MONKEYS CAN ASSOCIATE VISUAL-STIMULI WITH REWARD DELAYED BY 1 S EVEN AFTER PERIRHINAL CORTEX ABLATION, UNCINATE FASCICLE SECTION OR AMYGDALECTOMY

In the present experiment monkeys learned concurrent associations of t wo-dimensional objects (presented on a computer screen) with delayed r eward. Hypothetical mechanisms of associative memory, such as long-ter m potentiation (LTP), require coincidental activation of two populatio n of neurons: one representing the object and the other signalling the reward. In monkeys neurons in area TE of temporal cortex show object- specific activity during object presentation but only fraction of thos e neurons remain active after stimulus offset. In a delayed reward con dition the majority of object-specific neurons in TE cease firing befo re reward is given and can be detected. In the present study the rate of learning with 1000 ms delay of reward was no slower than learning w ith immediate reward. This indicates that information about the object is somehow retained across the delay, possibly somewhere outside TE. In the present study we tested that assumption. Area TE projects to th e perirhinal cortex and, via uncinate fascicle, to the prefrontal cort ex. In our hands, ablations of perirhinal cortex or disconnection of p refrontal cortex from TE (by transection of uncinate fascicle) did not impair learning with delayed reward. Ablation of amygdala, a structur e involved in reward-learning, slowed down learning equally with and w ithout delay. We conclude that retaining information about the visuall y perceived objects across a delay does not exclusively depend upon in tegrity of perirhinal cortex, or uncinate fascicle, or amygdala. Paral lel involvment of those structures remains a possibility and establish ment of the role of residual activity of TE neurons requires further n europhysiological investigation. (C) 1997 Elsevier Science B.V.