The rodent retrosplenial cortex (RSC) has been widely implicated in episodic memory tasks, including spatial learning tasks, like the Morris water maze (e.g., Vann et al., 2003) and incidental learning like object recognition (de Landeta et al., 2020), and tasks that incorporate both of these types of learning, like object location memory (Brunswick et al., 2023). More recently, the RSC has been studied for its role in associative learning with several studies indicating its direct involvement in both trace (Kwapis et al., 2017, Kwapis et al., 2015, Trask et al., 2021b) and context (Corcoran et al., 2018, Trask and Helmstetter, 2022) fear conditioning, as well as more complex forms of learning like sensory preconditioning (Fournier et al., 2020, Robinson et al., 2014; for review, see Fournier et al., 2021).

Despite its role in complex forms of associative learning, some work suggests that the RSC is not involved in a simpler version of associative learning known as delay fear conditioning (Kwapis et al., 2017, Trask et al., 2021a). In delay fear conditioning, a brief conditional stimulus (CS) like a tone or a light co-terminates with a footshock (unconditional stimulus, or US). This procedure shares many similarities with the RSC-dependent trace fear conditioning, in which a brief CS is paired with a footshock, but the two are separated by a period of time known as a trace interval. Interestingly, the retrosplenial cortex has been shown to be important for retrieval of a remotely-acquired delay fear memory (Fournier et al., 2020), when memory is typically less precise (as reviewed in Trask et al., 2021b).

However, another large discrepancy between trace and delay conditioning is that the time between the onset of the CS and the onset of the US in trace conditioning is typically longer than during a delay procedure, suggesting that the RSC might become important in encoding of conditional fear memory when the CS-US relationship is less precisely-timed and more ambiguous. One critical difference, therefore, between delay and trace conditioning is that the CS is a more precise predictor of the timing of US delivery in delay conditioning. In line with a potential role in less precisely-timed associative learning, the RSC has also been observed to be important in contextual fear conditioning (e.g., Trask and Helmstetter, 2022), where the context or conditioning chamber itself (and not a discrete stimulus, like a tone or a light) serves as the CS to signal footshocks.

Our recent work has demonstrated that the contributions of the retrosplenial cortex are likely not uniform along its rostrocaudal axis, with differential roles for the anterior (−1.6 to −3.6 mm relative to bregma) and posterior (−5.6 to −7.6 mm relative to bregma) subregions. While the anterior RSC (aRSC) is needed for encoding of cue- or event-driven information in associative learning (Trask et al., 2021c, Trask and Helmstetter, 2022), the posterior region (pRSC) is necessary for encoding (Trask et al., 2021c, Trask and Helmstetter, 2022) and retrieval (Trask & Helmstetter, 2022) of the contextual aspects of fear memory. This is in line with work demonstrating that the anterior RSC was necessary for object recognition memory (de Landeta et al., 2020). However, both regions are needed for the retrieval of a trace, but not delay, fear memory (Trask et al., 2021a) suggesting that a complete memory representation is needed for accurate recall.

Neural activity can be assessed using expression of the immediate early gene zif268. In our hands, zif268 expression tracks with general levels of neural activity and its expression is reduced by optogenetic inhibition in the retrosplenial cortex (Trask et al., 2021a, Trask et al., 2021b). In related work from our lab (Bonanno et al., 2023, Met Hoxha et al., 2024, Robinson et al., 2024) and others (Hoffman et al., 2015), zif268 expression in the basolateral amygdala is elevated in animals that show increased fear behavior. As such, the present experiments examined expression of zif268 as a proxy for neural activity.

Here, we predicted that zif268 expression would be elevated following a remote, but not recent, retrieval of a delay fear memory in the anterior retrosplenial cortex (Experiment 1) based on work demonstrating that reducing neural activity in the RSC impairs memory retrieval of a remotely-acquired delay fear memory and work demonstrating a specific role for the aRSC in CS-related memory. Additionally, we aimed to test if a delay fear memory could produce changes in RSC activity if the CS-US interval matches CS onset and US onset to that of a similar trace procedure (Experiment 2).