Fig. 2. Exposure to repeated stress impairs temporal and spatial codes of PNs in dCA1.
a Neuronal ensembles (one mouse, one exploration session) sorted according to their normalized reactivation strengths. The arrow denotes the threshold for reactivation strength = 1, the pink rectangle indicates the significant ensembles corresponding to reactivation strength > 1. b Stress exposure did not change the fraction of significant ensembles among all (left) and lower activity neurons (middle) but significantly increased the fraction of significant ensembles among higher activity neurons (right). All: pB-C > 0.99, pB-S > 0.99, pC-S > 0.99; nB = 84, nC = 28, nS = 45. Lower activity: pB-C > 0.99, pB-S > 0.99, pC-S > 0.99; nB = 83, nC = 28, nS = 43. Higher activity: pB-C > 0.99, pB-S = 0.03, pC-S = 0.17; nB = 50, nC = 15, nS = 32. c Neurons (one mouse, one exploration session) sorted according to their weights within an ensemble. The arrow indicates the threshold that demarcates the ensemble’s core (pink rectangle). d Stress exposure decreased the size of ensembles’ cores for all neurons (left: pB-C > 0.99, pB-S < 0.0001, pC-S = 0.0017; nB = 87, nC = 24, nS = 38), mainly due to higher activity neurons. Lower activity: pB-C = 0.84, pB-S = 0.38 pC-S = 0.09; nB = 83, nC = 28, nS = 40. Higher activity: pB-C = 0.25, pB-S = 0.0002, pC-S = 0.09; nB = 80, nC = 26, nS = 39. Box plots: medians and quartiles of mean fraction of significant ensembles b or of neurons participating in ensembles’ cores d distributions per mouse per session. e Participation of high activity rate neurons in ensembles’ cores increased with time, but stress exposure prevented this increase (p-2 < 0.001, p-1.5, -1, -0.5, 0, 0.5, 1, 1.5, 2 > 0.13, p2.5 < 0.0001, p3 = 0.96, p3.5 < 0.0001; n-2 > 68, n-1.5, -1, -0.5 > 702, n0 > 2461, n0.5, 1, 1.5, 2 > 285, n2.5 > 75, n3 > 18, n3.5 > 5); Pairwise, two-tailed t-tests between control and stress per activity rate bin. P-values corrected for multiple comparisons with Holm-Šidák Method. Circles: mean ensemble participation index binned over activity rates. Error bars: s.e.m. Dashed line indicates the threshold for neurons with higher activity rates. f Stress exposure decreased the mean participation of higher activity rate neurons to ensembles’ cores (pB-C > 0.99, pB-S = 0.031, pC-S = 0.32; nB = 4445, nC = 1201, nS = 5863). g Stress exposure increased the fraction of neurons in ensembles’ peripheries (pB-C > 0.99, pB-S = 0.046, pC-S = 0.28; nB = 40, nC = 28, nS = 35). Box plots: medians and quartiles of mean fraction of neurons participating in ensemble’s periphery distributions per mouse per session. h Reactivation strength of ensembles increased with time, but to a lower extent upon stress exposure (pB-C < 0.0001, pB-S < 0.0001, pC-S = 0.0072; nB = 1462, nC = 510, nS = 749; Two-Way ANOVA). Circles: mean reactivation strengths of ensembles index ranked according to decreasing values. Error bars: s.e.m. i Stress exposure increased the activity rates of neurons participating in ensembles with similar reactivation strengths (pB-C S < 0.0001, pB-S < 0.0001, pC-S < 0.0001; nB = 25230, nC = 9575, nS = 11668). Circles: mean reactivation strengths of the ensembles in which each neuron took part binned over activity rates. Error bars: s.e.m. Dashed line indicates the threshold for neurons with higher activity rates. j Stress exposure decreased the amount of SI per neuron (pB-C = 0.38, pB-S < 0.0001, pC-S < 0.0001; nB = 1152, nC = 396, nS = 540). k Stress exposure decreased the SI in neurons with lower activity rate and increased the SI in neurons with higher activity rate (pB-C = 0.21, pB-S = 0.074, pC-S < 0.0001; nB = 9575, nC = 25,230, nS = 1168). Circles: mean SI binned over activity rates. Error bars: s.e.m. l Stress exposure decreased the fraction of neurons with higher spatial information (pBcontrol-Bstress > 0.99, pBcontrol-C > 0.99, pBstress-S = 0.0011, pC-S = 0.0014; nBcontrol = 22, nBstress = 37, nC = 26, nS = 29). Circles: mean fraction of higher SI neurons per mouse per session. Solid symbols, stress group. Empty symbols, control group. Error bars: s.e.m. m Schematic description of the transformation from correlation adjacency matrices to connectivity networks and their meaning in terms of temporal sparseness of activity. n Stress exposure increased co-activity and thus prevented the increase in sparseness of activity of dCA1 PNs (pB-C S < 0.0001, pB-S < 0.0001, pC-S < 0.0001; nB = 14303, nC = 6054, nS = 9026). Circles: mean connectivity index binned over activity rates. Error bars: s.e.m. Dashed line indicates the threshold for neurons with higher activity rates. b, d, f, g, l Kruskal–Wallis test, p-values adjusted after Dunn’s corrections for multiple comparisons. i, j, k, n Two-way ANOVA, p-values adjusted after Tukey’s corrections for multiple comparisons.