Adolescent stress accelerates postpartum novelty recognition impairment in 5xFAD mice

Abstract

Pregnancy and the postpartum period induce physiological changes that can influence women's cognitive functions. Alzheimer's disease (AD) has a higher prevalence in women and is exacerbated by early life stress. In the present study, we found that late adolescent social isolation combined with the experience of pregnancy and delivery accelerates the onset of cognitive deficits in 5xFAD dams, particularly affecting their ability to recognize novelty. These cognitive deficits manifested as early as 16 weeks, earlier than the usual timeline for these mice, and were closely associated with increased levels of corticosterone, suggesting dysregulation of the hypothalamic-pituitary-adrenal (HPA) axis. Notably, the presence of β-amyloid plaques in brain regions associated with novelty recognition did not significantly contribute to these deficits. This highlights the potential role of stress and HPA axis dysregulation in the development of cognitive impairments related to AD, and underscores the need for further investigation.

Keywords: Alzheimer’s disease; acceleration; adolescent stress; corticosterone; novelty recognition impairment.

Figure 1

Experimental schedule. (A) Virgin 5xFAD females were group-housed and not mated (unstressed virgins). (B) Virgin 5xFAD females were isolated from 5 to 8 weeks of age and not mated (stressed virgins). (C) Virgin 5xFAD females were group-housed, mated with a male at 8 weeks of age, and gave birth to pups (unstressed dams). (D) Virgin 5xFAD females were isolated from 5 to 8 weeks of age, mated with a male at 8 weeks of age, and gave birth to pups (stressed dams). The novel object recognition test, followed by three-chamber social interaction test, was performed at 16 and 40 weeks of age (indicated by the arrows). Different cohorts of mice subjected to behavioral tests at 16 weeks and 40 weeks of age were prepared to avoid the repeated exposure to stressful behavioral procedures.

Figure 2

Cognitive behavioral changes in 5xFAD mice at 40 weeks of age. (A) Regardless of pregnancy and delivery history, both non-stressed and stressed 5xFAD mice exhibited behavioral deficits in the novelty object recognition test (NORT). (B) In the social interaction test (SIT), both non-stressed and stressed 5xFAD mice showed behavioral deficits in the sociability-trial (S-trial), irrespective of their pregnancy and delivery history. (C) In SIT, both non-stressed and stressed 5xFAD mice demonstrated behavioral deficits in the social novelty-trial (SN-trial), regardless of their pregnancy and delivery history. All data were non-normally distributed and analyzed using the Mann–Whitney U test. All comparisons were non-significant (p > 0.05), except stressed virgins vs. stressed dams for NORT (^*p = 0.034). N = 3–12. Data are presented as mean ± SEM. See Supplementary Table S1 for details on the sample size and statistical analyses.

Figure 3

Cognitive behavioral changes in 5xFAD mice at 16 weeks of age. (A) Only stressed 5xFAD dams displayed behavioral deficits in novelty object recognition when compared to the other three groups. Mann–Whitney U test: Unstressed virgins vs. stressed dams, ^***p < 0.001; stressed virgins vs. stressed dams, ^*p < 0.05; unstressed dams vs. stressed dams, ^*p < 0.05. (B) During the S-trial of SIT, stressed 5xFAD virgins, unstressed 5xFAD dams, and stressed 5xFAD dams showed deficits in sociability when compared to unstressed virgins. Mann–Whitney U test: Unstressed virgins vs. stressed virgins, ^**p < 0.01; unstressed virgins vs. unstressed dams, ^*p < 0.05; unstressed virgins vs. stressed dams, ^*p < 0.05. (C) In the SN-trial of SIT, stressed 5xFAD dams exhibited an accelerated behavioral deficit in social novelty recognition compared to the other three groups. Two-way ANOVA: For interaction, F_(3,56) = 0.325, p = 0.571, partial η² = 0.006; for stress, F_(1,56) = 12.972, p = 0.001, partial η² = 0.188; for pregnancy/delivery, F_(1,56) = 12.842, p = 0.001, partial η² = 0.187. Bonferroni: Unstressed virgins vs. stressed virgins, ^*p < 0.05; stressed virgins vs. stressed dams, ^**p < 0.01; unstressed dams vs. stressed dams, ^**p < 0.01; unstressed virgins vs. stressed dams, ^***p < 0.001. N = 9–17. Data are presented as mean ± SEM. See Supplementary Table S1 for details on the sample size and statistical analyses.

Figure 4

Changes in serum corticosterone levels and the quantity of β-amyloid plaques in 5xFAD mice at 16 weeks of age. (A) Serum corticosterone levels increased in response to both late adolescent social isolation and pregnancy/delivery, with a notably accelerated rise when both conditions co-occurred. Mann–Whitney U test: Unstressed virgins vs. stressed virgins, ^*p < 0.05; unstressed virgins vs. unstressed dams, ^**p < 0.01; unstressed virgins vs. stressed dams, ^***p < 0.001; stressed virgins vs. stressed dams, ^**p < 0.05; unstressed dams vs. stressed dams, ^*p < 0.05. (B–D) Late adolescent social isolation and/or pregnancy/delivery did not affect the number of β-amyloid plaques in the anterior insula (AI; B), prelimbic cortex (PrL; C) and dentate gyrus of the hippocampus (DG; D), as evaluated by a Two-Way ANOVA, except for the comparison between unstressed virgins vs. stressed dams using Bonferroni adjustments (p = 0.027). (E) Representative images of β-amyloid plaques in the dentate gyrus of the hippocampus. p > 0.05 for all main factors and interaction. N = 7–11. Data are presented as mean ± SEM. See Supplementary Table S1 for details on the sample size and statistical analyses.

Figure 5

Correlations between corticosterone levels and cognitive behaviors in 5xFAD mice at 16 weeks of age. (A) Serum corticosterone levels were found to be negatively correlated with the novelty recognition indexes for NORT. (B) There was no significant correlation between corticosterone levels and sociability indexes for SIT during the S-trial. (C) A negative correlation was observed between corticosterone levels and sociability indexes for SIT during the SN-trial. N = 7–11. R² and p-values were obtained from Pearson’s correlation test for all graphs. See Supplementary Table S1 for details on the sample size and statistical analyses.