Will talk about: Dopamine and learning mechanisms in the basal ganglia
John Reynolds is an Associate Professor in Neuroscience in the Department of Anatomy at University of Otago in New Zealand. His research team studies the application of neuroplasticity approaches to Parkinson’s disease and stroke. His interest is in understanding the natural conjunction and timing of neurmodulator signals underlying synaptic plasticity in affected brain areas. John graduated in Medicine in 1994, and then returned to University of Otago after a few years in practice to undertake PhD studies. He has received an international Brain Research Young Investigator Award and a National Tertiary Teaching Award, and he currently holds a Rutherford Discovery Fellowship from the Royal Society of NZ. He chairs the Scientific Advisory Committee of the Neurological Foundation of NZ.
Organisms with multifunctional capability but limited motor resources must decide which competing action will be best to perform in any given situation, to maximize the likelihood of a positive outcome and minimise negative consequences. This requires a brain system specialized to solve the problem of selection, ie. capable of deciding which functional system should be allowed access at any given time to the machinery driving behavioural output. The re-entrant loop architecture connecting the basal ganglia subcortical nuclei to external brain structures such as cortex and thalamus can be viewed as an ideal substrate providing a solution to this problem. A major control point within these loops is the synapses between the cortex and the major input structure of the basal ganglia, the striatum. Efficacy of these corticostriatal synapses can be modified via a three-factor learning rule, involving the interaction between cortical inputs influencing the firing of striatal neurons, and phasic activity in midbrain dopamine systems reporting the presence of salient events. Thus the basal ganglia can be viewed as a selection system with an integral reinforcement learning mechanism, biasing the system towards specific behavioural outcomes based on prior experience. In this talk, I will review the basic structure of the selection machinery of the basal ganglia, and the rules for synaptic plasticity within the striatum as they might relate to reinforcement learning. I will present in vivo data from our laboratory that is pointing to a critical timing requirement for dopamine in corticostriatal synaptic plasticity and links dopamine reinforcement on a second by second timescale with the millisecond sensitivity of spike-timing dependent plasticity.