Retinoic acid modulation of granule cell activity and spatial discrimination in the adult hippocampus

Abstract

Retinoic acid (RA), derived from vitamin A (retinol), plays a crucial role in modulating neuroplasticity within the adult brain. Perturbations in RA signaling have been associated with memory impairments, underscoring the necessity to elucidate RA's influence on neuronal activity, particularly within the hippocampus. In this study, we investigated the cell type and sub-regional distribution of RA-responsive granule cells (GCs) in the mouse hippocampus and delineated their properties. We discovered that RA-responsive GCs tend to exhibit a muted response to environmental novelty, typically remaining inactive. Interestingly, chronic dietary depletion of RA leads to an abnormal increase in GC activation evoked by a novel environment, an effect that is replicated by the localized application of an RA receptor beta (RARβ) antagonist. Furthermore, our study shows that prolonged RA deficiency impairs spatial discrimination-a cognitive function reliant on the hippocampus-with such impairments being reversible with RA replenishment. In summary, our findings significantly contribute to a better understanding of RA's role in regulating adult hippocampal neuroplasticity and cognitive functions.

Keywords: dentate gyrus; granule cells; hippocampal neuroplasticity; retinoic acid; spatial discrimination; vitamin A.

FIGURE 1

Retinoic acid (RA)-responsive subpopulation exists within the dentate GCs. (A) Experimental schematics for β-gal-labeling of RA-responsive cells in the hippocampus. (B) Representative images for the spatial distribution of β-gal-positive cells in the dorsoventral axis of the hippocampus. Scale bar, 200 μm. (C) Colocalization of β-gal expression with calbindin1, a GC marker, in the dorsal and ventral DG. Scale bar, 100 μm. (D) High-resolution images of β-gal-positive cell patterns in the dorsal and the ventral hippocampus. Scale bar, 50 μm. (E) Quantitative comparison of β-gal-positive GCs between the dorsal and the ventral hippocampus. (dGCL: n = 4, vGCL: n = 5 sections). (F) High-resolution images of β-gal-positive cell patterns on the suprablade and infrablade of the dorsal DG. Scale bar, 100 μm. (G) Quantitative comparison of β-gal-positive cells on the suprablade and the infrablade (Supra: n = 10, Infra: n = 10 sections). *p < 0.05.

FIGURE 2

Retinoic acid (RA)-responsive GCs are non-reactive to NE stimuli. (A) Experimental schematics for NE-induced GC activation in RARE-LacZ TG mice. (B) Representative images of c-Fos expression in RARE-LacZ mice, stained with c-Fos (green), β-gal (red), and DRAQ5 (blue). Scale bar, 200 μm. Quantitative comparison of c-Fos-positive GCs induced by (C) HC and NE (HC: n = 4, NE: n = 4 mice) and (D) on suprablade and infrablade in NE (Supra: n = 4, Infra: n = 4 mice). Quantitative comparison of β-gal-positive GCs induced by (E) HC and NE (HC: n = 4, NE: n = 4 mice), and (F) in suprablade and infrablade (Supra: n = 4, Infra: n = 4 mice). (G) High-resolution images of c-Fos, β-gal, and DRAQ5 immunolabeling results. Scale bar, 20 μm. (H) Colocalization assay between c-Fos- and β-gal-positive cells within GCs. *p < 0.05, **p < 0.01.

FIGURE 3

Inhibition of RA signals leads to an abnormal increase in GC activation evoked by NE stimuli. (A) Schematics to examine the effect of dietary RA depletion on NE-induced GC activation. (B) Representative images of c-Fos immunoreactivity with DAPI nuclear staining. Scale bar, 200 μm. (C) High-resolution images of the c-Fos-positive cell patterns within GCs. Scale bar, 50 μm. (D) Quantitative comparisons of c-Fos-positive cells between control groups and VAD groups (Control-HC: n = 5, Control-NE: 3, VAD-HC: n = 3, VAD-NE: n = 3 mice). (E) The representative traces of vehicle (VEH, black) and RARβ antagonist (LE135, red) in response to depolarizing step current injection. (F) The firing frequency as a function of injected current amplitude. Black and red shading represents the SEM for each dataset over current. (G) Resting membrane potentials (RMP) and input resistance (Rin) are comparable between groups (VEH: n = 13, LE135: n = 12 cells). *p < 0.05, **p < 0.01, ****p < 0.0001.

FIGURE 4

Dietary depletion of RA reversibly impairs spatial discrimination in the Intellicage™ paradigm. (A) Schematics for the effect of dietary RA depletion on behavioral phenotypes. (B) A behavioral paradigm of spatial discrimination test in the Intellicage™ device. (C) Sucrose-corner (corner 1) preference from 1 day before the place learning to the first day of place learning state (above), and the total success rate (below). Dotted line: preference 25%. (D) Sucrose-corner (corner 3) preference of reversal learning state (above) and the total success rate (below). Dotted line: preference 25%. (E) Sucrose-removed-corner (corner 1) preference of learning state. Dotted line: preference 25%. (F) Total visits to sucrose-corner (corner 1) or neutral corners in spatial learning. [(A–F) Control: n = 10, VAD: n = 10]. (G) Sucrose-corner preference (corner 2) of place learning state after VA replenishment (above) and the total success rate (below). Dotted line: preference 25%. (H) Sucrose-corner preference (corner 4) of reversal learning state. Dotted line: preference 25%. (I) Sucrose-removed-corner (corner 2) preference of learning state. Dotted line: preference 25%. (J) Total visits to sucrose-corner (corner 2) or neutral corners in spatial learning after VA replenishment. [(G–J) Control: n = 4, VAD: n = 4]. *p < 0.05, ***p < 0.001, ****p < 0.0001.