Effects of persistent cocaine self-administration on amygdala-dependent and dorsal striatum-dependent learning in rats

Abstract

Rationale: The influence of persistent cocaine self-administration on learning and memory has never been evaluated.

Objectives: Our objective was to isolate the effects of contingently administered cocaine from those of its general pharmacological or non-contingent actions on multiple memory system functioning.

Methods: A triad design was used to yoke passive cocaine and saline administration to the behavior of rats who were actively self-administering cocaine. Following 4 weeks of cocaine or saline exposure in 2-h sessions, six triads were tested in the amygdala-dependent conditioned cue preference task and dorsal striatum-dependent win-stay task in an eight-arm radial maze environment. Drug or saline sessions continued throughout task testing.

Results: Throughout task testing, rats actively and passively exposed to cocaine sustained a total daily intake of approximately 15 mg/kg. During the conditioned cue preference task, saline-exposed rats showed robust conditioned preference for a Froot Loops-paired cue. Rats actively and passively exposed to cocaine showed no evidence of conditioning despite normal exploration in the maze during preference testing. For the win-stay task, no significant differences were found among the three groups in terms of the number of sessions to acquire the task or task accuracy at criterion. Rats actively or passively exposed to cocaine, however, completed sessions more quickly than saline-exposed rats at criterion.

Conclusion: These findings suggest that contingent and non-contingent cocaine administration similarly disrupt stimulus-reward functions of the amygdala, but do not disrupt stimulus-response functions of the dorsal striatum. This dissociation may relate to differences in the rate by which dopamine is cleared from these tissues following cocaine exposure or possibly to cocaine-induced devaluation of natural rewards, which influences stimulus-reward learning, but not stimulus-response learning.