Measuring emotional memory in the elevated T-maze using a training-to-criterion procedure

The elevated T-maze, an ethologically based test, has been used to investigate the effects of anxiolytic drugs on memory and the relationships between neural systems involved in such modulation. This test allows the measurement in the same rat of two kinds of aversively motivated behaviors-inhibitory avoidance and one-way escape. The apparatus consists of three arms of equal dimensions, elevated 50 cm from the floor. One arm is enclosed by walls and stands perpendicular to the two open arms. Placing the rat at the end of the enclosed arm and recording the time to withdraw from this arm during three consecutive trials assesses inhibitory avoidance. Soon afterwards, the rat is placed at the end of one of the open arms and the time to leave this arm recorded as escape response. Three days later memory is assessed by reexposing the rats to the maze. One critical question raised by these studies is whether the anterograde amnesia induced by anxiolytic drugs could be due to insufficient learning during training or to amnesia. The present work investigated whether the introduction of a multitrial training-to-criterion procedure could overcome this question. For this purpose, rats were tested as many times as needed to stay in the enclosed arm continuously for 300 s (avoidance learning to criterion). Results from Experiment 1 showed that rats trained to a learning criterion shows significantly better retention performance. Experiment 2 evaluated the effects of pretraining diazepam (DZP) treatment on this training-to-criterion protocol. The results indicate that DZP did not affect acquisition performance but induced a dose-dependent impairment of the inhibitory avoidance in the memory test. One-way escape (latency to enter the enclosed compartment from the open arms) was not affected by DZP. These results rule out the possibility that the impairment of inhibitory avoidance memory in the elevated T-maze could be due to lack of learning during training, and support the hypothesis that the disruptive effects of DZP are on processes involved in long-term storage of information. (C) 1999 Elsevier Science Inc.