Striatal dynamics during cued action initiation


January 25, 2017 - 12:00pm
Northwest 243
About the Speaker
Konstantin Bakhurin
Speaker Affiliation: 

The striatum is the major input structure of the basal ganglia and is believed to play a role in the learning and generation of reward-guided actions. Striatal dynamics are thought to be involved in enabling animals to time their behavior during reward prediction. However, how such functions arise from the striatum’s interaction with external inputs is not well understood. I will introduce a silicon-based microelectrode technology which enables simultaneous recordings from up to hundreds of units, and multiple brain regions, in behaving mice. I used these tools to record large-scale activity patterns from the striatum and upstream prefrontal inputs, the orbitofrontal (OFC) and premotor cortex, in mice performing a reward prediction task in which they generate anticipatory licking behavior. I applied machine-learning approaches to extract the time of licking initiation from cortical and striatal dynamics on a single-trial basis, and found that the striatum outperforms the OFC in encoding several measures of time during the task. I also show that the premotor cortex and the striatum functionally interact during anticipatory behavior. Finally, I examined the influence of cortical input on striatal dynamics by combining large-scale recordings with transient optogenetic suppression of corticostriatal projections. I found that striatal encoding of action initiation is dynamically modulated by cortical input.