The representation of learning parameters in neural activity patterns in the nucleus accumbens

Abstract

Value-based decision-making is at the core of nearly all motivated behaviors and requires the ability to associate outcomes with specific actions and make adaptive decisions about future behavioral action. At the core of value-based decision-making and reinforcement is the nucleus accumbens (NAc), which is integrally involved in learning, selecting, and executing goal-oriented behaviors. The NAc is a heterogeneous population primarily composed of GABAergic Dopamine-1 (D1) and Dopamine-2 (D2) medium spiny projection (MSN) neurons that integrate glutamatergic and dopaminergic input to ultimately shape basal ganglia motor nuclei output. We thus set out to define how neuromodulators, individual neurons, and neuronal population dynamics in this brain region encode information critical to support adaptive behavior. We first show that D1 and D2 MSNs do not have opposing roles in behavior, as once thought, but instead encode learning parameters key for the development of associations. We then show that dopamine, widely thought to signal reward prediction error, responds in a valence-independent fashion to novelty and that habituation to a novel cue is reflected in the dopamine signal and in the weaker association with an aversive stimulus than an unfamiliar cue. Finally, we suggest how these findings can be integrated to provide a framework for how information encoded in the accumbens can ultimately drive adaptive behavior and disfunction can lead to disease states.