Extended Data Fig. 8: Place field remapping and Fos-induction across multiple virtual environments in the absence of a task.
From: Fos ensembles encode and shape stable spatial maps in the hippocampus
a) Three different virtual environments used to examine place field responses in the no-task condition. Mice were familiar with environment A before imaging began. b) Examples of place field remapping across environments. Each heatmap (25 heatmaps in total) shows the spatially binned activity of multiple cells, where each row is a cell. Each cell’s spatially binned activity is min-max normalized for the purposes of visualization. In the first row of five heatmaps, the same cells are shown in the same order in each column, ordered by their preferred position in the first environment A session. The second row is sorted by the second environment A session and so on. Each column in the grid corresponds to a different session. For each row in the grid, only cells with significant place fields on the sorting session and that are present in all sessions are shown. c) Average place field correlation across sessions. For each cell, only session-session pairs where at least one of the sessions has a significant place field are included in the analysis. The number of cells that went into each session-session comparison are shown in (d). Cells are pooled across 5 mice. d) Number of cells that went into each datapoint in (c). e) Schematic of session-session pairs analyzed in (f–j) within environments and across environments. To avoid within environment and across environment pairs being separated by different amounts of time, only consecutive session-session pairs were considered. f–j) Similarity metrics for place maps, place cell identity, and Fos-induction for session-pairs in the same environment or different environments. Same environment vs different environment: p = 7.3 x 10−11, n = 18 (same), 8 (different) pairs (f); p = 1.3 x 10−6, n = 16, 7 (g); p = 0.81, n = 18, 8 (h); p = 0.051, n = 14, 5 (i); p = 0.067, n = 14, 5 (j); two-sample two-tailed t-test. Black squares and error bars denote mean ± s.e.m. k–m) Three example cells tracked across environment switches. For each cell, the spatially binned activity is plotted (black line) along with the 1st and 99th percentiles of the shuffle distribution (gray). Red markers indicate significant place fields. n) Fraction of cells that are Fos-high on session n+1 out of cells that either continue to have no place field, gain, lose, or retain a place field within the same environment (only adjacent sessions are included). No PF vs Gain PF: p = 0.013, n = 16 session pairs. Lose PF vs Retain PF: p = 0.034, n = 16 session pairs. Paired-sample two-tailed t-test. Black squares and error bars denote mean ± s.e.m. o) Fraction of cells that are Fos-high on session n+1 out of cells that either continue to have no place field, gain, lose, or retain a place field across different environments. No PF vs Gain PF: p = 0.015, n = 6 session pairs. Lose PF vs Retain PF: p = 0.014, n = 6 session pairs; paired-sample two-tailed t-test. Black squares and error bars denote mean ± s.e.m.
