Abstract

Although the initial events of Alzheimer's disease (AD) are still not known, it is clear that the disease in its sporadic form results from the combination of genetic and environmental risk factors. Among the latter, behavioral stress has been increasingly recognized as an important factor in the propagation of AD. However, the mechanisms underlying this modulation remain to be fully investigated. Since stress up-regulates the ALOX5 gene product, 5-lipoxygenase (5LO), herein we investigated its role in modulating stress-dependent development of the AD phenotype. To reach this goal, triple transgenic (3xTg) mice and 3xTg genetically deficient for 5LO were investigated after undergoing a restraint/isolation paradigm. In the present paper, we found that 28 days of restraint/isolation stress worsened tau phosphorylation and solubility, increased glycogen synthase kinase 3β activity, compromised long-term potentiation and impaired fear-conditioned memory recall in 3xTg animals, but not in 3xTg animals lacking 5LO (3xTg/5LO−/−). These results highlight the novel functional role that the ALOX5 gene plays in the development of the biochemical, electrophysiological and behavioral sequelae of stress in the AD context. They provide critical support that this gene and its expressed protein are viable therapeutic targets to prevent the onset or delay the progression of AD in individuals exposed to this risk factor.

INTRODUCTION

Alzheimer's disease (AD) is an aging-associated neurodegenerative dementia classically defined by the presence of amyloid beta (Aβ) plaques and neurofibrillary tau tangles in the brain. While familial AD mutations have been discovered in the Aβ precursor protein (APP) and in proteins that cleave APP to produce Aβ, the overwhelming majority of those who develop AD lack such mutations. Thus, environmental factors that accelerate the pathogenesis of AD present an attractive target for intervention to delay or prevent pathology (1). Recent evidence has implicated stress as an environmental factor involved in the advancement the AD phenotype. Higher plasma or urinary cortisol in late life confers greater risk for AD development and a positive correlation exists between stress hormone levels and rate of AD progression (2–4). In animal models of AD, stress, either behaviorally or pharmacologically via stress hormones, worsens and accelerates the AD phenotype (5–8). However, the mechanisms by which stress modulates the onset and development of AD pathogenesis remain to be fully investigated. One of the actions of glucocorticoids in neuronal cells includes up-regulation of the ALOX5 gene product, the 5-lipoxygenase protein (5LO), a key enzyme for leukotriene biosynthesis (9–11). We have previously reported that genetic absence of 5LO protects against tauopathy resulting from a short course of dexamethasone treatment (12). However, stress leading to AD development progression in humans is likely due to chronic rather than acute exposure. Moreover, while dexamethasone is very specific to the glucocorticoid receptor, it fails to mimic the fuller repertoire of the effects of corticosterone, the endogenous corticosteroid in mice. To address these issues, we subjected triple transgenic AD mice (3xTg) which develop Aβ plaques and tangles, and 3xTg mice lacking 5LO (3xTg/5LO−/−) to 28 days of chronic restraint/isolation (R/I). We found that R/I stress was able to interfere with memory recall, induce pathologic phosphorylated tau species and impair long-term potentiation. This effect was not seen in 3xTg animals that lacked 5LO. Our work is the first to report that 5LO plays a functional role in a physiologically relevant model of chronic stress using transgenic AD animals.

RESULTS

R/I stress impairs fear-conditioned recall in 3xTg but not 3xTg/5LO−/− animals

We subjected both 3xTg and 3xTg/5LO−/− animals to R/I stress for 28 days, after which they were assessed behaviorally in the fear conditioning and Y-maze paradigms. The fear conditioning paradigm has been widely used to assess long-term memory in rodents, and the Y-maze paradigm has been used as a surrogate of working, or short-term memory. As shown in Figure 1 we found no change in the Y-maze paradigm both in number of entries or alternation percentage, regardless of presence of 5LO or exposure to R/I stress. In contrast, in the fear conditioning paradigm, while all animal groups displayed similar behavior in the training phase (data not shown), we found that R/I stress impaired both contextual and cued recall in 3xTg animals as shown in Figure 1C and D. In contrast, R/I-induced memory impairment was not seen in 3xTg/5LO−/− animals in either contextual or cued recall, with 3xTg, 3xTg/5LO−/− and 3xTg/5LO−/− R/I animals displaying similar levels of freezing.

Figure 1.

Genetic absence of ALOX5 rescues R/I-induced memory impairments in 3xTg mice. 3xTg (n = 6), 3xTg R/I stressed (n = 7), 3xTg/5LO−/− (n = 4) and 3xTg/5LO−/− R/I stressed (n = 5) animals were assayed for Y-maze total arm entries (A) and alternation behavior (B), as well as fear-conditioned contextual (C) and cued (D) recall. Results are mean ± SEM (*P < 0.05).

Figure 1.

Genetic absence of ALOX5 rescues R/I-induced memory impairments in 3xTg mice. 3xTg (n = 6), 3xTg R/I stressed (n = 7), 3xTg/5LO−/− (n = 4) and 3xTg/5LO−/− R/I stressed (n = 5) animals were assayed for Y-maze total arm entries (A) and alternation behavior (B), as well as fear-conditioned contextual (C) and cued (D) recall. Results are mean ± SEM (*P < 0.05).

Genetic absence of ALOX5 prevents stress-induced increases in plasma corticosterone

To assess whether our behavioral paradigm altered circulating corticosteroid levels in our model system, we used a sensitive and specific ELISA to assay for corticosterone. R/I stress elevated circulating plasma corticosterone levels in 3xTg animals, but failed to do so in 3xTg/5LO−/− animals as shown in Figure 2A. However, no differences were found in any genotype or treatment group with regard to brain levels of the glucocorticoid receptor as shown in Figure 2B. Since glucocorticoid administration in vitro has been reported to increase leukotriene B4 levels (11), we assayed brain levels of leukotriene B4 (LTB4) using a sensitive and specific ELISA. We found LTB4 to be elevated in the brains R/I stressed 3xTg animals, however, as expected, no detectable levels of LTB4 were found in 3xTg animals lacking 5LO as shown in Figure 2C.

Figure 2.

Genetic absence of ALOX5 blunts R/I-dependent hormone stress and Aβ peptides elevation in 3xTg mice. (A) Levels of plasma corticosterone was assayed using a sensitive and specific ELISA in 3xTg (n = 6), 3xTg R/I stressed (n = 7), 3xTg/5LO−/− (n = 4) and 3xTg/5LO−/− R/I stressed (n = 5) animals. (B) Brain cortex levels of glucocorticoid receptor; and (C) LTB4 levels in 3xTg, 3xTg receiving 28 days R/I, 3xTg5LO−/− and 3xTg5LO−/− receiving 28 days R/I. Levels of soluble (D) Aβ40 and (E) Aβ42 were assayed in brain cortex RIPA lysate using a sensitive and specific ELISA in 3xTg (n = 6), 3xTg R/I stressed (n = 7), 3xTg/5LO−/− (n = 4), and 3xTg/5LO−/− R/I stressed (n = 5) animals. Results are mean ± SEM (*P < 0.05).

Figure 2.

Genetic absence of ALOX5 blunts R/I-dependent hormone stress and Aβ peptides elevation in 3xTg mice. (A) Levels of plasma corticosterone was assayed using a sensitive and specific ELISA in 3xTg (n = 6), 3xTg R/I stressed (n = 7), 3xTg/5LO−/− (n = 4) and 3xTg/5LO−/− R/I stressed (n = 5) animals. (B) Brain cortex levels of glucocorticoid receptor; and (C) LTB4 levels in 3xTg, 3xTg receiving 28 days R/I, 3xTg5LO−/− and 3xTg5LO−/− receiving 28 days R/I. Levels of soluble (D) Aβ40 and (E) Aβ42 were assayed in brain cortex RIPA lysate using a sensitive and specific ELISA in 3xTg (n = 6), 3xTg R/I stressed (n = 7), 3xTg/5LO−/− (n = 4), and 3xTg/5LO−/− R/I stressed (n = 5) animals. Results are mean ± SEM (*P < 0.05).

Aβ levels are lower in animals that lack ALOX5 and are not influenced by R/I stress

Administration of stress-inducing paradigms or corticosteroids has increased Aβ species in some AD mouse models but not in others (5–8). To investigate whether Aβ was influenced in our mouse models under our experimental conditions, we assayed brain levels of Aβ40 and Aβ42using cortical lysate in 3xTg and 3xTg/5LO−/− animals that underwent R/I stress. As shown in Figure 2D and E, we found that Aβ was not elevated by R/I stress in either 3xTg or 3xTg/5LO−/− animals. However, we observed that Aβ was lower in 3xTg/5LO−/− animals than in 3Tg animals, confirming a finding that we have previously reported (12).

Paired helical filament tau levels are elevated in stressed 3xTg but not 3xTg/5LO−/− animals

In addition to Aβ, stress has been shown to modulate tau pathology in rodent models of AD. To see how tau could be modulated in our studies, we used radioimmunoprecipitation assay (RIPA)-soluble fractions of brain homogenate to assay for soluble tau species. As shown in Figure 3A and B, we found no differences among animal groups in total tau, as probed by the human tau HT7 antibody. Additionally, we could not detect any differences in the phosphorylation of tau at threonine 231 and 235, as recognized by the antibody AT180. However, when probing with PHF1 and PHF13 antibodies, which recognize tau phosphorylated at serine 396/serine 404 and serine 396, respectively, we found that chronic behavioral stress elevated the steady-state levels of these phosphoepitopes above that of 3xTg control animals. In contrast, 3xTg/5LO−/− even though they underwent the same R/I stress displayed no such elevation. Since the elevation of PHF1 and PHF13 epitopes signify advanced phosphorylation, we also investigated whether insoluble tau aggregation had begun in 3xTg animals that had undergone R/I stress. Using the FA extractable fractions of the brain homogenates, we assayed for HT7 tau. As shown in Figure 3C, we found that FA-soluble tau was elevated in R/I stressed 3xTg animals, while it was barely detectable in all other groups.

Figure 3.

Absence of ALOX5 prevents tau hyper-phosphorylation in R/I stressed 3xTg mice. (A). Representative western blot analysis of cortical RIPA lysates of 3xTg (n = 6), 3xTg R/I stressed (n = 7), 3xTg/5LO−/− (n = 4) and 3xTg/5LO−/− R/I stressed (n = 5) animals immunoblotted for total tau (HT7), phosphorylated tau at Th231/Th235 (AT180), at Ser396/Ser404 (PHF1) and at Ser396 (PHF13) tau species. (B). Densitometric analyses of the immunoreactivities shown in the previous panel. The densitometry for phosphorylated tau is and expressed a ratio with total soluble tau (HT7). (C). Representative western blot analysis for formic acid extractable tau in brain cortex lysates form 3xTg, 3xTg R/I, 3xTg5Lo−/− and 3xTg5LO−/− R/I. Results are mean ± SEM (*P < 0.05).

Figure 3.

Absence of ALOX5 prevents tau hyper-phosphorylation in R/I stressed 3xTg mice. (A). Representative western blot analysis of cortical RIPA lysates of 3xTg (n = 6), 3xTg R/I stressed (n = 7), 3xTg/5LO−/− (n = 4) and 3xTg/5LO−/− R/I stressed (n = 5) animals immunoblotted for total tau (HT7), phosphorylated tau at Th231/Th235 (AT180), at Ser396/Ser404 (PHF1) and at Ser396 (PHF13) tau species. (B). Densitometric analyses of the immunoreactivities shown in the previous panel. The densitometry for phosphorylated tau is and expressed a ratio with total soluble tau (HT7). (C). Representative western blot analysis for formic acid extractable tau in brain cortex lysates form 3xTg, 3xTg R/I, 3xTg5Lo−/− and 3xTg5LO−/− R/I. Results are mean ± SEM (*P < 0.05).

R/I stress elevation of GSK3β kinase activity depends on the presence of ALOX5

In order to understand the underlying mechanism of how R/I stress led to an elevation of phosphorylated tau isoforms, we investigated the major tau-associated kinases previously reported to act on tau in vivo. We used brain homogenates to probe for glycogen synthase kinase 3 β (GSK3β), cyclin-dependent kinase 5 (cdk5), as well as its co-activators p35 and p25. As shown in Figure 4A, we found that steady-state levels of these kinases were unchanged. To see how this steady-state protein levels were reflected in their enzymatic activity, next we performed ex vivo kinase activity assays for both kinases. As shown in Figure 4B, we found that GSK3β activity was elevated in the brains of 3xTg animals that were subjected to R/I stress but not in the brains of 3xTg/5LO−/− animals. In contrast, cdk5 activity was unchanged by genotype or treatment. Since 5LO knockout protected against stress-induced AD neuropathology as well as behavioral impairments, we assessed the influence of genetic absence of 5LO on synaptic integrity in 3xTg mice that had been R/I stressed. We assayed levels of the synaptic proteins synaptophysin (SYP) and postsynaptic density protein 95 (PSD95) and found no change in their levels as a function of genotype or treatment as shown in Figure 4C and D.

Figure 4.

Tau metabolism in 3xTg and 3xTg5LO−/− mice undergoing chronic R/I stress. (A) Representative western blot analysis of brain cortex lysates of 3xTg (n = 6), 3xTg R/I stressed (n = 7), 3xTg/5LO−/− (n = 4) and 3xTg/5LO−/− R/I stressed (n = 5) probed for GSK3β, pGSK3β, cdk5, p35 and p25. (B) GSK3β and cdk5 kinase ex vivo activities in brain cortex lysates from 3xTg, 3xTg undergoing R/I, 3xTg 5LO−/− and 3xTg5LO−/− undergoing R/I. (C). Representative western blot analyses for synaptophysin (SYP) and PSD95 in brain cortex homogenates from 3xTg (n = 6), 3xTg R/I stressed (n = 7), 3xTg/5LO−/− (n = 4) and 3xTg/5LO−/− R/I stressed (n = 5) animals. Results are mean ± SEM (*P < 0.05).

Figure 4.

Tau metabolism in 3xTg and 3xTg5LO−/− mice undergoing chronic R/I stress. (A) Representative western blot analysis of brain cortex lysates of 3xTg (n = 6), 3xTg R/I stressed (n = 7), 3xTg/5LO−/− (n = 4) and 3xTg/5LO−/− R/I stressed (n = 5) probed for GSK3β, pGSK3β, cdk5, p35 and p25. (B) GSK3β and cdk5 kinase ex vivo activities in brain cortex lysates from 3xTg, 3xTg undergoing R/I, 3xTg 5LO−/− and 3xTg5LO−/− undergoing R/I. (C). Representative western blot analyses for synaptophysin (SYP) and PSD95 in brain cortex homogenates from 3xTg (n = 6), 3xTg R/I stressed (n = 7), 3xTg/5LO−/− (n = 4) and 3xTg/5LO−/− R/I stressed (n = 5) animals. Results are mean ± SEM (*P < 0.05).

R/I stress long-term potentiation deficit in 3xTg animals is modulated by ALOX5

To assess synaptic plasticity, next we studied synaptic function by electrophysiological studies. Using hippocampal slices, we first established that basal synaptic transmission was similar between animals and treatment groups by generating input/output curves and measuring field excitatory postsynaptic potentials (fEPSPs) elicited in the CA1 by stimulating Schaffer collaterals as shown in Figure 5A. We also examined paired-pulse facilitation to assess for short-term plasticity as shown in Figure 5B. We found no differences among groups in either parameter. We then assessed long-term potentiation over 90 min as shown in Figure 5C–E, and found that compared with 3xTg baseline, R/I-stressed 3xTg animals displayed a reduced LTP response. In contrast, 3xTg/5LO−/− animals had an augmented LTP response but this was reduced to 3xTg baseline when subjected to R/I stress.

Figure 5.

Genetic absence of ALOX5 rescues synaptic dysfunction in 3xTg mice undergoing chronic R/I stress. For electrophysiology studies hippocampal slices were used (n = number of slices/number of animals; 3xTg, n = 13/3; 3xTg R/I stress, n = 13/3; 3xTg/5LO−/−, n = 14/3; 3xTg/5LO−/− R/I stress, n = 14/3). (A) Input/output curves and representative fEPSPs at increasing stimulus strengths. (B) Mean fEPSPs slopes as a function of interpulse interval between first and second fEPSPs evoked at CA3–CA1 synapses in slices from the same mice at 20, 50, 100, 200 and 1000 ms in the same animals. (C). 90 min recordings of fEPSP expressed as percentage of 3xTg pre-tetanus baseline. (D). LTP magnitudes at 0–10 min post-tetanus and at 80–90 min (E) showing reduced fEPSPs in 3xTg R/I stressed animals compared with all animal groups (*), and elevated fEPSPs in 3xTg/5LO−/− animals compared with all groups (#), with similar levels of fEPSPs in unstressed 3xTg and stressed 3xTg/5LO−/− animals. Results are mean ± SEM (*P < 0.05).

Figure 5.

Genetic absence of ALOX5 rescues synaptic dysfunction in 3xTg mice undergoing chronic R/I stress. For electrophysiology studies hippocampal slices were used (n = number of slices/number of animals; 3xTg, n = 13/3; 3xTg R/I stress, n = 13/3; 3xTg/5LO−/−, n = 14/3; 3xTg/5LO−/− R/I stress, n = 14/3). (A) Input/output curves and representative fEPSPs at increasing stimulus strengths. (B) Mean fEPSPs slopes as a function of interpulse interval between first and second fEPSPs evoked at CA3–CA1 synapses in slices from the same mice at 20, 50, 100, 200 and 1000 ms in the same animals. (C). 90 min recordings of fEPSP expressed as percentage of 3xTg pre-tetanus baseline. (D). LTP magnitudes at 0–10 min post-tetanus and at 80–90 min (E) showing reduced fEPSPs in 3xTg R/I stressed animals compared with all animal groups (*), and elevated fEPSPs in 3xTg/5LO−/− animals compared with all groups (#), with similar levels of fEPSPs in unstressed 3xTg and stressed 3xTg/5LO−/− animals. Results are mean ± SEM (*P < 0.05).

DISCUSSION

In this study, we found that 3xTg AD mice exposed to 28 days R/I stress paradigm manifest significant memory impairments and that the genetic absence of ALOX5 protected them from this effect. In addition, R/I stress significantly worsened tau phosphorylation, increased the amount of insoluble tau, which associated with an increased activity of GSK3β kinase, and that all of these effects were not seen in stressed 3xTg animals genetically deficient for ALOX5. Finally, we found that R/I stress significantly impaired LTP in 3xTg mice, but 3xTg/5LO−/− animals undergoing R/I had a LTP response indistinguishable from 3xTg LTP without stress. Taken together these results highlight the novel functional role that ALOX5 plays in the development of the biochemical, electrophysiological and behavioral sequelae of stress in the AD context.

Consistent evidence shows that in its sporadic form, AD results from the interaction between genetic and environmental risk factors (1). Therefore the identification and mitigation of risk factors that occur earlier in life are appealing for the purposes of therapeutic intervention and prevention. Environmental behavioral stressors can be unavoidable and omnipresent, and therefore have potential to dramatically alter AD risk. While many preclinical studies have clearly demonstrated a negative influence of stress on AD progression, the underlying mechanisms remain elusive. Though stress hormones have a multitude of central nervous system effects, one of their actions is to modulate the ALOX5 pathway (11). Its product, the 5LO, has been shown to be a crucial molecular contributor to the AD phenotype: overexpression of 5LO is linked to augmentation of Aβ plaque and neurofibrillary tau development while knockout or pharmacologic inhibition is linked to reduction in pathology (13,14). A 10-day course of corticosterone administration increases message and steady-state protein levels of 5LO in the rat brain, while antagonism of the glucocorticoid receptor reduces both in vitro (9,10). We have previously shown that genetic absence of 5LO prevents a short course of dexamethasone (7 days, 5 mg/kg) from advancing pathogenic tau phosphorylation (12). However in that study, we were unable to detect any memory impairment caused by our treatment in 3xTg mice, and thus were unable to establish that ALOX5 knockout-associated protection against tau phosphorylation had any behavioral relevance. In the present study, genetic deletion of ALOX5 not only blunted tau phosphorylation but also stress-mediated memory insult and synaptic impairment, further establishing its importance to an early stress-induced exacerbation of the AD phenotype.

While we anticipated that this gene knockout would reduce brain Aβ levels, we were surprised to find that our stressor did not increase levels of Aβ in either 3xTg animals as has been reported by others, most notably by Green et al. (15) who first showed stress-mediated pathology in the triple transgenic mouse. Although this discrepancy is very likely attributable to differences in stress paradigm, animal gender, extraction methods and tissue selection for homogenization, it is worthwhile to note that R/I stress has been unable to change Aβ in other mouse models of AD (13). By contrast, we clearly found that ALOX5 gene knockout reduced stress-mediated advanced phosphotau isoforms formation. This dichotomy is interesting given that amyloid β peptide is hypothesized to be the initiating factor in tau phosphorylation and synaptic dysfunction, and brings forward two potential explanations which may not be mutually exclusive. One explanation is that a reduction in soluble Aβ caused by knockout of ALOX5 precludes stress-induced tau phosphorylation, while another is that stress may affect Aβ and tau through different pathways. Since there are reports where certain stressors do not alter Aβ but hyperphosphorylate tau at AD-specific epitopes, the response of Aβ machinery to stress is probably stressor-specific, while tau hyperphosphorylation is probably a more general response to stress. However, further experimentation is necessary to dissect exactly which stressors require 5LO to modulate the AD phenotype and which do not. As evidence of such potential specificity, knockout of ALOX5 does not protect against tau phosphorylation induced by lipopolysaccharide exposure in transgenic AD models, but does protect against lipopolysaccharide-induced elevation of Aβ-forming machinery (16).

Almost all studies on 5LO in AD have focused on its importance in middle and late life in AD transgenic mice (∼8–16 months of age). While this age range represents a time when significant plaque and tangle burden can be appreciated, several reports have linked synaptic impairment preceding significant deposition of pathology in 3xTg mice (as well as most other transgenic AD animals) (17,18). Even before notable disruptions in pre- and postsynaptic markers manifest, LTP is altered in AD (19). At advanced ages we have found that ALOX5 gene knockout protects against loss of both synaptic markers as well as LTP but this is the first report of LTP-sparing effects of 5LO in the absence of synaptic marker loss (14,20). Interestingly, unstressed 3xTg/5LO−/− animals had enhanced LTP responses compared with unstressed 3xTg animals, which could imply that 5LO disruption may not only aid in neuroprotection but could also be used in synaptic integrity augmentation. Future work investigating the importance of 5LO in other disease states involving abnormal neuronal functioning and synaptic plasticity are warranted to provide further clarity on this issue.

Parenthetically, while our study on ALOX5 has highlighted its role in AD, ALOX5 has also been increasingly recognized in other complex neurological and psychiatric contexts, including in anxiety and depression in rodents (21–23). Since in late AD there can be behavioral changes associated with widespread cortical and subcortical insult, investigation of the role of ALOX5 with regard to these facets of advanced AD pathology would also be intriguing.

In summary, this report presents experimental evidence of the functional role that the ALOX5 plays in advancing the AD phenotype induced by a common environmental stressor. It provides preclinical rational for targeting this pathway in individuals that are vulnerable to this environmental factor, and more broadly, as further evidence to employ 5LO blockers therapies as preventatives for AD.

MATERIALS AND METHODS

Animals and restraint/isolation stress paradigm

All animal procedures were approved by the Animal Care and Usage Committee, in accordance with the US NIH guidelines. The 3xTg mice harboring mutations in APP, PS1 and tau, first described by Oddo et al.(24) were backcrossed 10 times with animals lacking 5LO−/− of the same genetic background to produce founder 3xTg/5LO−/− animals (25). Naïve female 3xTg and 3xTg/5LO−/− animals aged 4–4.5 months were used for our studies (3xTg, n = 6; 3xTg R/I stress, n = 7; 3xTg/5LO−/−, n = 4; 3xTg/5LO−/− n = 5). All animals were housed on a 12-h light/dark cycle in the Medical Research Building at the Temple University Health Science Campus, which is accredited by the Association for the Assessment and Accreditation of Laboratory Animal Care International. Standard mouse chow and water were provided ad libitum. Separate groups of mice were used for biochemical and electrophysiological analyses. Restraint/isolation (R/I) stress was administered as described by Carroll et al. (26) by placing mice in a well-ventilated chamber (12 cm length × 3 cm diameter), which prevented locomotion. Mice were placed into this chamber and singly housed 60 m/days, 6 days/week for 4 weeks. After the conclusion of R/I stress, mice were placed back into home cages. Body weight of animals was monitored at weekly intervals during our studies and was not found to deviate from the weight of unstressed controls. A week after completion of the behavioral testing peripheral blood was collected and mice were sacrificed. At sacrifice, animals were perfused with ice-cold 0.9% phosphate buffered saline containing 10 mm ethylenediaminetetraacetic acid (EDTA), brains immediately removed and stored at −80°C.

Behavioral paradigms

Following R/I stress, 3xTg and 3xTg/5LO−/− animals were assayed for memory using the Y-maze and fear conditioning paradigm as previously reported (12). Briefly, the Y-maze paradigm consists of three arms 32 cm long × 10 cm wide with 26 cm walls (San Diego Instruments, San Diego, CA, USA). Animals were placed in the central portion of the maze and allowed to freely explore for 5 min. Entries into each arm were recorded using a video camera and percent alternation was derived using the formula: total number of alternations/(total entries—2). An alternation was defined as three consecutive arm entries into different arms. For fear conditioning, tests were conducted in a fear conditioning chamber with black methacrylate walls, a speaker, and grid floor (Start Fear Systems; Harvard Apparatus, Holliston, MA, USA). Animals were trained to associate a neutral tone with a foot shock and freezing behavior was recorded. No difference in this baseline freezing behavior was found (data not shown). For the contextual recall, animals were placed back in the chamber without foot shocks or neutral tones. For cued recall, animals were placed back in the chamber with modified walls, lighting, and scent with only the neutral tone being sounded.

Biochemical assays

For all biochemical studies, cortical lysates were used

Mouse brain homogenates were sequentially extracted first in RIPA buffer containing EDTA-free protease inhibitor (Roche, Indianapolis, IN, USA) and phosphatase inhibitor (Thermo Fisher, Morris Plains, NJ, USA) for the soluble fraction, and the pellet in formic acid (FA) for the insoluble fraction as previously described (13). Aβ40 and Aβ42 levels were assayed by using a sensitive sandwich enzyme-linked immunosorbent assay (ELISA) kit (Wako Chemicals, Richmond, VA, USA) in accordance with the manufacturer's protocols. Plasma levels of corticosterone and leukotriene B4 were assayed using a sensitive sandwich enzyme-linked immunosorbent assays (ELISA) kit (Enzo Life Science, Farmingdale, NY, USA) in accordance with the manufacturer's protocols.

Immunoblotting

Extracts from brain homogenates were electrophoretically separated using 10% Bis–Tris gels according to the molecular weight of the target molecule and then transformed onto nitrocellulose membranes (Bio-Rad, Life Science Research; Hercules, CA, USA) as previously described (8). Membranes were blocked with Odyssey blocking buffer and incubated with primary antibodies overnight at 4°C. After three washes in TBS-T, membranes were incubated with IRDye secondary antibodies (Li-COR, Lincoln, NE, USA). Actin was always used as a loading control. Primary antibodies and dilutions used in the study are shown in Table 1.

Table 1.

Antibodies used in the study.

Antibody Immunogen Host Dilution Source 
Glucocorticoid receptor (GR) Partially purified rat glucocorticoid receptor Mouse 1:1000 Pierce 
HT-7 Amino acids 159–163 of human tau Mouse 1:750 Pierce 
AT-180 Partially purified phospho-T231/S235 tau Mouse 1:200 Pierce 
PHF-13 Partially purified phospho-Ser396 tau Mouse 1:200 Cell signaling 
PHF-1 Purified phospho-Ser396/S404 tau Mouse 1:200 Dr P. Davies 
Synaptophysin (SYP) Amino acids 221–313 of synaptophysin of human origin Mouse 1:200 Santa Cruz 
Postsynaptic density protein 95 (PSD95) Synthetic human PSD95 Rabbit 1:200 Cell signaling 
GSK3α/β Amino acids 1–420 full length GSK-3β of Xenopus origin Mouse 1:200 Santa Cruz 
p-GSK3α/β Amino acid Ser21 of human GSK-3a. Rabbit 1:200 Cell signaling 
Cdk5 C-terminus of Cdk5 of human origin Rabbit 1:200 Santa Cruz 
P35/25 C-terminus of p35/25 of human origin Rabbit 1:200 Santa Cruz 
Actin C-terminus of Actin of human origin Goat 1:1 000 Santa Cruz 
Antibody Immunogen Host Dilution Source 
Glucocorticoid receptor (GR) Partially purified rat glucocorticoid receptor Mouse 1:1000 Pierce 
HT-7 Amino acids 159–163 of human tau Mouse 1:750 Pierce 
AT-180 Partially purified phospho-T231/S235 tau Mouse 1:200 Pierce 
PHF-13 Partially purified phospho-Ser396 tau Mouse 1:200 Cell signaling 
PHF-1 Purified phospho-Ser396/S404 tau Mouse 1:200 Dr P. Davies 
Synaptophysin (SYP) Amino acids 221–313 of synaptophysin of human origin Mouse 1:200 Santa Cruz 
Postsynaptic density protein 95 (PSD95) Synthetic human PSD95 Rabbit 1:200 Cell signaling 
GSK3α/β Amino acids 1–420 full length GSK-3β of Xenopus origin Mouse 1:200 Santa Cruz 
p-GSK3α/β Amino acid Ser21 of human GSK-3a. Rabbit 1:200 Cell signaling 
Cdk5 C-terminus of Cdk5 of human origin Rabbit 1:200 Santa Cruz 
P35/25 C-terminus of p35/25 of human origin Rabbit 1:200 Santa Cruz 
Actin C-terminus of Actin of human origin Goat 1:1 000 Santa Cruz 

Kinase activity assays

GSK3β and cdk5 kinase activities were carried out as previously described (12,13). Briefly, GSK3β and cdk5 were immunoprecipitated from brain lysate using respective antibodies. The immunoprecipitates were washed with lysis buffer followed by HBS (10 mm HEPES (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid), pH 7.4, 150 mm NaCl). Activity of GSK3β was determined using phosphor-glycogen synthase peptide 2 (Millipore, Billerica, MA, USA) while cdk5 activity was determined using histone H1 (Santa Cruz Biotechnology, Dallas, TX, USA). After incubation in HBS buffer containing 15 mm MgCl2, 50 mm ATP, 1 mm dithiothreitol and 1 µCi of P32 ATP. After 30 min of incubation at 37°C, a liquid scintillation counter was used to measure reaction products.

Electrophysiology

Electrophysiological experiments were performed as previously described (14). Briefly, a separate group of 4–4.5-month old mice were used in this study (n = number of slices/number of animals; 3xTg, n = 13/3; 3xTg R/I stress, n = 13/3; 3xTg/5LO−/−, n = 14/3; 3xTg/5LO−/− R/I stress, n = 14/3). Mice were sacrificed by rapid decapitation and brains placed in ice-cold artificial cerebrospinal fluid (ACSF) in which sucrose (248 mm) was substituted for NaCl. Hippocampal slices (400 µm) were cut using a vibratome and placed in ACSF with bubbled 95% O2/5% CO2 at room temperature to recover for 1 h. Slices were transferred to a recording chamber and continuously perfused with ACSF maintained at 32–34°C. fEPSPs were recorded from the CA1 stratum radiatum using an extracellular glass pipette (3–5 MΩ) filled with ACSF. Schaffer collateral/commissural fibers in the stratum radiatum were stimulated with a bipolar tungsten electrode placed 200–300 µm from the recording pipette. Stimulation intensities were chosen to produce a fEPSP at 1/3 of the maximum amplitude based on an input/output (I/O) curve using stimulations of 0–300 µA. Paired-pulse facilitation experiments were performed using a pair of stimuli of the same intensity delivered 20, 50, 100, 200 and 1000 ms apart. Baseline was recorded for 20 min before tetanization with pulses every 30 s. Long-term potentiation (LTP) at CA3–CA1 synapses was induced by four trains of 100 Hz stimulation delivered in 20 s intervals. Recordings were made 30 s for 90 min following tetanization. The fEPSP rise/slope between 30 and 90% was measured offline using Clampfit 10.3 (Molecular Devices, Sunnyvale, CA, USA) and normalized to the mean rise/slope of the baseline. All tests were performed by an experimenter who was unaware of genotype or treatment.

Data analysis

One-way ANOVA with Bonferroni post hoc multiple comparison tests were performed using Prism 5.0 (GraphPad Software, La Jolla, CA, USA). All data are presented as mean + SEM. Significance was set at P < 0.05

FUNDING

This work was in part supported by grants from the National Institute on Health (NIH) AG33568 to D.P.; P30 DA13429 and T32 DA07237 to the Center for Substance Abuse Research. Additional funding was provided by a grant from the Alzheimer′s Art Quilt Initiative to D.P.

ACKNOWLEDGEMENTS

We would like to thank Dr Peter Davis for the generous gift of the PHF-1 antibody used in the study. We are thankful to Dr Jin-Guo Li for technical assistance with the LTB4 assay.

Conflict of Interest statement. The authors have no financial disclosures to declare. All authors report no actual or potential conflicts of interest.

REFERENCES

1
Villemagne
V.L.
Pike
K.E.
Chetelat
G.
Ellis
K.A.
Mulligan
R.S.
Bougeat
P.
Ackermann
U.
Jones
G.
Szoeke
C.
Salvado
O.
, et al.  . 
Longitudinal assessment of Aβ and cognition in aging and Alzheimer's disease
Ann. Neurol.
 , 
2011
, vol. 
69
 (pg. 
181
-
192
)
2
Csernansky
J.G.
Dong
H.
Fagan
A.M.
Wang
L.
Xiong
C.
Holtzman
D.M.
Morris
J.C.
Plasma cortisol and progression of dementia in subjects with Alzheimer-type dementia
Am. J. Psychiatry
 , 
2006
, vol. 
163
 (pg. 
2164
-
2169
)
3
Huang
C.W.
Lui
C.C.
Chang
W.N.
Lu
C.H.
Wang
Y.L.
Chang
C.C.
Elevated basal cortisol level predicts lower hippocampal volume and cognitive decline in Alzheimer's disease
J. Clin. Neurosci.
 , 
2009
, vol. 
16
 (pg. 
1283
-
1286
)
4
Wilson
R.S.
Arnold
S.E.
Schneider
J.A.
Kelly
J.F.
Tang
Y.
Bennett
D.A.
Chronic psychological distress and risk of Alzheimer's disease in old age
Neuroepidemiology
 , 
2006
, vol. 
27
 (pg. 
143
-
153
)
5
Jeong
Y.H.
Park
C.H.
Yoo
J.
Shin
K.Y.
Ahn
S.M.
Kim
H.S.
Lee
S.H.
Emson
P.C.
Suh
Y.H.
Chronic stress accelerates learning and memory impairments and increases amyloid deposition in APPV717I-CT100 transgenic mice, an Alzheimer's disease model
FASEB J.
 , 
2006
, vol. 
20
 (pg. 
729
-
731
)
6
Kang
J.E.
Cirrito
J.R.
Dong
H.
Csernansky
J.G.
Holtzman
D.M.
Acute stress increases interstitial fluid amyloid-β via corticotropin-releasing factor and neuronal activity
Proc. Natl. Acad. Sci. U. S. A.
 , 
2007
, vol. 
104
 (pg. 
10673
-
10678
)
7
Lee
K.W.
Kim
J.B.
Seo
J.S.
Kim
T.K.
Im
J.Y.
Baek
I.S.
Kim
K.S.
Lee
J.K.
Han
P.L.
Behavioral stress accelerates plaque pathogenesis in the brain of Tg2576 mice via generation of metabolic oxidative stress
J. Neurochem.
 , 
2009
, vol. 
108
 (pg. 
165
-
175
)
8
Joshi
Y.B.
Chu
J.
Pratico
D
Stress hormone leads to memory deficits and altered tau phosphorylation in a mouse model of Alzheimer's disease
J. Alzheimers. Dis.
 , 
2012
, vol. 
31
 (pg. 
167
-
168
)
9
Uz
T.
Dwivedi
Y.
Savani
P.D.
Imagnatello
F.
Pandey
G.
Manev
H.
Glucocorticoids stimulate inflammatory 5-lipoxygenase gene expression and protein translocation in the brain
J. Neurochem.
 , 
1999
, vol. 
73
 (pg. 
693
-
699
)
10
Uz
T.
Dwivedi
T.
Qeli
A.
Peters-Golden
M.
Pandey
G.
Manev
H.
Glucocorticoid receptors are required for up-regulation of neuronal 5-lipoxygenase (5LOX) expression by dexamethasone
FASEB J.
 , 
2001
, vol. 
15
 (pg. 
1792
-
1794
)
11
Puccio
S.
Chu
J.
Pratico
D.
Involvement of 5-lipoxygenase in the corticosteroid-dependent amyloid-β formation: in vitro and in vivo evidence
PLoS One
 , 
2001
, vol. 
6
 pg. 
e15163
 
12
Joshi
Y.B.
Chu
J.
Pratico
D.
Knockout of 5-lipoxygenase prevents dexamethasone-induced tau pathology in 3xTg mice
Aging Cell
 , 
2013
, vol. 
12
 (pg. 
706
-
711
)
13
Chu
J.
Giannopoulos
P.F.
Ceballos-Diaz
C.
Golde
T.E.
Pratico
D.
5-Lipoxygenase gene transfer worsens memory, amyloid, and tau brain pathologies in a mouse model of Alzheimer's disease
Ann. Neurol.
 , 
2012
, vol. 
72
 (pg. 
442
-
454
)
14
Giannopoulos
P.F.
Chu
J.
Joshi
Y.B.
Sperow
M.
Li
J.G.
Kirby
L.G.
Pratico
D.
Gene knockout of 5-lipoxygenase rescues synaptic dysfunction and improves memory in the triple-transgenic model of Alzheimer's disease
Mol. Psychiatry
 , 
2014
, vol. 
19
 (pg. 
511
-
518
)
15
Green
K.N.
Billings
L.M.
Roozendaal
B.
McGaugh
J.L.
LaFerla
F.M.
Glucocorticoids increase amyloid-β and tau pathology in a mouse model of Alzheimer's disease
J. Neurosci.
 , 
2006
, vol. 
26
 (pg. 
9047
-
9056
)
16
Joshi
Y.B.
Giannopoulos
P.F.
Chu
J.
Pratico
D.
Modulation lipopolysaccharide-induced memory insult, γ-secretase, and neuroinflammation in triple transgenic mice by 5-lipoxygenase
Neurobiol. Aging
 , 
2014
, vol. 
35
 (pg. 
1024
-
1031
)
17
Chakroborty
S.
Kim
J.
Schneider
C.
Jacobson
C.
Molgo
J.
Stuzmann
G.E.
Early presynaptic and postsynaptic calcium signaling abnormalities mask underlying synaptic depression in presymptomatic Alzheimer's disease mice
J. Neurosci.
 , 
2012
, vol. 
32
 (pg. 
8341
-
8353
)
18
Weeson
D.W.
Borkowski
A.H.
Landreth
G.E.
Nixon
R.A.
Levy
E.
Wilson
D.A.
Sensory network dysfunction, behavioral impairments, and their reversibility in an Alzheimer's β-amyloidosis mouse model
J. Neurosci.
 , 
2011
, vol. 
31
 (pg. 
5962
-
5971
)
19
Koch
G.
Di Lorenzo
F.
Bonni
S.
Ponzo
V.
Caltagirone
C.
Martorana
A.
Impaired LTP-but not LTD-like cortical plasticity in Alzheimer's disease patients
J. Alzheimers Dis.
 , 
2012
, vol. 
31
 (pg. 
593
-
599
)
20
Giannopoulos
P.F.
Chu
J.
Joshi
Y.B.
Sperow
M.
Li
J.G.
Kirby
L.G.
Pratico
D.
5-lipoxygenase activating protein reduction ameliorates cognitive deficit, synaptic dysfunction and neuropathology in a mouse model of Alzheimer's disease
Biol. Psychiatry
 , 
2013
, vol. 
74
 (pg. 
348
-
356
)
21
Joshi
Y.B.
Pratico
D.
Knockout of 5-lipoxygenase results in age-dependent anxiety-like behavior in female mice
PLoS One
 , 
2011
, vol. 
6
 pg. 
e29448
 
22
Higashi
Y.
Hoshijima
M.
Yawata
T.
Nobumoto
A.
Tsuda
M.
Shimizu
T.
Ueba
T.
Suppression of oxidative stress and 5-lipoxygenase activation by edaravone improves depressive-like behavior after concussion
J. Neurotrauma
 , 
2014
23
Dzitoyeva
S.
Imbesi
M.
Uz
T.
Dimitrijevic
N.
Manev
H.
Manev
R.
Caffeic acid attenuates the decrease of cortical BDNF transcript IV mRNA induced by swim stress in wild-type but not in 5-lipoxygenase-deficient mice
J. Neural Transm.
 , 
2008
, vol. 
115
 (pg. 
823
-
827
)
24
Oddo
S.
Caccamo
A.
Shepherd
J.D.
Murphy
M.P.
Metherate
R.
Mattson
M.P.
Golde
T.E.
Kayed
R.
Metherate
R.
Mattson
M.P.
Akbari
Y.
LaFerla
F.M.
Triple-transgenic model of Alzheimer's disease with plaques and tangles: intracellular Aβ and synaptic dysfunction
Neuron
 , 
2003
, vol. 
39
 (pg. 
409
-
421
)
25
Goulet
J.L.
Snouwaeert
J.N.
Latourt
A.M.
Coffman
T.M.
Koller
B.H.
Altered inflammatory responses in leukotriene-deficient mice
Proc. Natl. Acad. Sci. U. S. A.
 , 
1991
, vol. 
91
 (pg. 
12852
-
12856
)
26
Carroll
J.C.
Iba
M.
Bangasser
D.A.
Valentino
R.J.
James
M.J.
Brunden
K.R.
Lee
V.M.
Trojanowski
J.Q.
Chronic stress exacerbates tau pathology, neurodegeneration and cognitive performance through a corticotropin-releasing factor receptor-dependent mechanism in a transgenic mouse model of tauopathy
J. Neurosci.
 , 
2011
, vol. 
31
 (pg. 
14436
-
14449
)