Joseph J. Paton

4 mai 2012

Champilamaud Neuroscience Programme, Lisbon, Portugal

A representation of time for learning in the striatum of behaving rats

invité par Sabrina Ravel


Animals from insects to humans use time on a scale of seconds to minutes to guide their actions. Yet, there has not been a direct demonstration of how the brain might encode time on this scale. The basal ganglia (BG) are a site of convergence for information from a broad set of functional territories, and appear to be necessary for the behavioral expression of learned temporal information. We recorded neuronal activity from a major BG input structure, the striatum, in rats trained to press a lever under conditions of changing reward availability times. Rats demonstrated knowledge of changing intervals of reward through adjustments in lever pressing start times. Neural activity during task performance showed varied dynamics around the temporal cue, in this case reward. Reordering of neurons based on time of activation relative to reward revealed a slow-propagating wave of activity across the population. This pattern of activity produces a non-repeating trajectory in the state of the population over the range of tens of seconds to one minute that can then be used to read out time. The accuracy of time estimates derived from this pattern of firing appear to change along with behavior at block divisions where intervals changed, suggesting animals may use a similar representation to guide their actions. These findings suggest that the dynamics of activity patterns in BG circuits reflect temporal basis functions for representing time on the scale of tens of seconds to minutes. This type of temporal representation may be used within models of reinforcement learning to guide action selection.
CNRS logo université Aix Marseille logo | plan du site | mentions légales | contact | admin | intranet | intcloud |