Neural circuits for flexible audiomotor learning

The mammalian auditory system is remarkably adaptive; salient experiences and behavioral contexts can fundamentally alter the processing of sounds in order to sensitize neural circuits to behaviorally relevant information. How does the central auditory system learn to associate sounds to rewards, and relatedly, how does behavioral context mediate this plasticity? The formation of representations of sensory signals such as speech, music, and other forms of acoustic learning is critical for survival. And, yet, the formation of these representations during real-time learning remains largely unknown. In this proposal, we posit that learning can be dissociated into two distinct learning processes: the initial acquisition and subsequent expression of knowledge. Acquisition involves learning the core discrimination learning that underlie a behavior, and expression entails the use of this acquired discrimination in context. Acquisition and expression have typically been conflated in most laboratory tasks, leaving an important gap in our understanding of learning mechanisms in the central auditory system. Moreover, dissociating between acquisition and expression has important implications for development and language disorders. For example, soothing music can elicit neurotypical behavior in autism patients with otherwise severe symptoms. We aim to identify the separable neural mechanisms that enable sensorimotor acquisition versus contextual expression. Recently, we have shown that we can precisely dissociate acquisition from expression in a sensorimotor reward learning task. Thus, we now have a powerful behavioral approach to isolate acquisition from expression during learning. In this proposal, we will define the precise neural circuitry in the auditory cortex that enables these two aspects of learning. The auditory cortex is known to be a major site of plasticity; associative learning between sounds and rewards induce shifts in the “tuning” of cortical neurons. The cholinergic basal forebrain, moreover, has been implicated as a potent driver of receptive field plasticity in the central auditory system. These plasticity mechanisms likely reflect fundamental neural changes that are linked to acquisition of task knowledge. A1 is also heavily modulated by brain state and context, suggesting that A1 may also play a role in expression of task knowledge. Here, we propose to combine simultaneous real-time two-photon imaging of neurons in the auditory cortex (Aim 1) and cholinergic axons (Aim 2-3). We will perform causal manipulations of AC (Aim 1), cholinergic activity (Aims 2-3), in vivo whole-cell voltage clamp recordings (Aim 2), and detailed behavioral analysis (Aims 1-3) to determine the neural basis of audiomotor learning.