Reward in the cerebellum
Our initial research explored how the cerebellum processes rewards during eye movements. We discovered that certain inputs to the cerebellum, called climbing fibers, signal the expected reward size. In experiments where monkeys performed eye movement tasks, these fibers became more active when the monkeys were shown the reward size they could earn. However, this activity was not influenced by the actual delivery of the reward. This suggests that the cerebellum's role extends beyond error correction to include learning based on associations.
Organization of reward and movement signals in the basal ganglia and cerebellum:
Building on our earlier work, the next phase of the project focused on comparing activity in the basal ganglia and cerebellum. We recorded neural activity in both areas of the same monkeys while they performed tasks involving eye movements and rewards. This analysis revealed key differences in how these regions process movement and reward information.
In both the basal ganglia and the cerebellum, we examined how signals changed as they moved through different stages of processing. In the basal ganglia, task-related signals became stronger as they progressed from intermediate stages to output stages. In contrast, in the cerebellum, the output signals were weaker than those in earlier stages. This suggests that the basal ganglia and cerebellum process information in opposite ways.
We also observed that the output signals from the basal ganglia had a wide range of temporal patterns and activity, which we compared to those in the cerebellum and cortex. Interestingly, the basal ganglia output showed an unusually high level of complexity in its patterns, reflecting a broader range of information being processed at this stage.
Optogenetics in monkeys to study how the brain processes rewards
To investigate how reward signals arise in the basal ganglia, we used a cutting-edge method called optogenetics in monkeys. This approach allowed us to target specific neurons projecting from the frontal cortex to the basal ganglia. By introducing a light-sensitive protein into these neurons using a specialized virus, we could identify and activate them with light (a technique known as opto-tagging).
We recorded the activity of these neurons while monkeys performed tasks involving reward and decision-making. Our findings revealed that reward and choice information is already present in the inputs to the basal ganglia, suggesting that these signals are not created within the basal ganglia itself. Instead, our results highlight the basal ganglia's crucial role in processing, rather than generating, reward-related signals.
