"Development of the rat P300 ERP during backward chaining"

Abstract

Several experiments in our laboratory shown P300 amplitude to be an incremental function of proximity to conditioned reinforcers in simple behavioral chains. In this paradigm, one tone predicts lever insertion while a second tone signals the click of the food magazine. Since the P300 is consistently strongest to seeing the food pellet, and becomes progressively weaker to the click, pellet tone, and lever tones respectively, we concluded that the P300 is a function of the strength of the conditioned reinforcer. The present experiment was designed to further examine this hypothesis systematically by adding a second lever to the chain leading to food reinforcement. Rats were initially trained in a paradigm where lever insertion (LEV 1) was cued by a 4.5 KHz tone with a non-target (NT) 2.5 KHz tone presented on a variable time schedule at an 8:1 non-target to target ratio. Meeting criterion on a VR-6 schedule produced 5.5 KHz followed by the click of the pellet dispenser signaling delivery of the food pellet. After about seven weeks of training, a second lever (LEV 2) cued by a 3.5 KHz tone was introduced preceding LEV 1. LEV 2 responses, on a VR-6 schedule, produced the LEV 1 cue- sequence. This training continued for about 8 additional weeks. Contrary to what we expected, the P300 amplitude to the LEV 2 tone was consistently greater than the P300 to the LEV 1 tone. One explanation for this is that the LEV 1 cue represents redundant information about the sequence leading to food. LEV 2 is the most informative about the opportunity to engage in the entire sequence leading to food, compared to the NT. To test this we are introducing a third lever into the sequence to see if the greatest P300 amplitude shifts from LEV 2 to LEV 3.