The effects of cytotoxic entorhinal lesions and electrolytic medial septal lesions on the acquisition and retention of a spatial working memory task

Rats with lesions either of medial septal nucleus (MSN) or the entorhinal cortex (ECx) were compared postoperatively with unoperated controls in a discrete-trial, delayed matching-to-position (DMTP) task, conducted on an elevated T-maze. A DMTP trial consisted of two consecutive visits to the maze: an information run and a choice run. The animals were first forced to visit a randomly selected choice arm in the information run. In the choice run, the correct response was to match the choice arm that had been visited on the information run, regardless of whether the information run itself had been rewarded or not. MSN animals failed to succeed in this task, performing at close to chance level throughout training. On the other hand, ECx rats consistently perform at a level comparable with that of unoperated controls; both groups attained more than 90% correct after 192 trials. Long-term retention testing was carried out after an intermission of 4 weeks, when the same task was re-administered to the ECx and unoperated control animals. ECx animals showed significantly less saving than controls in the retention test. In contrast, when the retention interval within a DMTP trial was increased by the imposition of a 20-s delay between the information and choice runs, the ECx group was not selectively affected by this manipulation.