Noise-induced Cochlear Synaptopathy with and Without Sensory Cell Loss

Abstract

Prior work has provided extensive documentation of threshold sensitivity and sensory hair cell losses after noise exposure. It is now clear, however, that cochlear synaptic loss precedes such losses, at least at low-moderate noise doses, silencing affected neurons. To address questions of whether, and how, cochlear synaptopathy and underlying mechanisms change as noise dose is varied, we assessed cochlear physiologic and histologic consequences of a range of exposures varied in duration from 15 min to 8 h and in level from 85 to 112 dB SPL. Exposures delivered to adult CBA/CaJ mice produced acute elevations in hair cell- and neural-based response thresholds ranging from trivial (∼5 dB) to large (∼50 dB), followed by varying degrees of recovery. Males appeared more noise vulnerable for some conditions of exposure. There was little to no inner hair cell (IHC) loss, but outer hair cell (OHC) loss could be substantial at highest frequencies for highest noise doses. Synapse loss was an early manifestation of noise injury and did not scale directly with either temporary or permanent threshold shift. With increasing noise dose, synapse loss grew to ∼50%, then declined for exposures yielding permanent hair cell injury/loss. All synaptopathic, but no non-synaptopathic exposures produced persistent neural response amplitude declines; those additionally yielding permanent OHC injury/loss also produced persistent reductions in OHC-based responses and exaggerated neural amplitude declines. Findings show that widespread cochlear synaptopathy can be present with and without noise-induced sensory cell loss and that differing patterns of cellular injury influence synaptopathic outcomes.

Keywords: cochlear deafferentation; cochlear synaptopathy; hair cell; noise-induced hearing loss; sensorineural hearing loss.

Fig. 1.

Immunostaining for pre-synaptic (CtBP2-red), post-synaptic (GluA2-green) and hair cell markers (Myosin VIIa-blue) to assess noise-induced cochlear injury. (A) Confocal maximal projection of adjacent IHCs from the 32 kHz region of a control mouse. (B) Schematic showing one IHC with 2 of ~17 afferent fiber synapses. (C) High-power thumbnails of the voxel space showing juxtaposed ribbon/receptor puncta. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 2.

Quantification of post-noise shifts in OHC- and neural-based response thresholds. Threshold shifts (means ± SE), relative to unexposed 16 wk animals, as quantified for equal energy series of exposures (A–D) and equal level series (E–H) at the 24 h and 2 wk post-exposure time points. For both DPOAE and ABR wave 1 response metrics and both exposure series, acute threshold shifts are large, particularly at basal turn frequencies, and demonstrate variable degrees of recovery by 2 wks. Gray bars indicate noise exposure band. Animals in each group (n) were as follows: Controls: (22), 24 h: 100–15 m (11), 100–30 m (11), 100–1 h (10), 100–2 h (7), 100–4 h (8), 97–4 h (10), 103–1 h (8), 106–30 m (9), 2 wk: 100–15 m (11), 100–30 m (10), 100–1 h (8), 100–2 h (7), 100–4 h (13), 97–4 h (9), 103–1 h (10), 106–30 m (12).

Fig. 3.

Quantification of hair cells and IHC synapses. Shown are survivals (% re controls) as quantified at the 2 wk post-exposure time point. OHC counts were made separately by row and were similar for the 3 rows; averages are displayed in (A) and (D). Where present, OHC losses were small and restricted to highest frequencies. IHCs were well preserved at all frequencies for all groups (B, C). Synaptic losses in the base were increasing functions of exposure energy to ~50% maximum loss, then declined for ears with mixed (neural + sensory) loss (C, F). Gray bars indicate noise exposure band. Group sizes are as detailed for Fig. 2.

Fig. 4.

Response growth in ears with recovered thresholds vs PTS with OHC loss. For 97 dB – 4 h exposure (A–D), DPOAE and ABR wave 1 thresholds are acutely elevated and suprathreshold response amplitudes are reduced, especially at high frequencies (B, D). Thresholds and DPOAE amplitudes recover by 2 wks, but neural response amplitudes in cochlear regions of synaptic loss (see Fig. 3) do not (D). In comparison, 100 dB – 4 h exposure producing mild PTS and OHC loss also yields acute DPOAE and wave 1 response declines (E–H), but for this overt hearing loss-producing exposure, recovery is incomplete. Neural response declines at 30 kHz are particularly large, reflecting the mixed sensory and neural sites of lesion. Data are means ± SE; Group sizes are as detailed for Fig. 2.

Fig. 5.

Non-synaptopathic exposures. Post-noise (24 h) threshold elevations are mild, restricted in frequency and temporary for both response types and both exposures (A, B, F, G). OHC (averaged across 3 rows), IHC and synapse counts at 2 wks post exposure (C, D, H, I) are not significantly different from control values (p > 0.05). Wave 1 response growth functions in the region of maximum acute threshold shift (17.54 kHz) (E, J) show no persistent suprathreshold declines. Gray bars indicate noise exposure band. Group sizes: Control (22; n = 9 controls for the 85–8 h exposure groups); 94–15 m, 24 h (10), 94–15 m, 2w (7); 85–8 h, 24 h (6), 85–8 h, 2w (8).

Fig. 6.

High-level exposure with PTS and OHC loss. Post-noise threshold shifts (means ± SE) at 2 wks (DPOAE, ABR) (a, B). OHC survivals (C) and IHC synapse survivals (D) are shown 2 wks after exposure (8–16 kHz band, 2 or 4 h, at 112 dB SPL); group sizes 112–2 h (10), 112–4 h (10). Shown for comparison in D, are synapse survivals for 100 dB – 2 h exposure (dotted line; also see Fig. 3). Gray bars indicate noise exposure band.

Fig. 7.

Response growth in ‘mixed’ hearing loss ears. Two weeks after exposure (112 dB, 2 or 4 h), DPOAE response growth functions show recovery to control values for f2 frequencies below and within the noise exposure band (A, B) and frequency-dependent response declines in regions of permanent OHC injury/loss (C, D). In contrast, dramatic ABR wave 1 response declines are seen across the same range of frequencies (E–H). Data are means ± SE. The 112 dB exposed groups were added late in the experimental series; thus, a separate set of controls was assessed close in time to the 112 dB-exposed groups.

Fig. 8.

Wave 1 amplitude vs. IHC synapse loss. Linear regression analysis indicated that two weeks after noise, wave 1 amplitude declines (30 kHz, 80 dB SPL) were proportional to synapse losses (30 kHz region) for groups with recovered thresholds and no hair cell loss (black/gray symbols). Groups with PTS and OHC injury/loss (100 dB – 4 h, 106 dB – 30 m, 112 dB – 2 h, 112 dB – 4 h; red symbols) fall outside of the 95% confidence interval. In these cases, ABR amplitude declines are exaggerated and no longer proportional to declines in synapse counts. Corresponding values corrected for the OHC-based PTS are displayed in teal (see text for details). Regression line (line generated without PTS/OHC loss groups) is given by the thick black line and 95% confidence interval is bounded by blue dotted lines. Data points are group means; n’s as given in Figs. 2, 5, 6. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 9.

Sex differences in noise vulnerability. For certain conditions of exposure, 16 wk male CBA/CaJ mice showed greater ABR wave 1 threshold shifts compared to identically-exposed, strain- and age- matched females. In the same ears, synapse losses showed no consistent sex dependence. Data means ± SE are shown at 2 wks post exposure. Groups: 85 dB – 8 h (4M, 4F); 106 dB – 30 m (8M, 4F); 100 dB – 4 h (6M, 7F); 112 dB – 4 h (4M, 6F).

Fig. 10.

Exposure comparisons using TWA. ABR wave 1 threshold shifts (A) and synapse survivals (B) 2 wk post exposure are plotted at 12 kHz and 30 kHz as functions of an OSHA-based TWA with a criterion time of 8 h. Dotted line in each panel identifies the 100% 8 h Permissible Exposure Limit (PEL) for human (in dBA). Note however, that TWAs were calculated for exposures expressed in dB SPL, not dBA; at the frequencies of exposure used in mouse, the differences in these two scales are irrelevant. Sound levels measured in the animal holding rooms in the vivarium (see Sergeyenko et al., 2013) were considered, but did not materially impact the calculated TWAs, whether for an 8-h or 24-h sampling period. Thus, data are expressed relative to an 8-h TWA. Calculations were performed using the calculator found at: [https://noisemeters.com/apps/occ/twa-do]. Exposures producing permanent OHC injury/loss are noted by red data points for the 30 kHz plots.