Dopamine signaling and the distal reward problem.

ABSTRACT

Actions and their associated consequences, such as reward attainment, are often temporally distant. Animals nevertheless learn such associations thereby solving the 'distal reward' problem. We sought to determine whether dopamine signaling plays a role in such learning. Wild-type and dopamine type I receptor knockout mice executed three left/right choices leading to one of eight differentially rewarded goal sites. Compared with wild-type mice, knockouts exhibited selective impairments in decision making at choice points distal, but not proximal, to goal sites. We conclude that dopamine's role in reinforcement learning depends on the temporal relationship of actions to reward and that dopamine signaling through D1 receptors constitutes a component of those brain mechanisms responsible for solving the distal reward problem.