Dopamine dynamics during stimulus-reward learning in mice can be explained by performance rather than learning.

Area of Science:

• Neuroscience
• Behavioral Science
• Computational Neuroscience

Background:

• The reward prediction error (RPE) hypothesis suggests dopamine (DA) signals reward discrepancies to drive learning.
• Emerging evidence indicates DA might regulate behavioral execution rather than learning signals.

Purpose of the Study:

• To investigate the role of ventral tegmental area (VTA) DA activity in a Pavlovian stimulus-reward task.
• To determine if VTA DA encodes RPE or modulates behavioral performance.

Main Methods:

• Recorded and manipulated VTA DA activity in head-fixed mice using optogenetics.
• Measured subtle movements with force sensors during a Pavlovian task.
• Analyzed DA neuron activity in relation to force exertion, licking, learning, and reward parameters.

Main Results:

• Identified distinct VTA DA neuron populations tuned to force exertion, active during spontaneous and conditioned behaviors.
• Found that force and licking variations fully explain DA dynamics previously attributed to RPE.
• Demonstrated that optogenetic DA manipulation modulates force exertion and behavioral transitions in real-time without affecting learning.

Conclusions:

• VTA DA activity does not encode RPE but rather modulates the gain of motivated behaviors.
• DA neurons dynamically control the latency, direction, and intensity of behavioral output during performance.