Mesostriatal dopamine is sensitive to changes in specific cue-reward contingencies

Area of Science:

• Neuroscience
• Behavioral Science
• Computational Neuroscience

Background:

• Learning causal relationships is fundamental for survival and decision-making.
• Dopamine's role in learning is often explained by reinforcement learning models, but its function in encoding causal contingencies is less understood.

Purpose of the Study:

• To investigate how dopamine neuron activity and neurotransmitter release are affected when the learned causal relationship between a cue and a reward is degraded.
• To compare the explanatory power of standard reinforcement learning models versus causal learning models for dopamine responses.

Main Methods:

• Utilized outcome-selective Pavlovian contingency degradation in rats.
• Measured conditioned responding, dopamine neuron activity in the midbrain, and dopamine release in the ventral striatum.
• Employed optogenetic manipulations of dopamine during noncontingent rewards.

Main Results:

• Conditioned responding to the cue-reward contingency was attenuated following degradation.
• Dopamine neuron activity and release in response to the cue and reward were reduced.
• Contingency degradation eliminated the trial-by-trial history dependence of dopamine responses.
• A causal learning model better explained dopamine responses and behavior than standard reinforcement learning models.

Conclusions:

• Mesolimbic dopamine system activity reflects the learned causal contingencies between events.
• Dopamine's role extends beyond simple reward prediction error to encoding the structure of causal relationships.
• Findings challenge standard reinforcement learning frameworks and support causal inference mechanisms in dopamine function.