Effects of chronic cocaine self-administration and N-acetylcysteine on learning, cognitive flexibility, and reinstatement in nonhuman primates

RationaleCocaine use disorder (CUD) is associated with cognitive deficits that have been linked to poor treatment outcomes. An improved understanding of cocaine's deleterious effects on cognition may help optimize pharmacotherapies. Emerging evidence implicates abnormalities in glutamate neurotransmission in CUD and drugs that normalize glutamatergic homeostasis (e.g., N-acetylcysteine [NAC]) may attenuate CUD-related relapse behavior.ObjectivesThe present studies examined the impact of chronic cocaine exposure on touchscreen-based models of learning (repeated acquisition) and cognitive flexibility (discrimination reversal) and, also, the ability of NAC to modulate cocaine self-administration and its capacity to reinstate drug-seeking behavior.MethodsFirst, stable repeated acquisition and discrimination reversal performance was established. Next, high levels of cocaine-taking behavior (2.13-3.03mg/kg/session) were maintained for 150 sessions during which repeated acquisition and discrimination reversal performance was probed periodically. Finally, the effects of NAC treatment were examined on cocaine self-administration and, subsequently, extinction and reinstatement.ResultsCocaine self-administration significantly impaired performance under both cognitive tasks; however, discrimination reversal was disrupted considerably more than acquisition. Performance eventually approximated baseline levels during chronic exposure. NAC treatment did not perturb ongoing self-administration behavior but was associated with significantly quicker extinction of drug-lever responding. Cocaine-primed reinstatement did not significantly differ between groups.ConclusionsThe disruptive effects of cocaine on learning and cognitive flexibility are profound but performance recovered during chronic exposure. Although the effects of NAC on models of drug-taking and drug-seeking behavior in monkeys are less robust than reported in rodents, they nevertheless suggest a role for glutamatergic modulators in CUD treatment programs.