GABAergic neural activity involved in salicylate-induced auditory cortex gain enhancement.

Although high doses of sodium salicylate impair cochlear function, it paradoxically enhances sound-evoked activity in the auditory cortex (AC) and augments acoustic startle reflex responses, neural and behavioral metrics associated with hyperexcitability and hyperacusis. To explore the neural mechanisms underlying salicylate (SS)-induced hyperexcitability and "increased central gain," we examined the effects of GABA receptor agonists and antagonists on SS-induced hyperexcitability in the AC and startle reflex responses. Consistent with our previous findings, local or systemic application of SS significantly increased the amplitude of sound-evoked AC neural activity, but generally reduced spontaneous activity in the AC. Systemic injection of SS also significantly increased the acoustic startle reflex. S-baclofen or R-baclofen, GABA-B agonists, which suppressed sound-evoked AC neural firing rate and local field potentials, also suppressed the SS-induced enhancement of the AC field potential and the acoustic startle reflex. Local application of vigabatrin, which enhances GABA concentration in the brain, suppressed the SS-induced enhancement of AC firing rate. Systemic injection of vigabatrin also reduced the SS-induced enhancement of acoustic startle reflex. Collectively, these results suggest that the sound-evoked behavioral and neural hyperactivity induced by SS may arise from a SS-induced suppression of GABAergic inhibition in the AC.

Salicylate increases the gain of the central auditory system.

High doses of salicylate, the anti-inflammatory component of aspirin, induce transient tinnitus and hearing loss. Systemic injection of 250 mg/kg of salicylate, a dose that reliably induces tinnitus in rats, significantly reduced the sound evoked output of the rat cochlea. Paradoxically, salicylate significantly increased the amplitude of the sound-evoked field potential from the auditory cortex (AC) of conscious rats, but not the inferior colliculus (IC). When rats were anesthetized with isoflurane, which increases GABA-mediated inhibition, the salicylate-induced AC amplitude enhancement was abolished, whereas ketamine, which blocks N-methyl-d-aspartate receptors, further increased the salicylate-induced AC amplitude enhancement. Direct application of salicylate to the cochlea, however, reduced the response amplitude of the cochlea, IC and AC, suggesting the AC amplitude enhancement induced by systemic injection of salicylate does not originate from the cochlea. To identify a behavioral correlate of the salicylate-induced AC enhancement, the acoustic startle response was measured before and after salicylate treatment. Salicylate significantly increased the amplitude of the startle response. Collectively, these results suggest that high doses of salicylate increase the gain of the central auditory system, presumably by down-regulating GABA-mediated inhibition, leading to an exaggerated acoustic startle response. The enhanced startle response may be the behavioral correlate of hyperacusis that often accompanies tinnitus and hearing loss.

Auditory nerve fiber responses following chronic cochlear de-efferentation.

The aim of the present study was to examine the role of the olivocochlear system in auditory processing by examining the long-term effects of cochlear de-efferentation on auditory nerve response properties in adult chinchillas. Spontaneous rates, response thresholds, tuning curves, discharge rate-level functions, and adaptation of single auditory nerve fibers were measured in chinchillas with complete cochlear de-efferentation produced by sectioning the olivocochlear bundle in the internal auditory meatus. De-efferentation was verified as successful on the basis of acetylcholinesterase staining of surface preparations of the organ of Corti. Following chronic de-efferentation, there was a striking decrease in spontaneous rate, consistent with earlier observations in cats. In addition, the present study shows that complete de-efferentation results in: (1) increased driven discharge rates and decreased dynamic range of discharge rate-level functions, (2) larger onset-to-steady state ratio of discharge rate at moderate intensities, and (3) a hypersensitive tail of the tuning curve. These effects, largely confined to neurons that were most sensitive to frequencies between 2-8 kHz, indicate that the cochlear efferent system is important in maintaining normal function (e.g., frequency and intensity selectivity) of the auditory periphery by modulating auditory nerve fiber response properties.

The functional anatomy of the normal human auditory system: responses to 0.5 and 4.0 kHz tones at varied intensities.

Most functional imaging studies of the auditory system have employed complex stimuli. We used positron emission tomography to map neural responses to 0.5 and 4.0 kHz sine-wave tones presented to the right ear at 30, 50, 70 and 90 dB HL and found activation in a complex neural network of elements traditionally associated with the auditory system as well as non-traditional sites such as the posterior cingulate cortex. Cingulate activity was maximal at low stimulus intensities, suggesting that it may function as a gain control center. In the right temporal lobe, the location of the maximal response varied with the intensity, but not with the frequency of the stimuli. In the left temporal lobe, there was evidence for tonotopic organization: a site lateral to the left primary auditory cortex was activated equally by both tones while a second site in primary auditory cortex was more responsive to the higher frequency. Infratentorial activations were contralateral to the stimulated ear and included the lateral cerebellum, the lateral pontine tegmentum, the midbrain and the medial geniculate. Contrary to predictions based on cochlear membrane mechanics, at each intensity, 4.0 kHz stimuli were more potent activators of the brain than the 0.5 kHz stimuli.

The functional neuroanatomy of tinnitus: evidence for limbic system links and neural plasticity.

We used PET to map brain regions responding to changes in tinnitus loudness in four patients who could alter tinnitus loudness by performing voluntary oral facial movements (OFMs). Cerebral blood flow was measured in four patients and six controls at rest, during the OFM, and during stimulation with pure tones. OFM-induced loudness changes affected the auditory cortex contralateral to the ear in which tinnitus was perceived, whereas unilateral cochlear stimulation caused bilateral effects, suggesting a retrocochlear origin for their tinnitus. Patients, compared with controls, showed evidence for more widespread activation by the tones and aberrant links between the limbic and auditory systems. These abnormal patterns provide evidence for cortical plasticity that may account for tinnitus and associated symptoms. Although audiologic symptoms and examinations of these patients were typical, the unusual ability to modulate tinnitus loudness with an OFM suggests some caution may be warranted in generalizing these findings.

Incomplete recovery of chicken distortion product otoacoustic emissions following acoustic overstimulation.

Distortion product otoacoustic emissions (DPOAEs) were measured in chickens before and after exposure to a 525-Hz pure tone (120 dB SPL, 48 h). The exposure caused extensive hair cell loss and destroyed the tectorial membrane along the abneural edge of the basilar papilla in the low-to-mid-frequency region of the cochlea. Although the lesion was restricted, DPOAEs were greatly depressed at all frequencies immediately after the exposure. The high-frequency DPOAEs gradually recovered to preexposure values after the exposure; however, there was little or no improvement in DPOAEs at test frequencies equal to or slightly above the exposure frequency even after 16 weeks of recovery. By 28 days of recovery, the previously damaged region of the basilar papilla had been repopulated by hair cells and the lower honeycomb layer of the tectorial membrane had regenerated, but not the upper fibrous layer. The upper fibrous layer of the tectorial membrane was still missing after 16 weeks of recovery and the region of damage corresponded closely to the frequency regions where the DPOAEs were depressed.

Long-term effect of acoustic trauma on distortion product otoacoustic emissions in chickens.

Distortion product otoacoustic emissions (DPOAE), 2F1-F2, were measured in chickens before and after exposure to a 1.5-kHz pure tone presented at 120-dB sound-pressure level for 48 h. The low-level component of the DPOAE input/output function was shifted to the right at all frequencies immediately after the exposure with the greatest effect occurring at and above the exposure frequency. The slope of the high-level component of the input/output also increased significantly for test frequencies close to and above the exposure frequency so that the amplitude of the DPOAE was equal to or greater than normal at high stimulus levels. DPOAE obtained at the highest and lowest frequencies were almost normal after 8 weeks of recovery; however, the input/output functions near the exposure frequency showed almost no improvement over the 8-week recovery period. The lack of recovery could conceivably be due to residual damage to hair cells that survived the exposure or to incomplete regeneration of the tectorial membrane.

Changes in distortion product otoacoustic emissions and outer hair cells following interrupted noise exposures.

Changes in distortion product otoacoustic emissions (DPOAEs) were examined during and after interrupted noise exposures and compared to the condition of the outer hair cells (OHCs) and inner hair cells (IHCs) as assessed by scanning electron microscopy (SEM). Binaural, adult chinchillas were exposed to a 95 dB SPL, octave band noise centered at 0.5 kHz for 15 days using a 3 h on/9 h off schedule. DPOAEs were measured before, during and after the exposures. DPOAE amplitudes decreased significantly during the first few days of the interrupted noise exposures and then began to recover. At most frequencies, the emission amplitudes recovered completely to pre-exposure baseline values by five days after the last exposure. The results of the present study indicate that the changes in DPOAE amplitude paralleled the recovery in the amplitude and threshold of the compound action potentials as reported previously (Boettcher et al., 1992). Although the DPOAEs completely recovered, considerable OHC loss and stereocilia disarray was evident even four weeks after exposure.

Recovery of CAP threshold and amplitude in chickens following kanamycin ototoxicity.

Chickens were given a dose of kanamycin (400 mg/kg/d x 10 d) which destroyed hair cells over the basal 37-58% of the basilar papilla. Afterwards, the threshold and amplitude of the compound action potential were measured at recovery times ranging from 2 days to 10-20 weeks post-kanamycin treatment. At 2 days post-treatment, the thresholds at 1000, 2000 and 4000 Hz were elevated 40-60 dB while the thresholds at 250 and 500 Hz were elevated only 25 dB. By 10-20 weeks post-treatment, the threshold at 250 and 500 Hz had completely recovered whereas a residual threshold shift of 5 dB to 25 dB was present between 1000 to 4000 Hz. The maximum amplitude of the compound action potential was also reduced by more than 60% at all frequencies at 2 days post-treatment; however by 10-20 weeks post-treatment, the amplitude of the compound action potential had completely recovered at 500, 1000 and 2000 Hz. By contrast, the amplitude of the compound action potential at 4000 Hz was still reduced by more than 50% of its normal value 10-20 weeks post-treatment. The results of the present study indicate that the time course of recovery of the compound action potential is extremely slow and may lag behind the regeneration of hair cells by many weeks. The permanent deficits observed at the high frequencies could conceivably be due to functional deficits in regenerated hair cells, their afferent synapses or the loss of cochlear ganglion cells.

Physiological and histological changes associated with the reduction in threshold shift during interrupted noise exposure.

The compound action potential (AP) was recorded from one group of chinchillas exposed to interrupted noise (95 dB SPL, octave band centered at 500 Hz, 3 h on, 9 h off) for 15 days. A second group of chinchillas was exposed to the same interrupted noise for 1, 2 or 15 days and their cochleas were analyzed by scanning electron microscopy (SEM). During the first few days of the exposure, the AP threshold was elevated approximately 40 dB at the low-to-mid frequencies; however, the threshold shifts decreased with increasing exposure duration so that the threshold shift was only about 10 dB after the 15th day of exposure. The amplitude of the AP also recovered with exposure time. In contrast to the improvement in AP threshold, the number of missing hair cells increased and the condition of the stereocilia on inner and outer hair cells deteriorated between the first and 15th day of the exposure.

Neural correlates of gap detection in auditory nerve fibers of the chinchilla.

The neural correlates of gap detection were examined in a population of single auditory nerve fibers in the chinchilla. Acoustic stimuli consisted of 120 ms noise bursts (30-80 dB SPL) which contained silent intervals (gaps: 1, 2, 3, 4,5, 6 and 10 ms) at the midpoint. The neural response to the gap was quantified by the modulation index, (MAX-MIN)/AVE, which accounts for the steady state discharge rate before the gap (AVE), the minimum firing rate during the gap (MIN), and the maximum firing rate after the gap (MAX). In general, the modulation index increased as a function of gap width and stimulus level. Furthermore, there was a positive correlation between the modulation index and the characteristic frequency of the fiber. To estimate how detection could be based on the neuronal response, a criterion-free measure, analogous to d'. was calculated using z-scores obtained from the distributions of modulation index values collected before and during the gap and used to predict percent correct values for chinchilla psychophysical studies. The values increased with gap duration in a sigmoidal manner much like the psychometric functions in the chinchilla. In general, the neural gap thresholds obtained approximated those obtained psychophysically, although they were less affected by stimulus level.

Gap detection in hearing-impaired chinchillas.

Auditory temporal resolution is known to deteriorate with sensorineural hearing loss; however, there is considerable intersubject variability in human studies. The purpose of the present study was to obtain measures of temporal resolution in the chinchilla as the degree of noise-induced hearing loss was systematically varied. Gap-detection thresholds, a measure of temporal resolution, were evaluated at four levels of noise-induced asymptotic threshold shift (ATS). Gap thresholds were normal when the pure-tone thresholds were elevated approximately 15 dB. With a hearing loss of approximately 30 dB, the gap thresholds were longer than normal if compared at the same sound pressure level, but within normal limits if compared at the same sensation level. When the hearing loss exceeded 40 dB, gap thresholds were longer than normal both in terms of sound pressure level and sensation level. These results show that there is an orderly breakdown in temporal resolution as the degree of noise-induced ATS increases. The results are related to neural data and models of temporal resolution.

Detection of sinusoidally amplitude modulated noise by the chinchilla.

Amplitude modulation thresholds for sinusoidally amplitude modulated noise were obtained from four monaural chinchillas using shock-avoidance conditioning procedures. The noise was band limited at either 10 or 20 kHz, amplitude modulated at frequencies between 2 and 4096 Hz and presented at levels between 52 and 73 dB SPL. The modulation thresholds of the chinchilla were approximately 9% (-- 2 dB) at modulation frequencies below 32 Hz. At higher modulation frequencies, thresholds increased at the rate of 1.9 dB/octave. Modulation thresholds were also measured in human listeners using the same experimental apparatus. Amplitude modulation functions for both subject groups exhibited low-pass characteristics; however, the thresholds for humans were better than those of the chinchilla at modulation frequencies below 64 Hz.

Discharge patterns in the cochlear nucleus of the chinchilla following noise induced asymptotic threshold shift.

Chinchillas were exposed to an 86 dB SPL octave band of noise centered at 4.0 kHz for 3.5--5 days. The noise elevated the hearing thresholds between 4.0 and 16.0 kHz to between 60 and 75 dB SPL. Measurements from single neurons in the cochlear nucleus revealed abnormalities in the response properties of neurons with characteristic frequencies (CF) above 2.0 kHz. Units above 2.0 kHz had elevated thresholds (between 50 and 90 dB SPL) and broad tuning curves due to a greater loss in sensitivity near CF than at lower frequencies. The tuning curve Q10dB values for high frequency neurons were generally less than 3.0 and approached the Q10dB values for basilar membrane displacement. Spontaneous activity rates in units above 2.0 kHz were also low. In a few units, the threshold for single tone inhibition was significantly lower than that for excitation; the best inhibitory frequencies were always below 2.0 kHz. Two-tone inhibition was present in both low and high threshold neurons, but its strength was not assessed. Cochleagrams obtained 12 hours postexposure revealed discrete hair cell lesions in the basal third of the cochlea. The locations of the lesions were consistent with the frequencies of maximum hearing loss. The behavioral thresholds and the thresholds at CF of the most sensitive units were within 10--15 dB of each other. The results indicate that intense sounds reduce the sensitivity, frequency selectivity and spontaneous activity of units in the cochlear nucleus. The findings are similar to those obtained in auditory nerve fibers with ototoxic drugs and hypoxia.