Impaired acquisition in a 14-unit T-maze following medial septal lesions in rats is correlated with lesion size and hippocampal acetylcholinesterase staining

Abstract Septohippocampal cholinergic system involvement in acquisition of an aversively motivated 14-unit T-maze was evaluated in 4-month-old male Fischer-344 rats. Each rat was assigned to one of two groups that received either a bilateral electrolytic lesion to the medial septal area (MSA) or a sham operation. One week after surgery, each rat began pretraining in one-way active avoidance (footshock = 0.8 mA) consisting of 10 trials per day on each of 3 consecutive days. Criterion for successful completion of pretraining was 8/10 avoidances on the third day. On the day following completion of pretraining, each rat received 10 trials in a shock-motivated 14-unit T-maze. The performance requirement was to move through each of five maze segments within 10 s to avoid footshock (0.8 mA). A second 10-trial session was provided 24 h later. Performance measures included errors, alternation errors, runtime, shock frequency, and duration. Following maze training, each rat was sacrificed, and formalin-fixed brains were frozen for histology, which included procedures for thionin Nissl and acetylcholinesterase (AChE) staining. MSA-lesioned rats were observed to be significantly impaired on all measures of maze performance compared to sham-operated controls. Densitometric analysis of hippocampal AChE staining revealed a 30% reduction in relative AChE staining of MSA-lesioned rats compared to sham-operated controls. Lesion size was observed to be highly positively correlated with maze errors. A negative correlation of mean error score with density of AChE staining was observed for MSA-lesioned rats, but not for sham-operated rats. These results are consistent with previous studies implicating the cholinergic system in the acquisition of this task and further support the use of this maze for the evaluation of age-related cholinergic dysfunction and cognitive deficits in rodent models.