Population coding of complex auditory stimuli in auditory cortex

Project Summary The processing and categorization of complex sounds is a skill critical to interacting with one’s environment for many animals, the disruption of which can negatively affect quality of life for individuals suffering from auditory processing disorders. Despite this importance, the processing and categorization of complex auditory stimuli have yet to be fully understood. Existing work largely investigates the encoding of pure tone stimuli, which are scarce in nature and not behaviorally relevant for most animals. Moreover, few studies have attempted to prove causal relationships between neural activity correlated with stimulus presentation, and perception of a given sound. Some such studies have shown improvements in behavior following optogenetic activation strategies that stimulate broad subfields of auditory cortex; however, these strategies have been too spatially coarse to interrogate the precise circuitry underlying encoding of stimuli. In this project we thus propose to investigate the neural circuitry involved in categorization of auditory stimuli in mice performing a two-alternative forced choice (2AFC) discrimination task, at three different levels along the auditory cortical pathway. We will examine auditory cortical population dynamics first during categorization of unfamiliar pure tones (intermediate between trained categories), identifying networks using Granger causality analysis and probing causality using holographic stimulation. Given the importance of stimulus identity in 2AFC tasks, we expect to identify GC subnetworks that show an overrepresentation of cells tuned to the frequencies of trained stimuli, and expect that stimulation of strongly linked cells during presentation of intermediate stimuli will bias behavioral response towards the category associated with activity of those cells. We will subsequently examine network dynamics during presentation of more behaviorally relevant harmonic stimuli. Although harmonic structure has been shown to be necessary for perception of vocalizations by mice, the processing of periodicity by the mouse auditory system remains poorly understood. We hypothesize that primary auditory cortex (A1) layer 4 (L4) cells tuned to the components of harmonics carry stimulus information to harmonicity-sensitive cells in layer 2/3, which in turn is passed to cells sensitive to category membership as well as to periodicity-sensitive cells in secondary auditory cortex. We expect this work to lead to a more complete picture of the processing of harmonic sounds along the cortical hierarchy, as well as shed light on the neural encoding of categories through examination of the causal relationship between ensemble activity and behavior.