Synaptic plasticity in the medial superior olive of hearing, deaf, and cochlear-implanted cats

Abstract

The medial superior olive (MSO) is a key auditory brainstem structure that receives binaural inputs and is implicated in processing interaural time disparities used for sound localization. The deaf white cat, a proven model of congenital deafness, was used to examine how deafness and cochlear implantation affected the synaptic organization at this binaural center in the ascending auditory pathway. The patterns of axosomatic and axodendritic organization were determined for principal neurons from the MSO of hearing, deaf, and deaf cats with cochlear implants. The nature of the synapses was evaluated through electron microscopy, ultrastructure analysis of the synaptic vesicles, and immunohistochemistry. The results show that the proportion of inhibitory axosomatic terminals was significantly smaller in deaf animals when compared with hearing animals. However, after a period of electrical stimulation via cochlear implants the proportion of inhibitory inputs resembled that of hearing animals. Additionally, the excitatory axodendritic boutons of hearing cats were found to be significantly larger than those of deaf cats. Boutons of stimulated cats were significantly larger than the boutons in deaf cats, although not as large as in the hearing cats, indicating a partial recovery of excitatory inputs to MSO dendrites after stimulation. These results exemplify dynamic plasticity in the auditory brainstem and reveal that electrical stimulation through cochlear implants has a restorative effect on synaptic organization in the MSO.

Figure 1

Graphical analysis of SVs with respect to size (area) and roundness differentiate between three different vesicle types: large round (green), small round (blue), and pleomorphic (red). The left y-axis corresponds to the scatterplot and shows the mean roundness for SVs (error bars show standard deviation) for each area bin. The line plots show the percent of SVs with respect to each area division as indicated by the right y-axis. The figure shows the averaged data of all somatic endings on one representative MSO cell. Large round and small round SVs have statistically similar roundness values, but small round vesicles have significantly smaller mean areas. Pleomorphic SVs have a significantly smaller mean roundness compared with that of round SVs.

Figure 2

Three discrete SV types can be distinguished in each animal, regardless of variations due to fixation methods. Fixation with solutions containing 2% glutaraldehyde/2% paraformaldehyde (a,c,e) yields a more regular appearance of SV shape compared with that by fixation with 4% paraformaldehyde/0.1% glutaraldehyde (b,d,f). Regardless, pleomorphic (a,b), large round (c,d), and small round (e,f) SVs can still be distinguished within the same animal after quantifying area and roundness. Scale bar = 500 nm.

Figure 3

Immunohistochemical labeling of synaptic endings using VGLUT1 antibodies identifies excitatory synaptic contacts on an MSO dendrite of a cat with normal hearing. a: Cross-section through a dendrite shows four labeled endings (white asterisks) and one unlabeled ending (black asterisk). One labeled ending (double asterisks) is magnified in panel b, where an asymmetric postsynaptic density (arrow) and round SVs are consistent with the ending's inferred excitatory nature. Scale bars = 2 μm in a; 500 nm in b.

Figure 4

Immunohistochemical labeling of synaptic endings using GlyT2 antibodies identifies inhibitory synaptic contacts (white asterisks, c) on a representative MSO cell body of a cat with normal hearing (a). The immunoprecipitate darkens the mitochondria and forms a thin, dark halo around SVs, clearly revealing unstained terminals (black asterisk, b). b: A typical GlyT2-immunonegative ending has round SVs and asymmetric postsynaptic densities (arrows) characteristic of excitatory synapses. c: A typical GlyT2-immunopositive terminal has pleomorphic SVs and symmetric postsynaptic densities (arrow), characteristic of inhibitory synapses. Scale bars = 500 nm.

Figure 5

Representative electron micrographs of MSO principal cells from (a) a normal hearing cat, (b) a unilateral cochlear-implanted deaf cat, (c) an unstimulated congenitally deaf cat, and (d) a bilateral cochlear-implanted deaf cat. Inhibitory endings contain pleomorphic SVs and are highlighted in red, whereas excitatory endings contain small or large round SV and are highlighted in green. The representation of inhibitory axosomatic terminals diminishes with deafness, and is restored with stimulation through cochlear implants. Asterisks identify endings magnified in Figure 6. Scale bar = 5 μm.

Figure 6

Representative electron micrographs of inhibitory (a) and excitatory (b) axosomatic terminals in the normal MSO, magnified from Figure 5a. Symmetric postsynaptic densities and pleomorphic synaptic vesicles indicate the ending is inhibitory (a), whereas asymmetric postsynaptic densities and round synaptic vesicles indicate the ending is excitatory (b). These structural features were used to classify endings as shown in Figure 5.

Figure 7

The proportion of inhibitory axosomatic terminals is shown for all animals, grouped by cohort. Deaf cats have significantly smaller percentage of inhibitory contacts (DWC and DWK90; 27.9 ± 4.3%, n = 11) than hearing (PK90 and NORM; 50.1 ± 5.5%, n = 10) or cochlear-implanted animals (CIK and BCIK; 49.5 ± 7.0, n = 24; one-way ANOVA, P < 0.0001).

Figure 8

Light micrographs of VGLUT1 immunohistochemical-stained tissue show variations in the size of axodendritic endings in the MSO between hearing (a), deaf (b), and cochlear-implant stimulated (c) cats. Mean bouton size in hearing cats was largest, while the mean bouton size of deaf cats was statistically smallest among the cohorts. The mean bouton size of cochlear-implant stimulated cats was significantly different from deaf or hearing cats, and of an intermediate size. Arrows in each panel indicate a few examples of VGLUT1-positive boutons. Scale bar = 10 μm.

Figure 9

Schematic summary of the distribution and size of input terminals to the principal cells of the MSO. In normal-hearing cats there is approximately an even split of excitatory and inhibitory terminals on the cell body with mostly excitatory inputs to the dendrites. With congenital deafness, the size of the terminals shrinks and the relative number of inhibitory terminals is reduced on the cell body and vanishes on the dendrites. The introduction of activity to the auditory system via cochlear implants restores the distribution of inhibitory terminals to the neurons and partially restores terminal size.