Rationale: Cannabinoids such as delta(9)-tetrahydrocannabinol (delta(9)-THC) or WIN-55,212-2 (WIN-2) have psychoactive effects on cognition. As a result, the reinforcing properties of delta(9)-THC or WIN-2 may confound learning and memory tests with false negative results. It therefore seems advisable to assess the reinforcing properties of the drugs in the same behavioural model used for learning experiments.

Objective: We therefore developed conditioned place preference protocols in the open-field water maze and tested both delta(9)-THC (2 mg/kg) and WIN-2 (1 mg/kg and 3 mg/kg). Given that previous reports on cannabinoids have revealed conflicting data and that this was a novel behavioural test, we also tested the benzodiazepine receptor agonist diazepam (2.5 mg/kg). Some methodical refinements were appropriate in order to determine the behavioural strategy implemented by the animals.

Methods: All animals were injected intraperitoneally 30 min prior to training/testing. In experiment 1, male hooded Lister rats injected with drug were repeatedly placed on the drug-related platform and subsequently tested for place preference. In experiment 2, rats were trained to swim to the drug platform on drug days and to the vehicle platform on vehicle days. A series of probe trials was introduced to delineate what had been learned. Experiment 3 studied the effect of WIN-2 on spatial learning in the water maze.

Results: Neither WIN-2 nor delta(9)-THC induced place preference in the water maze. When trained in the swim procedure, however, WIN-2 was neutral, but Delta(9)-THC resulted in place aversion. Conversely, diazepam consistently produced place preference in both procedures. WIN-2 (3 mg/kg), however, produced a small learning deficit in the spatial water maze task.

Conclusion: It appears that the reinforcing properties of delta(9)-THC and WIN-2 in the doses used here are different, despite them both being agonists at cannabinoid receptors within the central nervous system. The fact that delta(9)-THC may be aversively related to a particular context has implications for previous work reporting deficits in spatial learning.