Extended Data Fig. 4: Experience-dependent, flexible space- and goal-referenced CA1 PC representation.

a, Schematic of the naming conventions for training experiences and reward switch tests. Two environments, belt A and belt B, are utilized. Lowercase ‘a’ and ‘b’ represent ≥4 days of fixed reward training on belt A or B, respectively. Uppercase ‘A’ and ‘B’ represent a single reward switching session on belt A or B, respectively. For instance, ‘abB’ indicates fixed reward training on belt A for ≥4 days (‘a’), an additional 4-day fixed reward training on belt B (‘b’), and a reward switch test on belt B (‘B’, Familiar, Day 1). ‘aB’ represents fixed reward training on belt A (‘a’) followed by a reward switch test on belt B (‘B’, Novel, Day 1). b, Cumulative PC fractions across PF position changes between Post and Pre for ‘aB’ (red, 425 PCs from 6 mice) versus ‘abB’ (blue, 849 PCs from 4 mice) on Day 1. p = 0.97. For ‘aB’ and ‘abB’, Day 1 is the first day on belt B with a reward switch. For ‘abB’, two out of four mice had prior reward switching experiences. c, Same as a but for groups ‘aAA’ on Day 2 (blue, 368 PCs from 6 mice) versus ‘bBA’ on Day 2 (red, 375 PCs from 3 mice). p = 3.3e–13. d, Same as a but for groups ‘aAB’ on Day 1 (blue, 503 PCs from 5 mice) versus Day 2 (red, 489 PCs from 5 mice). Group ‘aAB’ was trained on belt A (‘a’), followed by a reward switch on belt A (Day 1, Familiar ‘A’) and belt B (Day 2, Novel ‘B’). p = 2.4e–44. e, Same as d but for groups ‘bBA’ on Day 1 (blue, 403 PCs from 3 mice) versus Day 2 (red, 375 PCs from 3 mice). Group ‘bBA’ was initially trained on belt B, followed by a reward switch on belt B (Day 1, Familiar ‘B’) and belt A (Day 2, Novel ‘A’). p = 1.9e–6. Note: Two-sample two-sided Kolmogorov–Smirnov test for panels b–e.

Source data