Michigan State University - Research Groups

Coordination among neurons is believed to be a key feature in the brain’s ability to encode cognitive processes that involve multisensory integration that guides motor action. Neuroimaging and electrophysiological studies have shown that neurons in Prefrontal Cortex modulate their firing pattern during working memory tasks. The coordination mechanism among local neural circuits that accompany such modulation, however, remains poorly understood.
We designed an experimental task to decipher these mechanisms. Rats were trained on a working memory task that involves odor-tone matching, requiring them to associate an odor (vanilla/cherry) with a tone (low/high pitch) and make a Go/No-Go decision based on cue integration. The task required rats to discriminate between four distinct patterns of stimuli while poking their nose in a nosepoke hole and waiting for a fixed delay period. Each trial lasted for 5 seconds, comprising five epochs: pre-odor, odor presentation, delay period, tone presentation and action period. Rats were rewarded for pressing a right lever when integrating X with Y, pressing a left lever when integrating Z with W, and making no lever press when integrating either X with Z or Y with W.
On average, after 5-6 weeks of training (1 hour/ day, 5 days/week), rats performed the task at high proficiency (72% correct for the right go trials, 70% for the left go trials and 89% for no go trials). Rats were then chronically implanted with 32 channel silicon probes in their prelimbic area (PrL) and cingulate cortex area (Cg1) of mPFC. Spike sorting was performed using using custom built Matlab codes. PSTHs of each recorded unit was calculated using 100 msec bins, averaged across all successful trials. We performed three pairwise comparisons: trials with same odor, trials with same tone and trials with same outcome (Go/No Go). We found significant differences in response patterns near the end of the delay period (p<7*10^-5) and lasting until the middle of tone presentation epoch (p<0.002). In particular, changes in firing rates were significant in fixed odor trials, during tone epochs, while significant changes were observed during the delay period in fixed tone trials. Analysis of fixed outcome trials revealed a significant difference in firing rates throughout the delay period and tone presentation periods as well as during the action period in No Go trials. These findings suggest a multisensory integration role of mPFC units to maintain a persistent goal representation.