The interval between conditional stimulus onset and unconditional stimulus onset, not training-to-test interval, determines patterns of immediate early gene expression in the anterior retrosplenial cortex

Abstract

Prior work has found that the retrosplenial cortex (RSC) is necessary for formation and retrieval of trace, but not delay, fear conditioning. However, more recently, others have demonstrated that activity in the retrosplenial cortex is necessary for retrieval of a remotely-acquired delay fear memory, suggesting that as memory undergoes systems consolidation it becomes more dependent on neural activity in the RSC. Here, we aimed to examine expression of the immediate early gene zif268 in two distinct subregions of the retrosplenial cortex (anterior and posterior) following retrieval of either a recently-acquired or remotely-acquired delay fear memory. We found that while presenting the conditional stimulus either 1 day or 30 days following delay fear conditioning produced strong conditional responding, activity in either the anterior or posterior RSC assessed through expression of the immediate early gene zif268 was not elevated in the remote retrieval group, contrary to our hypothesis. Instead, activity in the pRSC was elevated in the group that received conditioning the day before. In line with some of our prior work, this suggests that animals in that group were showing neural activity in response to placement in a novel context. We then aimed to determine the circumstances under which delay fear retrieval could produce changes in the anterior RSC, which has been associated with conditional stimulus (CS) encoding and retrieval in a trace fear paradigm. We therefore compared delay and trace conditions to a delay conditioning procedure in which the CS and unconditional stimulus (US) onsets were matched to that of the trace procedure. We found that while both the trace and long-cue delay groups showed a similar behavioral pattern, with freezing that gradually extinguished throughout the 10-CS session, freezing in the standard delay group remained high. When examining zif268 activity, we found that while all three groups showed elevated zif268 expression in the pRSC, only the long-cue delay and trace groups showed increased aRSC activity. Interestingly, only the short-cue delay group showed increased zif268 activity in the basolateral amygdala, corresponding with their elevated fear behavior throughout the session. Together, these results suggest that zif268 activity in the RSC following conditioning is related to the interval between CS onset and US onset.