Release your inhibitions: how synaptic dynamics shape efferent activity in the auditory system

The Journal of Physiology, vol. 600, pp. 2543–2544

Abstract

The sense of hearing evolved to perform in complex listening situations, in which the presence of time-varying background noise is common. Under such circumstances it is easy to see that mechanisms to dynamically control the sensitivity along the spectral and temporal dimension are an advantageous feature for acoustic perception. One well known auditory gain control system is the olivocochlear efferent system (for a recent review see Lopez-Poveda, 2018). Cholinergic fibres originating in the auditory brainstem project back into the organ of Corti. Two major classes of olivocochlear projection neurons have been identified. Medial olivocochlear (MOC) neurons primarily reside in the ventral nucleus of the trapezoid body and predominantly project to the contralateral cochlea in a tonotopic manner; there they form synapses on outer hair cells and influence their electromotility, to eventually reduce the sensitivity of that ear. Lateral olivocochlear neurons, on the other hand, are found in or around the lateral superior olive and project unmyelinated fibres to the ipsilateral cochlea, where they contact auditory nerve terminals. Effects of lateral olivocochlear fibres in the organ of Corti are largely unknown. MOC neurons receive tonotopic input from neurons in the contralateral ventral cochlear nucleus (Romero & Trussell, 2021), possibly from T-stellate cells, and thus are part of a reflex pathway that carries frequency-specific negative feedback. Despite this, the ultimate function of the olivocochlear reflex in humans remains debated. Arguably most conclusive, and best supported by experimental evidence, is the idea that olivocochlear activation provides a mechanism to unmask salient narrow-band stimuli embedded in broad-band noise. This results in a rescue of the dynamic range of the response to the target, albeit at a higher detection threshold compared to listening in quiet (Lopez-Poveda, 2018). MOC neurons also receive descending excitation from the inferior colliculus and the auditory cortex, which challenges the view of a simple stimulus-driven reflex. Such descending projections prompt speculation about context-dependent modulation of specific spectral features of acoustic inputs. Numerous studies have tried to directly show the benefit of top-down control via olivocochlear efferents in various listening tasks. In a recent paper, Otsuka et al. (2022) showed that the intensity of the MOC reflex can be controlled by goal-oriented behaviour. Learned expectations about stimulus properties influenced the intensity of the MOC reflex. This finding provides a small glimpse of the complexity of functions and tasks that MOC-mediated modulation could be involved in (Otsuka et al., 2022). Overall, the MOC efferent fibres seem to represent a common final route of a number of pathways that assert descending control on the inner ear. To make matters even more complicated, Torres Cadenas et al. (2020) recently showed that MOC neurons also receive substantial sound-driven, glycinergic inhibitory inputs from the medial nucleus of the trapezoid body (MNTB). Thus, MOC neurons are clearly a site of input integration. How do the individual excitatory and inhibitory, ascending and descending inputs to the MOC neurons interact with each other? To answer this question it does not suffice to merely report the anatomical presence of a connection. Synaptic short-term dynamics need to be taken into account when forming a conceptual model of the MOC efferent function. One significant step in this direction was taken by Romero and Trussell (2021), who demonstrated that excitatory inputs to MOC neurons differed in short-term dynamics. At comparable input rates, inputs from the ventral cochlear nucleus showed short-term depression while those originating in the inferior colliculus showed augmentation. However, without taking the inhibitory inputs into account, predictions about the resulting efferent activity were very hard to make. To this important piece of the puzzle is now added a study by Torres Cadenas et al. (2022) published in this issue of The Journal of Physiology. Here the authors performed whole-cell recordings from genetically labelled MOC neurons in acute brain slices and electrically activated glycinergic inputs originating in the MNTB. They demonstrate that MOC inhibitory inputs showed rapid short-term depression upon high-frequency activation. However, for sustained activation a significant inhibitory drive builds through summation. Activity, similar to the MNTB baseline, did not cause suppression. Notably, these findings were corroborated in in vivo-like conditions regarding extracellular calcium concentration and temperature. Thus, the study by Torres Cadenas et al. (2022) allowed the authors to conclude that onset of acoustic activation in the intact animal will cause an initial suppression of MOC activity by rapid MNTB inhibition, followed by increasing MOC output due to short-term depression of MNTB inhibitory inputs. During ongoing stimulation a tonic inhibitory drive will develop that likely asserts crucial control on MOC output. Armed with this knowledge, a model of the interaction of sound-driven inputs into the MOC, and thus predictions about the dynamics of MOC efferent activation upon spectrally and temporally varying stimuli, seems in reach. However, the real-life conditions of activating the important top-down inputs to MOC neurons (Romero & Trussell, 2021) are still completely unknown. A stronger focus on experiments involving trained, awake animals responding to behaviourally relevant stimuli is necessary to resolve these questions. The most recent findings about the dynamics of input integration in MOC neurons, presented in this issue of The Journal of Physiology (Torres Cadenas et al., 2022), will undoubtedly be of great help with designing and understanding these future experiments. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article. None. Sole author. None. Open Access funding enabled and organized by Projekt DEAL.

Authors 1

1. Thomas Kuenzel corresponding Aachen Department of Chemosensation Institute for Biology 2 Department of Chemosensation Institute for Biology 2

   RWTH Aachen University

   Affiliation as printed

   Department of Chemosensation Institute for Biology 2 RWTH Aachen University Aachen Germany

   Department of Chemosensation, Institute for Biology 2, RWTH Aachen University, Aachen, Germany