PRELIMBIC-HIPPOCAMPAL CIRCUITRY DIFFERENTIATES MATCH FROM NON-MATCH ON A CONTINUOUS MATCH-TO-SAMPLE TASK

Goal-directed behavior requires context-appropriate decisions based on content in working memory, often proposed to be supported by prefrontal-hippocampal circuitry. However, how this is implemented by that circuitry remains an ongoing line of inquiry. Using electrophysiology, we recorded 868 prelimbic medial prefrontal cortex (mPFC) and 243 CA1 hippocampus (HPC) neurons, from 4 rats performing 51 sessions of a continuous match-to-sample (CMTS) task. Our prior work (Peters & Smith, 2020, Behav Neurosci) already demonstrated CMTS performance is disrupted by mPFC inhibition. On CMTS trials, rats were presented with a cup containing one of 12 odors mixed into a digging medium. They learned to dig for a buried reward when the odor on the current trial was the same as that on the previous trial (Match), but to refrain from digging on unrewarded trials where the current odor was different (Non-Match). Each trial, rats approached the cup, investigated the odor, and either dug for a reward or turned away to initiate the next trial. Not digging on Match or digging on Non-Match counted as an error. Rats performed well (82% correct), with errors almost exclusively made on Non-Match trials. The mPFC clearly differentiated Match from Non-Match, and showed cup-approach and at-the-cup signals. On Match trials exclusively, mPFC showed end-of-trial firing. Local field potential data reveal, on Match trials, beta-range power changes in mPFC during odor sampling, and Granger causality from mPFC to HPC at the end of a trial. In conclusion, our data elucidate several aspects of how mPFC-HPC circuitry underpin goal-directed behavior.