Environmental Enrichment Promotes Transgenerational Programming of Uterine Inflammatory and Stress Markers Comparable to Gestational Chronic Variable Stress

Abstract

Prenatal maternal stress is linked to adverse pregnancy and infant outcomes, including shortened gestation lengths, low birth weights, cardio-metabolic dysfunction, and cognitive and behavioural problems. Stress disrupts the homeostatic milieu of pregnancy by altering inflammatory and neuroendocrine mediators. These stress-induced phenotypic changes can be passed on to the offspring epigenetically. We investigated the effects of gestational chronic variable stress (CVS) in rats using restraint and social isolation stress in the parental F0 generation and its transgenerational transmission across three generations of female offspring (F1-F3). A subset of F1 rats was housed in an enriched environment (EE) to mitigate the adverse effects of CVS. We found that CVS is transmitted across generations and induces inflammatory changes in the uterus. CVS did not alter any gestational lengths or birth weights. However, inflammatory and endocrine markers changed in the uterine tissues of stressed mothers and their offspring, suggesting that stress is transgenerationally transmitted. The F2 offspring reared in EE had increased birth weights, but their uterine gene expression patterns remained comparable to those of stressed animals. Thus, ancestral CVS induced changes transgenerationally in fetal programming of uterine stress markers over three generations of offspring, and EE housing did not mitigate these effects.

Keywords: chronic variable stress; enriched environment; gene expression; inflammation; pregnancy; prenatal stress; preterm birth; resilience; rodents; uterus.

Figure 1

Maternal transgenerational prenatal stress experimental design. Flow chart illustrating that F0 pregnant dams were subjected to social isolation and restraint stress from GD12–18. EE housing included a combination of physical and sensorimotor enrichment and was implemented only in the F1 generation from weaning to GD20. The F1 (SN–EE) dams were then bred with nonstressed males to assess whether EE mitigated the negative effects of PNMS, and if these mitigative effects would be passed down to their F2 and F3 offspring. Therefore, the F2 and F3 generations were not directly exposed to EE housing or stress but experienced their transgenerational ancestral effects. F = filial generation. Created with Biorender.com (accessed on 15 July 2022).

Figure 2

Gestational lengths were unchanged in animals subjected to PNMS. (A) Gestational lengths among treatment groups. Gestational lengths in control and stress groups exposed to either SH or EE housing conditions in the (B) F1 and (C) F2 generations. Data are compared to F0N, mean ± SEM. Ordinary one-way analysis of variance (ANOVA) (A) and two-way ANOVA (B,C) analyses were used. N = 9–13 (A) or 7–10 (B,C). SH = standard housing; EE = enriched housing; F = filial generation; S = stressed; N = nonstressed.

Figure 3

Offspring weights remained unchanged between treatment groups, whereas EE housing significantly increased neonatal birth weights of females and males in the F2 generation. Offspring weights between treatment groups in (A) females and (B) males. Pup weights of control and stressed (C) females and (D) males subjected to SH or EE housing. Data are compared to F1NN (A,B) and F2NNN (C,D), mean ± SEM. Box plot mid-lines indicate medians, whiskers indicate min-max values, and boxes indicate interquartile ranges. Kruskal-Wallis test (A,B) or two-way ANOVA (C,D) analyses were used. SH = standard housing; EE = enriched housing; F = filial generation; S = stressed; N = nonstressed. N = 35–65 (females, A,C); 38–66 (males, B,D). Asterisks indicate significance: <0.05 (*); <0.002 (**); <0.001 (***).

Figure 4

Litter sizes were unchanged between treatments. Data are compared to F1NN and were analyzed using the Kruskal-Wallis test. Box plot mid-lines indicate medians, whiskers indicate min-max values, and boxes indicate interquartile ranges. N = 14–22 per group. F = filial generation; S = stressed; N = nonstressed.

Figure 5

Elevated CORT concentrations in the F1 and F2 stressed offspring despite enrichment therapy. (A) Plasma CORT levels in the F0–F3 stressed and control animals. Effects of housing and treatment on the (B) F1, (C) F2, and (D) F3 generations. Asterisks indicate significance: <0.05 (*); <0.002 (**); <0.001 (***). Data are compared to F0N, mean ± SEM. Ordinary one-way ANOVA (A) and two-way ANOVA (B–D) analyses were used. N = 4–5 per group. SH = standard housing; EE = enriched housing; F = filial generation; S = stressed; N = nonstressed.

Figure 6

Uterine ROS levels were unchanged between treatments and housing conditions across the F0–F3 generations. Four random regions of each uterus were used to measure mean fluorescence intensity (MFI) with DHE staining. (A) Representative images for each treatment and housing group. (B) Analysis of ROS levels (mean fluorescence intensity, MFI) in uterine samples from stressed dams compared to controls. Assessments of the effects of EE housing on uterine ROS levels in the (C) F2 and (D) F3 generations of animals subjected to ancestral stress. Data are compared to F0N, mean ± SEM. Ordinary one-way ANOVA (B) and two-way ANOVA (C,D) analyses were used. N = 5. SH = standard housing; EE = enriched housing; F = filial generation; S = stressed; N = nonstressed.

Figure 7

Uterine expression of Il1a was significantly downregulated in the F3 generation of stressed dams raised in SH and in F1 stressed animals exposed to both housing conditions. (A) Gene expression of Il1a in uteri of control and stressed dams across the F0–F3 generations. Uterine expression of Il1a in (B) F1 and (C) F2 animals exposed to different treatments and housing conditions. Asterisks indicate significance: <0.001 (***). Data are compared to F0N, mean ± SEM. Ordinary one-way ANOVA (A) and two-way ANOVA (B,C) analyses were used. N = 7–11 (A); 8–10 (B); or 6–10 (C). SH = standard housing; EE = enriched housing; F = filial generation; S = stressed; N = nonstressed.

Figure 8

Uterine mRNA expression of Il1b increased significantly in the F2 generation exposed to transgenerational stress and in the F1 and F2 controls exposed to EE. (A) Il1b expression in F0–F3 dams subjected to CVS under SH. Uterine gene expression of Il1b in the (B) F1 and (C) F2 generations of controls and stressed animals exposed to SH or EE housing. Asterisks indicate significance: <0.05 (*); <0.002 (**). Data are compared to F0N, mean ± SEM. Ordinary one-way ANOVA (A) and two-way ANOVA (B,C) analyses were used. N = 8–12 (A); 7–12 (B); or 6–12 (C). SH = standard housing; EE = enriched housing; F = filial generation; S = stressed; N = nonstressed.

Figure 9

Uterine protein abundance of IL-1RAP remained unchanged over the stressed F0–F3 generations regardless of housing type. (A) IL-1RAP protein abundance quantified using densitometry (representative blots included). (B) Uterine protein abundance of IL-1RAP in F0–F3 stressed dams compared to controls. (C–E) IL-1RAP abundance according to treatment and housing across the F1–F3 offspring. Blots were quantitated using Odyssey software. All groups were compared using the Kruskal-Wallis test (B), and the effects of treatment and housing were assessed using two-way ANOVA (C–E). Data are normalized to glyceraldehyde 3-phosphate dehydrogenase (GAPDH) and compared to the F0N, mean ± SEM. N = 4 (B) or 3–4 (C–E). SH = standard housing; EE = enriched housing; F = filial generation; S = stressed; N = nonstressed.

Figure 10

Uterine expression of Hsd11b2 is significantly decreased in F1–F3 stressed dams and in the animals exposed to EE housing. (A) Gene expression of Hsd11b2 in uteri of control and stressed dams across the F0–F3 generations. Uterine expression of Hsd11b2 in (B) F1, (C) F2, and (D) F3 animals exposed to different treatments and housing conditions. Asterisks indicate significance: <0.05 (*); 0.002 (**); <0.001 (***). Data are compared to F0N, mean ± SEM. Ordinary one-way ANOVA (A) and two-way ANOVA (B–D) analyses were used. N = 8–12 (A,B); 7–12 (C); or 6–12 (D) per group. SH = standard housing; EE = enriched housing; F = filial generation; S = stressed; N = nonstressed.

Figure 11

Uterine gene expression analysis of Nr3c2 showed significant increases in stressed F2 and F3 dams under SH, with similar effects shown when raised under enrichment. (A) Nr3c2 mRNA levels across the F0–F3 generations of stressed dams compared to controls. Analysis of Nr3c2 expression in the (B) F1 and (C) F2 generations of dams exposed to different treatments and housing types. Asterisks indicate significance: <0.05 (*); <0.002 (**); <0.001 (***). Data are compared to F0N, mean ± SEM. Ordinary one-way ANOVA (A) and two-way ANOVA (B,C) analyses were used. N = 8–12 (A,B) or 7–12 (C) per group. SH = standard housing; EE = enriched housing; F = filial generation; S = stressed; N = nonstressed.

Figure 12

Uterine expression levels of the antioxidant enzyme Sod1 were significantly increased in F1 stressed animals and in F1 controls subjected to EE conditions. (A) Expression of Sod1 in uteri of stressed F0–F3 dams compared to controls. (B) Uterine expression of Sod1 in F1 females exposed to stress and EE housing compared to controls and SH. Asterisks indicate significance: <0.05 (*). Data are compared to F0N, mean ± SEM. Ordinary one-way ANOVA (A) and two-way ANOVA (B) analyses were used. N = 8–12 per group. SH = standard housing; EE = enriched housing; F = filial generation; S = stressed; N = nonstressed.

Figure 13

The concentrations of IL-1α, IL-1β, and IL-10 were unchanged between control and stressed groups in uterine tissues. The IL-1β and IL-10 concentrations were increased in the control group compared to animals raised under enriched housing. (A–C) Uterine concentrations of IL-1α, IL-1β, and IL-10 between control and stressed animals raised under SH. Concentrations of IL-1β (D–F) and IL-10 (G–I) in the F1–F3 uteri of dams exposed to different treatments and housing types. Asterisks indicate significance: <0.05 (*); <0.002 (**); <0.001 (***). Data are compared to F0N, mean ± SEM. Ordinary one-way ANOVA (A–C) and two-way ANOVA (D–I) analyses were used. N = 3–6 per group. SH = standard housing; EE = enriched housing; F = filial generation; S = stressed; N = nonstressed.

Figure 14

Timeline illustrating the stress protocol, tissue collection, and offspring analyses. Gestational stress was implemented from GD12 to GD18 using restraint and social isolation stressors, creating our psychological and psychosocial chronic variable stress (CVS) model. Blood collection occurred on GD18 in the mothers and on P110 in the offspring. Dams were sacrificed at the weaning of their offspring (LD21) when uterine tissues were collected. Tested offspring were euthanized and had their tissues collected on P115. Created with BioRender.com (accessed on 19 July 2022).