Morphological and functional correlates of vestibular synaptic deafferentation and repair in a mouse model of acute-onset vertigo

Abstract

Damage to cochlear primary afferent synapses has been shown to be a key factor in various auditory pathologies. Similarly, the selective lesioning of primary vestibular synapses might be an underlying cause of peripheral vestibulopathies that cause vertigo and dizziness, for which the pathophysiology is currently unknown. To thoroughly address this possibility, we selectively damaged the synaptic contacts between hair cells and primary vestibular neurons in mice through the transtympanic administration of a glutamate receptor agonist. Using a combination of histological and functional approaches, we demonstrated four key findings: (1) selective synaptic deafferentation is sufficient to generate acute vestibular syndrome with characteristics similar to those reported in patients; (2) the reduction of the vestibulo-ocular reflex and posturo-locomotor deficits mainly depends on spared synapses; (3) damaged primary vestibular synapses can be repaired over the days and weeks following deafferentation; and (4) the synaptic repair process occurs through the re-expression and re-pairing of synaptic proteins such as CtBP2 and SHANK-1. Primary synapse repair might contribute to re-establishing the initial sensory network. Deciphering the molecular mechanism that supports synaptic repair could offer a therapeutic opportunity to rescue full vestibular input and restore gait and balance in patients.

Keywords: Excitotoxicity; Plasticity; Synapses; Vestibular disorders; Vestibule.

Fig. 1.

Operational diagram used in the study. Following TTK (T0), histological and functional analyses were performed at each of the indicated time points. Histological analysis of inner ear tissues was performed using TEM. Assessment of posturo-locomotor alterations was performed using specific behavioural testing. aVOR alterations were measured using video-oculography. BL, base line; VNS, videonystagmoscopy.

Fig. 2.

TEM observation of vestibular primary synapses following TTK in the utricle. (A,B) Sham vestibular epithelium. Type I hair cells (HCI) are recognised by their amphora-like shape and the calyx afferent (c) encasing the cell body. Type II hair cells (HCII) have a more cylindrical shape and no calyx afferent. (C) Afferents are greatly swollen and the structure of the epithelium is completely distorted 4 h after TTK. Some swollen afferents probably corresponded to calyx terminals (*), whereas others are more likely to correspond to bouton terminals (+). Note the row of supporting cell nuclei (sc) just above the basal membrane of the epithelium. (D) The epithelium has grossly recovered a control-like appearance 1 week after TTK. (E) Higher magnification of the area marked with the left box in D, showing control-like ultrastructure including a presynaptic ribbon (arrow) in a type II hair cell facing a postsynaptic bouton afferent (b). (F) Higher magnification of the area marked with the right box in D, showing a contact between a type I hair cell and the calyx afferent contacting it. The arrowhead points to the calyceal junction, a prominent characteristic of this contact in healthy and mature epithelia. Sample: n=3 in each group. Scale bars: 5 µm (A,B), 10 µm (C,D), 1 µm (E) and 2 µm (F).

Fig. 3.

Evaluation of synaptic protein expression following TTK administration. (A) Schematised vestibular bouton synapses with the respective locations of immunostained CtBP2 (red) and SHANK-1 (green) proteins. (B) Observation field (arrowed box) of 4.5e⁻⁰³ mm², which includes about 75-100 hair cells in the centre of a utricle. (C) CtBP2 and SHANK-1 fluorescent spots semi-automatically detected within the observation field using IMARIS software. (D) Characteristic expression of fluorescent spots at a calyx terminal (white arrows). (E) Distant and (F) colocalised spots at high magnification. (G) Repartition of the distances between all CtBP2 and SHANK-1 fluorescent spots highlighting a specific subpopulation of spots, the interdistances of which are located within 1 μm (arrow). Images in B-D,F,G: control utricles. (H) Cumulative distances between CtBP2 and SHANK-1 before (sham), 4 h and 3 weeks after TTK application. Note that the gray sham data overlap with the 3 week data. (I-N) Expression of CtBP2 (I,J), SHANK-1 (K,L) and the percentage of their colocalisation (M,N) at each time point before and after TTK application in both the utricle (black circles; n=6 for 48 h, 72 h and 2 weeks; n=5 for sham, 24 h and 3 weeks; n=4 for 4 h and 1 week) and crista ampullaris (white circles; n=6 for 4 h, 72 h and 3 weeks; n=5 for sham, 24 h, 48 h, 1 week and 2 weeks). The ribbon counts were normalised relative to the observation field described in the Materials and Methods section. Results are expressed as mean±s.e.m. Dotted lines represent sham values. ANOVA followed by Dunnett post-hoc analyses were performed to compare each time point to the sham: *P<0.05; **P<0.01; ***P<0.001; ^(§)P=0.6. Scale bars: 200 µm (B), 5 µm (C), 2 µm (D) and 0.5 µm (E,F).

Fig. 4.

Analysis of nystagmus, canal and otolith function alterations after TTK administration and UVN. (A) Spontaneous eye movements were recorded in the dark before lesion and 24 h after TTK administration or UVN. Recordings show spontaneous eye drift (slow phase) and multiple quick phases (arrows) corresponding to pathological nystagmus. (B) Comparison of the nystagmus quantified from the TTK (n=11) and UVN (n=6) mice groups during the 3 weeks following the lesion. (C) Raw traces of eye movements observed during sinusoidal table rotation in yaw plane at 0.5 Hz and 30°/s sinusoidal rotation, before and 24 h after TTK or UVN. Note the reduction in amplitude of the eye movements and asymmetry in the response towards the ipsilesional side or contralateral rotation (here ipsi is up). (D) Quantification of the aVOR gain following TTK and UVN. To ease comparison, data are normalised to the pre-lesion values (raw values are reported in Table 2). (E) MOR were measured during off-vertical axis rotation at 50°/s. At 24 h after TTK and UVN, note the absence of the horizontal beating of the eye normally observed. (F) Quantification of the MOR bias following TTK and UVN. Data are normalised to the pre-lesion values. Results are presented as mean±s.e.m. If present, statistical significance is shown comparing each value to the base line (BL) value (repeated measures ANOVA and post-hoc Tukey tests, see Materials and Methods section). In the grey area under B, D and F, statistical differences between UVN and TTK values are reported (two- or three-way ANOVA and post-hoc Tukey tests, see Materials and Methods section). Statistics are represented as *P<0.05; **P<0.01; ***P<0.001; n.s., not specified. In all cases, the TTK population improved after 1 week and fully recovered after 3 weeks, whereas UVN mice remained pathological.

Fig. 5.

Posturo-locomotor alterations and alterations of the general behaviour following TTK administration. Evolution of vestibular alteration over time in TTK-treated, UVN-operated and sham mice during the 2 min evaluation. (A) Vestibular signs and (B) general behaviour in the open field. (C) Swim deficit score in the pool. (D) Postural alterations during the tail-hanging landing paradigm. Results are expressed as mean±s.e.m. The sham mice are represented by open circles, the TTK-treated mice by dark triangles and the UVN-operated mice by grey squares (n=9 in each group). Repeated measures ANOVA followed by Bonferroni post-hoc analyses were used to observe TTK administration effects (i.e. TTK vs sham mice): **P<0.01; ***P<0.001. We performed a repeated measures ANOVA followed by Bonferroni post-hoc analyses to observe treatment effects (i.e. TTK vs UVN): ^§P<0.001, ^(§)P=0.06.

Fig. 6.

Correlation grid between vestibular primary synapses afferentation state and signs of vestibular functional alterations. Upper traces: schematic representation of the percentage of intact synapses in the utricle (light blue) and crista (deep blue), drawn on the basis of CtBP2 expression in the sensory epithelia at each of the investigated time points (Fig. 3). Middle traces: schematic representation of the percentage of aVOR [dashed line (or light green) for TTK and dash-dotted line (or lighter dark green) for UVN] and MOR [full line (or green) for TTK and dotted line (or darkest green) for UVN] gains, drawn on the basis of the aVOR measurements shown in Fig. 4. Lower traces: schematic representation of the percentage of posture-locomotor alterations in the UVN (light pink) and TTK (dark pink) mice, drawn on the basis of the vestibular signs displayed in Fig. 5A.