A comparison of multiple-unit activity in the medial prefrontal and agranular insular cortices during Pavlovian heart rate conditioning in rabbits.

Multiple-unit activity (MUA) was recorded from chronically implanted electrodes in the medial prefrontal cortex (PFCm) and the agranular insular cortex (Iag) in separate groups of rabbits during habituation training, followed by aversive Pavlovian conditioning and subsequent extinction training. Control animals received explicitly unpaired presentations of the tone conditioned stimulus (CS) and eye-shock unconditioned stimulus (US). Both the cardiac orientating reflex and the conditioned heart rate response (HR CR) consisted of bradycardia, whereas tone-evoked tachycardia was observed in animals that received unpaired stimuli. Short-latency (less than 20-60 ms), tone-evoked increases in PFCm MUA were observed during the initial trials of habituation training, with their magnitude declining predictably across repeated tone-alone presentations. Subsequent CS/US pairings, however, served systematically to reinstate and enhance this CS-evoked MUA, while both non-associative (unpaired CS/US) and extinction (CS alone) training resulted in significant attenuation of such activity. Unconditioned tone-evoked increases in MUA were also observed in the Iag during habituation; however, such unit responses appeared to be more variable than their PFCm counterparts and were of considerably lesser magnitude. Moreover, in striking contrast to the above PFCm findings, conditioning and non-associative training did not differentially affect overall mean evoked MUA in the Iag, although different post-tone patterns of activity were obtained with the two procedures. The contrasting training effects observed in animals with PFCm vs. Iag electrode placements did not appear to be attributable to differences in regional sensitivity to the US, since excitatory patterns of MUA were elicited by unsignalled presentations of eye-shock at most placements within each cortical field. Accordingly, the present findings are consistent with our previous lesion data in suggesting that, although training-induced changes in PFCm neuronal activity may contribute to the initial events in aversive Pavlovian conditioning, an involvement of the Iag in such processes, if any, remains to be demonstrated.