An allostatic epigenetic memory on chromatin footprints after double-hit acute stress

Abstract

Stress induces allostatic responses, whose limits depend on genetic background and the nature of the challenges. Allostatic load reflects the cumulation of these reponses over the course of life. Acute stress is usually associated with adaptive responses, although, depending on the intensity of the stress and individual differences , some may experience maladaptive coping that persists through life and may influence subsequent responses to stressful events, as is the case of post-traumatic stress disorder. We investigated the behavioral traits and epigenetic signatures in a double-hit mouse model of acute stress in which heterotypic stressors (acute swim stress and acute restraint stress) were applied within a 7-day interval period. The ventral hippocampus was isolated to study the footprints of chromatin accessibility driven by exposure to double-hit stress. Using ATAC sequencing to determine regions of open chromatin, we showed that depending on the number of acute stressors, several gene sets related to development, immune function, cell starvation, translation, the cytoskeleton, and DNA modification were reprogrammed in both males and females. Chromatin accessibility for transcription factor binding sites showed that stress altered the accessibility for androgen, glucocorticoid, and mineralocorticoid receptor binding sites (AREs/GREs) at the genome-wide level, with double-hit stressed mice displaying a profile unique from either single hit of acute stress. The investigation of AREs/GREs adjacent to gene coding regions revealed several stress-related genes, including Fkbp5, Zbtb16, and Ddc, whose chromatin accessibility was affected by prior exposure to stress. These data demonstrate that acute stress is not truly acute because it induces allostatic signatures that persist in the epigenome and may manifest when a second challenge hits later in life.

Fig. 1

Experimental timeline and plasma CORT measurements. Males (n = 8) and females (n = 9) were randomly assigned to A: CONT, B: ARS, C: FSS, and D: 2-h. At sacrifice (11:30–15:30 h), trunk blood was collected in control mice or 40-min after exposure to FSS (FSS, ARS, or 2-h mice) and the levels of plasmatic CORT were measured using RIA. Males and females exposed to FSS 40 min prior showed increased levels of CORT compared to controls, regardless of prior exposure to ARS (2-way ANOVA, stress: F(1,20) = 96.13, p < 0.0001). Bar chart shows the mean ± S.E.M. of 3 determinations per group. ***p < 0.001. CONT: control, ARS: acute restraint stress, FSS, forced swim stress, 2-h: double hit stress, vHPC: ventral hippocampus, CORT RIA: corticosterone radioimmunoassay, ♀: female, ♂: male.

Fig. 2

Gene set enrichment analysis of ATAC-seq for multiple comparisons. GSEA shows discrete biological pathways (thick marks) enriched when comparing males and females, each stressor to controls, or stressors between them. Aggregated biological pathways were assigned a common name that reflected similar gene functions across multiple GO terms (brackets). (See Supp. Data 5 for full ordered list of pathways). Enrichment results were based on ATAC peaks annotated to genes. NES score was calculated based on the log2FC(female/male) for the sex comparisons, as well as log2FC(FSS/C), log2FC(ARS/C), log2FC(2-h/C), log2FC(2-h/ARS), log2FC(2-h/FSS), log2FC(FSS/ARS) for the stress comparisons. All presented enrichment reached a significance of padj<0.05. FSS/C: forced swim stress versus control, ARS/C: acute restraint stress versus control, 2-h/C: double-hit stress versus control, 2-h/ARS: double-hit stress versus acute restraint stress, 2-h/FSS: double-hit stress versus forced swim stress, FSS/ARS: forced swim stress versus acute restraint stress, NES: normalized enrichment score, p.adj: adjusted p-value, ♀: female, ♂: male.

Fig. 3

Identification of the most variable motifs across stress conditions. A. The top fifty motifs with the most significant variability across stress groups are plotted in a heatmap. Differential chromatin accessibility of each motif was calculated from a deviation score representing how accessible each motif site was relative to the expectation based in equal chromatin accessibility profiles across samples, normalized by a set of background peak sets matched for GC and average accessibility. B. Clustering of motifs into transcription factor families based on similar binding motifs. Transcription factors AR, GR, and MR ranked the top three and clustered together into the same transcription factor family. ARS: acute restraint stress, FSS: forced swim stress, 2-h: double hit stress, AR: androgen receptor, GR/Nr3c1: glucocorticoid receptor, MR/Nr3c2: mineralocorticoid receptor, DEV: deviation score, ♀: female, ♂: male.

Fig. 4

Changes in open chromatin for TSS binding AR, GR, and MR. Meta plots of the mean ATAC signal at TSS sites with transcription factor motif binding sites for A. AR, B. GR, and C. MR. All plots were centered around the TSS ± 2 kb. Wilcox testing identified significant differences in TSS mean signaling across groups. ●p < 0.05 versus same stress between sex, *p < 0.05 versus control within sex. D. Fkbp5, E. Zbtb16, and F. Ddc show chromatin ATAC profiles of differentially accessible genes computed by paired comparison with controls in females (left) and males (right) (padj < 0.05; fold change > 1.3), where padj = adjusted p value. TSS: transcription start site, CONT: control, ARS: acute restraint stress, FSS, forced swim stress, 2-h: double-hit stress, ♀: female, ♂: male.