The reduced cochlear output and the failure to adapt the central auditory response causes tinnitus in noise exposed rats

Abstract

Tinnitus is proposed to be caused by decreased central input from the cochlea, followed by increased spontaneous and evoked subcortical activity that is interpreted as compensation for increased responsiveness of central auditory circuits. We compared equally noise exposed rats separated into groups with and without tinnitus for differences in brain responsiveness relative to the degree of deafferentation in the periphery. We analyzed (1) the number of CtBP2/RIBEYE-positive particles in ribbon synapses of the inner hair cell (IHC) as a measure for deafferentation; (2) the fine structure of the amplitudes of auditory brainstem responses (ABR) reflecting differences in sound responses following decreased auditory nerve activity and (3) the expression of the activity-regulated gene Arc in the auditory cortex (AC) to identify long-lasting central activity following sensory deprivation. Following moderate trauma, 30% of animals exhibited tinnitus, similar to the tinnitus prevalence among hearing impaired humans. Although both tinnitus and no-tinnitus animals exhibited a reduced ABR wave I amplitude (generated by primary auditory nerve fibers), IHCs ribbon loss and high-frequency hearing impairment was more severe in tinnitus animals, associated with significantly reduced amplitudes of the more centrally generated wave IV and V and less intense staining of Arc mRNA and protein in the AC. The observed severe IHCs ribbon loss, the minimal restoration of ABR wave size, and reduced cortical Arc expression suggest that tinnitus is linked to a failure to adapt central circuits to reduced cochlear input.

Figure 1. Development of tinnitus in equally sound-exposed animals correlates with altered structure of the IHC synapse.

(A) Following exposure to 120 dB SPL at 10 kHz for 1 h (assessed at 6 d after exposure) or for 1.5 h (assessed at 30 d after exposure), tinnitus occurred in ∼30% of the animals. Mean silence activity (± S.D.) as a measure for tinnitus for no-tinnitus (white) and tinnitus animals (black). The criterion for tinnitus was a silence activity above 0.1 (horizontal grey line, black triangle on ordinate). Number of animals is given in or below the bars. (B) A significant difference in the hearing threshold to click stimuli was only observed between animals with or without tinnitus after 1.5 h exposure. Mean (± S.D.) ABR thresholds for click stimuli of no-tinnitus (white) and tinnitus animals (black), tinnitus judged from their silence activity. Hearing threshold is depicted above each bar. Number of animals is given within each bar. The grey horizontal line and area illustrates the mean ABR threshold (± S.D.) before exposure. (C, D) High-frequency hearing is impaired following noise exposure, more pronounced in tinnitus animals than in no-tinnitus animals. Following 1.5 h exposure (D), also low-frequency hearing is impaired in tinnitus animals. Mean ABR threshold loss (dB, ± S.E.M.) to frequency-specific tone bursts for no-tinnitus (circles) and tinnitus animals (squares). The grey line at the top of each panel shows the normal hearing threshold before exposure. Frequencies with a significant hearing loss (>99% confidence interval of hearing threshold before exposure) are marked by crossed symbols. Frequencies with a significantly different loss between no-tinnitus and tinnitus animals are indicated by asterisks (Student’s t-Test with Bonferroni-Holms adjustments for alpha-shifts). Asterisks in brackets indicate descriptive statistics for p-values from t-tests that fail to meet the Bonferroni-Holms criterion. n.s. not significant. (E, F) Antibody staining for GluR4 (red, open arrowhead) and ribbon synapses (CtBP2, green, open arrow) are shown for the IHCs of the midbasal turn for the animals used in (A, B). Cell nuclei are stained with DAPI (blue). Scale bars, 10 µm. (G) IHC ribbon numbers of control and no-tinnitus animals were not significantly different. IHC ribbon numbers of hearing-impaired tinnitus animals were significantly reduced in the midbasal and basal turns in comparison to no-tinnitus animals. Ribbon counts were compared for statistical significance using the 1-way ANOVA, p-values were corrected for alpha-shift by multiple testing using the Bonferroni-Holms procedure, df = 8.

Figure 2. Changes in ABR waveforms following noise exposure is more pronounced in animals with tinnitus.

(A) Average ABR waveform before exposure at stimulation level 90 dB SPL (upper panel) and 40 dB above hearing threshold (lower panel). Mean (black line) ± S.D (grey area) of n = 32 animals. (B) ABR waves illustrating the difference in ABR waveform after 1 h (left panels) or 1.5 h (right panels) noise exposure for animals with (red) or without tinnitus (green) in comparison with the waveforms before noise exposure (mean ± S.D., black line and grey area) depicted for 90 dB SPL (upper panels) and 40 dB above the hearing threshold (lower panels). (C) Correlation of ABRs to click stimuli before and after 1 h noise exposure of individual animals (expressed as the correlation factor (CorF) for close to threshold (40 dB hearing level) and at high stimulation levels (90 dB SPL). Mean (± S.D.) derived from n = 10 (No-tinnitus, green circles) and n = 5 (Tinnitus, red squares) rats. The correlation factor of 1 (dashed horizontal line) indicates a perfect similarity of ABR waveforms before and after noise exposure. Correlation was significantly lower in tinnitus animals at 40 dB. At 90 dB SPL, the difference was not quite significant (p = 0.055). (D) Correlation of ABRs to click stimuli before and after 1.5 h noise exposure. At both 40 dB and 90 dB SPL, the correlation factor of ABRs from animals with tinnitus was significantly reduced. n = 12 and 5 for no-tinnitus and tinnitus animals, respectively. * p 0.05, ** p<0.01, 1-sided t-Test. (E) Correlation of averaged ABRs to click stimuli as a function of stimulus levels (dB SPL). In all four groups, the correlation factor steeply increased at supra-threshold levels. For tinnitus animals, the ABR waveform correlation did not reach the value of no-tinnitus animals, indicating reduced amplitude and waveform recovery after noise exposure.

Figure 3. Peak-to-peak amplitudes of late peaks of ABR waves remain reduced following noise exposure in animals with tinnitus.

Mean peak growth input/output (I/O) function (± S.D.) for early, delayed and late peaks before exposure (black line and grey shaded area) after 1 h or 1.5 h exposure. Three selected peak-to-peak amplitude growth functions (µV) with increasing stimulus levels (dB SPL) are shown for rats with tinnitus (green) or without tinnitus (red). In the rats with tinnitus, the peak-to-peak amplitudes remain reduced up to late peaks (right panel). The peak latencies are given in each panel for negative (n) and positive (p) peaks.

Figure 4. Silencing of Arc expression in auditory cortex (AC) in animals with tinnitus.

(A, B) Double detection of Arc mRNA (blue) and Arc protein (red) in the AC of equally noise-exposed rats shows a significantly reduced expression in animals with tinnitus in all cortical layers, quantified in (C, unpaired Student´s t-test, p<0.001, alpha = 0.05, df = 6). Scale bars, 50 µm. n = 3 animals per group in three independent experiments. Images correspond to coronal sections 2.5 and 3.6 mm posterior to Bregma. Hybridization with sense riboprobes plus omission of the primary antibody produced no signals (insert in A, Sense).