Fig. 4

Reorganization of SOM-IN activity with reward relocation. A Temporal sequence of surgical and behavioral procedures for relearning task with top view illustration of the virtual environment for the spatial learning task (reward location in green) and relearning task (new reward location in red; old location green dashed line). B–G Summary plots of behavioral changes during learning and relearning in control mice (n = 6; black) and SOM-Rptor-KO mice (n = 5; magenta) showing no changes in number of trials per minute (B), animal speed (C), and percent licks in reward zone (E) in both mice groups during relearning; increases in percent success trials (D), lick accuracy (F), and learning index (G) during relearning only in control SOM-IRES-Cre mice, showing that control mice, but not SOM-Rptor-KO mice, relearn a new reward location in the same environment. H Example of reorganization of SOM-IN responses for a cell with “reward off sustained” response at end of learning (left) and “reward off transient” response at end of relearning (right). For each session, top left is mean Ca²⁺ responses (grey) and speed (red) as function of position for all trials in the session with reward zone indicated (green for learning; red for relearning); bottom left is color-coded Ca²⁺ activity in each trial of the session; top right is place correlation matrix of all paired laps; and bottom right is distribution of r values (blue), mean r (red) versus r distribution obtained by shuffling position measures (gray). I Similar representation of reorganization for a SOM-IN with “reward on sustained” response at end of learning and “reward off sustained” response at end of relearning. J Mean place correlation with activity for all SOM-INs showing the increase during relearning in control but not SOM-Rptor-KO mice. K Diagram of Ca²⁺ activity analysis criteria for each type of response with > 30% changes (light gray shaded area) in Ca²⁺ activity in the new reward zone (red shading) or after it, relative to activity before the reward zone (black line), defining response types (“off sustained”, dark red; “on sustained”, dark green; “off transient”, light red; “on transient”, light green; or no modulation, black) in the relearning task. L Cell response identity matrix for all cells in control mice during learning (green, left) and relearning (red, right) ordered by response type at end of learning, showing a gradual acquisition of a new activity related to reward during relearning. M Distribution of cells with different response types at start and end of learning and relearning for control mice, showing a decrease in number of modulated cells at start of relearning relative to end of learning, and an increase during relearning. N Summary graph for all cells with response related to reward at start and end of learning and relearning, showing initial decrease and subsequent increase in SOM-INs with responses during relearning in control (n = 6; black) but not in SOM-Rptor-KO (n = 6; magenta) mice. Details of statistical tests provided in Additional file 5: Table S1. *p < 0.05, **p < 0.01, ***p < 0.001, ns not significant