Abstract

Epidemiological studies show that stimulating activities reduce therisk of dementia. In animal models of Alzheimer disease, there have been conflicting results of the effects of environmental enrichment (EE) on disease-related amyloid pathology. Here, we tested the direct effect of EE, independently of amyloid pathology, on brain neurofibrillary tangles (NFTs), which best correlate with dementia. We exposed transgenic mice (E257K/P301S-Tau-Tg driven by the natural tau promoter) to moderate nonstrained EE or regular environment. Concomitant with neurogenesis, we detected a decrease in NFT burden and a decrease in the activation of microglia in EE versus regular-environment mice. There was also a trend toward improvementin cognitive tasks in the EE mice. Increased immunoreactivity of brain-derived neurotrophic factor, which is involved in the regulation of tau phosphorylation, was detected in the EE mice, suggesting its possible involvement in the beneficial effects on NFTs and other parameters in the EE mice. These results suggest that NFTs may be directly responsive to environmental stimulating activities and that even nonstrained activities may mitigate tauopathies independent of theinvolvement of amyloid.

Introduction

Retrospective epidemiological studies suggest that stimulating activities, specifically a high level of education, a mentally demanding occupation, and challenging/active leisure and physical activity over a lifetime, are associated with a reduced incidence of dementia (1-3). Prospective studies have also shown that complex mental activity is associated with a reduced risk of developing dementia (4-6), and there is further support for the benefit of physical activity (7). These effects are generally attributed to brain cognitive reserve, that is, increased neuronal connectivity in brain regions involved in learning and memory that may compensate for and withstand detrimental symptoms of clinical dementia before they manifest.

Stimulating activities in animals using “environmental enrichment” (EE), which involves enhanced levels of sensory, cognitive, and motor stimulation, elicit various plastic responses in adult brains, including increased dendritic arborization and neurogenesis and improved learning (8). These effects are likely modulated by microglia (9-12). The effects of EE have been intensively investigated in transgenic (tg) models of Alzheimer disease (AD), particularly in mice with amyloid plaque neuropathology. These studies demonstrate cognitive improvement, but effects on amyloid pathology are variable (most studies report a significant decrease [13-16]), but involvement of both amyloid-related and -unrelated mechanisms (17) and no decrease in amyloid deposition (18) have been reported. Two studies have shown an increase in amyloid load in enriched animals, although cognitive deficits were mitigated (19, 20). These discrepancies suggest that amyloid pathology may not be the primary or only target responding to stimulating activities. The complexity of the responsiveness to EE has been further supported by the finding that ApoE4, the most prevalent AD genetic risk factor, had a detrimental effect of enhanced apoptosis, as opposed to the beneficial effect of EE in ApoE3 mice (21, 22).

Much less is known about the effect of EE on neurofibrillary tangles (NFTs), the aggregates of the phosphorylated microtubule-associated protein tau. Neurofibrillary tangles have been considered the best correlate of dementia in AD (23-25). Neurofibrillary tangles are also characteristic of other tauopathies that lack amyloid plaques, such as frontotemporal dementia, Pick disease, and others. In tg mice that express anti-nerve growth factor (NGF) antibodies, EE induced a decrease in amyloid burden without any affect on tau pathology (13); this result may be related to the NGF deficits (26-28). A recent study in APPswe/PS1{Delta}E9 mice (which have early tau pathology in the form of phosphorylated tau rather than mature aggregated NFTs in addition to amyloid pathology) showed a decrease in phosphorylated tau (29). This might be secondary to the decrease in amyloid oligomers because tau phosphorylation is regulated by pathways downstream of amyloid pathology (30).

Here, we studied the effects of EE on NFTs independent of other brain pathologies using the tauopathy mouse model that we recently generated. It expresses pathogenic human tau (E257T/P301S) under the regulation of the natural tau promoter (tolerated level of expression) and is characterized by pure NFT pathology with cognitive deficits (31). We found a decrease in NFT burden in the tauopathy mice exposed to EE accompanied by some cognitive improvement and an increase in the brain-derived neurotrophic factor (BDNF), the neuroprotective factor that has been reported to reduce tau phosphorylation via the tyrosine receptor kinase B (32). Leem et al (33) recently showed a decrease in phosphorylated tau (but not in aggregated NFTs or in cognitive deficits) after a long-term high-intensity exercise (possibly representing activities suitable for professional athletes) in mice that overexpress human tau. Our results suggest that NFTs developed under authentic tau regulation are reduced directly, and not in a manner depending on amyloid pathology, after a moderate degree of stimulating activities (which may reflect activity levels feasible in the general population).

Materials and Methods

Animals

E257T/P301S human tau protein (DM-Tau-tg) mice (31) were further crossed with C57BL mice for more than 6 generations to obtain tg offspring that could be identified by polymerase chain reaction analysis of tail genomic DNA. Experiments were approved by the animal ethics committee.

Environmental Enrichment and Study Design

Four-month-old DM-Tau-tg mice were housed for 9 months in EE or regular-environment (RE) cages. For moderate EE, the cages were large (610 × 435 × 215 mm) and equipped with a running wheel for tolerated voluntary exercise and differently shaped objects (tunnels, boxes, cubes, balls, shelters, ladder, labyrinth) that were substituted with others once a week in such a way that the same objects were used every other week. Regular environment consisted of standard laboratory cages (420 × 260 × 180 mm) without objects. Control groups consisted of RE DM-Tau-tg mice and RE wild-type (WT)-non-tg mice.

Two separate, independent experiments were performed. Experiment 1 included 13 DM-Tau-tg mice (5 males, 8 females) exposed to EE (EE-DM-Tau-tg mice), 11 DM-Tau-tg mice (5 males, 6 females) exposed to RE (RE-DM-Tau-tg mice), and 14 (6 males, 8 females) non-tg nonenriched mice (RE-WT). Experiment 2 included 9 EE-DM-Tau-tg (3 males, 6 females), 9 RE-DM-Tau-tg (3 males, 6 females), and 11 (5 males, 6 females) RE-WT mice.

Tissue Collection

Animals were killed at the end of each experiment under deep anesthesia and quickly transcardially perfused with phosphate-buffered saline. Brains were quickly removed; 1 hemisphere was frozen at −80°C, and the other was postfixed for 20 hours in 4% paraformaldehyde in phosphate-buffered saline (pH 7.2, ice-cold) and processed for sagittal paraffin sectioning at 6 μm.

Histology and Immunohistochemistry

Paraffin-embedded sections were silver-impregnated by Gallyas's (34) silver method. Neurofibrillary tangles were also detected by immunohistochemistry (IHC) using the AT8 mouse monoclonal Ab (Innogenetics, Ghent, Belgium) that recognizes tau phosphorylated at 202/205, as well as tau pathology in DM-Tau-tg mice (31). Microglial cells were stained with biotinylated tomato lectin (3 mg/mL Lycopersicon esulentum tomato; Sigma, St Louis, MO) and the Iba-1 antibody (WAKO, Osaka, Japan). Neuroblasts were detected using the doublecortin marker (DCX; Santa Cruz Biotechnology, Santa Cruz, CA). Brain-derived neurotrophic factor immunoreactivity was detected using a rabbit antibody (N-20; Santa Cruz Biotechnology).

Immunostaining was performed using the Mouse-on-Mouse system (Vector Laboratories, Burlingame, CA) for AT8, the EnVision System HRP Kit (DAKO Cytomation, Glostrup, Denmark) for BDNF and Iba-1, and the LSAB technique (LSAB2 System HRP; DAKO Cytomation) for lectin. Paraffin sections were deparaffinized and rehydrated in graded alcohols, and antigen retrieval was performed with citrate buffer pH 6 in a food steamer device (Braun, Kronberg, Germany) for 60 minutes. Endogenous peroxidase was blocked with 0.3% H2O2 in methanol followed by incubation in the appropriate blocking buffer (10% fetal bovine serum in Tris-buffered saline for all primary antibodies used except lectin; 0.3% Triton X in Tris-buffered saline for lectin) for 30 minutes. Sections were incubated overnight at 4°C with the primary antibodies AT8 (1:50), Iba-1 (1:2000), DCX (1:200), and BDNF (1:800) for 2 hours for lectin; 3,3′-diaminobenzidine tetrahydrochloride (DAB; DAKO) was used as a chromogen, and sections were counterstained with hematoxylin. Slides were then dehydrated in graded ethanols and covered with entelan.

Double immunofluorescence staining for combinations of Iba-1 and M1 or M2 microglia/macrophage phenotypes were performed using similar protocols. Sections from experimental autoimmune encephalomyelitis experiments were used as positive controls in these studies. Chemicals were purchased from Sigma unless specified otherwise.

Neuropathologic Evaluation

Neuropathologic evaluation was performed on 4 brain sections, spaced approximately 60 to 100 μm apart, under light microscope (Zeiss Axioplan 2; Carl Zeiss MicroImaging GmbH, Gottingen, Germany), with the aid of a CCD camera (Nikon, Tokyo, Japan) (35-37). Evaluation of neuropathology was blindly performed by 2 independent observers under 20× optical fields using a rectangular grid applied into the prefrontal lenses (prefrontal grid), with dimensions 610 × 610 μm for the 20× magnification. Higher magnification (40×) was used where necessary. The surface of each section was scanned, and the mean corresponding positive counted cells per surface area of the grid were used to calculate their mean densities per squared millimeter. Lectin- or Iba-1-positive microglia/macrophages, Gallyas-positive dark neurons, and AT8-positive cells per squared millimeter were evaluated in the cortex, hippocampus, thalamus, striatum, and brainstem. Cells positive for DCX were counted under 40× optical fields in the subgranular zone of the hippocampus and subventricular zone/rostral migratory stream. Four sagittal sections containing these structures and spaced approximately 60 to 80 μm apart were used. Data from DCX-positive cells were expressed as “cells per field."

Microglial activation states were quantified by determining the ramification index (RI) ranging from 0 for ramified “resting” cells to 1 for “active” amoeboid cells (38, 39), using the ImageJ software. Briefly, digital images from lectin- and Iba-1-positive microglia in the cortex, hippocampus, striatum, and brainstem were captured under 40× optical fields; 20 cells per section in 3 sections were evaluated per animal. These digital images were then separately processed with the binary mode of the ImageJ software (ImageJ ver.1.38; National Institutes of Health) to calculate the RI, using the cell perimeter and area in the following formula: RI = (4π × area)/perimeter2 (38).

To evaluate BDNF immunoreactivity, digital images were captured under 20× optical fields (3 sections per animal, 20 images per section evenly and repetitively distributed among the cortex, hippocampus, and brainstem) and processed in the ImageJ software using an adopted method previously described for semiautomated quantification of immunostained specimens (40). Briefly, digital images of DAB-stained sections were analyzed into their red-green-blue components; the blue channel was selected because DAB-stained areas demonstrate the greatest intensity levels for this channel (41, 42). The blue channel was then thresholded to segment the DAB-stained regions, binary transformed into black-white images by the ImageJ software, and compared with the original 3-colored image to validate its transformation. Automatic calculation of the total “black” pixels extracted the BDNF-immunostained area in each image. These data were used to express the mean BDNF relative signal in pixels.

Enzyme-Linked Immunosorbent Assay

The mouse BDNF enzyme-linked immunosorbent assay kit (Acris Antibodies, Herford, Germany) was used according to the manufacturer's instructions for analyzing BDNF levels in hippocampal homogenates of the EE-DM-Tau-tg mice and RE-DM-Tau-tg mice obtained as described (31).

Behavioral Examinations

T-Maze

The T-maze test was used for assessing the spatial short-term memory and alternation behavior, that is, determining the mouse's ability to recognize and differentiate between a new unknown and a familiar compartment (31, 43). The T-shaped maze was made of plastic with 2 arms 45 cm in length that extended at a right angle from a 57-cm-long alley. The arms had a width of 10 cm and were surrounded by 10-cm-high walls. The test consists of 2 trials with an intertrial interval of 1 hour, during which time the animals were put back to their home cages. During an 8-minute acquisition trial, one of the short arms was closed. In a 3-minute retention trial, mice had access to both arms and to the alley. Numbers of entries into the unfamiliar arm and the time spent in the unfamiliar arm were recorded. Mice normally tend to enter more times and spend more time in the new unknown arm than in the familiar one or in the alley.

Eight-Arm Maze

For the evaluation of spatial memory related to hippocampal-cortical function, we used the 8-arm radial maze scaled for mice (44, 45). The animals were introduced to the radial maze and were observed until they made entries to all 8 arms or until they completed 25 entries, whichever came first. The number of entries needed to complete a full round of 8 arms (once to each of the arms) within the 25 trials was recorded. The lower the number of entries needed, the better the cognitive score (best = 8, worse = 25). Maze performance was calculated each day for 5 consecutive days. Results were presented as area under the curve using the formula as follows: (Day 2 + Day 3 + Day 4 + Day 5) − 4 × (Day 1) with the entries indicated by day number (31, 46, 47).

Data Analysis

Results are presented as mean ± SEM. For comparisons of the quantity or intensity of stained cells and behavioral performance in the mazes between the study groups, the unpaired parametric Student t-test or nonparametric Mann-Whitney U test were used when appropriate.

Results

Decreased NFT Burden in EE-DM-Tau-tg Mice

We exposed the DM-Tau-tg mice to an EE paradigm that involved sensory, cognitive, and motor stimulation in the form of a moderate/nonstrained paradigm (substituting the objects with others only once a week, in such a way that the same objects were used every other week, as opposed to other reported protocols where objects were changed or completely substituted more often [48]), together with a tolerated voluntary nonforced physical exercise. The EE started at 4 months of age (ie, 2 months before the onset of NFT pathology in this model) and continued for approximately 9 months. This long-term exposure to EE was used to allow high effectiveness, in a manner similar to that of O'Callaghan et al (49) who showed long-term EE protection against age-related hippocampal changes.

In the first experiment, the EE-DM-Tau-tg mice showed a significant decrease of tau pathology/ NFT burden, as indicated by the significant lower burden of cells stained for the Gallyas staining for NFTs/tau pathology in the brain relative to the RE-DM-Tau-tg mice (5.46 ± 0.34 vs 9.6 ± 0.6, a 43.3% lower burden in enriched-mice [p < 0.0001]): the decrease was 53.6% (p < 0.0001), 56.6% (p = 0.006), and 18.9% in the cortex, hippocampus, and brainstem, respectively (Figs. 1A, B). This decrease in NFTs in the brain was further confirmed by IHC with AT8 (0.34 ± 0.05 vs 0.54 ± 0.06, a decrease of 37.1% from RE- to EE-DM-Tau-tg mice [p = 0.01]): the decrease was 34.4% and 45.7% (p = 0.05) in the cortex and brainstem, respectively (Figs. 2A, B). In the second experiment, there was a significantly lower burden of Gallyas-positive cells in the brain of EE- versus RE-DM-Tau-tg mice (7.44 ± 1.1 vs 15.3 ± 1.7, a 51.4% lower burden [p < 0.0001]): the decrease was 50.2% (p < 0.0001), 35.7%, and 35.2% (p = 0.08) in the cortex, hippocampus, and brainstem, respectively. This decrease in NFTs was confirmed by IHC with AT8 (0.8 ± 0.04 vs 1.04 ± 0.06, a decrease of 23.1% from RE- to EE-DM-Tau-tg mice [p = 0.003]): the decrease was 14% and 31.9% (p = 0.002) in the cortex and brainstem, respectively.

FIGURE 1.

Reduced neurofibrillary tangle (NFT) burden assessed by the Gallyas method in environmentally enriched (EE)-DM-Tau-tg mice. (A) There are fewer Gallyas-positive cells (arrows) in the cortex, hippocampus, and brainstem of EE-DM-Tau-tg versus regular-environment (RE)-DM-Tau-tg mice. Scale bars = 100 μm. Insets are higher-magnification images from the same fields. (B) Semiquantitative assessment of Gallyas staining (p < 0.0001, p < 0.0001, and p = 0.006 for total brain, cortex, and hippocampus, respectively).

FIGURE 1.

Reduced neurofibrillary tangle (NFT) burden assessed by the Gallyas method in environmentally enriched (EE)-DM-Tau-tg mice. (A) There are fewer Gallyas-positive cells (arrows) in the cortex, hippocampus, and brainstem of EE-DM-Tau-tg versus regular-environment (RE)-DM-Tau-tg mice. Scale bars = 100 μm. Insets are higher-magnification images from the same fields. (B) Semiquantitative assessment of Gallyas staining (p < 0.0001, p < 0.0001, and p = 0.006 for total brain, cortex, and hippocampus, respectively).

FIGURE 2.

Reduced neurofibrillary tangle (NFT) burden assessed by AT8 immunohistochemistry in environmentally enriched (EE)-DM-Tau-tg mice. (A) A decrease in AT8-positive cells is evident in the cortex, hippocampus, and brainstem of EE-DM-Tau-tg versus regular-environment (RE)-DM-Tau-tg mice. Scale bars = 100 μm. Insets are higher-magnification images from the same fields. (B) Semiquantitative assessment of AT8 staining. p = 0.01 and p = 0.05 for total brain and brainstem, respectively.

FIGURE 2.

Reduced neurofibrillary tangle (NFT) burden assessed by AT8 immunohistochemistry in environmentally enriched (EE)-DM-Tau-tg mice. (A) A decrease in AT8-positive cells is evident in the cortex, hippocampus, and brainstem of EE-DM-Tau-tg versus regular-environment (RE)-DM-Tau-tg mice. Scale bars = 100 μm. Insets are higher-magnification images from the same fields. (B) Semiquantitative assessment of AT8 staining. p = 0.01 and p = 0.05 for total brain and brainstem, respectively.

Evidence for Neurogenesis in the Enriched-DM-Tau-tg

Neurogenesis occurs in adult mammalian brains, mainly in the hippocampal dentate gyrus and in the subventricular zone (50, 51), We assessed neurogenesis using the microtubule-associated protein DCX, which is expressed in newly formed neurons (52). More DCX-positive cells were detected in the EE-DM-Tau-tg mice versus RE-DM-Tau-tg mice, particularly in the subgranular zone in hippocampus (1.53 ± 0.15 vs 0.47 ± 0.1 [p < 0.0001]). Differences were less evident in the subventricular zone and rostral migratory stream (42.69 ± 10.7 vs 37.59 ± 10.24 [p = 0.76]) (Fig. 3). This result suggests that our EE paradigm was effective with respect to inducing neurogenesis.

FIGURE 3.

Neurogenesis in environmentally enriched (EE)-DM-Tau-tg mice. (A) Greater numbers of doublecortin marker (DCX)-positive cells are seen mainly in the hippocampal subgranular zone (SGZ) and, to a lesser extent, in the subventricular zone (SVZ) and the rostral migratory stream (RMS) of EE-DM-Tau-tg versus regular-environment (RE)-DM-Tau-tg mice. (B, C) Semi-quantitative assessment of DCX in the SGZ in hippocampus (p < 0.0001) (B), and in SVZ and RMS (p = 0.76) (C). LV indicates lateral ventricle.

FIGURE 3.

Neurogenesis in environmentally enriched (EE)-DM-Tau-tg mice. (A) Greater numbers of doublecortin marker (DCX)-positive cells are seen mainly in the hippocampal subgranular zone (SGZ) and, to a lesser extent, in the subventricular zone (SVZ) and the rostral migratory stream (RMS) of EE-DM-Tau-tg versus regular-environment (RE)-DM-Tau-tg mice. (B, C) Semi-quantitative assessment of DCX in the SGZ in hippocampus (p < 0.0001) (B), and in SVZ and RMS (p = 0.76) (C). LV indicates lateral ventricle.

A Slight Decrease in the Activated State of Microglia in the EE-DM-Tau-tg Mice

By lectin staining, no significant difference in microglial burden was detected between EE- and RE-DM-Tau-tg mice (5.24 ± 0.35 vs 5.02 ± 0.32). Interestingly, there was lower RI (thinner processes) in EE- versus RE-DM-Tau-tg mice (0.0635% ± 0.03% vs 0.071% ± 0.03%, a decrease of 10.6% [p = 0.01]) (Fig. 4). Similar results were obtained with Iba-1 Ab, that is, there was no difference in microglial numbers between EE- and RE-DM-Tau-tg mice (11.40 ± 0.31 vs 10.98 ± 0.25), whereas there was a significant decrease in RI of EE- versus RE-DM-Tau-tg mice (0.0795% ± 0.003% vs 0.092% ± 0.004%, a decrease of 13.6% [p = 0.02]) (Fig. 5). These results suggest that, although microglial cell counts were not affected, they were less activated in the EE- DM-Tau-tg mice. Specific staining for discrimination of either an M1 or M2 phenotype of these microglia did not reveal any positivity of these cells, possibly indicating an M0 phenotype (data not shown). As expected, no perivascular macrophages were observed in any mice.

FIGURE 4.

There is a slight decrease in activated microglia (assessed by lectin staining) (arrows) in environmentally enriched (EE)-DM-Tau-tg mice. (A) Microglial burden and activation in the cortex, hippocampus, and brainstem of EE-DM-Tau-tg and regular-environment (RE)-DM-Tau-tg mice. Although there are similar numbers of microglia (arrows) in the groups, they have a lower ramification index (RI) (less activated, thinner processes) in EE-DM-Tau-tg versus RE-DM-Tau-tg mice. Scale bars = 100 μm. Insets are higher-magnification images from the same fields. (B, C) Quantitative assessment of total burden of microglia (B) and of activated microglia (C) (p = 0.01).

FIGURE 4.

There is a slight decrease in activated microglia (assessed by lectin staining) (arrows) in environmentally enriched (EE)-DM-Tau-tg mice. (A) Microglial burden and activation in the cortex, hippocampus, and brainstem of EE-DM-Tau-tg and regular-environment (RE)-DM-Tau-tg mice. Although there are similar numbers of microglia (arrows) in the groups, they have a lower ramification index (RI) (less activated, thinner processes) in EE-DM-Tau-tg versus RE-DM-Tau-tg mice. Scale bars = 100 μm. Insets are higher-magnification images from the same fields. (B, C) Quantitative assessment of total burden of microglia (B) and of activated microglia (C) (p = 0.01).

FIGURE 5.

There is a slight decrease in activated microglia (assessed by Iba-1 staining) in environmentally enriched (EE)-DM-Tau-tg mice. (A) Microglial burden and activation in the cortex, hippocampus, and brainstem of EE-DM-Tau-tg and regular-environment (RE)-DM-Tau-tg mice. Although there are similar numbers of microglia (arrows) in both groups, they have a lower ramification index (RI) (thinner processes) in EE-DM-Tau-tg versus RE-DM-Tau-tg mice. Scale bars = 100 μm. Insets are higher-magnification images from the same fields. (B, C) Quantitative assessment of total burden of microglia (B) and of activated microglia (C) (p = 0.02).

FIGURE 5.

There is a slight decrease in activated microglia (assessed by Iba-1 staining) in environmentally enriched (EE)-DM-Tau-tg mice. (A) Microglial burden and activation in the cortex, hippocampus, and brainstem of EE-DM-Tau-tg and regular-environment (RE)-DM-Tau-tg mice. Although there are similar numbers of microglia (arrows) in both groups, they have a lower ramification index (RI) (thinner processes) in EE-DM-Tau-tg versus RE-DM-Tau-tg mice. Scale bars = 100 μm. Insets are higher-magnification images from the same fields. (B, C) Quantitative assessment of total burden of microglia (B) and of activated microglia (C) (p = 0.02).

Increased BDNF in EE-DM-Tau-tg Mice

Immunohistochemistry with anti-BDNF Ab revealed that EE-DM-Tau-tg mice had increased BDNF staining compared with that in RE-DM-Tau-tg mice (increase of 19.2% [p = 0.01]). This was observed in cell processes and axons (Fig. 6). The increase was 35.7% (p = 0.001), 29.5%, and 9.7% (p = 0.003) in the cortex, hippocampus, and brainstem, respectively. Analysis of the BDNF levels in hippocampal homogenates by enzyme-linked immunosorbent assay revealed a trend of increased level in the EE-DM-Tau-tg mice relative to RE-DM-Tau-tg mice (29.52 ± 4.47 vs 24.08 ± 4.59 pg/μg protein, a nonsignificant increase of 18.4%). RE-DM-Tau-tg mice expressed lower BDNF levels than RE-WT mice (15.5% [p = 0.047], 27.6% [p = 0.017], 41.6% [p = 0.037], and 7.2% [p = 0.027], respectively) in total brain, cortex, hippocampus, and brainstem, respectively.

FIGURE 6.

Increased brain-derived neurotrophic factor (BDNF) in environmentally enriched (EE)-DM-Tau-tg mice. (A) There is greater BDNF immunoreactivity in the cortex, hippocampus, and brainstem of EE-DM-Tau-tg versus regular-environment (RE)-DM-Tau-tg mice. Boxes in the photos indicate areas of magnified insets displaying the BDNF-positive processes and axons. Differences are indicated as “more and denser brown signals.” (B) Semiquantitative assessment of BDNF immunoreactivity. p = 0.01, p = 0.0001, and p = 0.003 for total brain, cortex, and brainstem, respectively.

FIGURE 6.

Increased brain-derived neurotrophic factor (BDNF) in environmentally enriched (EE)-DM-Tau-tg mice. (A) There is greater BDNF immunoreactivity in the cortex, hippocampus, and brainstem of EE-DM-Tau-tg versus regular-environment (RE)-DM-Tau-tg mice. Boxes in the photos indicate areas of magnified insets displaying the BDNF-positive processes and axons. Differences are indicated as “more and denser brown signals.” (B) Semiquantitative assessment of BDNF immunoreactivity. p = 0.01, p = 0.0001, and p = 0.003 for total brain, cortex, and brainstem, respectively.

Behavioral Tests

In the T-maze, EE-DM-Tau-tg mice showed a better performance than the RE-DM-Tau-tg mice, with a higher number of entries into the unfamiliar arm, reaching close to the number of the RE-WT mice (3.62 ± 0.28, 2.89 ± 0.3, and 3.71 ± 0.2 entries in EE-DM-Tau-tg, RE-DM-Tau-tg, and RE-WT mice, respectively). The difference between RE-DM-Tau-tg and RE-WT was significant (p = 0.04), but the EE-DM-Tau-tg mice (albeit showing a similar performance to that of RE-WT mice) were not significantly different from RE-DM-Tau-tg mice (p = 0.1). However, the similar performances of EE- DM-Tau-tg and RE-WT mice (3.62 ± 0.28 and 3.71 ± 0.2, respectively, p = 0.8) support the idea that enrichment restored cognitive function to the RE-DM-Tau-tg (Fig. 7A). A similar improvement in EE-DM-Tau-tg mice was found when calculating the time in seconds spent in the unfamiliar arm of the T-maze (not shown).

FIGURE 7.

Cognitive tasks in environmentally enriched (EE)-DM-Tau-tg mice. (A) In the T-maze, EE-DM-Tau-tg mice show a nonsignificant better performance versus regular-environment (RE)-DM-Tau-tg mice, that is, there are higher numbers of entries into the unfamiliar arm (p = 0.1). The similar performance of the EE-DM-Tau-tg and the RE wild-type (WT) mice (nonsignificant difference [NS], p = 0.8) suggests that the low performance of the RE-DM-Tau-tg (vs RE WT mice, p = 0.04) was restored in EE-DM-Tau-tg mice. (B) In the 8-arm maze, EE-DM-Tau-tg mice needed fewer entries into the arms (calculated as area under the curve [AUC]) versus RE-DM-Tau-tg mice; their performance was similar to that of RE-WT mice, although there were no significant differences among the groups.

FIGURE 7.

Cognitive tasks in environmentally enriched (EE)-DM-Tau-tg mice. (A) In the T-maze, EE-DM-Tau-tg mice show a nonsignificant better performance versus regular-environment (RE)-DM-Tau-tg mice, that is, there are higher numbers of entries into the unfamiliar arm (p = 0.1). The similar performance of the EE-DM-Tau-tg and the RE wild-type (WT) mice (nonsignificant difference [NS], p = 0.8) suggests that the low performance of the RE-DM-Tau-tg (vs RE WT mice, p = 0.04) was restored in EE-DM-Tau-tg mice. (B) In the 8-arm maze, EE-DM-Tau-tg mice needed fewer entries into the arms (calculated as area under the curve [AUC]) versus RE-DM-Tau-tg mice; their performance was similar to that of RE-WT mice, although there were no significant differences among the groups.

In the 8-arm maze, there was a similar trend of better performance in EE- versus RE-DM-Tau-tg mice, approaching the area under the curve of the RE-WT mice (6.55 ± 3.8, 10.67 ± 2.82, and 6.75 ± 2.49 entries in EE-DM-Tau-tg, RE-DM-Tau-tg, and RE-WT mice, respectively) but without reaching statistical significance (Fig. 7B).

Discussion

Although firm epidemiological evidence points to the beneficial effect of stimulating activities on symptoms of dementia and AD, mechanistic support from experimental animal studies is limited. Despite intensive investigation, there is still controversy regarding the possibility that the protective effect of EE is mediated via the amyloid pathology, including the possibility that amyloid plaques may be exacerbated after EE (13, 14, 16-20, 29, 53, 54). Less focus has been placed on the responsiveness of NFT pathology to EE. We used an EE paradigm that may have a therapeutic and applicable advantage because it involves moderate sensory and cognitive stimulation and includes tolerated voluntary exercise. Moreover, the DM-Tau-tg mice model is driven by the natural tau promoter, leading to a tolerated level of expression of the tg mutant tau, which contrasts with some other tauopathy tg models in which unrelated promoters may lead to very high levels of tau transgene overexpression and are accompanied by motor deficits that are not necessarily related to tauopathy (55-58). Our finding that NFTs are reduced in various brain regions by EE indicates that the NFT pathology developed under natural physiological regulation responds directly to EE, that is, without involvement of amyloid pathology; moreover, this seemed to be accompanied by some cognitive benefit. The data also suggest that EE may protect not only against AD but also against other tauopathies in which only NFTs are detected. Neurofibrillary tangle reduction in response to a tolerated EE paradigm may have a broader applicability because exposure to moderate stimulating activities would be applicable to most of the population. In contrast, exposing tau pathology mouse model (under the regulation of the enolase-promoter) (59) to a long-term high-intensity exercise reduced the degree of tau phosphorylation (33). However, it is not clear whether this intensive physical activity (which may correspond more to professional athletes) also reduced mature NFTs or whether intensive physical activity had any beneficial effects on cognitive deficits.

The EE DM-Tau-tg mice showed evidence of neurogenesis, indicating that EE likely elicited plastic responses in the adult brain, specifically by neurogenesis (8, 51). Our finding that neurogenesis in EE DM-Tau-tg mice was particularly evident in the hippocampus is similar to the report that neurogenesis specifically occurred in the hippocampus of enriched WT mice (52). It will be interesting to investigate whether the extent of neurogenesis taking place in the DM-Tau-tg mice is similar to that in WT- mice exposed to EE. This will indicate whether the pathologic tau (double-mutant) alters/impairs the neurogenesis and plastic responses.

No differences in numbers of microglia were detected between EE-DM-Tau-tg and RE- DM-Tau-tg mice, possibly indicating that the increase in microglia reported in EE WT mice may be impaired in the EE-DM-Tau-tg mice because of some alteration/deficit in the tau function. This has been reported in enriched PS1ΔE9- and PS1M146L mice (9, 11), and there is a report on the lack of increase in microglial burden in enriched APPswe/PS1ΔE9 mice (29). Yet, although no change in microglial count was detected, a decrease in the microglial activation state was detected with both lectin staining and Iba-1 IHC. Morphology provides some indication about the activation state of microglia, although it cannot predict function--either proinflammatory, M1 type, or anti-inflammatory, M2 type (60). The slight, although significant, decrease in the activation state of microglia by EE treatment may suggest a reduction of noxious stimuli for microglia (61, 62). However, the parenchymal microglia did not acquire a typical M1 or M2 phenotype (data not shown), a finding that may not support a direct role of the microglial cells in reducing NFTs by the EE. On the basis of the reports of a positive correlation between NFT burden and microglial-burden in brains of both tauopathy patients (63, 64) and NFT Tg mice (65), a lower microglial activation that accompanied the lower NFT burden in our EE-DM-Tau-tg mice may not be surprising. Although activation of microglia has been reported to precede the formation of NFTs in naive NFT Tg mice (51), we speculate that NFTs may also affect the microglial activation, that is, once a decrease in tangles occurs, there may be a lower activation signal for microglia. In other words, the decrease in microglial activation might be an epiphenomenon reflecting the decrease in the neurodegenerative process in the EE-DM-Tau-tg mice than a causative effect.

We observed an increase in BDNF in neuronal and glial processes in the EE-DM-Tau-tg versus RE-DM-Tau-tg mice. This indicates that the pathogenic DM-Tau did not block this BDNF elevation reported in WT mice exposed to EE (66-68). This increase in BDNF after EE may shed some light on the mechanisms involved in the reduction in NFT burden in the enriched tauopathy mice because BDNF has been reported to reduce tau phosphorylation via the tyrosine receptor kinase B, involving the GSK-3β and PI-3 kinase, which participate in the signaling transduction cascade regulating tau phosphorylation (32). Other studies also showed that the regulation of GSK-3b/PI-3 is associated with BDNF (69-73). Moreover, Berardi et al showed that EE of mice that cannot express NGF, another neurotrophic factor that induces tau dephosphorylation (27, 28), reduced amyloid pathology but did not reduce tau phosphorylation, lending further support to the notion that expression of neuroprotective neurotrophic factors under EE may be involved in tau dephosphorylation. Some additional indirect support to the relevance of the increase in BDNF to the decrease in NFTs in the enriched mice may originate from our finding that RE DM-Tau-tg mice also showed a decrease in BDNF level relative to RE WT mice; EE seemed to rescue both deficits.

Our DM-Tau-tg mice also showed a trend of improvement in cognitive tasks after exposure to EE. This improvement in performance of the EE-DM-Tau-Tg, which approached the performance of the RE-WT mice, may be related to a reversal/prevention, at least partially, of the detrimental effects characteristic to the DM-Tau-tg mice. It is possible that not only anti-NFT processes but also other general EE-induced processes (e.g. neurogenesis and synaptic remodeling) are involved in the improvement in cognitive tasks in our NFT model. Further studies are needed to dissect the contribution of such processes, but our present results indicate that mice with NFTs are responsive to EE and that their cognitive status can be improved.

Environmental enrichment was beneficial when it was started at an early age, thereby raising the possibility that EE may be beneficial for the prevention of NFT pathology and cognitive deficits. While being exposed to stimulating activities from a young age (specifically of nonextreme regimes, yet long-term) seems feasible to the general population, it would be important to investigate whether exposure to stimulating activities at older ages (when the tauopathy disease is already evident) would also be beneficial. Further experiments are also needed to address how long is the beneficial effect preserved after cessation of EE. In addition, the effect of EE on biochemical properties of tau, such as the response of soluble oligomers (toxic forms 74), as well as on the detergent solubility of tau, are important issues to be investigated.

In conclusion, our results showed a decrease in NFT burden accompanied by an improvement in memory in an authentic tauopathy model after a tolerated paradigm of EE, pointing to a direct effect of EE on NFTs rather than a secondary anti-NFT effect of EE on amyloid pathology. This anti-NFT effect correlated with increased BDNF. The responsiveness of NFTs even under moderate/nonextreme stimulating activities may have promising implications to the general population.

Acknowledgments

The authors thank Fanny Baitscher from the Department of Neurology, Hadassah, and Evangelia Nousiopoulou from the Department of Neurology, AHEPA, for technical assistance.

References

1.
Friedland
RP
Fritsch
T
Smyth
KA
.
Patients with Alzheimer's disease have reduced activities in midlife compared with healthy control-group members
.
Proc Natl Acad Sci U S A
 
2001
;
98
:
3440
45
2.
Scarmeas
N
Levy
G
Tang
MX
.
Influence of leisure activity on the incidence of Alzheimer's disease
.
Neurology
 
2001
;
57
:
2236
42
3.
Stern
Y
Gurland
B
Tatemichi
TK
.
Influence of education and occupation on the incidence of Alzheimer's disease
.
JAMA
 
1994
;
271
:
1004
10
4.
Letenneur
L
Gilleron
V
Commenges
D
.
Are sex and educational level independent predictors of dementia and Alzheimer's disease? Incidence data from the PAQUID project
.
J Neurol Neurosurg Psychiatry
 
1999
;
66
:
177
83
5.
Verghese
J
Lipton
RB
Katz
MJ
.
Leisure activities and the risk of dementia in the elderly
.
N Engl J Med
 
2003
;
348
:
2508
16
6.
Wilson
RS
Bennett
DA
Bienias
JL
.
Cognitive activity and incident AD in a population-based sample of older persons
.
Neurology
 
2002
;
59
:
1910
14
7.
Lautenschlager
NT
Cox
KL
Flicker
L
.
Effect of physical activity on cognitive function in older adults at risk for Alzheimer disease: A randomized trial
.
JAMA
 
2008
;
300
:
1027
37
8.
Nithianantharajah
J
Hannan
AJ
.
The neurobiology of brain and cognitive reserve: Mental and physical activity as modulators of brain disorders
.
Prog Neurobiol
 
2009
;
89
:
369
82
9.
Choi
SH
Veeraraghavalu
K
Lazarov
O
.
Non-cell-autonomous effects of presenilin 1 variants on enrichment-mediated hippocampal progenitor cell proliferation and differentiation
.
Neuron
 
2008
;
59
:
568
80
10.
Ehninger
D
Kempermann
G
.
Regional effects of wheel running and environmental enrichment on cell genesis and microglia proliferation in the adult murine neocortex
.
Cereb Cortex
 
2003
;
13
:
845
51
11.
Villeda
S
Wyss-Coray
T
.
Microglia-a wrench in the running wheel?
Neuron
 
2008
;
59
:
527
29
12.
Ziv
Y
Ron
N
Butovsky
O
.
Immune cells contribute to the maintenance of neurogenesis and spatial learning abilities in adulthood
.
Nat Neurosci
 
2006
;
9
:
268
75
13.
Berardi
N
Braschi
C
Capsoni
S
.
Environmental enrichment delays the onset of memory deficits and reduces neuropathological hallmarks in a mouse model of Alzheimer-like neurodegeneration
.
J Alzheimers Dis
 
2007
;
11
:
359
70
14.
Cracchiolo
JR
Mori
T
Nazian
SJ
.
Enhanced cognitive activity-over and above social or physical activity-is required to protect Alzheimer's mice against cognitive impairment, reduce Aβ deposition, and increase synaptic immunoreactivity
.
Neurobiol Learn Mem
 
2007
;
88
:
277
94
15.
Lazarov
O
Robinson
J
Tang
YP
.
Environmental enrichment reduces Aβ levels and amyloid deposition in transgenic mice
.
Cell
 
2005
;
120
:
701
13
16.
Pardon
MC
Sarmad
S
Rattray
I
.
Repeated novel cage exposure-induced improvement of early Alzheimer's-like cognitive and amyloid changes in TASTPM mice is unrelated to changes in brain endocannabinoids levels
.
Neurobiol Aging
 
2009
;
30
:
1099
113
17.
Costa
DA
Cracchiolo
JR
Bachstetter
AD
.
Enrichment improves cognition in AD mice by amyloid-related and unrelated mechanisms
.
Neurobiol Aging
 
2007
;
28
:
831
44
18.
Arendash
G
Garcia
MF
Costa
DA
.
Environmental enrichment improves cognition in aged Alzheimer's transgenic mice despite stable β-amyloid deposition
.
Neuroreport
 
2004
;
15
:
1751
54
19.
Jankowsky
JL
Melnikova
T
Fadale
DJ
.
Environmental enrichment mitigates cognitive deficits in a mouse model of Alzheimer's disease
.
J Neurosci
 
2005
;
25
:
5217
24
20.
Jankowsky
JL
Xu
G
Fromholt
D
.
Environmental enrichment exacerbates amyloid plaque formation in a transgenic mouse model of Alzheimer disease
.
J Neuropathol Exp Neurol
 
2003
;
62
:
1220
27
21.
Levi
O
Jongen-Relo
AL
Feldon
J
.
ApoE4 impairs hippocampal plasticity isoform-specifically and blocks the environmental stimulation of synaptogenesis and memory
.
Neurobiol Dis
 
2003
;
13
:
273
82
22.
Levi
O
Michaelson
DM
.
Environmental enrichment stimulates neurogenesis in apolipoprotein E3 and neuronal apoptosis in apolipoprotein E4 transgenic mice
.
J Neurochem
 
2007
;
100
:
202
10
23.
Arriagada
PV
Growdon
JH
Hedley-Whyte
ET
.
Neurofibrillary tangles but not senile plaques parallel duration and severity of Alzheimer's disease
.
Neurology
 
1992
;
42
:
631
39
24.
Bierer
LM
Hof
PR
Purohit
DP
.
Neocortical neurofibrillary tangles correlate with dementia severity in Alzheimer's disease
.
Arch Neurol
 
1995
;
52
:
81
88
25.
Braak
H
Braak
E
.
Neuropathological stageing of Alzheimer-related changes
.
Acta Neuropathol
 
1991
;
82
:
239
59
26.
Fisher
A
Heldman
E
Gurwitz
D
.
M1 agonists for the treatment of Alzheimer's disease
.
Novel properties and clinical update. Ann N Y Acad Sci
 
1996
;
777
:
189
96
27.
Nuydens
R
Dispersyn
G
de Jong
M
.
Aberrant tau phosphorylation and neurite retraction during NGF deprivation in PC12 cells
.
Biochem Biophys Res Commun
 
1997
;
240
:
687
91
28.
Shelton
SB
Johnson
GV
.
Tau and HMW tau phosphorylation and compartmentalization in apoptotic neuronal PC12 cells
.
J Neurosci Res
 
2001
;
66
:
203
13
29.
Hu
YS
Xu
P
Pigino
G
.
Complex environment experience rescues impaired neurogenesis, enhances synaptic plasticity, and attenuates neuropathology in familial Alzheimer's disease-linked APPswe/PS1{Delta}E9 mice
.
FASEB J
 
2010
;
24
:
1667
81
30.
Small
SA
Duff
K
.
Linking Aβ and tau in late-onset Alzheimer's disease: A dual pathway hypothesis
.
Neuron
 
2008
;
60
:
534
42
31.
Rosenmann
H
Grigoriadis
N
Eldar-Levy
H
.
A novel transgenic mouse expressing double mutant tau driven by its natural promoter exhibits tauopathy characteristics
.
Exp Neurol
 
2008
;
212
:
71
84
32.
Elliott
E
Atlas
R
Lange
A
.
Brain-derived neurotrophic factor induces a rapid dephosphorylation of tau protein through a PI-3 kinase signalling mechanism
.
Eur J Neurosci
 
2005
;
22
:
1081
89
33.
Leem
YH
Lim
HJ
Shim
SB
.
Repression of tau hyperphosphorylation by chronic endurance exercise in aged transgenic mouse model of tauopathies
.
J Neurosci Res
 
2009
;
87
:
2561
70
34.
Gallyas
F
.
Silver staining of Alzheimer's neurofibrillary changes by means of physical development
.
Acta Morphol Acad Sci Hung
 
1971
;
19
:
1
8
35.
Boimel
M
Grigoriadis
N
Lourbopoulos
A
.
Efficacy and safety of immunization with phosphorylated tau against neurofibrillary tangles in mice
.
Exp Neurol
 
2010
;
224
:
472
85
36.
Rosenmann
H
Grigoriadis
N
Karussis
D
.
Tauopathy-like abnormalities and neurologic deficits in mice immunized with neuronal tau protein
.
Arch Neurol
 
2006
;
63
:
1459
67
37.
Shiryaev
N
Jouroukhin
Y
Giladi
E
.
NAP protects memory, increases soluble tau and reduces tau hyperphosphorylation in a tauopathy model
.
Neurobiol Dis
 
2009
;
34
:
381
88
38.
Wilms
H
Hartmann
D
Sievers
J
.
Ramification of microglia, monocytes and macrophages in vitro: Influences of various epithelial and mesenchymal cells and their conditioned media
.
Cell Tissue Res
 
1997
;
287
:
447
58
39.
Eder
C
Schilling
T
Heinemann
U
.
Morphological, immunophenotypical and electrophysiological properties of resting microglia in vitro
.
Eur J Neurosci
 
1999
;
11
:
4251
61
40.
King
TW
Brey
EM
Youssef
AA
.
Quantification of vascular density using a semiautomated technique for immunostained specimens
.
Anal Quant Cytol Histol
 
2002
;
24
:
39
48
41.
Goto
M
Nagatomo
Y
Hasui
K
.
Chromaticity analysis of immunostained tumor specimens
.
Pathol Res Pract
 
1992
;
188
:
433
37
42.
Kuyatt
BL
Reidy
CA
Hui
KY
.
Quantitation of smooth muscle proliferation in cultured aorta
.
A color image analysis method for the Macintosh. Anal Quant Cytol Histol
 
1993
;
15
:
83
87
43.
Zueger
M
Urani
A
Chourbaji
S
.
Olfactory bulbectomy in mice induces alterations in exploratory behavior
.
Neurosci Lett
 
2005
;
374
:
142
46
44.
Olton
D
.
Rememberance of places passed: Spatial memory in rats
.
J Exp Psychol Anim Behav Processes
 
1976
;
2
:
179
216
45.
Yanai
J
Pick
CG
.
Studies on noradrenergic alterations in relation to early phenobarbital-induced behavioral changes
.
Int J Dev Neurosci
 
1987
;
5
:
337
44
46.
Avraham
Y
Bonne
O
Berry
EM
.
Behavioral and neurochemical alterations caused by diet restriction-the effect of tyrosine administration in mice
.
Brain Res
 
1996
;
732
:
133
44
47.
Dagon
Y
Avraham
Y
Link
G
.
The synthetic cannabinoid HU-210 attenuates neural damage in diabetic mice and hyperglycemic pheochromocytoma PC12 cells
.
Neurobiol Dis
 
2007
;
27
:
174
81
48.
Nithianantharajah
J
Hannan
AJ
.
Enriched environments, experience-dependent plasticity and disorders of the nervous system
.
Nat Rev Neurosci
 
2006
;
7
:
697
709
49.
O'Callaghan
RM
Griffin
EW
Kelly
AM
.
Long-term treadmill exposure protects against age-related neurodegenerative change in the rat hippocampus
.
Hippocampus
 
2009
;
19
:
1019
29
50.
Kempermann
G
Wiskott
L
Gage
FH
.
Functional significance of adult neurogenesis
.
Curr Opin Neurobiol
 
2004
;
14
:
186
91
51.
Van
Praag H
Kempermann
G
Gage
FH
.
Neural consequences of environmental enrichment
.
Nat Rev Neurosci
 
2000
;
1
:
191
98
52.
Brown
J
Cooper-Kuhn
CM
Kempermann
G
.
Enriched environment and physical activity stimulate hippocampal but not olfactory bulb neurogenesis
.
Eur J Neurosci
 
2003
;
17
:
2042
46
53.
Herring
A
Ambree
O
Tomm
M
.
Environmental enrichment enhances cellular plasticity in transgenic mice with Alzheimer-like pathology
.
Exp Neurol
 
2009
;
216
:
184
92
54.
Mirochnic
S
Wolf
S
Staufenbiel
M
.
Age effects on the regulation of adult hippocampal neurogenesis by physical activity and environmental enrichment in the APP23 mouse model of Alzheimer disease
.
Hippocampus
 
2009
;
19
:
1008
18
55.
Allen
B
Ingram
E
Takao
M
.
Abundant tau filaments and nonapoptotic neurodegeneration in transgenic mice expressing human P301S tau protein
.
J Neurosci
 
2002
;
22
:
9340
51
56.
Dawson
HN
Cantillana
V
Chen
L
.
The tau N279K exon 10 splicing mutation recapitulates frontotemporal dementia and parkinsonism linked to chromosome 17 tauopathy in a mouse model
.
J Neurosci
 
2007
;
27
:
9155
68
57.
Lewis
J
McGowan
E
Rockwood
J
.
Neurofibrillary tangles, amyotrophy and progressive motor disturbance in mice expressing mutant (P301L) tau protein
.
Nat Genet
 
2000
;
25
:
402
5
58.
Terwel
D
Lasrado
R
Snauwaert
J
.
Changed conformation of mutant Tau-P301L underlies the moribund tauopathy, absent in progressive, nonlethal axonopathy of Tau-4R/2N transgenic mice
.
J Biol Chem
 
2005
;
280
:
3963
73
59.
Shim
SB
Lim
HJ
Chae
KR
.
Tau overexpression in transgenic mice induces glycogen synthase kinase 3β and β-catenin phosphorylation
.
Neuroscience
 
2007
;
146
:
730
40
60.
Colton
CA
Wilcock
DM
.
Assessing activation states in microglia
.
CNS Neurol Disord Drug Targets
 
2010
;
9
:
174
91
61.
Tynan
RJ
Naicker
S
Hinwood
M
.
Chronic stress alters the density and morphology of microglia in a subset of stress-responsive brain regions
.
Brain Behav Immun
 
2010
;
24
:
1058
68
62.
Perry
VH
.
Contribution of systemic inflammation to chronic neurodegeneration
.
Acta Neuropathol
 
2010
;
120
:
277
86
63.
Boimel
M
Grigoriadis
N
Lourbopoulos
A
.
Statins reduce the neurofibrillary tangle burden in a mouse model of tauopathy
.
J Neuropathol Exp Neurol
 
2009
;
68
:
314
25
64.
Ishizawa
K
Dickson
DW
.
Microglial activation parallels system degeneration in progressive supranuclear palsy and corticobasal degeneration
.
J Neuropathol Exp Neurol
 
2001
;
60
:
647
57
65.
Yoshiyama
Y
Higuchi
M
Zhang
B
.
Synapse loss and microglial activation precede tangles in a P301S tauopathy mouse model
.
Neuron
 
2007
;
53
:
337
51
66.
Angelucci
F
De
Bartolo P
Gelfo
F
.
Increased concentrations of nerve growth factor and brain-derived neurotrophic factor in the rat cerebellum after exposure to environmental enrichment
.
Cerebellum
 
2009
;
8
:
499
506
67.
Cotman
CW
Berchtold
NC
.
Exercise: A behavioral intervention to enhance brain health and plasticity
.
Trends Neurosci
 
2002
;
25
:
295
301
68.
Falkenberg
T
Mohammed
AK
Henriksson
B
.
Increased expression of brain-derived neurotrophic factor mRNA in rat hippocampus is associated with improved spatial memory and enriched environment
.
Neurosci Lett
 
1992
;
138
:
153
56
69.
Chen
MJ
Russo-Neustadt
AA
.
Exercise activates the phosphatidylinositol 3-kinase pathway
.
Brain Res Mol Brain Res
 
2005
;
135
:
181
93
70.
Mai
L
Jope
RS
Li
X
.
BDNF-mediated signal transduction is modulated by GSK3β and mood stabilizing agents
.
J Neurochem
 
2002
;
82
:
75
83
71.
Park
SW
Lee
JG
Ha
EK
.
Differential effects of aripiprazole and haloperidol on BDNF-mediated signal changes in SH-SY5Y cells
.
Eur Neuropsychopharmacol
 
2009
;
19
:
356
62
72.
Lee
JG
Cho
HY
Park
SW
.
Effects of olanzapine on brain-derived neurotrophic factor gene promoter activity in SH-SY5Y neuroblastoma cells
.
Prog Neuropsychopharmacol Biol Psychiatry
 
2010
;
34
:
1001
6
73.
Ortega
F
Perez-Sen
R
Morente
V
.
P2X7, NMDA and BDNF receptors converge on GSK3 phosphorylation and cooperate to promote survival in cerebellar granule neurons
.
Cell Mol Life Sci
 
2010
;
67
:
1723
33
74.
Lasagna-Reeves
CA
Castillo-Carranza
DL
Guerrero-Muoz
MJ
.
Preparation and characterization of neurotoxic tau oligomers
.
Biochemistry
 
2010
;
49
:
10039
41

Author notes

This study was supported by Grant No. 5721 from the Chief Scientist Office of the Ministry of Health, Israel.