Corticofugal Circuits for Active Listening

项目总结

Project Summary During active listening, sound features that are distracting, irrelevant, or totally predictable are suppressed and do not rise to perceptual awareness. By contrast, inputs selected for amplification convey behaviorally relevant auditory signals used to guide ongoing perceptual decision making. The neural circuit mechanisms that selectively suppress or amplify bottom-up inputs to support active listening remain largely mysterious. Logically, neurons that support active listening would have inputs from cognitive signals that encode expectation, attentional selection and task demands, yet would also be able to adjust the gain and tuning of low-level auditory neurons that encode or compute bottom-up sound features. The massive network of descending auditory corticofugal neurons fit the bill because their cell bodies are embedded in highly plastic centers for cortical sound processing, yet their axons innervate subcortical auditory nuclei in the thalamus, midbrain and brainstem. Addressing the involvement of corticofugal neurons in active listening behaviors has been challenging due to the technical difficulty of isolating and manipulating specific classes of auditory cortex neurons in awake, actively listening animals. Here, we describe an approach to overcome these technical obstacles and address the hypothesis that a specific sub-class of auditory corticofugal neuron, the layer 6 corticothalamic neuron (L6 CT), plays an essential role in sculpting enhanced cortical and perceptual processing of expected sounds. In Aim 1, we will use cutting-edge methods for cell type-specific imaging and electrophysiology in awake, behaving mice to make targeted recordings from two classes of auditory subcerebral projection neurons: layer 5 corticocollicular neurons (L5 CCol) and L6 CTs. We expect to find stark differences in the auditory tuning, sensitivity to internal state variables, local outputs and monosynaptic inputs of L5 CCol and L6 CT neurons (Aim 1a-1d, respectively). In Aim 2, we will record from targeted subtypes of auditory cortex neurons as mice learn to form a spatiotemporal filter for processing expected sounds. We will address the hypothesis that L6 CT neurons modify their activity shortly before the onset of expected sounds to optimize cortical processing of behaviorally relevant signals. In Aim 3, we will test the causal involvement of L6 CT spike patterning for enhanced processing of expected sounds by optogenetically silencing their activity at key times in well-trained mice (to test necessity) or activating them in naïve mice (to test sufficiency). Collectively, these experiments will reveal neural circuit mechanisms that support the selection of bottom-up inputs for enhanced perceptual processing during active listening. By extension, improper regulation of this circuit could underlie the irrepressible awareness of unwanted or distracting sounds (e.g., attention deficit hyperactivity disorder) or the perception of sounds that do not exist in the environment (e.g., tinnitus and schizophrenia). 