Testing if and how 3-factor learning rules operate at cortico-basal ganglia synapses to drive motor skill learning - How transient patterns of synaptic activity result in the learning of rapid, sequential motor skills is unknown. In sequentially organized skills, the brain must somehow keep track of what movements were made when and selectively reinforce only the “correct” motor gestures. How such a record is created and selectively reinforced is unclear. An influential idea is that the brain solves this problem using a 3-factor learning rule at cortico-basal ganglia (C-BG) synapses: coincident presynaptic (1) and postsynaptic (2) activity generates a time series of transient eligibility traces that provide a record of what movements were made when. Next, a dopamine (DA) signal encoding performance quality acts on the trace to selectively reinforce synapses associated with correct gestures (3). Nonetheless, if and how 3-factor rules operate during the learning of rapid, sequential movements remains to be rigorously tested. A juvenile male zebra finch learns to imitate the song of an adult male tutor to generate a rapid sequence of elementary gestures, providing a powerful system for investigating the neural mechanisms of sequential skill learning. In the context of birdsong, a 3-factor learning model posits that two different cortical inputs separately encode “what movement” and “when,” and coincident activity of both inputs (1) is necessary to create an eligibility trace in the BG (2) for DA to act on (3). In preliminary studies, I localized the when-encoding C-BG synapses as the site where DA acts to drive learning. Here I will exploit this localization to test my central hypothesis: that a 3-factor rule operates at C-BG synapses to drive synaptic plasticity and song learning. In parallel, I will explore the intracellular cascades that transduce C-BG and DA activity to generate the synaptic eligibility trace. To test this hypothesis, I will combine cutting-edge computational, in vivo molecular-genetic, and ex vivo slice electrophysiology. My Specific Aims are 1) To isolate the factors that drive song learning and C-BG plasticity. 2) To investigate how dopamine is transduced intracellularly in a cell-type-specific manner. 3) To establish the necessity, sufficiency, and time constant of a candidate eligibility trace. Individually, each Aim will unravel fundamental synaptic and molecular mechanisms that enable the brain to translate transient patterns of synaptic activity into long-lasting plastic changes capable of driving natural learning. I will conduct this research under the supervision of Dr. Richard Mooney, with critical support from Drs. John Pearson, Tianyi Mao, and Ryohei Yasuda. This interdisciplinary team of accomplished mentors will provide me with complementary expertise in electrophysiological (Mooney), computational (Pearson), and molecular genetic (Mao, Yasuda) approaches. I bring my expertise in behavioral analysis and closed-loop optogenetic approaches and will fill a critical gap in my training: in vivo single-cell imaging and ex vivo electrophysiological techniques. These are vital approaches for me to achieve the long-term goal of my independent research aimed at investigating the neural mechanisms of motor learning.
