Impairments in reversal learning following short access to cocaine self-administration

Cocaine use disorder is characterized by compulsive drug-seeking that persists long into abstinence. Work using rodent models of cocaine addiction has found evidence for reversal learning deficits 21 days after non-contingent cocaine administration and 60 days after self-administration. Here we sought to determine if a deficit in reversal learning is present 3-4 weeks after cessation of cocaine self-administration, when relapse to cocaine-seeking is robust. Conversely, we hypothesized that reversal learning training would protect against relapse, similar to other forms of environmental enrichment.

Male rats underwent short access (ShA, 2 h/10d) or long access (LgA, 1 h/7d then 6 h/10d) cocaine self-administration, followed by 21-29 days of abstinence. During abstinence, a subset of rats underwent training in a plus-maze that required an egocentric strategy to earn a sucrose reward. Following response acquisition and retention, the ability to reverse the spatial navigation strategy was tested.

Total trials to criteria and total errors made did not differ between the groups during response acquisition, retention, or reversal. On the first reversal test, ShA rats performed better than LgA and control rats. ShA rats' performance worsened over time. There were no effects of cognitive training or length of cocaine access on context-primed relapse of cocaine-seeking.

The present data indicate that perhaps LgA cocaine self-administration does not produce adaptations to regions mediating context-primed relapse as it does for cocaine and cocaine-associated cue-induced reinstatement of drug-seeking. A time-dependent deficit in reversal learning was found only in ShA rats. Reversal learning training did not protect against cocaine relapse.