摘要
Neurodegenerative diseases of the basal ganglia have been associated with impairments in 鈥榠mplicit learning鈥? failing to detect repeated sequences embedded in long series of random stimuli. The present study has introduced a novel 5-step sequence learning task (ESLeT) for rats, in which the animals learn to respond to randomly occurring stimuli in a horizontal spatial array, and to chain 5 sequential stimulus-response choices to gain food reward. The rats exhibit improved performance in both speed and accuracy of responding when the sequence is predictable following the first stimulus in the chain, suggesting a comparable capacity for 鈥榠mplicit learning鈥? Performance was more rapid but less accurate when the rats were pre-treated with the dopamine indirect agonist amphetamine, markedly disrupted in both dependent variables when pre-treated with the cholinergic antagonist, scopolamine, and modestly disrupted following bilateral striatal lesions, with distinctive patterns of error and changes in the precise patterns of stimulus-by-stimulus responding associated with each experimental treatment. The ESLeT task provides a new test with distinctive advantages for the analysis of vigilance, serial responding and procedural motor learning in animal models of human neurodegenerative disease.