Extended Data Fig. 3: In vivo single cell resolution calcium recordings in dmPFC during OFL.
a, PC1 (upper) and PC2 (lower) neuronal trajectories associated with Shock-to-DEM and Shock-to-OBS. b, When freezing was equated across Observed and Direct conditioning by taking a subset of the 10 highest-freezing OFL trials (72 ± 12%) and the 10 lowest-freezing Direct trials (61 ± 11%) for each mouse (paired t-test, t(6) = 1.795, P = 0.1227), Observed and Direct related activity trajectories still occupy distinct places in low-dimensional state-space after PCA (compare with Fig. 2e,f). c, PC1 trajectory distance remained different between Observed and Direct when freezing was equated across OFL and DFL conditioning. Black line denotes periods that are significantly different from rotated data (Benjamini-Hochberg corrected permutation tests). d, Uniform manifold approximation and projection (UMAP) dimensionality reduction state-space visualization of neuronal data during shock and inter-shock periods on Shock-to-DEM and Shock-to-OBS test sessions. e, Percent neurons exhibiting changes in activity during freeze or move. f, Example of spatially intermingled freeze and move neurons within the field of view. Neither freeze-related (mean distance to cluster-centroid: 61.2 a.u.) nor move-related (mean distance to cluster-centroid: 61.9 a.u.) neurons were clustered to a greater or lesser extent than expected by chance in any mouse. The distance between the centroid of freeze-related and move-related clusters (11.3 a.u.) was not to a greater or lesser extent than expected by chance in any mouse. g, Freeze and move state-related activity trajectories in low-dimensional state-space following PCA. h, PC2 neuronal trajectories associated with freeze and move state. i, Weak relationship between neuronal activity related to stopping during pre-CS baseline and freezing during OFL, as measured by Pearson’s r and linear regression adjusted R2. j, PC1 + 2 neuronal trajectories differentially anticipate upcoming switches to freeze state during Observed (centre) and Direct (right) but not switches to stopping during pre-CS baseline (left). Black line denotes period significantly different from rotated data (Benjamini-Hochberg corrected permutation tests). Data shown in c and j are means ± SEM. N = 8 mice for all panels.
