Self-directed exploration provides a Ncs1-dependent learning bonus

Abstract

Understanding the mechanisms of memory formation is fundamental to establishing optimal educational practices and restoring cognitive function in brain disease. Here, we show for the first time in a non-primate species, that spatial learning receives a special bonus from self-directed exploration. In contrast, when exploration is escape-oriented, or when the full repertoire of exploratory behaviors is reduced, no learning bonus occurs. These findings permitted the first molecular and cellular examinations into the coupling of exploration to learning. We found elevated expression of neuronal calcium sensor 1 (Ncs1) and dopamine type-2 receptors upon self-directed exploration, in concert with increased neuronal activity in the hippocampal dentate gyrus and area CA3, as well as the nucleus accumbens. We probed further into the learning bonus by developing a point mutant mouse (Ncs1(P144S/P144S)) harboring a destabilized NCS-1 protein, and found this line lacked the equivalent self-directed exploration learning bonus. Acute knock-down of Ncs1 in the hippocampus also decoupled exploration from efficient learning. These results are potentially relevant for augmenting learning and memory in health and disease, and provide the basis for further molecular and circuit analyses in this direction.

Figure 1. Exploratory rearing promotes spatial learning.

(A) Diagram showing object recognition testing procedure. (B–E) Relationship between exploratory rearing and spatial memory in the object recognition test. Mice were trained in an arena with four objects and under dim (n = 14) or bright (n = 15) lighting while rearing (B) and horizontal travel distance (C) were recorded. (D) Later, preference towards two displaced objects (DO) over stationary objects (SO) was evaluated. (E) Ability to recognize a novel object (NO) over a familiar object (FO) was evaluated. (F–I) Relationship between exploratory rearing and spatial memory in the object recognition test for mice injected with saline or botulinum toxin A (BoNT) into hindlimb muscle, the latter of which caused focal hindlimb paresis. Rearings (F) and horizontal travel distance (G) were recorded during object habituation and subsequent testing for DO preference (H) and NO preference (I). (J) Plot of rearing data (B,F) against DO memory (D,H) (K) Plot of travel distance data (C,G) against DO memory (D,H). Data are expressed as mean ± SEM. Pearson correlation (r²) and slope (m) were determined. *P < 0.05, **P < 0.01, ***P < 0.001. See also Supplementary Figure S1 and Table S1.

Figure 2. Novel dim environments that enhance rearing induce c-Fos and elevate Ncs1 and Drd2 expression.

(A) Counts of c-Fos-positive cells in brain regions at 90 minutes post-exposure to a novel dim, novel bright, or familiar dim environment. Three sections per mouse, n = 3 per group. (B,C) Comparison of various mRNA transcripts in the hippocampus 30 minutes following exposure to testing environments. (B) Representative images of electrophoresed RT-PCR amplicons for indicated mRNA. (C) Ratio of Ncs1, Drd2, Fos, Creb1, and Actin mRNA levels normalized to Gapdh (n = 3 per group, triplicate). Data are expressed as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001. See also Supplementary Figure S2.

Figure 3. Ncs1^P144S/P144S mice show reduced exploration and impaired spatial memory.

(A) Models of wild-type (left) and P144S (right) NCS-1 protein. (B) Ncs1 mRNA levels between genotypes. (C) Representative immunoblot (upper) of NCS-1 using tubulin as loading control with densitometric quantification (lower). (D–G) Exploratory activity for Ncs1^P144S/P144S mice (n = 12 per group) in the open field (D–F) and New Frontier tests (G) were conducted under dim or bright lighting and the following measures were recorded: rearing (D), ambulation in the center of (E) or over the whole arena (F), and crossing events (G). (H–K) Object recognition testing showing rearings (H) and travelled distance (I) during training phase, and preference for displaced (J) or novel objects (K). (L) Correlation plot for displaced object preference as a function of exploratory rearing. (M) Correlation plot between horizontal movement and displaced object preference. Ncs1^P144S/P144S, n = 16; +/+, n = 15; Data are expressed as mean ± SEM. Pearson correlation (r²) and slope (m) are shown. * p < 0.05, ** p < 0.01. DO; displaced objects, SO; stationary objects, NO; novel object, FO; familiar object. See also Supplementary Figures S3,S4, and Table S1.

Figure 4. Ncs1 knock-down in dentate gyrus (DG) reduces exploration.

(A) Representative immunoblot for hippocampal NCS-1 and tubulin three days after the DG was infused with negative siRNA or siRNA against Ncs1, as compared to untreated control mice on the left with densitometric analysis (normalized to tubulin) showing that Ncs1 siRNA reduced NCS-1 protein level (two representative samples per group). Mice were tested in the open field (B,C) and hole-board test (D,E) under dim lighting and measured for rearing (B,D), distance (C) and holepoke exploration (E). (F–I) Relationship between exploratory rearing and spatial memory in the object recognition test for Ncs1 siRNA- and negative siRNA-treated mice. Reduced rearing (F) and normal travel distance (G) during object habituation and subsequent testing for displaced object preference (DO) over stationary objects (SO) (H). (I) Intact novel object (NO) over former object (FO) preference in all siRNA-treated mice. (J) Plot of exploratory rearing data (F) against displaced object memory (H). (K) Correlation plot between horizontal movement and displaced object preference. Negative siRNA, n = 13; Ncs1 siRNA, n = 11. Data are expressed as mean ± SEM. Pearson correlation (r²) and slope (m) are shown. *p < 0.05, **p < 0.01. See also Supplementary Figure S5 and Table S1.