual rate of phase locking decline. The decline is similar across studies and only 40 dB per decade, corresponding to the asymptotic decline of a second order filter.

Frequency dependence of sensitivity to interaural phase differences in pure tones

It is well established that in normal-hearing humans, the threshold of interaural time differences for pure tones increases dramatically above about 1300 Hz, only to become unmeasurable above 1400 Hz. However, physiological data and auditory models suggest that the actual decline in sensitivity is more gradual and only appears to be abrupt because the maximum of the psychometric function dips below the threshold proportion correct, e.g., 0.794. Published data only report thresholds at certain proportions correct but not the decline of proportions correct or of the sensitivity index d' with increasing frequencies. Here, we present pure-tone behavioral data obtained with a constant stimulus procedure. Seven of nine subjects showed proportions correct above 0.9 at 1300 Hz and virtually no sensitivity at 1500 Hz (proportion correct within 0.07 of chance level). This corresponds to a sensitivity decline of 46-78 dB/oct, much steeper than predicted by existing models or by the decline of phase locking of the auditory nerve fibers in animal data.

The upper frequency limit for the use of phase locking to code temporal fine structure in humans: A compilation of viewpoints

The relative importance of neural temporal and place coding in auditory perception is still a matter of much debate. The current article is a compilation of viewpoints from leading auditory psychophysicists and physiologists regarding the upper frequency limit for the use of neural phase locking to code temporal fine structure in humans. While phase locking is used for binaural processing up to about 1500 Hz, there is disagreement regarding the use of monaural phase-locking information at higher frequencies. Estimates of the general upper limit proposed by the contributors range from 1500 to 10000 Hz. The arguments depend on whether or not phase locking is needed to explain psychophysical discrimination performance at frequencies above 1500 Hz, and whether or not the phase-locked neural representation is sufficiently robust at these frequencies to provide useable information. The contributors suggest key experiments that may help to resolve this issue, and experimental findings that may cause them to change their minds. This issue is of crucial importance to our understanding of the neural basis of auditory perception in general, and of pitch perception in particular.

The Calyx of Held: A Hypothesis on the Need for Reliable Timing in an Intensity-Difference Encoder

The calyx of Held is the preeminent model for the study of synaptic function in the mammalian CNS. Despite much work on the synapse and associated circuit, its role in hearing remains enigmatic. We propose that the calyx is one of the key adaptations that enables an animal to lateralize transient sounds. The calyx is part of a binaural circuit that is biased toward high sound frequencies and is sensitive to intensity differences between the ears. This circuit also shows marked sensitivity to interaural time differences, but only for brief sound transients ("clicks"). In a natural environment, such transients are rare except as adventitious sounds generated by other animals moving at close range. We argue that the calyx, and associated temporal specializations, evolved to enable spatial localization of sound transients, through a neural code congruent with the circuit's sensitivity to interaural intensity differences, thereby conferring a key benefit to survival.

Synaptopathy in the Aging Cochlea: Characterizing Early-Neural Deficits in Auditory Temporal Envelope Processing

Aging listeners, even in the absence of overt hearing loss measured as changes in hearing thresholds, often experience impairments processing temporally complex sounds such as speech in noise. Recent evidence has shown that normal aging is accompanied by a progressive loss of synapses between inner hair cells and auditory nerve fibers. The role of this cochlear synaptopathy in degraded temporal processing with age is not yet understood. Here, we used population envelope following responses, along with other hair cell- and neural-based measures from an age-graded series of male and female CBA/CaJ mice to study changes in encoding stimulus envelopes. By comparing responses obtained before and after the application of the neurotoxin ouabain to the inner ear, we demonstrate that we can study changes in temporal processing on either side of the cochlear synapse. Results show that deficits in neural coding with age emerge at the earliest neural stages of auditory processing and are correlated with the degree of cochlear synaptopathy. These changes are seen before losses in neural thresholds and particularly affect the suprathreshold processing of sound. Responses obtained from more central sources show smaller differences with age, suggesting compensatory gain. These results show that progressive cochlear synaptopathy is accompanied by deficits in temporal coding at the earliest neural generators and contribute to the suprathreshold sound processing deficits observed with age.SIGNIFICANCE STATEMENT Aging listeners often experience difficulty hearing and understanding speech in noisy conditions. The results described here suggest that age-related loss of cochlear synapses may be a significant contributor to those performance declines. We observed aberrant neural coding of sounds in the early auditory pathway, which was accompanied by and correlated with an age-progressive loss of synapses between the inner hair cells and the auditory nerve. Deficits first appeared before changes in hearing thresholds and were largest at higher sound levels relevant to real world communication. The noninvasive tests described here may be adapted to detect cochlear synaptopathy in the clinical setting.

Assessment of the limits of neural phase-locking using mass potentials

In the diverse mechanosensory systems that animals evolved, the waveform of stimuli can be encoded by phase locking in spike trains of primary afferents. Coding of the fine structure of sounds via phase locking is thought to be critical for hearing. The upper frequency limit of phase locking varies across species and is unknown in humans. We applied a method developed previously, which is based on neural adaptation evoked by forward masking, to analyze mass potentials recorded on the cochlea and auditory nerve in the cat. The method allows us to separate neural phase locking from receptor potentials. We find that the frequency limit of neural phase locking obtained from mass potentials was very similar to that reported for individual auditory nerve fibers. The results suggest that this is a promising approach to examine neural phase locking in humans with normal or impaired hearing or in other species for which direct recordings from primary afferents are not feasible.

Estimation of neural phase locking from stimulus-evoked potentials

The frequency extent over which fine structure is coded in the auditory nerve has been physiologically characterized in laboratory animals but is unknown in humans. Knowledge of the upper frequency limit in humans would inform the debate regarding the role of fine structure in human hearing. Of the presently available techniques, only the recording of mass neural potentials offers the promise to provide a physiological estimate of neural phase locking in humans. A challenge is to disambiguate neural phase locking from the receptor potentials. We studied mass potentials recorded on the cochlea and auditory nerve of cat and used several experimental manipulations to isolate the neural contribution to these potentials. We find a surprisingly large neural contribution in the signal recorded on the cochlear round window, and this contribution is in many aspects similar to the potential measured on the auditory nerve. The results suggest that recording of mass potentials through the middle ear is a promising approach to examine neural phase locking in humans.

Response characteristics in the apex of the gerbil cochlea studied through auditory nerve recordings

In this study, we analyze the processing of low-frequency sounds in the cochlear apex through responses of auditory nerve fibers (ANFs) that innervate the apex. Single tones and irregularly spaced tone complexes were used to evoke ANF responses in Mongolian gerbil. The spike arrival times were analyzed in terms of phase locking, peripheral frequency selectivity, group delays, and the nonlinear effects of sound pressure level (SPL). Phase locking to single tones was similar to that in cat. Vector strength was maximal for stimulus frequencies around 500 Hz, decreased above 1 kHz, and became insignificant above 4 to 5 kHz. We used the responses to tone complexes to determine amplitude and phase curves of ANFs having a characteristic frequency (CF) below 5 kHz. With increasing CF, amplitude curves gradually changed from broadly tuned and asymmetric with a steep low-frequency flank to more sharply tuned and asymmetric with a steep high-frequency flank. Over the same CF range, phase curves gradually changed from a concave-upward shape to a concave-downward shape. Phase curves consisted of two or three approximately straight segments. Group delay was analyzed separately for these segments. Generally, the largest group delay was observed near CF. With increasing SPL, most amplitude curves broadened, sometimes accompanied by a downward shift of best frequency, and group delay changed along the entire range of stimulus frequencies. We observed considerable across-ANF variation in the effects of SPL on both amplitude and phase. Overall, our data suggest that mechanical responses in the apex of the cochlea are considerably nonlinear and that these nonlinearities are of a different character than those known from the base of the cochlea.

Effect of sampling frequency on the measurement of phase-locked action potentials

Phase-locked spikes in various types of neurons encode temporal information. To quantify the degree of phase-locking, the metric called vector strength (VS) has been most widely used. Since VS is derived from spike timing information, error in measurement of spike occurrence should result in errors in VS calculation. In electrophysiological experiments, the timing of an action potential is detected with finite temporal precision, which is determined by the sampling frequency. In order to evaluate the effects of the sampling frequency on the measurement of VS, we derive theoretical upper and lower bounds of VS from spikes collected with finite sampling rates. We next estimate errors in VS assuming random sampling effects, and show that our theoretical calculation agrees with data from electrophysiological recordings in vivo. Our results provide a practical guide for choosing the appropriate sampling frequency in measuring VS.

Phase locking of auditory-nerve fibers to the envelopes of high-frequency sounds: implications for sound localization

Although listeners are sensitive to interaural time differences (ITDs) in the envelope of high-frequency sounds, both ITD discrimination performance and the extent of lateralization are poorer for high-frequency sinusoidally amplitude-modulated (SAM) tones than for low-frequency pure tones. Psychophysical studies have shown that ITD discrimination at high frequencies can be improved by using novel transposed-tone stimuli, formed by modulating a high-frequency carrier by a half-wave-rectified sinusoid. Transposed tones are designed to produce the same temporal discharge patterns in high-characteristic frequency (CF) neurons as occur in low-CF neurons for pure-tone stimuli. To directly test this hypothesis, we compared responses of auditory-nerve fibers in anesthetized cats to pure tones, SAM tones, and transposed tones. Phase locking was characterized using both the synchronization index and autocorrelograms. With both measures, phase locking was better for transposed tones than for SAM tones, consistent with the rationale for using transposed tones. However, phase locking to transposed tones and that to pure tones were comparable only when all three conditions were met: stimulus levels near thresholds, low modulation frequencies (<250 Hz), and low spontaneous discharge rates. In particular, phase locking to both SAM tones and transposed tones substantially degraded with increasing stimulus level, while remaining more stable for pure tones. These results suggest caution in assuming a close similarity between temporal patterns of peripheral activity produced by transposed tones and pure tones in both psychophysical studies and neurophysiological studies of central neurons.

Auditory neuropathy/dys-synchrony and its perceptual consequences

Auditory neuropathy/dys-synchrony is a form of hearing impairment in which cochlear outer hair cell function is spared but neural transmission in the auditory pathway is disordered. This condition, or group of conditions with a common physiologic profile, accounts for approximately 7% of permanent childhood hearing loss and a significant (but as yet undetermined) proportion of adult impairment. This paper presents an overview of the mechanisms underlying auditory neuropathy/dys-synchrony-type hearing loss and the clinical profile for affected patients. In particular it examines the perceptual consequences of auditory neuropathy/dys-synchrony, which are quite different from those associated with sensorineural hearing loss, and considers currently available, and future management options.

Distribution of estrogen receptor alpha mRNA in the brain and inner ear of a vocal fish with comparisons to sites of aromatase expression

Among vertebrates, teleost fish have the greatest capacity for estrogen production in the brain. Previously, we characterized the distribution of the estrogen-synthesizing enzyme aromatase in the brain of the midshipman fish. Here, we investigated the distribution of estrogen receptor alpha (ERalpha). A partial cDNA of ERalpha was cloned and used to generate midshipman-specific primers for RT and real-time PCR which identified transcripts in liver and ovary, the CNS, and the sensory epithelium of the main auditory endorgan (sacculus). In situ hybridization revealed abundant expression throughout the preoptic area, a vocal-acoustic site in the hypothalamus, amygdala homologs of the dorsal pallium, the pineal organ, the inner ear, the pituitary, and the ovary. Weaker expression was found in the midbrain's nucleus of the medial longitudinal fasciculus and in the dimorphic vocal motor nucleus. ERalpha expression in the pineal, gonad, and pituitary axis may function to time seasonal abiotic cues to reproductive state, while expression in the vocal motor and auditory systems support neurophysiological evidence for estrogen as a modulator of vocal motor and auditory encoding mechanisms in midshipman fish. While ERalpha is restricted to specific nuclei, aromatase expression is abundant in glial cells throughout the entire forebrain, and high in midbrain and hindbrain - spinal vocal regions. The only site of aromatase-containing neurons is in the peripheral auditory system, where it is localized to ganglion cells in the auditory nerve. Estrogen production proximal to ERalpha-positive neurons may provide for focal sites of estrogen effects on reproductive-, vocal-, and auditory-related neurons.

Hair-cell versus afferent adaptation in the semicircular canals

The time course and extent of adaptation in semicircular canal hair cells was compared to adaptation in primary afferent neurons for physiological stimuli in vivo to study the origins of the neural code transmitted to the brain. The oyster toadfish, Opsanus tau, was used as the experimental model. Afferent firing-rate adaptation followed a double-exponential time course in response to step cupula displacements. The dominant adaptation time constant varied considerably among afferent fibers and spanned six orders of magnitude for the population ( approximately 1 ms to >1,000 s). For sinusoidal stimuli (0.1-20 Hz), the rapidly adapting afferents exhibited a 90 degrees phase lead and frequency-dependent gain, whereas slowly adapting afferents exhibited a flat gain and no phase lead. Hair-cell voltage and current modulations were similar to the slowly adapting afferents and exhibited a relatively flat gain with very little phase lead over the physiological bandwidth and dynamic range tested. Semicircular canal microphonics also showed responses consistent with the slowly adapting subset of afferents and with hair cells. The relatively broad diversity of afferent adaptation time constants and frequency-dependent discharge modulations relative to hair-cell voltage implicate a subsequent site of adaptation that plays a major role in further shaping the temporal characteristics of semicircular canal afferent neural signals.

Steroid-Dependent Auditory Plasticity Leads to Adaptive Coupling of Sender and Receiver

For seasonally breeding vertebrates, reproductive cycling is often coupled with changes in vocalizations that function in courtship and territoriality. Less is known about changes in auditory sensitivity to those vocalizations. Here, we show that nonreproductive female midshipman fish treated with either testosterone or 17beta-estradiol exhibit an increase in the degree of temporal encoding of the frequency content of male vocalizations by the inner ear that mimics the reproductive female's auditory phenotype. This sensory plasticity provides an adaptable mechanism that enhances coupling between sender and receiver in vocal communication.

Pitch discrimination of diotic and dichotic tone complexes: harmonic resolvability or harmonic number?

Three experiments investigated the relationship between harmonic number, harmonic resolvability, and the perception of harmonic complexes. Complexes with successive equal-amplitude sine- or random-phase harmonic components of a 100- or 200-Hz fundamental frequency (f0) were presented dichotically, with even and odd components to opposite ears, or diotically, with all harmonics presented to both ears. Experiment 1 measured performance in discriminating a 3.5%-5% frequency difference between a component of a harmonic complex and a pure tone in isolation. Listeners achieved at least 75% correct for approximately the first 10 and 20 individual harmonics in the diotic and dichotic conditions, respectively, verifying that only processes before the binaural combination of information limit frequency selectivity. Experiment 2 measured fundamental frequency difference limens (f0 DLs) as a function of the average lowest harmonic number. Similar results at both f0's provide further evidence that harmonic number, not absolute frequency, underlies the order-of-magnitude increase observed in f0 DLs when only harmonics above about the 10th are presented. Similar results under diotic and dichotic conditions indicate that the auditory system, in performing f0 discrimination, is unable to utilize the additional peripherally resolved harmonics in the dichotic case. In experiment 3, dichotic complexes containing harmonics below the 12th, or only above the 15th, elicited pitches of the f0 and twice the f0, respectively. Together, experiments 2 and 3 suggest that harmonic number, regardless of peripheral resolvability, governs the transition between two different pitch percepts, one based on the frequencies of individual resolved harmonics and the other based on the periodicity of the temporal envelope.

Asymmetry of masking between noise and iterated rippled noise: evidence for time-interval processing in the auditory system

This study describes the masking asymmetry between noise and iterated rippled noise (IRN) as a function of spectral region and the IRN delay. Masking asymmetry refers to the fact that noise masks IRN much more effectively than IRN masks noise, even when the stimuli occupy the same spectral region. Detection thresholds for IRN masked by noise and for noise masked by IRN were measured with an adaptive two-alternative, forced choice (2AFC) procedure with signal level as the adaptive parameter. Masker level was randomly varied within a 10-dB range in order to reduce the salience of loudness as a cue for detection. The stimuli were filtered into frequency bands, 2.2-kHz wide, with lower cutoff frequencies ranging from 0.8 to 6.4 kHz. IRN was generated with 16 iterations and with varying delays. The reciprocal of the delay was 16, 32, 64, or 128 Hz. When the reciprocal of the IRN delay was within the pitch range, i.e., above 30 Hz, there was a substantial masking asymmetry between IRN and noise for all filter cutoff frequencies; threshold for IRN masked by noise was about 10 dB larger than threshold for noise masked by IRN. For the 16-Hz IRN, the masking asymmetry decreased progressively with increasing filter cutoff frequency, from about 9 dB for the lowest cutoff frequency to less than 1 dB for the highest cutoff frequency. This suggests that masking asymmetry may be determined by different cues for delays within and below the pitch range. The fact that masking asymmetry exists for conditions that combine very long IRN delays with very high filter cutoff frequencies means that it is unlikely that models based on the excitation patterns of the stimuli would be successful in explaining the threshold data. A range of time-domain models of auditory processing that focus on the time intervals in phase-locked neural activity patterns is reviewed. Most of these models were successful in accounting for the basic masking asymmetry between IRN and noise for conditions within the pitch range, and one of the models produced an exceptionally good fit to the data.

Mechanical bases of frequency tuning and neural excitation at the base of the cochlea: comparison of basilar-membrane vibrations and auditory-nerve-fiber responses in chinchilla

We review the mechanical origin of auditory-nerve excitation, focusing on comparisons of the magnitudes and phases of basilar-membrane (BM) vibrations and auditory-nerve fiber responses to tones at a basal site of the chinchilla cochlea with characteristic frequency approximately 9 kHz located 3.5 mm from the oval window. At this location, characteristic frequency thresholds of fibers with high spontaneous activity correspond to magnitudes of BM displacement or velocity in the order of 1 nm or 50 microm/s. Over a wide range of stimulus frequencies, neural thresholds are not determined solely by BM displacement but rather by a function of both displacement and velocity. Near-threshold, auditory-nerve responses to low-frequency tones are synchronous with peak BM velocity toward scala tympani but at 80-90 dB sound pressure level (in decibels relative to 20 microPascals) and at 100-110 dB sound pressure level responses undergo two large phase shifts approaching 180 degrees. These drastic phase changes have no counterparts in BM vibrations. Thus, although at threshold levels the encoding of BM vibrations into spike trains appears to involve only relatively minor signal transformations, the polarity of auditory-nerve responses does not conform with traditional views of how BM vibrations are transmitted to the inner hair cells. The response polarity at threshold levels, as well as the intensity-dependent phase changes, apparently reflect micromechanical interactions between the organ of Corti, the tectorial membrane and the subtectorial fluid, and/or electrical and synaptic processes at the inner hair cells.

The perceptual tone/noise ratio of merged iterated rippled noises

Iterated rippled noise (IRN) is constructed by delaying a random noise by d ms, adding it back to the same noise, and repeating the process iteratively. When two IRNs with the same power but slightly different delays are added together, the perceptual tone/noise ratio of the "merged" IRN is markedly reduced with respect to that of either of the component IRNs. In this paper, the reduction in the perceptual tone/noise ratio is measured for IRNs in which one of the delays is always 16 ms and the other is either 16 +/- 0.1 ms or 16 +/- 1.1 ms. The component IRNs have the same number of iterations, and the number varies across conditions from 4 to 256. The perceptual tone/noise ratio is measured using a discrimination matching procedure developed for single IRNs; each merged IRN is compared with a range of "standard" stimuli having varying proportions of a complex tone and a broadband noise [Patterson et al., J. Acoust. Soc. Am. 100, 3286-3294 (1996)]. For single IRNs, the function relating the signal-to-noise ratio of the matching standard to the number of iterations in the IRN was found to be essentially straight. This relationship was explained in terms of the height of the first peak in the autocorrelation of the stimulus wave, or by the first peak in the summary autocorrelogram produced by a time-domain auditory model. For the merged IRNs in the current experiment, the matching-point functions are found to have pronounced downward curvature, in addition to being well below the function for single IRNs. To account for the reduction in the perceptual tone/noise ratio of merged IRNs, the autocorrelation model was extended to include a simple rule for combining adjacent peaks in the autocorrelation function of the wave, and the autocorrelogram model was revised to improve the simulation of the "loss of phase locking" at higher frequencies in the autocorrelogram.

Potassium currents in auditory hair cells of the frog basilar papilla

The whole-cell patch-clamp technique was used to identify and characterize ionic currents in isolated hair cells of the leopard frog basilar papilla (BP). This end organ is responsible for encoding the upper limits of a frog's spectral sensitivity (1.25-2.0 kHz in the leopard frog). Isolated BP hair cells are the smallest hair cells in the frog auditory system, with spherical cell bodies typically less than 20 microm in diameter and exhibiting whole-cell capacitances of 4-7 pF. Hair cell zero-current resting potentials (Vz) varied around a mean of -65 mV. All hair cells possessed a non-inactivating, voltage-dependent calcium current (I(Ca)) that activates above a threshold of -55 mV. Similarly all hair cells possessed a rapidly activating, outward, calcium-dependent potassium current (I(K)(Ca)). Most hair cells also possessed a slowly activating, outward, voltage-dependent potassium current (I(K)), which is approximately 80% inactive at the hair cell Vz, and a fast-activating, inward-rectifying potassium current (I(K1)) which actively contributes to setting Vz. In a small subset of cells I(K) was replaced by a fast-inactivating, voltage-dependent potassium current (I(A)), which strongly resembled the A-current observed in hair cells of the frog sacculus and amphibian papilla. Most cells have very similar ionic currents, suggesting that the BP consists largely of one homogeneous population of hair cells. The kinetic properties of the ionic currents present (in particular the very slow I(K)) argue against electrical tuning, a specialized spectral filtering mechanism reported in the hair cells of birds, reptiles, and amphibians, as a contributor to frequency selectivity of this organ. Instead BP hair cells reflect a generalized strategy for the encoding of high-frequency auditory information in a primitive, mechanically tuned, terrestrial vertebrate auditory organ.

Predicting the temporal responses of non-phase-locking bullfrog auditory units to complex acoustic waveforms

Axons from the basilar papilla of the American bullfrog (Rana catesbeiana) do not phase lock to stimuli within an octave of their best frequencies. Nevertheless, they show consistent temporal patterns of instantaneous spike rate (as reflected in peristimulus time histograms) in response to repeated stimuli in that frequency range. We show that the second-order Wiener kernels for these axons, derived from the cross-correlation of continuous (non-repeating), broad-band noise stimulus with the spike train produced in response to that stimulus, can predict with considerable precision the temporal pattern of instantaneous spike rate in response to a novel, complex acoustic waveform (a repeated, 100-ms segment of noise, band-limited to cover the single octaves above and below best frequency). Furthermore, we show that most of this predictive power is retained when the second-order Wiener kernel is reduced to the highest-ranking pair of singular vectors derived from singular-value decomposition, that the retained pair of vectors corresponds to a single auditory filter followed by an envelope-detection process, and that the auditory filter itself predicts the characteristic frequency (CF) of the axon and the shape of the frequency-threshold tuning curve in the vicinity of CF.

THE ROLE OF TIMING IN THE BRAIN STEM AUDITORY NUCLEI OF VERTEBRATES

Vertebrate animals gain biologically important information from environmental sounds. Localization of sound sources enables animals to detect and respond appropriately to danger, and it allows predators to detect and localize prey. In many species, rapidly fluctuating sounds are also the basis of communication between conspecifics. This information is not provided directly by the output of the ear but requires processing of the temporal pattern of firing in the tonotopic array of auditory nerve fibers. The auditory nerve feeds information through several parallel ascending pathways. Anatomical and electrophysiological specializations for conveying precise timing, including calyceal synaptic terminals and matching axonal conduction times, are evident in several of the major ascending auditory pathways through the ventral cochlear nucleus and its nonmammalian homologues. One pathway that is shared by all higher vertebrates makes an ongoing comparison of interaural phase for the localization of sound in the azimuth. Another pathway is specifically associated with higher frequency hearing in mammals and is thought to make use of interaural intensity differences for localizing high-frequency sounds. Balancing excitation from one ear with inhibition from the other in rapidly fluctuating signals requires that the timing of these synaptic inputs be matched and constant for widely varying sound stimuli in this pathway. The monaural nuclei of the lateral lemniscus, whose roles are not understood (although they are ubiquitous in higher vertebrates), receive input from multiple pathways that encode timing with precision, some through calyceal endings.

Response phase: a view from the inner hair cell

Inner hair cell (IHC) responses are recorded from the apical three turns of the guinea pig cochlea in order to define the relationship between hair cell depolarization and position of the basilar membrane. At low frequencies, inner hair cell depolarization is generally observed near basilar membrane velocity to scala vestibuli, reflecting the putative freestanding nature of the IHC's stereocilia. While this is consistent with previous IHC results, independent of location, and with neural responses for fibers with low best frequencies, it is inconsistent with single-unit results from the base of the cochlea, where response phase is associated with basilar membrane velocity to scala tympani. Results suggest that the temporal disparity between IHC and neural data from the base of the cochlea may relate to several factors that influence transmembrane voltage in IHCs. First, extracellular voltages (Ingvarsson, 1981; Sellick et al., 1982; Russell and Sellick, 1983) can potentially affect low- and high-frequency regions differently because electrical interactions are more likely in the base of the cochlea than in the apex (Dallos, 1983, 1985). Second, waveform distortion and kinetic properties associated with voltage-dependent ion channels in the IHC's basolateral membrane can both influence response phase by adding harmonic components and lagging the receptor potential by as much as 90 deg. Third, the velocity dependence of IHCs in the apex appears to extend to higher frequencies than the velocity dependence demonstrated for IHCs in the base of the cochlea. These features, which influence the timing of discharges in the auditory nerve, are compared and evaluated.

Cochlear mechanisms of frequency and intensity coding. II. Dynamic range and the code for loudness

Our preceding paper described SPL-dependent changes in the shape of transfer functions recorded from inner and outer hair cells as well as supporting cells, in the 500-2500 Hz regions of the Mongolian gerbil cochlea. As SPL was increased, large shifts were observed in the peak of the transfer function. A strongly compressive nonlinearity was also observed at CF. This paper examines the data from the perspective of intensity coding in the auditory periphery. Based on the data, we offer a new explanation for the mechanisms underlying the different dynamic ranges of low and high threshold auditory neurons. We also find that, for pure tone stimuli, the growth of excitation at the characteristic place saturates rapidly, and cannot encode the wide dynamic range of loudness. The data are analyzed to explore other excitation pattern candidates for loudness coding. The growth of the peak of the IHC transfer function, as well as the growth of the response area, have been found to be linearly related to loudness growth over most of its dynamic range. Implications of the data for auditory intensity coding are discussed.

Temporal aspects of the effects of cooling on responses of single auditory nerve fibers

During an investigation of the effects of cochlear cooling on frequency tuning and input/output relations of single auditory nerve fibers in gerbil (Ohlemiller and Siegel (1994) Hear. Res. 80, 174-190), cooling-related changes in post-stimulus time histogram (PSTH) shape and phase-locking to tonebursts were characterized in a small sample of neurons. Local cochlear cooling by 5-10 degrees C below normal core temperature did not alter overall PSTH shape, although some evidence was found for a reduction in the time constants of rapid and short term rate adaptation. The relative contributions of rapid and short term response components appeared unaltered. Effects of cooling on phase-locking were assessed by calculating the synchronization index for responses to intense ( > 70 dB SPL) tonebursts at 0.5, 1.0, and 2.0 kHz. Synchronization filter functions exhibited modest reductions in both magnitude and the upper frequency limit of phase-locking. The effects of cooling on the temporal character of responses appear distinct from those of a simple reduction in stimulus intensity. Results are interpreted in terms of cooling-related changes in responses of cochlear hair cells and afferent neurons, and suggest that temperature artifacts are unlikely to underlie reported species differences in PSTH shape and phase-locking.

Spontaneous classification of complex tones at high and ultrasonic frequencies in the bat, Megaderma lyra

Megaderma lyra, a bat species using harmonically structured calls for echolocation, exploits the spectral content of its echoes for texture discrimination. It is the aim of the present study to test according to which sensory qualities harmonic complex tones are spontaneously classified by this bat. The applied experimental paradigm is especially adapted to the preference of M. lyra to use absolute pitch cues. Three animals were trained in a 2-AFC procedure to classify three-component stimuli as low or high, with all their harmonics below or above a pure tone reference of 33 kHz, respectively. Later, the original tones were interspersed with "incomplete" test stimuli, with their fundamentals (and lower harmonics) missing. These were ambiguous in that their possible virtual, i.e., collective pitches were below the reference whereas their pure tone pitches were above it. Bat 1 classified 22 of 23 test stimuli with missing fundamentals between 5.3 and 28.3 kHz according to their collective pitches, whereas bat 2 judged all presented ambiguous tones on the basis of their pure tone pitches. Bat 3 failed the pitch control criterion which is why in this case results cannot be interpreted unequivocally. The implications of these findings are discussed with respect to the bats' behavioral context, as well as to psychoacoustical models of the formation of the pitch of complex tones.

Dissecting the frog inner ear with Gaussian noise. I. Application of high-order Wiener-kernel analysis

Wiener kernel analysis was used to characterize the auditory pathway from tympanic membrane to single primary auditory nerve fibers in the European edible frog, Rana esculenta. Nerve fiber signals were recorded in response to white Gaussian noise. By cross-correlating the noise stimulus and the nerve fiber response, we computed (1) the full second-order Wiener kernel, and (2) the diagonals of the zeroth- to fourth-order Wiener kernels. These diagonals are usually referred to as polynomial correlation functions. The measured Wiener kernels were fitted with a 'sandwich' model. A new fitting procedure was used to compute the response characteristics of (1) the first filter, (2) the static nonlinearity, and (3) the second filter, which form the functional components of the model. The first filter is a bandpass filter. In the majority of low frequency fibers, with best excitatory frequency (BEF) < 800 Hz, this filter was tuned to two frequencies. This dual tuning mechanism gives rise to 'off-diagonal' components in the second-order Wiener kernel. The static nonlinearity resembles a rectifier, and is dominated by second-order (quadratic) nonlinearity. As a function of BEF, the shape of the nonlinearity changes systematically. Finally, the last filter in the model was a low pass filter. Across fibers, its cutoff frequency f-3dB ranged from 106 to 434 Hz.

Dissecting the frog inner ear with Gaussian noise. II. Temperature dependence of inner ear function

The temperature dependence of the response of single primary auditory nerve fibers (n = 31) was investigated in the European edible frog, Rana esculenta (seven ears). Nerve fiber responses were analyzed with Wiener kernel analysis and polynomial correlation. The responses were described with a cascade model, consisting of a linear bandpass filter, a static nonlinearity, and a linear lowpass filter. From the computed Wiener kernels and the polynomial correlation functions, the characteristics of the three model components were obtained. With increasing temperature (1) tuning of the first filter increased in the majority (n = 16) of amphibian papilla fibers (best excitatory frequency, BEF < 1 kHz, n = 21) but remained unchanged in the majority (n = 10) of basilar papilla fibers (BEF > 1 kHz, n = 11), (2) the gain of the first filter remained unchanged, (3) the shape of nonlinear IO function remained unchanged, (4) the combined gain of the static nonlinearity and the second filter usually increased, but displayed considerable scatter across fibers (from -0.7 dB/degrees C to 3 dB/degrees C), and (5) the cutoff frequency of the second lowpass filter increases, with average 0.13 oct/degrees C. The immunity of the shape of the nonlinearity is considered evidence of a temperature independent gating mechanism in the transduction channels. The temperature dependence of the second filter may have resulted from a decrease of the hair cell membrane resistance, but may also reflect changes in subsequent staging of nerve fiber excitation.

Cochlear mechanisms of frequency and intensity coding. I. The place code for pitch

In the past, several researchers have reported a substantial shift in the peak of the tone-evoked excitation pattern toward the base of the cochlea following an increase in the SPL of the stimulating tone. Evidence for such peak shifts has been found in the responses of auditory nerve fibers, cochlear microphonics, and the responses of outer hair cells and supporting cells in the cochlea, as well as in basilar membrane vibration measurements, and indirectly, in psychophysical data. However, direct evidence for such a peak shift in inner hair cell (IHC) responses has been relatively sparse. If the peak shift is preserved in the information conveyed to the auditory nerve fibers by the IHCs, the classical 'place theory' for frequency coding in the cochlea requires modification. In this study, the nature and extent of the SPL-dependent peak shift is examined with the help of recordings in the IHCs and other cells of the organ of Corti in the 0.5-2.5 kHz region of the Mongolian gerbil cochlea. It is demonstrated that the peak shift is a universal phenomenon in the diverse cell types in this region of the cochlea. Most importantly, a large SPL-dependent peak shift is demonstrated in IHC responses. On the other hand, the recordings indicate that the apical cutoff of the spatial excitation pattern is SPL-independent. We conclude, therefore, that the place theory of pitch perception must be abandoned or at least modified.

Phase locking to high frequencies in the auditory nerve and cochlear nucleus magnocellularis of the barn owl, Tyto alba

The auditory system of the barn owl is an important model for temporal processing on a very fast time scale and for the neural mechanisms and circuitry underlying sound localization. Phase locking has been shown to be the behaviorally relevant temporal code. This study examined the quality and intensity dependence of phase locking in single auditory nerve fibers of the barn owl to define the input to the known brainstem circuit for temporal processing. For direct comparison in the same individuals, recordings were also obtained from the relevant next higher center, the nucleus magnocellularis (NM). Phase locking was regularly seen at sound pressure levels (SPL) below those eliciting an increase in spike rate, thus providing an additional cue for signal detection. The quality of phase locking, expressed as vector strength, decreased with increasing frequency. Auditory nerve fibers showed an unusual step-like decline with a prominent plateau in the mid-frequency range (1.5-3 kHz), indicating that some specialization enables the owl to halt the deterioration and extend phase locking to frequencies up to 10 kHz, above the range commonly observed in other species. Phase locking in the NM was consistently inferior to that of auditory-nerve fibers at frequencies above 1 kHz, suggesting that the synapse plays a limiting role in temporal precision. The response delays, or group delays, derived from the phase-versus-frequency functions of auditory nerve fibers were not consistent with the unusual spatial frequency representation in the owl cochlea. This questions the common assumption that group delays reflect cochlear wave travel times.

Intracellular distributions and putative functions of calcium-binding proteins in the bullfrog vestibular otolith organs

Hair cells in the bullfrog vestibular otolith organs were immunolabeled by monoclonal and polyclonal antisera against calbindin (CaB), calmodulin (CaM), calretinin (CaR), and parvalbumin (PA). S-100, previously shown to immunolabel striolar hair cells in fish vestibular organs, only weakly immunolabeled hair cells in the bullfrog vestibular otolith organs. Immunolabeling was not detected in supporting cells. With the exception of CaR, myelinated axons and unmyelinated nerve terminals were immunolabeled by all of the above antisera. Immunolabeling was seen in all saccular hair cells, although hair cells at the macular margins were immunolabeled more intensely for CaB, CaM, and PA than more centrally located hair cells. As the macula margins are known to be a growth zone, this labeling pattern suggests that marginal hair cells up-regulate their calcium-binding proteins during hair cell development. In the utriculus, immunolabeling for CaM and PA was generally restricted to striolar hair cells. CaR immunolabeling was restricted to the stereociliary array. Immunolabeling for other calcium-binding proteins was generally seen in both the cell body and hair bundles of hair cells, although this labeling was often localized to the stereociliary array and the apical portion of the cell body. CaM and PA immunolabeling in the stereociliary array in saccular and utricular striolar cells suggests a functional role for these proteins in mechanoelectric transduction and adaptation.

Temperature-dependence of saccular nerve fiber response in the North American bullfrog

A clinical microwave device was used to heat the head and ear of the North American bullfrog in order to observe the temperature dependence of tuning in the sacculus, an organ known to possess the capability of electrical resonance in its hair cells. In tuning curves derived from reverse correlation analysis with noise stimuli, the temperature dependencies of the frequencies of tuning peaks and notches typically exhibited Q10s less than 1.1; whereas the frequencies of electrical resonances are expected to have Q10s of the order of 1.7. Therefore we conclude that electrical resonances are not significantly involved in tuning in the bullfrog sacculus.

Longitudinal comparisons of IHC ac and dc receptor potentials recorded from the guinea pig cochlea

Recordings were made from inner hair cells (IHC) at three locations distributed in the apical half of the guinea pig cochlea. Longitudinal variations in ac and dc components of receptor potentials produced in response to single-tone inputs were studied to further understand the ways in which IHCs communicate with their innervating afferent dendrites. While neural synchrony probably depends on the ac receptor potential, discharge rate may be controlled by the dc receptor potential generated by the IHC transducer plus an ac component derived from the phasic receptor potential. The latter reflects low-pass filtering inherent in the hair cell's basolateral membrane and calcium-dependent synaptic processes. By comparing the frequency dependence of ac and dc components in cells with different characteristic frequencies, it may be possible to learn how neural response areas are formed and why their shapes change along the cochlear spiral.

Cochlear function reflected in mammalian hair cell responses

This report describes response patterns recorded in inner and outer hair cells in the apical three turns of the guinea pig cochlea. Characteristic frequencies (CF) in these regions are approximately 270 Hz in turn four, 1000 Hz in turn three and 4000 Hz in turn two. Although the two receptor types exhibit differences in resting membrane potentials and in response phase at low stimulus frequencies, they both produce ac and dc receptor potentials in response to sound. When measured around CF, both cell types produce a depolarizing dc response at low and moderate levels. This contrasts with results published for basal-turn outer hair cells (Russell and Sellick, 1983; Cody and Russell, 1987) whose responses become asymmetrical only at very high levels.

Two-tone suppression in inner hair cell responses: correlates of rate suppression in the auditory nerve

Inner hair cell (IHC) recordings were made from second turn of the guinea pig cochlea where characteristic frequencies are approximately 4000 Hz. In order to compare IHC responses with rate suppression measured in the auditory nerve, suppressors were introduced that produced little or no response in the hair cell. The effects of a variable-frequency suppressor on a constant-frequency probe, placed near characteristic frequency, were also investigated since this paradigm is commonly used in single unit experiments. Resulting magnitude changes were measured in the fundamental component of the ac receptor potential and/or in the total dc produced in the region of temporal overlap between the two stimulus inputs. This latter component is especially important when considering how changes in IHC responses relate to decreases in discharge rate in single auditory nerve fibers. Since the ac receptor potential is filtered by the hair cell's basolateral membrane, the dc component probably controls transmitter release at the characteristic frequency of these second-turn IHCs. Based on results from these and previous experiments, a proposal is advanced to explain the evolution of two-tone suppression in the peripheral auditory system. The paper also discusses the use of excitatory versus non-excitatory suppressors and includes a description of two-tone suppression areas at the mechanical, IHC and single unit levels. The explanation of low-side suppression areas is of special interest since hitherto they have been difficult to model (Kim, 1985).

Physiological correlates of off-frequency listening

Recordings are made from inner hair cells (IHC) in the second turn of the guinea pig cochlea where characteristic frequencies (CF) are approximately 4000 Hz. Results from experiments using two stimulus inputs suggest that the characterization of two-tone suppression at this more-basal recording location is similar to the reported for IHCs in the third turn (Cheatham and Dallos, 1989, 1990a, 1990b). For example, introduction of a suppressor causes IHC frequency response functions to become narrower with the smallest magnitude reductions occurring between 1/2 to 1 octave below CF. In this frequency region, where suppression is minimal, it was also observed that suppressor magnitude was reduced by the probe. In other words, the mutual suppression of probe and suppressor may contribute to the sharpening of these functions. Since the peak of the frequency response function shifts to a lower frequency in the presence of the suppressor, these results may provide a physiological correlate of the psychophysical phenomenon known as 'off-frequency listening.'

Phase-locking of auditory-nerve discharges to sinusoidal electric stimulation of the cochlea

The activity of auditory-nerve fibers was recorded in anesthetized cats in response to sinusoidal electric stimuli applied through a bipolar electrode pair inserted about 5 mm into the cochlea through the round window. The synchronization index was calculated from period histograms for frequencies ranging from 0.2 to over 10 kHz. The stimulus artifact was largely eliminated through the use of differential micropipettes and an adaptive digital filter. Measured synchronization indices were many times larger than the indices that could be attributed to the residual stimulus artifact. Synchronization indices at each stimulus frequency varied considerably from fiber to fiber, even in the same animal. The dependence of synchrony on stimulus frequency was also variable, decreasing monotonically in some fibers and nonmonotonically in others. The average electric synchronization index for all fibers did not fall as steeply with frequency as does the average synchrony for acoustic stimuli. The finding of significant phase locking to electric stimuli well above 1 kHz suggests that the poor frequency discrimination of cochlear-implant recipients for single-channel stimulation above this frequency may be due to the inability of the central processor to make effective use of the available phase-locking information for monaural stimulation.

Postsynaptic inhibition can explain the concentration of short inter-spike-intervals in avian auditory nerve fibres

Spontaneous and sound-evoked single-unit activity was recorded from afferent neurones in the cochlear ganglion of the anaesthetized pigeon. The histogram of successive intervals of spontaneous activity of 51% of neurones exhibited more short intervals than expected from a Poisson point-process description of spike times; for another 43% of neurones the point-process was Poisson. A model of spike generation was developed to account for the concentration of short spike-intervals. The proposed model contains inhibitory postsynaptic potentials at the afferent dendrite, in addition to the excitatory postsynaptic potentials. Not only does the model reproduce the first-order interval statistics of neural activity, but it provides a mechanism for improving phase-locking to the fundamental frequency of a sinusoid, and also offers an explanation for the presence of reciprocal synapses in the human cochlea.

First order temporal properties of spontaneous and tone-evoked activity of auditory afferent neurones in the cochlear ganglion of the pigeon

Spontaneous and tone-evoked single-unit activity was recorded from afferent neurones in the cochlear ganglion of the anaesthetized pigeon, and the data analysed in a way that allowed the physics of underlying mechanisms to be described. The periodicity of neural activity was quantified by Fourier analysis of the histogram of successive spike intervals. Spontaneous activity was quasiperiodic for 57% of neurones (average rate: 74 s-1); it was irregular for the remainder of neurones (average rate: 55 s-1). The preferred frequency (PF) of the quasiperiodic spontaneous activity was, on average, equal to the characteristic frequency (CF) of the neurone (70% of cases) or CF/2 (30%). This observation can be explained by supposing that preferred intervals of spontaneous activity are generated by noise passing through a filter tuned to the CF of the neurone; in most cases (70%) discharge was synchronized to CF, but in the others the neurone fired to every second cycle of the filtered signal. Consistent with this interpretation, for 79% of neurones, the modal interval of spontaneous activity was, on average, directly proportional to the CF-period, irrespective of whether preferred intervals were detected. The synchronization index at the PF was inversely related to the PF, and was quantified by the amplitude response of a first-order low-pass filter with cutoff frequency of 48 +/- 18 Hz. The spontaneous activity of 9% of neurones exhibited a second-harmonic component of the PF. For both tone-evoked and spontaneous activity, the observed synchronization indices of harmonics of the stimulus frequency or of the PF were consistent with an underlying exponential spike-generator function. If such a function does indeed govern spike generation, then it implies that the Shannon entropy of the probability density function of the instantaneous firing rate is near its maximum value and suggests that the system is close to statistical equilibrium. Single-tone rate-suppression was detected for 53% of those neurones that exhibited multiple preferred intervals of spontaneous activity. It is conjectured that the phenomena of quasiperiodic spontaneous activity and single-tone rate-suppression are different aspects of a single presynaptic process. According to this model, we would expect to find these two phenomena in animals that have auditory fibres innervating electrically tuned hair cells, and that have stereocilia firmly coupled to a tectorial membrane.

Mechanisms that degrade timing information in the cochlea

Action potentials of cochlear nerve fibers are synchronized to the temporal variations of sounds, but this synchronization is attenuated for high-frequency sounds. In cochleas from a number of vertebrates, the frequency dependence of synchronization can be represented as a lowpass filter process whose order is at least three (Weiss and Rose, 1988a); i.e. at least three first-order kinetic processes may be responsible for the loss of synchronization. In this paper we assess the extent to which calcium processes, that are essential for chemical transmission at the hair-cell neuron junction, contribute to this attenuation of synchronization. We analyze a model of calcium processes in hair cells (Lewis, 1985; Hudspeth and Lewis, 1988a) for sinusoidal receptor potentials. We show that: (1) the relation between the receptor potential and the calcium current, which is nonlinear, acts approximately as a first-order lowpass filter whose cut-off frequency decreases with increasing receptor potential magnitude; (2) the relation between calcium current and calcium-concentration is a first-order, lowpass filter with constant cutoff frequency. These two calcium processes plus the lowpass-filter process resulting from the electrical resistance and capacitance of the hair-cell membrane - which limits the rate at which the receptor potential can change (Weiss and Rose, 1988b) - can account for much, although perhaps not for all, of the loss of synchronization of cochlear nerve fibers.

Basic properties of auditory-nerve responses from a "simple' ear: the basilar papilla of the frog

Spike discharges initiated by mammalian inner hair cells are produced by a complicated system involving both mechanical and neural components that normally operate in a bi-directional configuration involving multiple feedback loops. In contrast, the frog basilar papilla has the equivalent of inner hair cells, but lacks outer hair cells; it has no efferent system, and no basilar membrane. This suggests that the frog basilar papilla lacks some of the mechanical and neural feedback paths characteristic of the mammalian system. Detailed measurements of tuning curves, spontaneous activity and responses to tones an clicks reveal large parametric differences between frog and mammals in spontaneous rate, absolute refractory time, long-term adaptation and phase locking. Responses to tone bursts are qualitatively similar, but parametrically quite different. More focused examinations of these effects will be able to exploit the differences in adaptation to long- versus short-duration stimuli could be caused by depletion of afferent neurotransmitter or by activation of feedback loops involving the efferent system. In the basilar papilla, any differences in adaptation must result from changes in the afferent pathway alone.

On the determination of input sound frequencies by the auditory central processor

There are two schools of thought with regard to how the auditory central processor achieves good frequency resolution. The first is based on the steep 300 dB/oct rolloff; the second on the fact that the action potential (AP) output of neurons associated with hair cells is partially synchronized to the incoming fin or its subharmonics. The main objection to this proposal is that synchronization seems to fail for high-frequency audio inputs. It is shown that this failure may be due to experimental difficulties. It is impossible to avoid trauma to the cochlea and/or auditory nerve. To study synchronization at 10,000 Hz, the interspike-interval (ISI) histogram requires a timing accuracy of 4 microseconds or better despite input AP's that have a rise time of 200 microseconds. Synthetic AP ISI histograms are derived for a) unstimulated; b) fully synchronized; c) low-frequency; and d) high-frequency audio input conditions. The latter are compared with typical experimentally derived data. Hypothetical processing by reverberatory neurons is also considered.