Subsecond fluctuations in extracellular dopamine encode reward and punishment prediction errors in humans

Area of Science:

• Neuroscience
• Neurobiology
• Computational Neuroscience

Background:

• Midbrain dopamine neurons are thought to encode reward prediction errors, influencing learning and behavior.
• Evidence for dopamine's role in human punishment learning is limited.
• Understanding dopamine's function in both reward and punishment is crucial for brain research.

Purpose of the Study:

• To investigate whether dopamine release in the human striatum encodes reward prediction errors, punishment prediction errors, or both.
• To test the valence-specific encoding of prediction errors by dopamine in humans.
• To provide direct evidence for dopamine's role in human learning and decision-making.

Main Methods:

• Intracranial, subsecond measurements of dopamine release in the human striatum.
• Volunteers (patients undergoing deep brain stimulation surgery) performed a probabilistic reward and punishment learning choice task.
• Analysis of dopamine levels in relation to prediction errors during the task.

Main Results:

• Extracellular dopamine levels in the human striatum encode reward prediction errors.
• Extracellular dopamine levels also encode punishment prediction errors.
• These distinct signals are processed via independent valence-specific pathways within distinct time intervals.

Conclusions:

• Human dopamine release signals both reward and punishment prediction errors.
• Dopamine plays a critical role in both reward-based and punishment-based learning in humans.
• Findings support a more complex role for dopamine in adaptive learning than previously demonstrated.