Background
Alzheimer's disease (AD) is a progressive neurodegenerative disorder and the most common cause of dementia in the elderly. One of the hallmark pathologies of AD is the senile plaque that is constituted of amyloid-β (Aβ) peptides. Although mutations in three different genes favoring the overproduction of Aβ are known to cause early-onset familial AD (FAD) [
1,
2], the etiology of sporadic AD that accounts for the majority of AD cases remains unclear. It is hypothesized that complex interactions between the genetic background and various environmental factors underlie sporadic AD [
3,
4] and a stressful lifestyle may represent one of the important risk factors for AD [
5]. Elderly individuals prone to psychological distress are more likely to develop AD than those not prone to distress, and this trait is also associated with a more rapid progression of disease [
6,
7]. Consistent with the clinical observations, recent studies demonstrated that exposure to adverse behavioral stress accelerates the development of amyloid pathology and worsens memory decline in transgenic mouse models of AD [
8‐
11], although the underlying molecular mechanisms have not been investigated in detail.
Meanwhile, it is also known that women have a higher risk of developing AD than do men. Although the longevity effect might be a factor in the preponderance of women with AD, the sex difference in AD prevalence remains even after age adjustment [
12‐
14]. Interestingly, not only the sex but also the brain region represents a determinant of neuronal responses to stressful experience. For example, the hippocampus is a structure highly sensitive to stressor and is enriched with glucocorticoid receptors [
15]. Furthermore, opposite effects of stress on hippocampal functions have been reported; an acute stressful event facilitates learning and increases dendritic spine density in male rats, whereas learning and spine density deteriorate after exposure to the same stressor in female rats [
16‐
18]. Therefore, it is important to determine whether and how gender and adverse stress may interact to modify disease progression in different brain regions of AD transgenic mice.
The β-cleavage of amyloid precursor protein (APP) by BACE1 (β-site APP cleaving enzyme 1) initiates the generation of neurotoxic Aβ peptides. Notably, evidence is accumulating that increased levels of cerebral BACE1 and/or APP expression may be crucial contributing factors in developing sporadic AD [
19‐
22]. In this study, we tested the hypothesis that behavioral stress may trigger the upregulation of BACE1 and/or APP, which may lead to differential acceleration of β-amyloidogenesis in the hippocampus and cerebral cortex of male and female subjects. Specifically, 5XFAD transgenic model mice at 3 months of age, which exhibit little or only faint amyloid pathology under normal conditions [
23], were exposed to 5-day restraint stress, and we compared levels of BACE1, APP, its β-cleavage products and plaque burden between stressed and non-stressed subjects. We clearly demonstrate that the hippocampus of female 5XFAD mice shows the dramatic acceleration of β-amyloidogenesis with significantly elevated levels of both BACE1 and APP expression following the relatively brief stress treatment, providing a molecular basis for the higher prevalence and incidence of sporadic AD in women. Furthermore, our results also suggest that not only transcriptional but also translational mechanisms through phosphorylation of eukaryotic initiation factor-2α (eIF2α) may underlie BACE1 elevation associated with adverse stress during AD progression.
Discussion
Very few AD cases can be attributable to genetic causes, while the etiology of sporadic AD that constitutes the majority of AD cases remains unclear [
3,
4]. Nevertheless, most transgenic models of AD are created based on a simple genetic association between the rare inherited form of FAD and excessive Aβ production and do not encompass acquired characteristics [
27‐
30]. Many non-genetic factors including age, lifestyle (e.g., daily stress and diets), medical history and education have been reported to contribute to increasing the risk for AD [
31,
32]. Epidemiological investigations also show gender differences in the incidence and prevalence of AD with females being at higher risk [
12‐
14], although a biological foundation for gender differences remains to be determined. In this study, we applied a relatively brief behavioral stress to male and female 5XFAD transgenic model mice at the pre-pathological stage of disease that shows little or only faint amyloid deposition, and tested the hypothesis that sex and stress interactions may represent a key mechanism underlying sporadic AD and rendering women more prone to develop AD.
Our results clearly demonstrated that 5-day exposure to restraint stress increased levels of Aβ42 peptides, a pathogenic Aβ species that is more hydrophobic and has the propensity to assemble into neurotoxic oligomers and aggregates [
33‐
35], in the hippocampus of female 5XFAD mice but not in male 5XFAD mice. Meanwhile, the same stress treatment did not significantly affect Aβ42 levels in the cerebral cortex of male or female 5XFAD mice. Accordingly, amyloid plaque formation was also accelerated specifically in the female 5XFAD hippocampus following restraint stress. Interestingly, previous studies applied much longer stress treatments (e.g., several months of immobilization and/or isolation) to APP transgenic mice and showed that such prolonged stress exposures resulted in elevated Aβ concentrations and pathology in both the hippocampus and cortex without distinction based on sex [
8‐
11]. Therefore, it should be noted that only 5-day exposure to restraint stress was sufficient to significantly elevate levels of Aβ42 and plaque load in 5XFAD mice in this study. Of particular importance, such a brief stress treatment revealed that the hippocampus of females is vulnerable and prone to develop amyloid deposits in response to adverse behavioral stressors. These findings support the idea that the higher prevalence of sporadic AD in women may be, at least in part, attributable to the vulnerability of female brain, especially the hippocampus, to stress mechanisms that favor β-amyloidogenic processing of APP. This is also consistent with the observation that hippocampal Aβ deposition is one of the earliest features of AD [
36].
What mechanisms underlie the sex- and brain region-specific acceleration of Aβ accumulation following brief stress exposure? Previous studies demonstrate a sex difference in stress effects and estrogens are known to potentiate the glucocorticoid secretion under stress conditions [
37,
38]. This mechanism may represent the key component of the stress response that promotes β-amyloidosis more profoundly in female 5XFAD brains. Alternatively, recent studies have started to reveal the involvement of gender-specific molecular signaling processes or sex chromosome gene expression in addition to estrogen effects in causing sex differences in neuronal function [
39,
40]. Further study is needed to address the precise mechanisms including interactions between sex hormones and different stress mediators linking with changes in neuronal amyloidogenic processing of APP associated with AD.
Importantly, we found that BACE1 expression was elevated specifically in the hippocampus of stressed female 5XFAD mice in accordance with increased levels of Aβ42 and plaque burden. It has been reported that protein and/or activity levels of BACE1 become elevated in brains of sporadic AD patients [
19‐
21,
41] and 5XFAD mice [
26,
42‐
44] as disease progresses into the severer pathological stage with established amyloid plaques. Recent studies including ours demonstrate that phosphorylation of the translation initiation factor eIF2α plays a critical role in mediating the post-transcriptional upregulation of BACE1 associated with AD [
25,
26]. Increases in phospho-eIF2α levels occur in brains of sporadic AD and advanced pathological stages of APP transgenic mice including the 5XFAD model [
24‐
26,
45,
46] and are shown to correlate with BACE1 elevation [
25,
26]. In the present study, baseline phospho-eIF2α levels of non-stressed 5XFAD mice at the pre-pathological stage were not significantly different from those of wild-type control mice, while they were increased by 5-day restraint stress exposure specifically in the hippocampus of female 5XFAD mice in line with BACE1 elevation. Therefore, these results suggest that the phospho-eIF2α-dependent translational upregulation of BACE1 in response to behavioral stressors may represent an important molecular mechanism by which environmental factors initiate β-amyloidogenesis before significant Aβ deposition occurs during the early phase of sporadic AD. This hypothesis is strongly supported by our recent observation that the increase in phospho-eIF2α induced by Sal 003, a specific inhibitor of its phosphatase, elevates BACE1 levels in younger 5XFAD mice, which have not yet showed BACE1 upregulation at basal levels concomitant with only marginal increases in eIF2α phosphorylation [
26].
In addition to the translational mechanism, transcriptional control of BACE1 may also be implicated in AD pathogenesis [
47,
48]. In this study, we showed the increase of BACE1 mRNA level in the hippocampus of stressed female 5XFAD mice compared with that of non-stressed controls, suggesting a possibility that transcriptional mechanisms may also contribute to the BACE1 elevation associated with adverse behavioral stress. Our results are consistent with the findings that the promoter region of BACE1 gene contains glucocorticoid responsive elements [
49] and that glucocorticoid administration facilitates Aβ production possibly via increases in transcription of the BACE1 gene through this binding site [
50]. Some studies with postmortem human brains report elevations in BACE1 mRNA levels associated with sporadic AD [
21,
51], while others show no changes in mRNA despite the increased levels of BACE1 activity and protein [
22,
52‐
54]. Therefore, the mechanisms underlying BACE1 elevation in the sporadic AD brain remain controversial, which may be accounted for by differences in complex environmental factors mainly responsible for the disease progression. Our mouse model study suggests that both transcriptional and translational mechanisms may underlie BACE1 elevations associated with adverse stress during the development of AD.
Intriguingly, we also found that stress-responsive increases in APP expression levels occurred only in the hippocampus of female 5XFAD mice. Therefore, it seems likely that elevations in both BACE1 and its substrate APP work cooperatively to enable dramatic increases in the intermittent β-cleaved C-terminal fragment C99 and the acceleration of Aβ42 production and plaque formation in the hippocampus of female 5XFAD mice following 5-day exposure to behavioral stress. Given that Aβ and C99 peptides are amyloidogenic and can induce synaptic failure, neurodegeneration and memory loss [
34,
55‐
59], behavioral stress-dependent elevations in both β-cleavage products through BACE1 and APP upregulation have important implications for the pathogenesis of sporadic AD and the progression of neuronal dysfunction. Our findings are in agreement with recent reports showing that exposure to stress-level glucocorticoids (daily injections of dexamethasone for 7-21 days) elevates both BACE1 and APP levels leading to accelerated C99 production and Aβ accumulation [
50,
60,
61]. However, these pharmacologically induced stress responses seem more robust than the behavioral stress regimen applied in this study as the changes are observed in brains of male subjects including 3xTg-AD transgenic model mice and middle-aged wild-type mice or rats. In any case, our present study combined with others indicates that elevated levels of glucocorticoids found in sporadic AD brains [
62‐
64] may not only be a consequence of the pathology but also play a causal role in triggering β-amyloidogenesis through BACE1 and APP elevations during earlier stages of disease progression.
Conclusions
Our mouse model study clearly demonstrates that the responsiveness of brains (especially, hippocampal neurons) to stress conditions, which shift APP processing toward β-amyloidogenesis by upregulating BACE1 and its substrate APP, represents a crucial contributing factor in the development of sporadic AD and may account for a mechanism underlying the increased prevalence of women to develop AD. Moreover, our data also suggest that transcriptional and translational mechanisms may underlie BACE1 elevation in response to adverse stressors, supporting the idea that therapeutic interventions aimed at suppressing stress-related signaling pathways (e.g., reduction of glucocorticoids or eIF2α phosphorylation) may be beneficial for slowing down AD progression.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
LD performed a majority of the experiments, analyzed the data and wrote the manuscript. MJA and SDG performed the qPCR experiment and analyzed the data. MO designed the experiments and wrote the manuscript. All authors read and approved the final manuscript.