Single unit hyperactivity and bursting in the auditory thalamus of awake rats directly correlates with behavioural evidence of tinnitus

Abstract

Tinnitus is an auditory percept without an environmental acoustic correlate. Contemporary tinnitus models hypothesize tinnitus to be a consequence of maladaptive plasticity-induced disturbance of excitation-inhibition homeostasis, possibly convergent on medial geniculate body (MGB, auditory thalamus) and related neuronal networks. The MGB is an obligate acoustic relay in a unique position to gate auditory signals to higher-order auditory and limbic centres. Tinnitus-related maladaptive plastic changes of MGB-related neuronal networks may affect the gating function of MGB and enhance gain in central auditory and non-auditory neuronal networks, resulting in tinnitus. The present study examined the discharge properties of MGB neurons in the sound-exposure gap inhibition animal model of tinnitus. MGB single unit responses were obtained from awake unexposed controls and sound-exposed adult rats with behavioural evidence of tinnitus. MGB units in animals with tinnitus exhibited enhanced spontaneous firing, altered burst properties and increased rate-level function slope when driven by broadband noise and tones at the unit's characteristic frequency. Elevated patterns of neuronal activity and altered bursting showed a significant positive correlation with animals' tinnitus scores. Altered activity of MGB neurons revealed additional features of auditory system plasticity associated with tinnitus, which may provide a testable assay for future therapeutic and diagnostic development.

Figure 1. Example of thionin-stained coronal section containing medial geniculate body (MGB)

White dashed line represents the electrode track passing through dorsal (dMGB) and ventral MGB (vMGB) demarcated by dashed red lines. Scale bar = 500 μm. CA = Cornu Ammonis.

Figure 2. Characteristic frequencies (CFs) and thresholds of MGB units recorded from unexposed control animals and sound-exposed animals with behavioural evidence of tinnitus (tinnitus animals)

A, scatter plot of individual MGB unit thresholds at their CF from control (blue) and tinnitus (red) rats. B, average threshold at CF of MGB units recorded from control (blue) and tinnitus (red) rats did not exhibit any significant difference (P = 0.234; unpaired t test, two-tailed). +, Mean firing threshold. Error bars represent SEM.

Figure 3. Spontaneous firing rates (SFRs) of MGB units recorded from unexposed control (control animals) and sound-exposed animals with behavioural evidence of tinnitus (tinnitus animals)

A, average SFRs of MGB units recorded from tinnitus animals (9.09 ± 0.9423; red, six animals) were significantly elevated (P = 0.0005; unpaired t test, two-tailed) in comparison to controls (4.68 ± 0.48; blue, nine animals). +, Mean SFR. B, scatter plot of average SFR of MGB units recorded from individual animals (y-axis) Showing a significant correlation (y = 1.830x + 4.812, P = 0.004; Pearson correlation coefficient) with the animals’ tinnitus z-scores (x-axis). Blue and red squares represent control and sound-exposed animals, respectively. Error bars represent SEM.

Figure 4. Spontaneous bursting activity of MGB units recorded from unexposed control (control animals) and sound-exposed animals with behavioural evidence of tinnitus (tinnitus animals)

A, C and E, MGB units recorded from tinnitus animals (red; six animals) exhibited significantly elevated, average bursts per minute (P = 0.0001) (A), average spikes in a burst (P = 0.009) (C) and average burst duration (P = 0.043) (E) compared to control rats (blue; nine animals). Unpaired t test was used to determine the statistical significance between the two groups. +, Mean of the test variable. B, D and F, scatter plots of average burst variables (y-axis) of MGB units from individual animals against that an animal's tinnitus z-score (x-axis). A significant correlation (y = 11.98x + 9.619, P = 0.001) was observed between average bursts per minute and animals’ tinnitus scores (B). However, average spikes in burst (P = 0.082) (D) and average burst duration (P = 0.237) (F) did not exhibit significant correlation with the animals’ tinnitus z-scores. Pearson correlation coefficient was used to determine statistical significance of the correlations. Blue and red squares represent control and sound-exposed animals, respectively. Error bars represent SEM.

Figure 5. Rate-level functions comparing average responses of MGB units to broadband noise (BBN) from unexposed control animals (blue; eight animals) with sound-exposed animals with behavioural evidence of tinnitus (red; six animals)

A–E, rate-level function of MGB units in response to BBN stimuli. A, significant (unpaired t test) increase in average firing rates of MGB units was observed in tinnitus rats (red) at intensities of 0 dB (P = 0.001), 20 dB (P = 0.006), 50 dB (P = 0.025), 60 dB (P = 0.002), 70 dB (P = 0.024) and 80 dB (P = 0.049) compared to control rats (blue). Average firing rates were not significantly different at other intensities. B–E, scatter plots of average firing rates of MGB units (y-axis) from individual animals in response to BBN stimuli at different intensities against that animal's tinnitus z-score. A significant correlation was observed between average firing rates and animals’ tinnitus scores at 60 dB (C) (y = 10.21x + 20.53, P = 0.002), 70 dB (D) (y = 9.707x + 29.13, P = 0.014) and 80 dB (E) (y = 10.39x + 38.92, P = 0.017) but not at 30 dB (B). In all the plots, blue and red squares represent control and sound-exposed animals, respectively. Pearson correlation coefficient was used to determine statistical significance of the correlations. Error bars represent SEM.

Figure 6. Rate-level functions comparing average responses of MGB units to individual unit's CF from unexposed control animals (blue; nine animals) with sound-exposed animals with behavioural evidence of tinnitus (red; six animals)

A–E, rate-level function of MGB units in response to unit's CF. A, significant increase (unpaired t test) in average firing rates of MGB units was observed at 10 dB (P = 0.027), 40 dB (P = 0.048), 50 dB (P = 0.025), 60 dB (P = 0.035), 70 dB (P = 0.036) and 80 dB (P = 0.012) in tinnitus rats (red) compared to control rats (blue). Average firing rates were not significantly different at intensities of 30 dB and below. B–E, scatter plots of average firing rate of MGB units recorded from individual animals in response to CF stimuli (y-axis) at different intensities against that animal's tinnitus z-score. Significant correlations were observed at 60 dB (C) (y = 10.47x + 20.71, P = 0.011), 70 dB (D) (y = 10.23x + 31.18, P = 0.015) and 80 dB (E) (y = 11.64x + 39.75, P = 0.012) but not at 30 dB (B). In all the plots, blue and red squares represent control and sound-exposed animals, respectively. Pearson correlation coefficient was used to determine statistical significance of the correlations. Error bars represent SEM.