Validation of a radial maze test for assessing learning and memory in rats.

Choice behavior of rats in a radially symmetrical 6-arm maze, without food reward, was validated for assessing changes in learning and memory following treatment with 4 psychoactive agents. The test is designed for future use in routine toxicity studies with laboratory rodents. Each radial main arm of the maze leads to a T-shaped choice-point with a blind alley on the left and a long angled alley on the right. Order of choice of the radial main arms served to score within-session working memory, by evaluating relative recency of arm reentries. The choice between blind alley and long-angled alley at the T-intersections provided a measure of between-session reference memory. Maze behavior as an indicator for impairment in learning and memory was validated by testing rats treated with d-amphetamine, chlorpromazine, scopolamine and physostigmine. Based on the above evaluations, working memory was found to be severely impaired by 0.3 and 1.0 mg/kg scopolamine, and reference memory to be improved by 0.02 mg/kg physostigmine and 1.0 mg/kg amphetamine. Locomotor activity, in terms of the total number of arm choices per test session, was altered by all substances as expected from previous reports in the literature. The test appears to be a valid and sensitive method for assessing learning and memory in the rat without the use of food reward, and thus well suited for implementation in routine toxicity studies with rodents.