Fig. 2: Predominant goal-referenced spatial coding in a novel environment.
From: Mechanisms of experience-dependent place-cell referencing in hippocampal area CA1
a, Running (left) and licking (right) behavior of an example animal on a novel belt before (pre; black arrow depicts reward at 50 cm) and after (post; red arrow depicts reward at 140 cm) a reward switch (top and middle). Mean velocity (left) and licking probability (right) for all mice (n = 6) (bottom). Shading indicates s.e.m. b, Spatial Δf/f activity across laps during pre (top) and post (middle) from two goal PCs on a familiar belt: one with a PF before and the other after the reward. Vertical dashed white lines indicate reward locations. Peak-normalized mean Δf/f across space in pre and post (bottom). c, PC Δf/f activity on a novel belt. Peak-normalized mean Δf/f of all space (left) and goal PCs (right) in pre and post (top and middle). Maps are sorted by their peak locations in pre. PC numbers across space in pre and post (bottom). d, The fraction of different PC profiles (n = 425 PCs from 6 mice). e, PC fractions across PF shift distance on familiar (black, n = 6 mice) and novel (red, n = 6 mice) belts (P = 1.8 × 10−17, two-sample two-sided Kolmogorov–Smirnov (KS) test). f, Proportion of space and goal PCs across space. The vertical dashed line indicates the reward location. The horizontal dashed line indicates the expected fraction (~0.167, random remapping). g, The Pearson’s correlation matrix between PVs at all locations from pre and post on a novel belt. Black and white lines indicate the diagonal and off-diagonal. h, Same as g, but for PVs from odd and even laps in pre. Pos., position. i, Mean PV correlation along the diagonal and off-diagonal on the novel belt (n = 6 mice for all conditions; pre versus post, −0.028 ± 0.033 versus 0.49 ± 0.052; P = 5.1 × 10−4; pre-odd versus pre-even, 0.93 ± 0.011 versus −0.048 ± 0.048; P = 5.9 × 10−6). Paired two-sample two-sided t-test for both comparisons. Data are presented as mean ± s.e.m.
