Fig. 6: Silencing 4–7-week-old abDGCs impairs novel object preference, increases CA3–CA1 co-firing and reduces population dimensionality.
From: Adult-born dentate granule cells promote hippocampal population sparsity
a,b, cNOR task arena (a) and layout (b). Letters depict object locations (novel objects in blue). Each day, one laser-off alternated with one laser-on test to provide within-day comparison. c, Estimation plot showing the percentage of time spent by mice in the abDGC::GFP and abDGC::ArchT groups with novel objects in laser-off and laser-on tests. Upper plot: each data point represents the mean percentage time spent with the novel object; chance performance is shown by the dashed line. Lower plot: mean difference in novel object exploration time (%) for laser-on relative to laser-off tests. Mice exhibited significantly lower novel object preference during 4–7-week-old abDGC silencing (abDGC::GFP: n = 10 sessions in 2 mice; abDGC::ArchT (4–7 wpi): n = 8 sessions in 4 mice; abDGC::ArchT (9–12 wpi): n = 8 sessions in 2 mice). d, CA3–CA1 ensemble co-firing. Upper panel: each data point represents the correlated activity of one pair of simultaneously recorded CA3–CA1 PCs. Lower plot: median difference between co-firing in laser-on versus laser-off tests (abDGC::GFP: n = 400 CA3–CA1 cell pairs in 2 mice; abDGC::ArchT (4–7 wpi): n = 1,080 CA3–CA1 cell pairs in 3 mice; abDGC::ArchT (9–12 wpi): n = 150 CA3–CA1 cell pairs in 2 mice). e, Dimensionality of PC population theta co-firing vectors between laser-off and laser-on tests (dimensionality score = estimated dimensionality/number of PCs in each vector; abDGC::GFP: n = 10 sessions in 2 mice; abDGC::ArchT (4–7 wpi): n = 6 sessions in 3 mice; abDGC::ArchT (9–12 wpi): n = 8 sessions in 2 mice). Panels c–e show Cumming estimation plots as in Fig. 1f. ***P < 0.001, **P < 0.01.
