Abstract
The basal ganglia, and the striatum in particular, are critical for action reinforcement [1,2]. The dorsal striatum, which can be further subdivided into dorsomedial (DMS) and dorsolateral (DLS) striatum, is mainly composed of two subpopulations of striatal medium spiny projection neurons (MSNs): dopamine D1 receptor-expressing MSNs that constitute the striatonigral or direct pathway (dMSNs); and dopamine D2 receptor-expressing MSNs that constitute the striatopallidal or indirect pathway (iMSNs) [3]. It has been suggested that each pathway has opposing roles in reinforcement, with dMSNs being important to learn positive reinforcement and iMSNs to learn to avoid undesired actions (Go/No-Go) [1]. Furthermore, optogenetic self-stimulation of dMSNs in DMS leads to reinforcement of actions, while self-stimulation of iMSNs leads to avoidance of actions [2]. However, in DLS, which has been implicated in the consolidation of well-trained actions and habits in mice [4,5], both pathways are active during lever-pressing for reward [6]. Furthermore, extensive skill training leads to long-lasting potentiation of glutamatergic inputs into both dMSNs and iMSNs [4]. We report here that, in DLS, both dMSNs and iMSNs are involved in positive reinforcement, but support different action strategies.