Memory of occasional events in rats: individual episodic memory profiles, flexibility, and neural substrate

Abstract

In search for the mechanisms underlying complex forms of human memory, such as episodic recollection, a primary challenge is to develop adequate animal models amenable to neurobiological investigation. Here, we proposed a novel framework and paradigm that provides means to quantitatively evaluate the ability of rats to form and recollect a combined knowledge of what happened, where it happened, and when or in which context it happened (referred to as episodic-like memory) after a few specific episodes in situations as close as possible to a paradigm we recently developed to study episodic memory in humans. In this task, rats have to remember two odor-drink associations (what happened) encountered in distinct locations (where it happened) within two different multisensory enriched environments (in which context/occasion it happened), each characterized by a particular combination of odors and places. By analyzing licking behavior on each drinking port, we characterized quantitatively individual recollection profiles and showed that rats are able to incidentally form and recollect an accurate, long-term integrated episodic-like memory that can last ≥ 24 d after limited exposure to the episodes. Placing rats in a contextually challenging recollection situation at recall reveals the ability for flexible use of episodic memory as described in humans. We further report that reversible inactivation of the dorsal hippocampus during recall disrupts the animal's capacity to recollect the complete episodic memory. Cellular imaging of c-Fos and Zif268 brain activation reveals that episodic memory recollection recruits a specific, distributed network of hippocampal-prefrontal cortex structures that correlates with the accuracy of the integrated recollection performance.

Keywords: episodic-like memory; hippocampus; olfactory memory; prefrontal cortex; recollection; rodent.

Figure 1.

The Episodicage: an experimental device to study episodic-like memory in rats. A, B, The Episodicage is a rectangular chamber with four odor ports (OP), each associated with a drinking pipette (P). Detection of rat's nose poke triggers odor delivery (OI, odor injection; OE, odor extraction) and introduction/withdrawal of the pipette by a motor (M). Rat behavior is tracked by four cameras above the odor ports and one central camera above the arena. The appearance of Episodicage could be modified by placing floors with different tactile characteristics, introducing various objects, using a video projector to project visual patterns on different versions of the floor, and diffusing environmental sounds via loudspeakers. C, A nose poke initiates the trial and the onset of the odor. After a delay, the pipette protrudes into the cage for a fixed amount of time. During this period, each lick to the pipette is detected and triggers a pump for delivery of drinking solutions during odor stimulation. The offset of the odor corresponds to the pipette withdrawal. After a given intertrial interval, the rat can activate the system again by a nose poke and a new trial is initiated.

Figure 2.

Time line of the episodic-like memory paradigm. A, First, 11 daily sessions of shaping were used to train the animal to the introduction of the pipette by making a nose poke into any of the four odor ports and drink after a delay (apparatus, 7 d) and to receive odor stimulation associated with sugar or quinine solution from the pipette (odors, 4 d). All trials were systematically initiated by the rats during a 20 min/d session. Rats then underwent three daily routine sessions (Routine, lower box, middle configuration) with no odor, no enriched context, and only water available from the four ports. Episode exposures consisted of two distinct episodes [lower box: left, Episode 1 (E1) configuration; right, Episode 2 (E2) configuration] with a 1 d routine (R) session in between. Episodes were presented once or twice (as illustrated here) in independent groups of rats. They were each characterized by a unique combination of odor–place–context associations rewarded with sugar solution (green labels; odor A at port 2 for E1; odor C at port 4 for E2) while the three other incorrect odor–place associations were associated with quinine (red crosses). During this episodic session, in a given context (“in which context”), rats encoded at which port location (“where”), one of the odors (“what”) was associated with sugar solution. B, During the retrieval test 24 h or 24 d after the last episode session, rats were placed again in the E2 context to evaluate what type of information they were able to recollect (P+O+: correct place, correct odor; P+O−: correct place, incorrect odor; P−O+: incorrect place, correct odor; P−O−: incorrect place, incorrect odor). The two-port test completely matched the episode, except that only water was delivered whatever place–odor configuration was experienced. The challenging four-port test was a more complex situation since it took place in the enriched context E2, but with four accessible ports [2 previously associated with context E2 (IC) and 2 with context E1 (OC)], each port associated with the pair of odors corresponding to its respective episode. For each port visit, the number of licks was expressed for each rat as a licks index and was used for analyzing encoding and recollection performance.

Figure 3.

Episodic-like memory performance after one or two presentations of the episodes. A–I, One (A–C) and two (D–I) presentations of each episode. Licks index for each odor–place–context association (P+O+: correct odor, correct place; P+O−: correct port, incorrect odor; P−O+: incorrect port, correct odor; P−O−: incorrect odor, incorrect port) during encoding of E1 and E2 (A, D) and during E2 retrieval at test 24 h (B, E) or 24 d later (G). C, F, I, Individual licks index values during E2 retrieval test (C, n = 7; F, n = 6; I, n = 13). Licks index values for each configuration are color coded for each individual rat. For example, the rat corresponding to the orange circle is classified as able to recall what–where information from the right context, whereas the light blue one is able to recall where information. Group data are expressed as means ± SEM in this and subsequent figures. Median is reported as horizontal bar. *p < 0.05; **p < 0.01; ***p < 0.005, Friedman test followed by Wilcoxon test.

Figure 4.

Episodic-like memory performance in E2 context during the challenging four-port test. A, Individual licks index for each configuration IC or OC (n = 10). B, Distribution of individual behavioral profiles during recollection of the episodic association. Histograms represent What-Where (WW), Where, What profiles IC or OC E2 (n = 10). C, Licks index of rats expressing a What-Where-IC profile (n = 5 of 10). The results of the P+O+/IC configuration was significantly different from all the other configurations, *p < 0.05, Friedman test followed by Wilcoxon test.

Figure 5.

Episodic-like memory performance during transient pharmacological inactivation of the dorsal hippocampus. A, Licks index for each odor–place–context configuration during encoding of the second session of E2 [E2(2)] and during E2 retrieval test (TE2) in control rats (CONT; n = 10) and rats injected with muscimol (MU; n = 7) before the retrieval test. B, C, Individual licks index values of CONT (B) and MU (C) rats during E2 retrieval test. Medians for each configuration are reported on the graphs. *p < 0.05, **p < 0.01. ns, Nonsignificant. Friedman test followed by Wilcoxon test or Mann–Whitney test for comparisons of independent groups.

Figure 6.

Recruitment of selective brain areas during recollection of the episodic association. A, Schematic drawings of rat brain coronal sections showing the regions of interest (filled areas) selected for counting of IEG-positive nuclei. Distance from bregma is indicated for each coronal section. B–M, c-Fos (B–G) and Zif268 (H–M) cell densities in dorsal (B, H) and ventral (C, I) hippocampus; medial (MO), ventral (VO), lateral (LO), and dorsolateral (DLO) parts of orbitofrontal (OFC) cortex (E, K); and PFC, aCC, mCC (F, L). Histograms represent c-Fos and Zif268 cell counts in control (white bars, ROUTINE; n = 6) and episode-experienced rats (black bars, EPISODIC; n = 6). *p < 0.05, **p < 0.01, ***p < 0.005, Student's t test. D, G, J, M, Photomicrographs showing increased c-Fos and Zif268 expression in episodic rats compared with control rats in dorsal hippocampus (D, J) and in aCC (G, M). Scale bars: G, 250 μm; D, J, M, 500 μm.

Figure 7.

Episodic-like memory performance correlates with the recruitment of prefrontal and hippocampal areas. A, B, D, Pearson's r correlation between licks index and c-Fos expression in aCC (A) and dDG (B) and between licks index and Zif268 expression in OFC (D). C, Pearson's cross-correlation between c-Fos expression in aCC and dDG and licks index. E, Pearson's cross-correlation between c-Fos expression in aCC and Zif268 expression in OFC and licks index. Control (ROUTINE) rats are represented with circles and episodic rats (EPISODIC) by squares. Right color scales in C and E indicate licks index level during the E2 recall test for each individual data point. Black line, Regression line for episodic rats.