Localized disorganization of the cochlear inner hair cell synaptic region after noise exposure

Abstract

The prevalence and importance of hearing damage caused by noise levels not previously thought to cause permanent hearing impairment has become apparent in recent years. The damage to, and loss of, afferent terminals of auditory nerve fibres at the cochlear inner hair cell has been well established, but the effects of noise exposure and terminal loss on the inner hair cell are less known. Using three-dimensional structural studies in mice we have examined the consequences of afferent terminal damage on inner hair cell morphology and intracellular structure. We identified a structural phenotype in the pre-synaptic regions of these damaged hair cells that persists for four weeks after noise exposure, and demonstrates a specific dysregulation of the synaptic vesicle recycling pathway. We show evidence of a failure in regeneration of vesicles from small membrane cisterns in damaged terminals, resulting from a failure of separation of small vesicle buds from the larger cisternal membranes.

Keywords: Inner hair cell; Noise exposure; Pre-synaptic; Vesicle.

Fig. 1.

Threshold shifts in ABR as result of noise exposure. Changes to whole cell morphology and afferent terminals after noise damage. (A) Threshold shift as a result of octave-band noise exposure as measured from ABRs. A threshold shift at frequencies ≥16 kHz is present 1 day (solid red line, circular symbols) after noise exposure, but resolves after 4 weeks (green line, triangular symbols). Grey line indicates no change. Whole cell analysis from SBF-SEM images. (B) Slices from SBF-SEM image stacks for two control and two noise damaged inner hair cells. Arrow indicates increasing depth through the cell. White lines show contour tracing of the cytoplasmic membrane. (*) Indicates example of the vacuolation occurring beneath the cells in noise exposed animals. (C) Reconstructed cell bodies and afferent terminal positions from the cytoplasmic membrane contours. The distortion of noise damaged cells caused by afferent terminal expansion can be clearly seen. Afferent terminal positions on control and noise damaged cells, represented by spheres. (Yellow) Intact terminals, (pink) damaged terminals, (purple) lost terminals. (D) Examples of synaptic ribbons and postsynaptic densities observed in SBF-SEM datasets, top pair: intact terminals, bottom pair: damaged terminals. (E) Number afferent terminals in modiolar and pillar cell facing hemispheres. There were no significant differences in control cells or in noise-exposed cells in the total population of terminals (P=0.292 control, 0.429 noise exposed). [Two-tailed Mann–Whitney U. n=6 cells from one (control) or two (damaged) animals. Error bars represent mean±s.d.] Scale bars: (B–C) 2 µm.

Fig. 2.

Intact and damaged terminals on the same cell showed differences in pre-synaptic ultrastructure. (A) TEM image of IHC from 4–8 kHz noise damaged animal, coloured boxes represent regions from which tomographic data was collected on an intact terminal (B, yellow), and a damaged terminal (C, red). Colours represent: white, border of reconstructed region ‘*’ indicates the presynaptic membrane; orange, mitochondria; yellow, endoplasmic reticulum; light blue, small membrane cisterns (uncoated); red, small membrane cisterns (coated); light-blue/red, mixed membranes; purple spheres (small), putative synaptic vesicles; violet, synaptic ribbon; pink, other endosome. Despite the close proximity, these terminals had very different morphologies. Scale bars: (A) 5 µm, (B–C) 200 nm.

Fig. 3.

Pre-synaptic regions adjacent to damaged terminals showed ultrastructural changes compared to control cells. (A–D) Tomographic slice and reconstruction from pre-synaptic region in control and noise exposed animals. Colours and asterisks as Fig. 2. (A) Control animal (intact terminal), (B) 8–16 kHz noise damaged animal (intact terminal), (C) 4–8 kHz noise damaged animal (damaged terminal), (D) 8–16 kHz noise damaged animal (damaged terminal). Scale bars: (A–D) 200 nm.

Fig. 4.

Uncoated membrane cisterns and cisterns with coated and uncoated regions were both increased in pre-synaptic regions from damaged terminals. (A–D) Example images and distributions for different membrane populations in pre-synaptic regions of intact and noise damaged terminals. (A) Endoplasmic reticulum. (B) Uncoated membrane cisterns, (C) coated membrane cisterns (red patches on pre-synaptic membrane represent endocytic events), (D) mixed membranes. Scale bars: (A, slice) 50 nm, (B–D, slices) 20 nm, (A–D, reconstructions) 200 nm.

Fig. 5.

Significant differences in membrane area for some membrane types between intact and damaged terminals. White bars represent intact terminal population; grey bars represent damaged terminal population. (A) Total membrane (sum of all populations) was significantly increased in the damaged terminals (P=0.009), but the endoplasmic reticulum component was not changed. (B) Uncoated membrane was significantly increased, in both uncoated membrane cisterns (P=0.002) and in the proportion forming part of mixed membrane cisterns (P=0.002). (C) Coated membrane was increased in total proportion, but this increase was only significant in the proportion forming part of mixed membranes (P=0.002), when this was excluded, the proportion of coated membrane cisterns was not significantly different (P=0.093). **P=<0.01 [two-tailed Mann–Whitney U. n=6 terminals from four (intact) or three (damaged) animals. Error bars represent mean±s.d.].

Fig. 6.

Synaptic vesicles in pre-synaptic regions were reduced in regions adjacent to damaged terminals. (A) Cytoplasmic vesicles (purple spheres) in pre-synaptic regions of intact terminals, compared with (B) cytoplasmic vesicles in pre-synaptic regions of damaged terminals, showed a reduction. (C) Synaptic ribbons (purple/violet) from pre-synaptic regions of intact terminals, (D) synaptic ribbons from pre-synaptic regions of damaged terminals. The number of vesicles appeared to be reduced, but the effect was variable. (E,F) tomographic slices showing vesicles (black arrow tethered at endoplasmic reticulum (ER), by filamentous tethers (white arrows). Inset shows reconstruction of ER and vesicle tether. (G) ER attached vesicles in pre-synaptic region of intact terminals. (H) ER attached vesicles in pre-synaptic region of damaged terminals. (I,J) tomographic slice showing synaptic membrane (SM) tethered vesicles, arrows as (E,F). (K) Synaptic membrane tethered vesicles at the synapse of an intact terminal. (L) Synaptic membrane tethered vesicles at the synapse of a damaged terminal. (M) Change in vesicle population number between intact (clear bars) and damaged (shaded bars). All populations showed a reduction, but this was only significant in the ER attached population. *P=0.015. [two-tailed Mann–Whitney U. n=6 terminals from four (intact) or three (damaged) animals. Error bars represent mean±s.d.]. Scale bars: (A–D,G,H) 200 nm, (K,L) 100 nm, (C,D) 50 nm, (E,F,I,J) 20 nm.