Auditory neurons concerned with temporal processing are the most precise time analyzing units in the mammalian brain. Some auditory neurons exhibit time resolutions of only a few µs. We are interested in the neuronal mechanisms of temporal auditory processing and their evolution in mammals. In particular, our studies are concerned with the role of neural inhibition in temporal processing. Inhibition has been more or less neglected as a possible player in neuronal filtering of temporal cues. However, recent results from several groups indicate a link of age related hearing deficits in temporal processing, age related down-regulation of the inhibitory transmitters GABA and glycine, and the role of inhibition in temporal filtering as found in the bat and gerbil auditory brainstem. The analysis of temporal cues of sounds is important for two basic tasks: (1) sound localization and (2) sound recognition.
1. Sound localization
Differences in arrival time of a sound at the two ears (interaural time differences, ITD) are the main cue for localizing low frequency sounds (in humans: frequencies <1400 Hz). The primary ITD-coding structure in the mammalian auditory system is the medial superior olive (MSO). MSO neurons in cats, dogs, and gerbils have been shown to be ITD-sensitive with a time resolution in the range of only tens of µs. We investigate the mechanism underlying this ITD sensitivity and its possible evolution.
In the classical view MSO neurons function as coincidence detectors for excitatory inputs that are systematically arranged in the form of delay-lines, thereby creating a topographical map of azimuthal space. However, MSO neurons receive prominent inhibitory inputs mainly from glycinergic neurons in the medial nucleus of the trapezoid body, which in turn are driven via the calyx of Held, the most temporally precise chemical synapse in the mammalian brain. The role of this inhibition in ITD processing, however, was unknown. Additionally, a recent study raised doubts concerning the existence of a map of ITDs in the mammalian auditory system.
Our in vivo recordings from gerbil MSO neurons show that many ITD-functions peak outside the physiological range of ITDs for the gerbil (gray area in Fig.1a,b) and that these peaks depend on the best frequency of a given neuron. The best ITD of neurons tuned to high frequencies is shorter compared to that of low frequency neurons. This allows for the maximal slope to be within the physiological range, regardless of frequency. Blocking the glycinergic inhibition with strychnine lead to a shift of the best ITD towards zero ITD and reduced the change in response magnitude across the physiological range of ITDs (Fig. 1b). These findings indicate that exquisitely timed glycinergic inhibition tunes ITD sensitivity, which is preset by coincidence of excitatory inputs, to the physiologically relevant range. Additionally, our data confirms that in the mammalian auditory system, ITDs are not represented in the form of a topographical map.
2. Sound recognition
The spectral and temporal structure of a sound is important for sound recognition. For instance, in speech perception spectral cues (lower panel in the figure) are necessary to recognize a voice or the emotional state of a speaker. The characteristic modulations in amplitude (upper panel in the figure) and frequency are necessary and sufficient to understand spoken language. Our studies show that specific membrane properties as well as inhibitory inputs contribute to the temporal selectivity of single neurons.
Approaches and Techniques
We use a comparative approach investigating animals living in different
'auditory worlds' (high frequency specialists like bats; models for ancient mammalian hearing like short tailed opossums; low frequency specialists like gerbils; mice and rats as “standard” models for hearing in modern placental mammals. Additionally, we are using transgenic mice to study developmental mechanisms.
- extracellular recording single unit techniques in vivo combined with acoustic stimulation (dichotic and free-field)
- manipulation of early auditory experience (combined with anatomical and/or physiological investigations)
- multi-electrode recordings
- patch-clamp recording techniques in vitro (acute brain slices and in vivo)
- immunohistochemistry (confocal microsopy)
- classical anatomical techniques (e.g. tracing studies)
- transgenic mice (collaboration with Wolfgang Wurst, HelmholtzZentrum München)
- modeling of spatial and temporal auditory processing
- human and animal psychoacoustics
- behavoral approaches