Fig. 4
Soluble Aβ42 oligomers overlap with the effect of Nr3c1ki/ki on spine maintenance. a Experimental timeline in NR3C1ki/ki-thy1-YFP double transgenic mice. b Corticosterone blood levels. Quartiles and median of dataset from Min-to-Max (N = 5 mice/group). Three-way ANOVA: effect of Aβ42 F1,16 = 123, P < 0.0001; effect of genotype F1,16 = 1.5, P = 0.2; effect of time F1,16 = 36.1, P < 0.0001 post-hoc Tukey test P < 0.05. c Percentage of alternance between arms of the Y-maze. Quartiles and median of dataset from Min-to-Max (N(+/+,ki/ki) = 9,9 vehicle; 8,8 Aβ42 mice). Two-way ANOVA analyses show no effect of genotype or oligomers P > 0.05. d Time exploring the new object over the old one presented 24 h earlier in the novel object recognition test. Quartiles and median of dataset from Min-to-Max expressed as ratio index in N(+/+,ki/ki) = 9,9 vehicle; 8,8 Aβ42 mice. Two-way ANOVA: effect of Aβ42 F1,30 = 20.1, P < 0.0001; interaction with Nr3c1ki/ki F1,30 = 6.6, P = 0.01 post-hoc Tukey test *P < 0.05, ***P < 0.0001. e Time exploring the new mouse over the old one presented 24 h earlier in the 3-chamber test. Quartiles and median of dataset from Min-to-Max expressed as ratio index in N(+/+,ki/ki) = 9,9 vehicle; 8,8 Aβ42 mice. Two-way ANOVA: effect of Aβ42 F1,30 = 17.8, P = 0.0002; effect of Nr3c1ki/ki F1,30 = 13.5, P = 0.0009 post-hoc Tukey test *P < 0.05, ***P < 0.0001. f Dendritic spine remodeling in the motor cortex of mice trained on the rotarod then injected with soluble Aβ42 oligomers before recall of behavioral performance one week later. g Dynamic changes between images 1 and 2 (effect of training), and images 2 and 3 (effect of Aβ42). h Proportion of dendritic spines per categories of dynamic events between time points. Quartiles and median of dataset from Min-to-Max (N(+/+,ki/ki) = 6,6 vehicle; 7,7 Aβ42 mice). Three-way ANOVA: effect of Aβ42 F1,22 = 15.78, P = 0.0006; effect of Nr3c1ki/ki F1,22 = 14, effect of dynamic events F4,88 = 879, P < 0.0001, interaction of 3 factors F4,88 = 5.9, P = 0.0003. Multiple comparisons by Mann Whitney test between genotypes *P < 0.05, **P < 0.005 or between vehicle and Aβ42.##P < 0.001. i Restoration of lost spines is not random. Quartiles and median of dataset from Min-to-Max (N(+/+,ki/ki) = 6,6 vehicle; 7,7 Aβ42 mice). Two-tailed unpaired t-test for comparing the observed versus 10,000 simulations of the restored lost spine at any position in the dendrite P < 0.05