One of the most outstanding challenges in systems neuroscience is to understand the neural circuits that implement motor learning. The brain uses sensory inputs to guide its motor outputs, but sensory feedback also can cause changes in motor behavior. The specialized neural circuitry of the cerebellum is thought to play a critical role in sensory-motor learning. The present application will understand how individual features of cerebellar circuitry implement sensory-motor learning by studying direction learning during smooth pursuit eye movements. Pursuit is a simple sensory motor behavior that exhibits cerebellum-dependent motor learning. In the direction-learning paradigm, subjects move their eyes to follow the path of a moving visual target. After a fixed duration, the pursuit target suddenly but predictably changes directions. Repeated presentations of the instruction for direction-learning causes smooth eye movements that predict the upcoming direction change before it occurs.
The first aim of my proposal is to dissect the learning process into multiple components that operate on different time scales. By following learning for up to 1500 trials, and probing for short-term versus consolidated components of learning, we will reveal how different aspects of the pursuit system may contribute to different components of learning. Next, single unit recordings from Purkinje cells in the floccular complex of the cerebellum during 1500 trials of direction-learning will test the hypothesis that early learning in the cerebellar cortex is transferred to the deep nuclei. These recordings will test the hypothesis that rapid, early changes in Purkinje cell firing rate are responsible for early timescales of behavioral learning, but that the early changes decay as learning is consolidated outside of Purkinje cells. The last aim of the proposal pushes the study of cerebellar motor learning beyond the classic focus on Purkinje cells. We will use multiple-contact electrodes to make simultaneous recordings from multiple, nearby elements in the cerebellar cortex. We will examine the roles of non-Purkinje cells in cerebellar learning and test for specific sites of cerebellar plasticity during learning. Analysis of neuron-neuron spike-count and spike timing correlations between interneurons and Purkinje cells will probe for changes in circuit connectivity during learning. Careful measurement of the properties of neurons that connect to Purkinje cells will show the types of cerebellar inputs that are being learned. Together this proposal will push the understanding of cerebellar motor learning beyond Purkinje cells and toward a full neural circuit explanation.
The study of cerebellar motor learning in the context of smooth pursuit affords an opportunity to elucidate sensory motor control in a complete neural circuit. An understanding of how neural circuits implement learning and motor behavior will move us towards understanding how visual sensory inputs guide movement, and also will provide a knowledge base for addressing the numerous neurological disorders that arise from deficits in learning and motor control.
