Homeostatic plasticity drives tinnitus perception in an animal model

Abstract

Hearing loss often results in tinnitus and auditory cortical map changes, leading to the prevailing view that the phantom perception is associated with cortical reorganization. However, we show here that tinnitus is mediated by a cortical area lacking map reorganization. High-frequency hearing loss results in two distinct cortical regions: a sensory-deprived region characterized by a decrease in inhibitory synaptic transmission and a normal hearing region showing increases in inhibitory and excitatory transmission and map reorganization. Hearing-lesioned animals displayed tinnitus with a pitch in the hearing loss range. Furthermore, drugs that enhance inhibition, but not those that reduce excitation, reversibly eliminated the tinnitus behavior. These results suggest that sensory deprivation-induced homeostatic down-regulation of inhibitory synapses may contribute to tinnitus perception. Enhancing sensory input through map reorganization may plausibly alleviate phantom sensation.

Conflict of interest statement

The authors declare no conflict of interest.

Figures

Fig. 1.

Cortical map reorganization after high-frequency hearing loss separates the map into two distinct areas. (A) ABR threshold shift after hearing lesion. (B) CF and threshold (color-coded) as a function of the position along the tonotopic axis. Note the threshold increase for frequencies >4 kHz. C, caudal; R, rostral. (C) Cortical maps and receptive fields in naïve and hearing-lesioned animals. Underneath are receptive fields labeled according to recording sites on the map. (D) Maps of response magnitude to a 2-kHz and 50-dB tone. Both responsive area and response magnitude increased in the hearing-lesioned animal.

Fig. 2.

High frequency hearing loss induces potentiation of excitatory synaptic transmission in the low-CF area. (A) Example traces of miniature excitatory postsynaptic currents (mEPSCs) recorded from low-CF and high-CF areas of naïve and hearing-lesioned animals. (B) Cumulative histograms of mEPSC amplitude and frequency. (C) Mean mEPSC amplitude and frequency. *P < 0.05; **P < 0.01; n.s., not significant. The number of recorded neurons is indicated on each bar. See also Fig. S1.

Fig. 3.

High frequency hearing loss differentially affects inhibitory synaptic transmission in the low-CF and high-CF areas. (A) Example traces of miniature inhibitory postsynaptic currents (mIPSCs) recorded from low-CF and high-CF areas of naïve and hearing-lesioned animals. (B) Cumulative histograms of mIPSC amplitude and frequency. (C) Mean mEPSC amplitude and frequency. *P < 0.05; **P < 0.01; n.s., not significant.

Fig. 4.

High-frequency hearing loss alters both phasic and tonic inhibition. (A) An example trace of spontaneous inhibitory postsynaptic currents (sIPSCs). Bath application of Gabazine abolishes phasic sIPSCs and reveals a tonic inhibitory current. (BI) Example traces showing phasic sIPSCs. (BII) Amplitude and frequency of sIPSCs. (CI) Tonic inhibition measured as the shift in baseline current after Gabazine application. (CII) Hearing loss reduces tonic inhibition in the high-CF area. (D) Charge transfer through phasic and tonic inhibitory currents (QPC and QTC, repectively). (EI) Tonic inhibition activated by an agonist, THIP. The amplitude of the THIP-activated tonic inhibitory current was not altered by hearing loss. *P < 0.05; n.s., not significant.

Fig. 5.

High frequency hearing loss reduces GAD65 protein level in the high-CF area. (A) Immunohistochemical staining of auditory cortical neurons (with NeuN green) and GAD65 protein. Note the reduced GAD65 level in high-CF zone of the hearing-lesioned animals. (Scale bar: 100 μm; Inset, 20 μm.) (B) Density of GAD65 puncta was reduced in the high-CF zone of hearing-lesioned animals (n = 8). (C) GAD65 expression level, as measured with Western blot analysis, was reduced in high-CF areas of the hearing-lesioned animals. **P < 0.01; n.s., not significant. See also Fig. S2.

Fig. 6.

Hearing lesion-induced tinnitus is reversibly abolished by an enhancement in GABA-mediated inhibition. Both previously trained (n = 8) and naïve animals (n = 5) naturally prefered the dark, low-pitch side in silent probe trials. After the hearing lesion, trained animals switched their preference to the light, high-pitch side in probe trials, indicating their perception of high-pitch tinnitus. Untrained animals did not switch their preferred side after the hearing lesion. Enhancing GABA-mediated inhibition by Vigabatrin or NO711 reversibly abolished tinnitus behaviors. Ketamine, a NMDA receptor antagonist, did not alter tinnitus behaviors through an excitatory mechanism. **P < 0.01; n.s., not significant.