Intrastriatal Shifts Mediate the Transition from Drug-Seeking Actions to Habits

Abstract

s e b n u v m s d A ddictions have been suggested psychologically to result from loss of executive control over maladaptive habits (1). Indeed, over the course of the development of addiction, rug seeking becomes progressively controlled by drug-associated timuli, acting as both conditioned reinforcers and as pavlovian eliciors of relapse and/or craving. This may reflect the subversion by drugs f natural instrumental learning processes, such as response-outcome nd stimulus-response instrumental learning mechanisms as well s pavlovian-instrumental interactions, thereby facilitating the deelopment of habitual control over drug seeking and taking (1,2). he paper by Corbit et al. (3) directly addresses this issue by showng that alcohol seeking becomes habitual over an extended period f access, being resistant to reinforcer devaluation and that this hift from goal-directed to habitual control is underpinned by a hift from dorsomedial to dorsolateral striatal control over behavor. Early theoretical arguments for a role of habits in the developent of drug addiction were based on the observation that “drugse behaviors tend to be relatively fast and efficient, readily enbled by particular stimulus configurations (i.e., stimulus bound), nitiated and completed without intention, difficult to impede in he presence of triggering stimuli, effortless, and enacted in the bsence of awareness” (4). These behavioral features resonate with he main characteristics of behavioral automaticity, often assumed o reflect habits, that Tiffany (4) emphasized: speed, autonomy, lack f control, effortlessness, and absence of conscious awareness. hus, in humans, habitual responses do not depend on awareness ut are directly triggered by conditioned stimuli without any reruitment of higher cognitive processes, such as intention or deciion making, and are therefore difficult to inhibit in the presence of he eliciting stimuli. Additionally, habitual responses are relatively nsensitive to variations in their consequences and impervious to hanges in the value of the goal (for review see [2]). Thus, drug use hat is initially goal directed, controlled by the reinforcing value of he drug, progressively becomes divorced from reward value to be ore controlled by response-eliciting, drug-associated stimuli in he environment. Perhaps the first evidence that animal models can provide inights into the psychobiologic processes that govern the transition rom controlled to uncontrolled alcohol drinking was the demon-