A principle aim of the NINDS is to determine how motor sequences are constructed by the nervous system. Dopamine (DA)-basal ganglia (BG) circuits are required for motor sequence learning, but it remains unclear how these circuits guide the trial-and-error learning process. Remarkably, our current understanding of these pathways comes largely from studies of animals learning simple actions for external rewards such as food or juice. Yet symptoms of BG diseases such as Parkinson's, Huntington's and dystonia include degradation of motor behaviors unrelated to reward seeking. And most human behaviors, such as learning a sport or an instrument, are not simple actions in pursuit of external rewards but are instead complex motor sequences learned by matching performance to internal goals. The songbird model system offers a unique opportunity to study how internally guided motor sequences are constructed. Zebra finches learn their song by matching a complex vocal sequence to an auditory memory of a tutor song. This sensorimotor learning requires a DA-BG circuit that is part of a tractable 'song system.' We will apply our core strengths in awake- behaving electrophysiology to the tractable songbird model system to decipher how motor performance is evaluated during practice. First, to test if DA neurons evaluate motor performance (the 'error' part of learning) we will conduct the first-ever recordings of BG-projecting DA neurons while controlling song 'error' with distorted auditory feedback (Aim 1). Preliminary recordings support the hypothesis that DA neurons encode 'performance prediction error' signals during singing. To determine how upstream sensorimotor signals compute 'error,' we will record from auditory cortical and BG projections to DA neurons in singing birds during the error-feedback task (Aim 2). Finally, zebra finches sing in two DA-dependent motor states: a variable practice mode when alone and a female-directed, stereotyped performance mode. To test if DA can both evaluate performance and also control its variability, we will record DA neurons during the error feedback task during undirected-to-directed song state transitions (Aim 3). Altogether, these studies will identify the neural correlates of the internal evaluation system that construct motor sequences. A major impediment to understanding pathological activity patterns observed in BG-related diseases is a limited understanding of signal propagation through the healthy circuit. The proposed work aims to understand the functions of DA-BG signals and how they are processed at successive stages of the circuit. At stake in this issue is the potential to tailor therapies, such as neural circuit re-programming and deep brain stimulation for movement disorders, based on detailed knowledge of normal brain physiology.
The basal ganglia are a set of brain nuclei highly conserved among vertebrates, including humans, critical for motor control and trial-and-error learning. This proposal will leverage advantages of a highly tractable model system, the songbird, to determine how the basal ganglia control and evaluate motor performance, two processes that go awry in a wide variety of basal ganglia diseases.
| Gadagkar, Vikram; Puzerey, Pavel A; Chen, Ruidong et al. (2016) Dopamine neurons encode performance error in singing birds. Science 354:1278-1282 |
