New behavioral protocols to extend our knowledge of rodent object recognition memory

Abstract

Animals often show an innate preference for novelty. This preference facilitates spontaneous exploration tasks of novelty discrimination (recognition memory). In response to limitations with standard spontaneous object recognition procedures for rodents, a new task ("bow-tie maze") was devised. This task combines features of delayed nonmatching-to-sample with spontaneous exploration. The present study explored aspects of object recognition in the bow-tie maze not amenable to standard procedures. Two rat strains (Lister Hooded, Dark Agouti) displayed very reliable object recognition in both the light and dark, with the Lister Hooded strain showing superior performance (Experiment 1). These findings reveal the potential contribution of tactile and odor cues in object recognition. As the bow-tie maze task permits multiple trials within a session, it was possible to derive forgetting curves both within-session and between-sessions (Experiment 1). In Experiment 2, rats with hippocampal or fornix lesions performed at normal levels on the basic version of the recognition task, contrasting with the marked deficits previously seen after perirhinal cortex lesions. Next, the training protocol was adapted (Experiment 3), and this modified version was used successfully with mice (Experiment 4). The overall findings demonstrate the efficacy of this new behavioral task and advance our understanding of object recognition.

Figure 1.

(A) Schematic of the bow-tie maze. A sliding door separates the two ends of the maze in which two objects are placed. (B) General procedure showing the presentation order of the objects in the standard object recognition task. All objects are rewarded (+). (Arrow) Rat movements. (Black print) Novel objects, (gray print) familiar objects.

Figure 2.

Experiment 1, Stage 1, object recognition in light and dark (<1 min retention interval). Graphs depict mean performance in the light (top), dark (middle), and light (bottom): (left) cumulative D1 score; (middle) updated D2 ratio; (right) cumulative total exploration for all objects. (Black symbols) Dark Agouti rats (DA), (white symbols) Lister Hooded rats (LH). For D1 and D2 scores, a score of zero reflects a failure to discriminate novel from familiar (chance). (Vertical bars) Standard error of the mean (although when small, they are obscured by the symbols). Group differences: (**) P < 0.01, (***) P < 0.001.

Figure 3.

Experiment 1, Stage 2: Object recognition with Short (top; 5-, 13-, 21-min), Mid (middle; 3-h), and Long (bottom; 24-h) retention delays. (Left) D1 score; (middle) D2 ratio; (right) total exploration for all objects. For the Short Delay, each mean D1 and D2 score is taken from a block of four trials. Retention intervals increased by 2 min for every trial after Trial 12. The data were, therefore, blocked into three groups of four trials (Trials 13–16, 17–20, 21–24), with respective mean retention intervals of 5, 13, and 21 min. For the Mid and Long delays, each mean D1 and D2 score is taken from 10 trials. (Black symbols) Dark Agouti rats (DA), (white symbols) Lister Hooded rats (LH). For D1 and D2 scores, a score of zero reflects a failure to discriminate novel from familiar (chance). (Vertical bars) Standard error of the mean (although when small, they are obscured by the symbols).

Figure 4.

Experiment 1, Stage 3: Improving levels of object recognition after a 3-h retention interval. The bar charts depict the mean performance of four trials for the “repeated” objects (presented six times) and four trials for the “single” objects (presented only once) in the delay phase. (Left) D1 score; (middle) D2 ratio; (right) total exploration for all objects. (Black bars) Dark Agouti rats (DA), (white bars) Lister Hooded rats (LH). For D1 and D2 scores, a score of zero reflects a failure to discriminate novel from familiar (chance). (Vertical bars) Standard error of the mean. Group differences: (*) P < 0.05.

Figure 5.

Diagrammatic reconstructions of the hippocampal (upper panel) and fornix (lower panel) lesions showing the cases with the largest (gray) and smallest (black) lesions. The numbers refer to the distance (in millimeters) from bregma (based on Paxinos and Watson 2005).

Figure 6.

Experiment 2: Impact of hippocampal (Hpc, upper) and fornix (Fnx, lower) lesions on object recognition (retention interval <1 min). (Left) Cumulative D1 score, (middle) updated D2 ratio, (right) cumulative total exploration for all objects. (Black symbols) Performance of the lesioned animals, (white symbols) performance of the sham controls. For comparison purpose, the final D1 and D2 scores of rats with perirhinal lesions from Aggleton et al. (2010) have been added (gray symbols). For D1 and D2 scores, a score of zero reflects a failure to discriminate novel from familiar (chance). (Vertical bars) Standard error of the mean (although when small, they are obscured by the symbols). Group differences: (*) P < 0.05.

Figure 7.

Experiments 3 (upper) and 4 (lower). (Upper) Mean performance in Experiment 3 of the “one-well” (white) group and the “two-wells” (black) group. (Lower) Performance of C57Bl/6 mice (Experiment 4). (Left) Cumulative D1 score; (middle) updated D2 ratio; (right) cumulative total exploration for all objects. For D1 and D2 scores, a score of zero reflects a failure to discriminate novel from familiar (chance). (Vertical bars) Standard error of the mean (although when small, they are obscured by the symbols). Group differences: (*) P < 0.05.

Figure 8.

Summary of recognition performance across the delays used in Experiment 1, Stage 2. The graph depicts the mean performance (D2) of the rats, plotted against the retention interval (log). Performance declined with longer intervals. (Black symbols) Dark Agouti rats (DA), (white symbols) Lister Hooded rats (LH). A score of zero reflects a failure to discriminate novel from familiar (chance). (Vertical bars) Standard error of the mean.