Consistent with the Opponent Process Theory of motivated behavior, cocaine administration produces dual and opposing affective states: an initial rewarding "high" followed by a dysphoric/anxiogenic "crash". It therefore seems likely that the motivation to seek cocaine is dependent upon the organism's assessment of the positive relative to the negative consequences of its use. While the neurobiology of the reinforcing aspects of cocaine has been well established, less is known about the systems responsible for the drug's negative actions. In this context, the current study involved an assessment of serotonergic (5-HT) function, which is enhanced by cocaine administration and has been linked to the presence of anxiogenic and depressive states in human and animal studies. In particular, we investigated the role of 5-HT within the bed nucleus of the stria terminalis (BNST) --

a structure within the extended amygdala that is activated during periods of stress and during the negative affective state associated with the withdrawal from drugs of abuse. The present study tested the hypothesis that increased 5-HT release in the BNST contributes to the anxiogenic effects of cocaine. A runway self-administration paradigm was employed in which animals were trained to traverse a straight alley in order to earn an infusion of IV cocaine (1.0mg/kg) delivered upon goal box entry. Testing consisted of 16 single daily trials. In this task, animals develop ambivalence about goal box entry (reflected by the development of approach/avoidance "retreat" behaviors) that we have shown to reflect the dual positive (rewarding) and negative (anxiogenic) associations that subjects form with the cocaine-paired goal-box.

To assess the involvement of 5-HT signaling within the BNST on this conflict behavior, prior to each trial rats received bilateral intracranial injections of CP94,253 (0.5microg and 1.0microg/side in 0.5microl or aCSF vehicle), a potent and selective 5-HT1B agonist that inhibits local 5-HT release via activation of terminal autoreceptors. Results indicated that CP94,253 did not alter the positive incentive properties of cocaine (start latencies were unaffected) nor did it alter gross motor behavior (as revealed in subsequent locomotor activity testing). Treatments did, however, selectively attenuate the negative effects of cocaine, as indicated by a dose-dependent decrease in the frequency of approach-avoidance "retreat" behaviors. This effect was independent of whether the manipulation occurred before (10 mins prior) or after (5 mins post-infusion) cocaine self-administration.

We therefore conclude that 5-HT signaling within the BNST likely contributes to the negative/anxiogenic effects of cocaine. This work supported by NIDA grant DA-033370 awarded to AE.