The long-term goal of this research is to understand the neural basis of learning and memory, especially how the brain learns complex motor behaviors, guided by sensory information. Vocal learning in songbirds provides a useful model system for this purpose, with special relevance to human speech learning. Songbirds learn to produce a copy of a previously memorized tutor song during a period of """"""""sensorimotor"""""""" learning, in which they use auditory feedback of their own voice to refine their vocal output until it matches the memorized song. The work proposed here focuses on a particular part of the system of brain areas devoted to song learning and production, a specialized cortical-basal ganglia circuit known as the anterior forebrain pathway (AFP), because it plays a crucial but ill-understood role both in song learning and in adult vocal plasticity. Moreover, cortical-basal ganglia circuits, which are well conserved evolutionarily, are thought to function in motor and reinforcement learning in many vertebrates, and to be a critical site of dysfunction in a number of neuropsychiatric disorders. Because the songbird AFP is a discrete cortical-basal ganglia circuit controlling a specific behavior, it may prove a particularly tractable system for elucidating the very general functions of such pathways, both normally and in disease. The APP develops song-selective auditory responses during learning, and shows motor-related activity during singing, but how these sensory and motor responses relate to each other is not clear. Activity in this circuit is also very variable from trial to trial, raising the question of how it could reliably encode information or guide song. With simultaneous recordings from multiple neurons in the output nucleus of the APP, LMAN, during both singing and song playback, the first aim will test the hypothesis that the APP encodes relevant song- and singing-related information in the form of a distributed """"""""population"""""""" code. A further hypothesis is that specific patterns of APP neural activity are critical for normal song development, perhaps guiding the formation of connections in the vocal motor nucleus RA. This will be tested with simultaneous recordings of neurons in both LMAN and RA, so that the covariance of their activity and how it relates to vocal output can be analyzed. LMAN-RA interactions will be studied first in normal birds at different stages of learning, and then after disruptions of the AFP activity pattern, in ways that will shed light both on normal synaptic processing within the APP, as well as on how this circuit influences the song motor pathway.
Showing the most recent 10 out of 41 publications
