Extended Data Fig. 2: OFL-related neuronal activation and in vivo optogenetic manipulations.
a, More c-fos-positive neurons in BLA (*P = 0.0043), vHPC (*P = 0.0027) and cACC (*P = 0.0135), not S1 (P = 0.5022) in OBS than unconditioned controls (CON). P values from Bonferroni post hoc tests following mixed-model ANOVA. n = 12 OBS mice, n = 15 CON mice. b, Viral spread (green shading) and optic fibre tips (‘X’) directing light at eArch3.0-expressing dmPFC neurons during OFL conditioning. c, No difference in CS-related freezing during conditioning or retrieval between eArch3.0 and YFP controls when light was shone during retrieval (not conditioning). Higher CS-related freezing versus pre-CS in eArch3.0 group (conditioning #P = 0.0001, retrieval #P = 0.0058, n = 8 mice) and YFP controls (conditioning #P = 0.0011, retrieval #P = 0.0083, n = 12 mice). P values from Bonferroni post hoc tests following mixed-model ANOVA. d, Viral spread (green shading) and optic fibre tips (‘X’) directing light at eArch3.0-expressing dmPFC neurons during OFL retrieval. e, Location of GRIN lenses in dmPFC. f, Heat-map of calcium activity aligned to CS during OFL conditioning. g, Peak Z-score (left, P = 0.9121) and decay time constant (right, P = 0.1148) of calcium activity during Shock-to-DEM or Shock-to-OBS. P values from paired t-tests. h, Increased population calcium activity during Shock-to-OBS in a random subset of 100 neurons (compare with population trace from all neurons, depicted in Fig. 2a). This shows that different population responses to Observed and Direct (depicted in Fig. 2a) is unlikely to be due to the smaller % of observed than Direct-responsive neurons. i, Percent neurons responsive to Observed (Shock-to-DEM) and Direct (Shock-to-OBS). j, K-means clustering of neuronal subpopulations with affinity for both Observed, Direct, both or neither. Affinity defined as the strength of the correlation between an individual neuron’s activity and the population-average activity of Observed or Direct related neurons (as depicted in Fig. 2c). k, Microendoscope dmPFC neuronal calcium imaging in OBS during 2 OFL conditioning sessions, followed by a DFL session. l, Location of GRIN lenses in dmPFC. m, Percent neurons responsive to Observed (Shock-to-DEM) on two OFL episodes and to Direct (Shock-to-OBS) in the same animals. Adjusted mutual information estimation indicated a high degree of Shock-to-DEM neuronal population overlap across the two OFL episodes (0.0179, adjusted *P < 0.0001 versus chance), whereas there was less overlap between the Shock-to-DEM population on OFL session 1 and the Shock-to-OBS population (0.0023, adjusted P = 0.0680). n, Decoding Observed (OFL1) versus Direct (*P = 0.0059) and failure to decode two sessions of OFL (P = 0.1372). P values from paired t-tests versus rotated. N = 8 mice panels e-j, N = 5 mice panels l-n. Two-tailed statistical tests were used. Data shown as means ± SEM.
