Striatal dopamine supports reward expectation and learning: A simultaneous PET/fMRI study

Area of Science:

• Neuroscience
• Cognitive Neuroscience
• Neuroimaging

Background:

• Mesolimbic dopamine pathways are implicated in reward learning via reward prediction errors.
• Dopamine signaling in the basal ganglia may also contribute to higher-order learning heuristics.
• Human neuroimaging and animal studies provide converging evidence for these roles.

Purpose of the Study:

• To investigate the role of striatal dopamine release in generating optimal learning strategies.
• To examine the interaction between dopamine signaling and cortical networks during reward learning.
• To differentiate the roles of dopamine in reward outcome signaling versus strategy generation.

Main Methods:

• A large multimodal neuroimaging study (N=81) using simultaneous task fMRI and [11C]Raclopride PET.
• Assessed striatal dopamine (DA) D2/3 receptor binding and fMRI activation during a probabilistic reward learning task.
• Utilized reinforcement learning models to quantify learning strategies (e.g., learning rates).

Main Results:

• Both fMRI and PET showed ventral striatum involvement in reward signaling.
• Increased dopamine release was uniquely linked to more task-optimal learning strategies.
• Dopamine release correlated with BOLD activation in a network including ACC, mPFC, thalamus, and PPC during expectation epochs.

Conclusions:

• Striatal dopaminergic signaling interacts with cortical networks to generate task-optimal learning strategies.
• Dopamine's role extends beyond isolated reward outcome representation to strategy optimization.
• Provides novel human in vivo evidence for dopamine's role in higher-order cognitive functions related to learning.