Figure 5. N-methyl-d-aspartate (NMDA) receptor knockout in V1 before first visuomotor experience impaired learning of a visuomotor task later in life.
(A) Experimental approach and timeline. Three groups of mice were trained: the first was composed of 6 ΔGrin1juv dark-reared mice, the second was composed of 13 ΔGrin1adult normally reared mice, and the third was composed of 6 C57BL/6 dark-reared control mice. Mice were water-restricted and subsequently trained to perform a virtual navigation task (see Materials and methods). (B) Left: schematic of virtual reality setup. Mice controlled forward translational motion and rotation in a virtual corridor by rotating a spherical treadmill and were trained to navigate to the end of a corridor for a water reward. As performance increased, the task difficulty was increased by lengthening the virtual corridor. Right: top-down view of the virtual corridor showing the trajectories of the mouse in three example trials (different gray levels) on days 1 (top) and 7 (bottom). The ratio of virtual corridor length to width is not drawn to scale. (C) Task performance as a function of training day (see Materials and mthods) of ΔGrin1juv mice (red), ΔGrin1adult mice (dark red), and dark-reared control mice (Controljuv, black) over the course of 7 days. Error bars indicate the standard error of the mean (SEM) across mice. ΔGrin1adult and Controljuv mice exhibited performance improvements over the course of training, while ΔGrin1juv mice did not. Performance on day 7 was different between ΔGrin1juv and both ΔGrin1adult and Controljuv mice. Here and in subsequent panels, n.s.: p>0.05, *p<0.05, **p<0.01, ***p<0.001. For all details of statistical testing, see Supplementary file 1A. (D) Turning in response to a perturbation that consisted of a sudden heading displacement of 30° to the left (yellow) or to the right (purple) of ΔGrin1juv, ΔGrin1adult, and Controljuv mice, early (top row) and late (bottom row) in training. Shading indicates SEM across trials. Gray shading indicates analysis window (+1 s to +3 s) used for quantification in (E). (E) Quantification of perturbation offset responses shown in (D) as the difference between average left and right perturbation turning responses, late (bottom row in D) minus early (top row in D) in training. Boxes show median and quartiles, all data are shown as dots (individual mice) to the right. ΔGrin1adult and Controljuv mice learned to initiate corrective turns in response to visual offset perturbations, while ΔGrin1juv mice did not.