Acetylcholine demixes heterogeneous dopamine signals for learning and moving
Jang, H. J.; McMahon Ward, R.; Golden, C. E. M.; Constantinople, C. M.
Show abstract
Midbrain dopamine neurons promote both reinforcement learning and movement vigour1-12. A major outstanding question is how dopamine-recipient neurons in the striatum parse these heterogeneous signals. Previous work suggests that cholinergic striatal interneurons may play a role, perhaps by gating dopamine-dependent plasticity13-16, but this has not been tested in behaving animals. Here we studied rats performing a decision-making task with reward- and movement-related events at distinct time points. Optical measurement of dopamine and acetylcholine release in the dorsomedial striatum (DMS) revealed distinct dynamics at task events. Reward cues evoked cholinergic pauses with different phase relationships relative to dopamine. When dopamine slightly lagged cholinergic dips, dopamine predicted future behaviour, and DMS firing rates on subsequent trials. In contrast, when dopamine slightly preceded cholinergic dips, there was no observable relationship between dopamine and learning. Finally, when dopamine was coincident with cholinergic bursts, it preceded and predicted the vigour of upcoming contralateral orienting movements. Our findings suggest that the precise phase relationship between dopamine and acetylcholine allows dopamine to be used for either movement or learning depending on instantaneous behavioural context.
Matching journals
The top 3 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Layer 6 ensembles can selectively regulate the behavioral impact and layer-specific representation of sensory deviants 98%
- Response outcome gates the effect of spontaneous cortical state fluctuations on perceptual decisions 98%
- Mixed representations of choice direction and outcome by GABA/glutamate cotransmitting neurons in the entopeduncular nucleus 98%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.