|Home | About | Journals | Submit | Contact Us | Français|
Abnormally phosphorylated tau protein is a key component of the pathology seen in neurodegenerative tauopathies, such as Alzheimer's disease (AD). Despite its association with disease, tau phosphorylation (tau-P) also plays an important role in neuroplasticity, such as dendritic/synaptic remodeling seen in the hippocampus in response to environmental challenges, such as stress. To define the boundaries between neuroplasticity and neuropathology, studies have attempted to characterize the paradigms, stimuli, and signaling intermediates involved in stress-induced tau-P. Supporting an involvement of stress in AD are data demonstrating alterations in stress pathways and peptides in the AD brain and epidemiological data implicating stress exposure as a risk factor for AD. In this review, the question of whether stress-induced tau-P can be used as a model for examining the relationship between functional neuroplasticity and neuronal vulnerability will be discussed.
Alzheimer's disease (AD) is a progressive neurodegenerative disorder affecting approximately 40% of the population over 80 years of age  and is defined neuropathologically by the presence of amyloid-β (Aβ) plaques and neurofibrillary tangles (NFT). NFTs are intracellular inclusions composed of aggregated, hyperphosphorylated forms of the microtubule-associated protein, tau. Hyperphosphorylated tau has reduced ability to bind and stabilize microtubule networks and can self-aggregate into paired helical filaments (PHFs), which comprise NFTs. The prevalence of NFTs in the brain is correlated with cognitive decline and neuronal loss in AD, suggesting that pathological changes in tau may be involved in neurodegeneration.
With less than 1% of AD cases having identified genetic loci, significant research has focused on environmental factors that may be involved in sporadic AD. Of the numerous environmental influences that are under investigation, propensity to experience “psychological distress” in humans has been demonstrated to confer a three-fold increase in risk of AD [2,3]. These epidemiological studies provided important groundwork for rodent studies attempting to define causal links between stress exposure and AD neuropathological endpoints. Many of these rodent studies have focused on the ability of acute physiological or psychological stress episodes to induce tau-P in the hippocampus [4–11].
In terms of disease relevance, acute stress-induced tau-P appears to be a rapid and transient phenomenon, suggesting an involvement in stress adaptation and neuroplasticity. Given the propensity of phosphorylated tau to dissociate from microtubules, the transient nature of tau-P induced by acute stress can be seen as an attempt to promote modulation of networks to form new memories or strengthen existing synapses.
Tau is a soluble phospho-protein that stabilizes axonal microtubules to maintain cellular morphology and facilitate neurite outgrowth and differentiation [12–14]. Six isoforms of tau exist in brain and are produced by alternative mRNA splicing of a single gene on chromosome 17. Isoforms of tau differ by the inclusion of inserts in the N-terminal domain and tandem repeats within the C-terminal microtubule-binding domain. Recent studies demonstrate that tau mediates Golgi membrane and dynein binding to microtubules to facilitate neuronal integrity [15,16]. Tau binding/stabilization of axonal microtubules is also crucial for trafficking of intracellular organelles and vesicles . One of the primary events that regulate tau-microtubule interactions is phosphorylation of tau, which alters tau binding and function [12,13].
Pathological alterations in tau occur in several neurodegenerative disorders, termed tauopathies, and include AD, frontotemporal dementia, Pick's disease, supranuclear palsy, and corticobasal degeneration. While the tau pathology in many of these diseases is of sporadic origin, rare mutations in the tau gene have also been identified and can lead to neurodegenerative tauopathies .
In AD, tau becomes hyperphosphorylated and aggregates, forming PHFs within the somatodendritic compartment [19–21]. PHF bundles accumulate as insoluble NFTs in the entorhinal cortex and CA1 subfield of the hippocampus before appearing in cortical areas. The incidence of NFTs is positively correlated with cognitive deficit and neuronal loss in AD, which has positioned pathological changes in tau as critical in neurodegeneration. Although the specific links between tau alterations and neurotoxicity are controversial, tau phosphorylation (tau-P) is considered one of the earliest cytoskeletal changes in AD and a critical step in the formation of NFTs. Whereas tau is normally phosphorylated at 2–3 moles/mole of protein, tau from the AD brain is hyperphosphorylated at a 7–10 molar ratio, and at more than 25 distinct sites .
With the development of a variety of specific antibodies, studies have proposed a sequence of tau alterations, which include conformation, truncation and phosphorylation changes that lead to NFT. A folded conformational change in the N and C-termini on tau is thought to be the earliest pathological event and can be detected using conformational-specific tau antibodies (e.g., MC1) . Caspase-cleavage of tau at the C-terminus has also been demonstrated to be an early pathologic change in tau and can be detected with neoepitope antibodies to truncated tau [24–28]. Other antibodies are directed to immature NFTs by recognizing tau-P at several N-terminal serines and threonine, which are thought to be the first residues to be hyperphosphorylated . This is in contrast to antibodies directed toward phosphorylation sites in the C-terminus of tau, which recognize hyperphosphorylated tau found within more mature, late-stage NFTs .
Corticotropin-releasing factor (CRF) is best known as the hypothalamic neuropeptide that activates the stress response (reviewed in ). CRF exerts its biological effects by binding to two G protein-coupled receptors (CRFRs) that activate adenylate cyclase. CRFR1 is expressed in the anterior pituitary and can account for the neuroendocrine actions of the peptide. CRFR1 is also expressed widely in brain, including AD-relevant regions as isocortex, hippocampus, and amygdala . CRFR2 is also expressed in brain, but has a distinct and more limited distribution than CRFR1 .
Reduced cortical CRF and increased hypothalamic immunoreactivity is a prominent neurochemical change in AD, which occurs early in disease progression, and is focused in areas vulnerable to AD neuropathology (for review ). In addition, marked increases in CRF binding have been described in specific cortical regions of the AD brain, suggesting upregulation of CRFRs in areas vulnerable to AD pathology . It is thought that some of these changes may result from sequestration of CRF by a specific binding protein (CRF-BP) . In concordance with this hypothesis, competitive antagonists to CRF-BP enhance performance on learning and memory tasks . Although studies have documented alteration of the CRF system, clinical correlations and isolation of molecular intermediates are still needed to identify the precise consequences of these changes during AD.
Another crucial variable in the stress cascade are glucocorticoids (GC), dominant stress hormones that are released as a result of CRF signaling and that mediate physiological changes during stress. The idea that excessive amounts of GC could be causal in endangering hippocampal neurons was first suggested in the 1970s, with the demonstration that the magnitude of GC hypersecretion could predict the extent of neuronal loss in the hippocampus in aged rats (reviewed in [36,37]). These studies provided the foundation for subsequent studies demonstrating direct links between GC exposure and neurotoxicity (reviewed in ). In several studies where GC concentrations have been raised into the stress range for extended periods, significant hippocampal neuron loss has been observed, particularly in the pyramidal cells of CA3 . In concordance with these rodent studies, increased circulating levels of GC have been reported with increasing age, are a prominent feature of AD, and may play a prominent role in AD pathology [36,37,40,41]. Although studies on changes in the CRF system and the ability of excess GC to endanger hippocampal neurons have provided important groundwork for identifying the role of stress in AD, the specific molecular intermediates connecting stress to development of neuropathology have remained elusive.
Another important link between the stress axis and AD came with the discovery that exposure to physiological [4–10] and psychological stress could induce tau-P in rodents . Because the circuitry and patterns of neuronal activity that mediate responses to physiological and psychological stressors differ [44– 47], the mechanisms underlying tau phosphorylation induced by different stress models have not yet been deciphered thoroughly. For example, both physiological and psychological stress can induce tau-P independently of glucocorticoid secretion [4,11], although most stressors have not been tested. Other mechanisms underlying stress-induced tau-P have been suggested to involve modulation of temperature sensitive tau phosphatases [8,9], which is particularly relevant to strong physiological stressors such as starvation, cold-water stress, and heat shock. In addition, this mechanism has been used to explain tau-P seen during hibernation  or exposure to anesthesia [49,50]. Recent studies that have found that dexamethasone treatment and/or stress exposure can increase Aβ production and tau accumulation in transgenic mouse models of AD [42,43].
Although the data thus far appears to provide a convincing case for a temperature-dependent mechanism for some stressors, more mild stressors appear to have distinct underlying mechanism that may not be directly tied to the kinase-phosphatase balance. In the case of psychological stress, such as restraint, genetic and pharmacologic studies of CRFRs have demonstrated that CRFR1 is required for stress-induced tau-P and that CRFRs regulate tau kinases and stress-induced tau-P in an opposing manner . The potential AD relevance of stress-induced tau-P was also investigated, with studies investigating the role of repeated (chronic) emotional stress on tau-P and tau solubility. When a repeated stress paradigm was administered, stress-induced tau-P was not seen to be transient and resulted in sustained increases in tau-P even 24 h after the final stress exposure . Furthermore, as a result of repeated stress, portions of phosphorylated tau were also sequestered to detergent-soluble fractions (cf acute stress), which is where the bulk of PHFs in the AD brain are found [51,52]. Whether or not the development of detergent-soluble tau species as a result of stress exposure can lead to toxicity and contribute to pathology is currently under investigation.
Because known genetic links comprise a low percentage of AD cases, identifying potential environmental contributors in AD is critical. Although age is considered the primary risk factor for AD, numerous studies have also implicated chronic stress in this regard, with changes in stress steroids and CRF system intermediates being altered early in AD progression. These studies, taken together with the recently discovered relationship between the CRF system in stress-induced tau-P, provide a more complete picture of the potential mechanisms linking stress and AD endpoints.
From the relevant literature, there seems little question that stress can be an exacerbating influence in combination with genetic and other factors (e.g., aging) but the limits of this have not been explored. Although it remains to be determined whether prolonged stress exposure can lead to bona-fide tau pathology, recent data demonstrating tau-P and solubility changes with repeated stress exposure certainly begs the question of whether an animal can be stressed into senility. Given that the balance between stress-induced tau-P and development of insoluble tau aggregates may be quite delicate, using stress-induced tau-P as a model to examine these dynamics may prove useful for further determining the boundaries between of the relationship between stress, AD tau pathology and neuroplasticity.
Supported by the National Institute on Aging AG032755, the Alzheimer's Disease Cooperative Study, and UCSD School of Medicine. Thanks to A. Roe, Dr. M. Staup and Dr. E. Masliah (UCSD) for critical reading of the manuscript.
Authors’ disclosures available online (http://www.jalz.com/disclosures/view.php?id=66).