Hippocampal-prefrontal dynamics in spatial working memory: interactions and independent parallel processing

Abstract

Memory processes may be independent, compete, operate in parallel, or interact. In accordance with this view, behavioral studies suggest that the hippocampus (HPC) and prefrontal cortex (PFC) may act as an integrated circuit during performance of tasks that require working memory over longer delays, whereas during short delays the HPC and PFC may operate in parallel or have completely dissociable functions. In the present investigation we tested rats in a spatial delayed non-match to sample working memory task using short and long time delays to evaluate the hypothesis that intermediate CA1 region of the HPC (iCA1) and medial PFC (mPFC) interact and operate in parallel under different temporal working memory constraints. In order to assess the functional role of these structures, we used an inactivation strategy in which each subject received bilateral chronic cannula implantation of the iCA1 and mPFC, allowing us to perform bilateral, contralateral, ipsilateral, and combined bilateral inactivation of structures and structure pairs within each subject. This novel approach allowed us to test for circuit-level systems interactions, as well as independent parallel processing, while we simultaneously parametrically manipulated the temporal dimension of the task. The current results suggest that, at longer delays, iCA1 and mPFC interact to coordinate retrospective and prospective memory processes in anticipation of obtaining a remote goal, whereas at short delays either structure may independently represent spatial information sufficient to successfully complete the task.

Fig. 1

Shows 8-arm maze and testing procedure for the delay non-match to sample task using two different time delays, 10 s and 5 min.

Fig. 2

Shows study design divided into pre-training, surgery and post-training periods. During pre-training subjects were trained at the 10 s and 5 min delay until they reached a criterion of 100% correct performance for both delays. Surgery followed pre-training criterion performance, with bilateral cannula implantation of both iCA1 and mPFC. Subjects were then retrained to a criterion of at least 100% correct performance for both delays, at which time saline was infused bilaterally into either iCA1 or mPFC. After reaching at least 75% correct performance for both delays under the saline condition for 2 consecutive days, the following day subjects received bilateral infusions of muscimol into either iCA1 or mPFC prior to testing. Following reacquisition of criterion performance with either ipsilateral or contralateral saline infusions into iCA1 and mPFC, subjects received ipsilateral or contralateral muscimol infusions into iCA1 or mPFC. Lastly, subjects were retrained to a criterion of at least 75% correct performance at both delays while receiving bilateral infusions of saline into both iCA1 and mPFC iCA1. After reaching criterion performance under the saline condition, bilateral infusions of iCA1 and mPFC were performed. iCA1 = hippocampus intermediate CA1; mPFC = medial prefrontal cortex.

Fig. 3

Shows dorsal view of the rat brain with patterns of inactivation used to test for interactions and independent parallel processing indicated by black dots. Stage 1-top left = bilateral mPFC; top right = bilateral iCA1; Stage 2 – middle left = contralateral iCA1and mPFC; middle right = ipsilateral iCA1 and mPFC; Stage 3-bottom middle = bilateral iCA1 and mPFC. iCA1 = hippocampus intermediate CA1; mPFC = medial prefrontal cortex.

Fig. 4

(A) Shows mean percent correct performance with standard error at the 5 min delay for subjects receiving bilateral infusions of saline (S) or muscimol (M) into iCA1 or mPFC (B) Shows mean percent correct performance with standard error at the 5 min delay for subjects receiving ipsilateral or contralateral infusions of saline (S) or muscimol (M) into iCA1 and mPFC (C) Shows mean percent correct performance with standard error at the 5 min delay for subjects receiving bilateral infusions of saline (S) or muscimol (M) into both iCA1 and mPFC.

Fig. 5

(A) Shows mean percent correct performance with standard error at the 10 s delay for subjects receiving bilateral infusions of saline (S) or muscimol (M) into iCA1 or mPFC. (B) Shows mean percent correct performance with standard error at the 10 s delay for subjects receiving ipsilateral or contralateral infusions of saline (S) or muscimol (M) into iCA1 and mPFC. (C) Shows mean percent correct performance with standard error at the 10 s delay for subjects receiving bilateral infusions of saline (S) or muscimol (M) into both iCA1 and mPFC.

Fig. 6

Histological plates showing locations of bilateral cannula tips for implantation in both iCA1 and mPFC for all subjects (N = 9). Some points overlap. Histological plates were adapted from [59].