Author's Department
Biology
Language
English (en)
Date Submitted
5-2018
Research Mentor and Department
Roland Ferger, Michael V Beckert, Keanu Shadron, José L Peña -- Dominick P. Purpura Department of Neuroscience, Albert Einstein College of Medicine
Restricted/Unrestricted
Unrestricted
Abstract
Barn owls (Tyto furcata) are a model organism to study sound location and provide access to a neural population representing auditory space. They use interaural time differences (ITD) and interaural level differences (ILD) to localize sound. Neurons in the barn owl’s optic tectum (OT), a part of the midbrain, which respond to combinations of these binaural cues, form a neural map of auditory space which supports sound-orienting behavior. Understanding population responses in the auditory system involves identifying the frequency components of acoustic signals which evoke spikes. Together, these frequency components and the time between their occurrence in the stimulus and a spike form the spectrotemporal receptive fields (STRFs) of auditory neurons. Using spike times to measure STRFs can elucidate if and how neurons along the sound localization pathway are uniquely tuned to spectrotemporal features. This question holds importance in light of a previous study which showed that neurons within the owl’s map of auditory space are selectively tuned to frequencies carrying the most reliable spatial cues (Cazettes et al. 2014). Therefore, in order to achieve a more complete understanding of population responses in the OT, we investigated the neurons’ STRFs.
To compute STRFs, we presented dichotic stimuli via headphones to anaesthetized barn owls while recording OT neuron responses with a microelectrode array. We analyzed the recorded spike trains in conjunction with the stimuli using custom written MATLAB scripts. Analysis of the neurons' STRFs required identification of the stimulus features which most frequently precede a spike. Therefore, we captured 20 ms of the stimulus waveform which precedes each spike and band-pass filtered these waveforms into 80 frequency channels. Next, we calculated the power over time within each frequency channel and averaged over all spikes to yield a STRF plot for the neuron. A baseline, produced analogously with randomized spike trains, was subtracted to account for global stimulus features triggering spikes. A peak in the STRF indicates high power within a specific frequency preceding the spikes at a given latency.
To test our method, we first computed STRFs in auditory nerve (AN) neurons, known to have narrow spectrotemporal tuning. STRFs of AN neurons indeed showed a single peak, indicating precise tuning to frequency, and confirmed that our MATLAB code was able to determine a neuron’s STRF efficiently. The AN neurons had firing rates of 100-400 spikes/s and STRFs were calculated with 4,000-10,000 spikes. The same analysis of OT neurons showed that STRF peaks in the OT were not as precisely localized in a single spectrotemporal region as in AN. A reason for this difference could be that OT neurons have lower firing rates (20-50 spikes/s) than AN and STRFs were based on a smaller number of spikes (500-1,000 spikes). To verify whether more defined STRFs were achievable in OT, we selected cells showing reproducible responses across trials of identical stimulus (frozen noise). This method yielded more precise STRFs.
Our preliminary results indicate that neurons in the owl’s OT are tuned to spectrotemporal features of sound. In addition, we identified response properties that generally lead to well-defined STRFs (low jitter to frozen noise, high number of spikes and/or high spike rates). Future directions in this area may include (a) screening more neurons in the OT with these properties in mind, (b) continuing to compare STRFs across neural populations and across varying stimulus parameters (ITD, ILD, frequency, …), and (c) investigating if OT neurons follow the previously shown relationship between spatial cues and frequency tuning (Cazettes et al. 2014).
Comments
Funding: Brain Initiative grant NS104911
References: Cazettes F, Fischer BJ, Peña JL (2014) Spatial cue reliability drives frequency tuning in the barn owl’s midbrain. eLife 3:e04854