The impairment of retention induced by insulin in mice may be mediated by a reduction in central cholinergic activity.

Immediate posttraining intraperitoneal injection of nonconvulsive doses of insulin (2-20 IU/kg) significantly impaired retention of male Swiss mice tested 24 h after training in a one-trial step-through inhibitory avoidance task. The dose-response curve showed a U-shaped form. However, of the doses tested, only 8 IU/kg was effective. Insulin did not affect response latencies in mice not given the footshock on the training trial, indicating that the actions of insulin on retention performance were not due to nonspecific proactive effects on response latencies. The impairing effects of insulin (8 IU/kg) on retention were time-dependent, which suggests that insulin impaired memory storage. The simultaneous administration of glucose (10-1000 mg/kg) antagonized, in a dose-related manner, the actions of insulin (8 IU/kg) on retention, suggesting that the hormone may have produced a hypoglycemic response leading to a decrease in CNS glucose availability with a subsequent memory impairment. Low subeffective doses of atropine (0.5 mg/kg) or mecamylamine (5 mg/kg), but not methylatropine (0.5 mg/kg) or hexamethonium (5 mg/kg), given immediately after training but 10 min before an ineffective dose of insulin (4 IU/kg), interacted with and impaired retention. The central anticholinesterase physostigmine (35 or 70 micrograms/kg), but not its quaternary analog neostigmine (35 or 70 micrograms/kg), prevented the memory impairment induced by insulin (8 IU/kg). Considered together, these findings are consistent with the view that a decrease in the CNS glucose availability impairs the synthesis and/or release of acetylcholine in brain regions critically involved in memory storage.