- Email: [email protected]
Reactive astrocytes associated with plaques in TgCRND8 mouse brain and in human Alzheimer brain express phosphoprotein enriched in astrocytes (PEA-15)
FEBS Letters 587 (2013) 2448–2454
journal homepage: www.FEBSLetters.org
Reactive astrocytes associated with plaques in TgCRND8 mouse brain and in human Alzheimer brain express phosphoprotein enriched in astrocytes (PEA-15) Lynsie A.M. Thomason a,1, Laura J. Smithson b,1, Lili-Naz Hazrati c, JoAnne McLaurin a, Michael D. Kawaja b,d,⇑ a
Department of Laboratory Medicine and Pathobiology, University of Toronto, Toronto, Ontario, Canada Centre for Neuroscience Studies, Queen’s University, Kingston, Ontario, Canada Tanz Centre for Research in Neurodegenerative Diseases, University of Toronto, Toronto, Ontario, Canada d Department of Biomedical and Molecular Sciences, Queen’s University, Kingston, Ontario, Canada b c
a r t i c l e
i n f o
Article history: Received 22 March 2013 Revised 30 April 2013 Accepted 6 June 2013 Available online 19 June 2013
Edited by Jesus Avila
a b s t r a c t To identify potential biomarkers associated with Alzheimer’s disease (AD)-like neuropathologies in the murine brain, we conducted proteomic analyses of neocortices from TgCRND8 mice. Here we found that phosphoprotein enriched in astrocytes 15 kDa (PEA-15) is expressed at higher levels in the neocortical proteomes from 6-month old TgCRND8 mice, as compared to non-transgenic mice. Immunostaining for PEA-15 revealed reactive astrocytes associated with the neocortical amyloid plaques in TgCRND8 mice and in post-mortem human AD brains. This is the first report of increased PEA-15 expression in reactive astrocytes of an AD mouse model and human AD brains. Ó 2013 Published by Elsevier B.V. on behalf of the Federation of European Biochemical Societies.
Keywords: Alzheimer’s Astrocyte PEA-15 TgCRND8 Human
1. Introduction Alzheimer’s disease (AD), a neurodegenerative disease characterized by the presence of amyloid deposits, neurofibrillary tangles and neuronal loss [1], is currently the leading cause of dementia, estimated to affect approximately 35.6 million people worldwide [2]. While no treatment exists for AD, studies have shown that active medical management of individuals diagnosed with AD can improve their quality of life [3], suggesting that early diagnosis of AD would allow for a potentially better outcome for patients. There remains, however, a lack of reliable and early diagnostic biomarkers for AD, with definitive diagnosis still relying on postmortem neuropathological findings of amyloid plaques and neurofibrillary tangles [4].
Abbreviations: AD, Alzheimer’s disease; DED, death effector domain; PEA-15, phosphoprotein enriched in astrocytes ⇑ Corresponding author at: Department of Biomedical and Molecular Sciences, Queen’s University, Kingston, ON, Canada K7L 3N6. Fax: +1 613 533 2022. E-mail address: [email protected] (M.D. Kawaja). 1 These two authors contributed equally to this project.
The TgCRND8 mouse model shows both amyloid deposition and cognitive deficits by 3 months of age, as well as amyloid deposition patterns similar to that seen in human AD patients [5]. These mice express a mutated form of the human amyloid precursor protein (APP) under the control of the hamster prion promoter. This APP contains 2 common mutations seen in individuals with familial AD, the Swedish (K670N, M617L) and Indiana (V717F) mutations [5]. These mutations near the b-secretase and c-secretase cleavage sites on APP are believed to enhance its cleavage, and thus lead to an accelerated generation and accumulation of the amyloid-b peptide (Ab). The TgCRND8 mouse model is useful as an AD research tool due to its robust and early onset Ab pathology [5]. The aim of the present study was to conduct an unbiased broad-spectrum proteomic analysis of the neocortices from TgCRND8 mice, with the expectation of revealing potential biomarkers or therapeutic targets for AD. We have identified and validated that PEA-15 (phosphorylated protein enriched in astrocytes, 15 kDa) is found at higher levels in the neocortices of 6-month old TgCRND8 mice, as compared to age-matched non-transgenic littermates. Importantly, we show that PEA-15-immunopositive reactive astrocytes surround those plaques in the brains of TgCRND8 mice and those amyloid plaques in the postmortem brains of AD patients.
0014-5793/$36.00 Ó 2013 Published by Elsevier B.V. on behalf of the Federation of European Biochemical Societies. http://dx.doi.org/10.1016/j.febslet.2013.06.015
L.A.M. Thomason et al. / FEBS Letters 587 (2013) 2448–2454
2. Methods 2.1. Animals and tissue collection TgCRND8 mice and non-transgenic littermates (both males and females) were maintained on a C3H/C57Bl6 background [5] with food and water ad libitum. All methods were approved by the University of Toronto’s Animal Care Committee, following the guidelines set forth by the Canadian Council on Animal Care. Under deep anesthesia with sodium pentobarbital, these animals were perfused transcardially with either (1) 0.1 M phosphate buffered saline (PBS; pH 7.4, 50 ml volume) to allow for harvest of fresh tissues for proteomic analyses; (2) 4% paraformaldehyde in 0.1 M PBS (pH 7.4) to obtain fixed tissues for immunolocalization at the light microscope level; and (3) 2% paraformaldehyde and 2% glutaraldehyde in 0.1 M PBS (pH 7.4) to obtain fixed tissues for examination at the transmission electron microscope level. The fresh or fixed neocortices were removed from each animal. The fresh tissues were placed in labeled tubes and snap frozen in liquid nitrogen, and the fixed tissues were stored in fixative. For fresh tissue collection, 6-month old transgenic and non-transgenic mice (n = 10 mice per genotype) were sacrificed, and for fixed tissue collection, both 3- and 6-month old transgenic and non-transgenic mice (n = 14 per age and genotype) were sacrificed for fixation. 2.2. Isoelectric focusing and 2 dimensional (2D) gel electrophoresis Individual samples of frozen neocortices were homogenized in whole tissue homogenization buffer containing 8 M urea, 2 M thiourea, 4% w/v CHAPS in 15 mM Tris (pH 7.0). Samples were centrifuged and supernatants were stored at 80 °C. Protein concentration was determined using the 2D-Quantification Kit (Amersham Biosciences, Baie d’Urfe, QC, Canada). 200 lg of sample protein was reduced using 1% (w/v) DTT for 15 min and diluted to 450 ll in rehydration buffer containing 8 M urea, 2 M thiourea, 4% (w/v) CHAPS, 2% (v/v) HED and 0.5% ampholytes. 24-cm (pH range 4–7) immobilized pH gradient (IPG) strip gels (Amersham) were rehydrated overnight at room temperature in the buffer containing the sample protein. Isoelectric focusing was carried out on a Protein IEF cell machine (BioRad; Mississauga, ON, Canada, Queen’s University Protein Function Discovery Facility) according to the following program: 8 h at 250 V; 1 h at 500 V; 1 h at 1000 V; 3 h linear ramping to 10 000 V; 65 kV h at 10 000 V. Focused IPG strips were stored at 80 °C to prevent protein diffusion. A minimum of four IPG (pI 4–7) strips were prepared for each sample (three pools of cortices per phenotype). 2D SDS–PAGE was carried out according to Graham and colleagues [6]. Focused IPG strips were treated for 20 min in equilibration buffer containing 8 M urea, 30% (v/v) glycerol and 2% SDS in 50 mM Bis–Tris (pH 8.8), supplemented with 1% (w/v) DTT and again for 20 min in equilibration buffer supplemented with 2.5% (w/v) iodoacetamide. Equilibrated strips were placed on top of 1 mm thick 10% acrylamide gels (with Bis–Tris (pH 7.4)) and large format 2D SDS–PAGE electrophoresis was carried out in a DALT6 apparatus (Amersham) in 1 MOPS electrophoresis buffer and run at 80 V at 4 °C overnight. Gels were stained with a standard silver stain [7] and scanned using a UMAX Powerlook2100XL. Gels were dried between porous cellophane sheets for spot excision and long-term storage. 2.3. Spot identification and mass spectrometry Protein samples were prepared for mass spectroscopy according to McDonald and colleagues [8]. Scanned images of gels were ana-
2449
lyzed manually, and overlaid images were compared visually to identify differences in location or intensity of corresponding spots (relative to background and surrounding spots). Spots were selected as candidates for further processing if the difference in abundance was present in every gel from transgenic compared to non-transgenic proteome (i.e., three pools of cortices per phenotype, four gels per phenotype, and four spots picked from every phenotype). Candidate spots were then manually excised from the gels, rehydrated and de-stained with aqueous 15 mM potassium ferricyanide and 50 mM sodium thiosulfate. Gel plugs digested with 6 ng/ml trypsin (sequencing Grade-modified trypsin; Promega, Napean, ON, Canada) for 5 h at 37 °C. Peptides were extracted using 1% formic acid/2% acetonitrile followed by 50% acetonitrile. Extraction solutions were combined and evaporated down to 10 ll. Peptides were spotted onto a MALDI target plate on top of a re-crystallized spot of alpha-cyanno-4-hydroxy-cinnamic acid matrix and spectra were collected using a Voyager DE-Pro matrixassisted laser desorption ionization time of flight massspectrometer (MALDI-TOF MS; PerSeptive Biosystems, Framingham, MA, USA). Peaks were analyzed using Data Explorer software; spectra were adjusted to baseline, noise filtered, externally calibrated against a spectrum of a known 4-peptide mix and internally calibrated against Trypsin I and Trypsin III peaks, when present. Search parameters allowed for one missed cleavage as well as peptide oxidation and carbamidomethylation. Peptide masses were manually selected and the resulting peak mass lists were searched against the NCBI database using the peptide mass fingerprinting database search tool MASCOT PMF search engine (http://www.matrixscience.com). Spectra which did not contain adequate peaks to get a statistically significant hit (significance as determined by MASCOT) were ZIP-TIPPED (Millipore Corporation, Bedford MA, USA) and eluted into 70% acetonitrile in 0.1% TFA in ddH2O to purify and concentrate the peptides and new MS spectra were collected. Only protein identifications with statistically significant MOWSE scores and which approximately matched the predicted molecular weight and pI values were included. 2.4. Immunoblotting For 2D immunoblotting, 200 lg of protein per sample was diluted to 450 ll in rehydration buffer and run out on 24-cm IPG strips (pH 4–7; Amersham) and equilibrated as described above. Equilibrated strips were placed on top of 1 mm thick 10% acrylamide gels (with Bis–Tris (pH 7.4)) and large format 2D SDS–PAGE electrophoresis was carried out as also described above. A 7.0 cm section of the 10% acylamide gel which contained the spot identified as phosphoprotein enriched in astrocytes 15 kDa (PEA-15) was cut out and transferred (for 1 h at 0.4 Å) onto nitrocellulose membranes. Membranes were blocked and then incubated in primary antibody against rabbit anti-PEA-15 IgG (dilution 1:500; Abcam, Cambridge, MA) at room temperature for 2 h. After rinsing, the membranes were incubated in goat anti-rabbit IgG (1:10,000; HRP conjugated; Jackson Immunoresearch, West Grove, PA, USA). Membranes were then treated with Chemi-Illuminescence (Amersham) according to the manufacturer’s protocol. 2.5. Electron microscopy Neocortical mouse tissues fixed with 2% paraformadehyde and 2% gluataraldehyde were cut at 100 lm thickness using a Vibratome. The sections were trimmed into 1 1 cm2 pieces, which were subsequently fixed overnight in 1% phosphate-buffered osmium tetroxide. After rinsing and dehydrating through a graded series of methanols, the sections were cleared with propylene oxide and embedded Epon. Both semithin and ultrathin sections were cut using a LKB Ultramicrotome, the latter of which were col-
2450
L.A.M. Thomason et al. / FEBS Letters 587 (2013) 2448–2454
lected onto copper grids and stained with uranyl acetate and lead citrate. Stained ultrathin sections were viewed and photographed with a Hitachi 7000 transmission electron microscope. 2.6. Post-mortem human neocortical tissues Human specimens were obtained from a collection of classified paraffin-embedded frontal and temporal neocortical tissues. NonAD tissues included those from the following three adults: (1) 72-year old female who died of coronary artery disease; (2) 76year old male who died of coronary artery disease; and (3) 88-year old male who died of pneumonia. None of these tissues had ADassociated changes. AD tissues included those from the following four adults: (1) 71-year old male with an AD Braak score of 6/6; (2) 73-year old male with an AD Braak score of 4/6; (3) 78-year old female with an AD Braak score 5/6; and, 88-year old female with an AD Braak score of 5/6. 2.7. Immunohistochemistry Fixed mouse brains were sectioned at 40 lm thickness using a freezing microtome. The sections were collected in 0.1 M Tris buffered saline (TBS; pH 7.4) and processed for the immunohistochemical detection of PEA-15, as well as glial fibrillary acidic protein (GFAP; astrocyte marker; DakoCytomation), Ionized calcium binding adaptor molecule 1 (Iba1, microglial marker), and neuronal nuclei (NeuN; neuronal marker; Millipore). Sections were first incubated in 0.3% hydrogen peroxide in 0.1 M TBS for 1 h and then 10% bovine serum albumin (BSA) in 0.1 M Tris and 0.25% TritonX-100 (TBX) for 1 h at room temperature. Sections were then incubated in avidin and biotin for 1 h each at room temperature. Next, sections were incubated in rabbit anti-GFAP IgG (1:1000), rabbit anti-Iba1 IgG (1:1000), or mouse anti-NeuN IgG (1:1000) (each diluted in 3% BSA in TBX) for two days at room temperature. After rinsing in 0.1 M TBS for 15–30 min, the sections were incubated with biotinylated goat anti-rabbit IgG (1:100; Vector Laboratories) for 2 h at room temperature. The sections were then incubated with avidin–biotin complex (EliteÒ ABC Peroxidase kit; Vector Laboratories) for 2 h at room temperature. The sections were then reacted in a solution containing 0.05% diaminobenzidine (DAB) tetrahydrochloride, 0.04% nickel chloride, and 0.015% hydrogen peroxide in 0.1 M TBS until the brown reaction product was clearly visible, as viewed with a stereomicroscope. The sections were then immediately transferred into buffer to stop the reaction. The sections were rinsed thoroughly (3) in 0.1 M TBS for no less than 30 min. For the benzidine dihydrochloride (BDHC) blue reaction, sections were first rinsed in 0.01 M phosphate buffer (pH 6.6) for 30 min, and then incubated in 10% BSA in 0.25% TBX for 1 h at room temperature. Next the sections were incubated in rabbit anti-PEA-15 IgG (1:500) or GFAP (1:1000) diluted in 3% BSA in 0.25% TBX for 2 days at room temperature. After rinsing in 0.1 M TBS, the sections were incubated with biotinylated goat anti-rabbit IgG (1:100; Vector Laboratories) for 2 h at room temperature. The sections were then incubated with avidin–biotin complex (EliteÒ ABC Peroxidase kit; Vector Laboratories) for 2 h at room temperature. The sections were then reacted in a solution containing 0.01% BDHC, 0.025% sodium nitroferricyanide, and 0.03% hydrogen peroxide in 0.01 M phosphate buffer (pH 6.6). Once the blue crystalline reaction product was clearly visible, the reaction was stopped by rinsing the sections in 0.01 M PBS (pH 6.6). Sections were then mounted onto slides, dried overnight at room temperature, dehydrated through a graded series of ethanols, cleared in HistoClearÒ, and then coverslipped in Permount mounting media. Sections were viewed under bright field optics, and images were acquired using a Zeiss microscope and Axiovision software.
Procedures for immunostaining of human neocortical tissues were the same as that described above, with the only exception being that the paraffin-embedded tissues (already mounted on slides) were cleared of paraffin and hydrated to water before incubating in 0.3% hydrogen peroxide.
3. Results Neocortical proteins were fractionated by 2D gel electrophoresis. No fewer than three individual samples of neocortices isolated from 6-month old TgCRND8 and non-transgenic littermates were resolved and silver-stained to reveal protein spot differences. Four proteins were found to be consistently present at more abundant levels in the TgCRND8 proteomes, as compared to age-matched non-transgenic proteomes (Fig. 1). Using mass spectrometry, two of these proteins in the TgCRND8 proteomes, labeled 3 and 4 (Fig. 1A, B) were identified as PEA-15 and Ab, respectively. The spot for PEA-15 on the silver-stained gels was consistently more intense in the TgCRND8 proteomes, as compared to the non-transgenic proteomes. The spot for Ab on the silver-stained gels was consistently found only in the TgCRND8 proteomes, providing direct evidence of the ectopic expression of Ab in these transgenic mice. To validate the presence of PEA15 in the neocortical proteomes of 6-month old TgCRND8 (as compared to the neocortical proteomes of non-transgenic mice), protein samples were again fractioned by 2D gel electrophoresis (with no subsequent silver staining) and then transferred to membranes for immunoblotting. Membranes were incubated with PEA-15 those having TgCRND8 neocortical tissues yielded two immunopositive spots, where as the non-transgenic neocortical tissues did not (insets of Fig. 1A and B). These spots had molecular weights of 15 kDa and isoelectric points of pH 5, and were consistent with previously reported 2D gelmigration patterns for PEA-15 [9]. Since PEA-15 can exist in non-phosphorylated, single-phosphorylated, or double-phosphorylated forms [9], the identification of two immunopositive spots may represent shifts in the phosphorylation status of PEA-15 in the brains of 6month old TgCRND8 mice. Regarding the described phenotype of 3- and 6-month old TgCRND8 mice, the single most conspicuous difference is the increased incidence of Ab plaques throughout the neocortex of the older aged mice. Non-transgenic littermates at both ages lacked such Ab plaques (Fig. 1C; double immunostaining for NeuN (pan neuronal marker) using the chromogen DAB to reveal positive immunoreactivity – brown reaction product, and for PEA-15 using the chromogen BDHC to reveal positive immunoreactivity – blue reaction product). In 6-month old TgCRND8 mice, these plaques found in the grey matter (Fig. 1D) and white matter (Fig. 1E) of the neocortices displayed prominent immunostaining for PEA-15 (blue reaction product). The intensity of this plaque-associated immunostaining was greater in the 6-month old TgCRND8 mice as compared to their 3-month old transgenic counterparts (data not shown). This immunostaining was deemed to be positive because no staining was seen in either the absence of the primary anti-PEA-15 IgG or the presence of the incorrect secondary IgG (e.g., using biotinylated rabbit-anti-goat IgG). Positive immunostaining for PEA-15 (blue reaction product) was also seen among capillaries of both transgenic and non-transgenic mice, likely confined to the end-foot processes of astrocytes. Double immunostaining for GFAP (brown reaction product) and PEA-15 (blue reaction product) yielded GFAP-immunopositive reactive astrocytes associated with plaques in the neocortices of 6-month old TgCRND8 mice; importantly, these GFAP-immunopositive astrocytes displayed positive immunoreactivity for PEA-15 (Fig. 1F). Double immunostaining for Iba1 (using the chromogen DAB) and GFAP
L.A.M. Thomason et al. / FEBS Letters 587 (2013) 2448–2454
2451
Fig. 1. Silver-stained 2D gels of neocortical proteins from 6 month old non-transgenic mice (A) and age-matched TgCRND8 mice (B) revealed 4 protein differences between these two proteomes. Mass spectrometry identified these proteins as small nuclear RNA-activating complex polypeptide 1 (spot 1), IgG-dependent histamine-releasing factor (spot 2), phosphoprotein enriched in astrocytes (PEA-15, spot 3), and amyloid-b (spot 4). Immunostaining for PEA-15 on 2D gels of neocortical tissues from 6 month old nontransgenic and TgCRND8 mice (insets in A and B) revealed two protein spots only in the proteomes of transgenic mice. Neocortical tissues of 6 month old non-transgenic mice (C) displayed PEA-15-immunopositive blood vessels among NeuN-immunopositive neurons. Neocortical tissues of 6 month old TgCRND8 mice (D, E) displayed PEA-15immunopositive blood vessels among NeuN-immunopositive neurons, as well as PEA-15-immunopositive plaques (D, box; E, circle) in both the grey and deep white matter. These neocortical plaques (⁄ in F) were surrounded by reactive astrocytes displaying positive immunostaining for GFAP (brown reaction product) and PEA-15 (blue reaction product). These neocortical plaques (⁄ in G) were composed of Iba1-immunopositive microglia (brown reaction product) and were surrounded by GFAP-immunopositive reactive astrocytes (blue reaction product). Scale bars = 100 lm (C, D) and 20 lm (F, G).
(using the chromogen BDHC) yielded Iba1-immunopositive reactive microglial cells in the core of the neocortical plaques of 6month old TgCRND8 mice; in contrast to the reactive astrocytes, these Iba1-immunopositive microglial cells displayed no positive immunoreactivity for GFAP; double immunostaining with Iba1 and PEA-15 failed to label PEA-15-positive astrocytes in the neocortices of 6-month old TgCRND8 mice (Fig. 1G).
The presence of PEA-15 immunostaining in those reactive astrocytes associated with neocortical plaques of 6-month old TgCRND8 mice warranted further examination of these glial cells at the ultrastructural level. Reactive astrocytes in immediate proximity to dystrophic neurites in the plaques had fragmented rough endoplasmic reticulum, pieces of which were found in smaller astrocytic processes throughout the plaques (Fig. 2A and B).
2452
L.A.M. Thomason et al. / FEBS Letters 587 (2013) 2448–2454
Fig. 2. Transmission electron microscopy of neocortical plaques from 6 month old TgCRND8 mice revealed astrocytic processes (shaded blue) having fragmented pieces of rough endoplasmic reticulum (arrowheads) (A, B); these astrocytic processes were in close proximity to dystrophic neurites (⁄). Neocortical tissues (frontal lobe) from an AD patient having a Braak score of 4/6 had plaques surrounded by reactive astrocytes immunopositive for both GFAP (brown reactive product) and PEA-15 (blue reaction product) (C, D). Some of these reactive astrocytes were also seen extending processes to adjacent blood vessels (BV). Though control neocortical tissues (frontal lobe) from a non-AD patient lacked plaques, astrocytes in grey and white matter displayed positive immunostaining for GFAP (brown reactive product) and PEA-15 (blue reaction product) (E, F). Scale bars = 1 lm (A), 0.8 lm (B), 20 lm (C, D), and 40 lm (E, F).
Postmortem human tissues revealed GFAP-immunopositive reactive astrocytes associated with neocortical (frontal and temporal) plaques in specimens diagnosed with AD (Braak scores ranging from 4/6 to 6/6). These plaque-associated reactive astrocytes displayed positive immunostaining for PEA-15 (Fig. 2C and D). These same PEA-15-immunopositive reactive astrocytes could be seen extending processes in and around the plaques, as well as toward adjacent capillaries. It is important to note that PEA-15 immunoreactivity was found through the neocortices of both AD and non-AD human specimens, including both neuropil and adjacent white
matter tracts (Fig. 2C–F). This pattern was not seen in either transgenic or non-transgenic mice, in which PEA-15 immunoreactivity was restricted to those reactive astrocytes associated with the plaques of TgCRND8 mice. 4. Discussion PEA-15 is a death effector domain (DED) containing protein found predominantly in the central nervous system, with expression previously observed in astrocytes and neurons alike [10]. This
L.A.M. Thomason et al. / FEBS Letters 587 (2013) 2448–2454
protein is highly conserved among mammals, with 96% similarity between the mouse and human forms [11]. PEA-15 was originally described as a protein related to diabetes mellitus and is now known to be involved in regulating numerous other cell processes [9]. The N-terminal DED in PEA-15 has led to investigations into its role in regulating apoptosis [9,10,12]. More recently it has been discovered that PEA-15 has a diverse set of regulatory functions, including roles in proliferation, glucose metabolism, cell cycle control, cell adhesion, migration and autophagy and its expression has been shown to be altered in various cancers [9,10,12–14]. Given the diversity of functions attributed to PEA-15, its enhanced detection among plaque-associated reactive astrocytes in TgCRND8 mice might have beneficial effects, exacerbate glial pathology, and/or merely serve as a by-product of the disease state. Though the scope of our study did not include a functional analysis of increased PEA-15 expression in the neocortices of TgCRND8 mice, it is tempting to speculate that PEA-15 expression in plaque-associated reactive astrocytes may be playing an anti-apoptotic role in this glial sub-population. It has been previously shown that PEA-15 expression decreases the susceptibility of astrocytes to tumor necrosis factor alpha (TNFa)-induced apoptosis, providing a possible mechanism as to why astrocytes are often resistant to apoptosis during inflammation or disease states [10]. Neuroinflammation plays a large role in the pathogenesis of AD [15] and increased TNFa levels have been found in patients with AD [16,17]. Further, in the TgCRND8 mouse model, intense TNFa immunostaining has been observed in reactive microglia and astrocytes surrounding Ab plaques in the neocortex and hippocampus [18]. Therefore, it is possible that increased PEA-15 expression in plaque-associated reactive astrocytes may protect the astrocytes from apoptosis induced by TNF in the vicinity of the Ab plaques. It is also possible that a more general relationship may exist between astrocyte PEA-15 expression and the diseased central nervous system, given that ‘‘marked‘‘ PEA-15 expression was found in reactive astrocytes located in the grey matter of the spinal cord of an ALS mouse model (SOD1 G93A mutant) [12]. As previously mentioned, our report provides direct and validated evidence that detection of PEA-15 is increased in the brains of an AD mouse model. Contrary to our study, Takano et al. [19] reported a decrease in PEA-15 expression in the hippocampus of APPE693D mice (as determined solely by quantitative comparison between wild type and transgenic hippocampal proteomes). This discrepancy may be explained by certain differences between the two studies, including the different mouse models and brain regions examined. Our experiments used the TgCRND8 model of AD, which shows the presence of Ab oligomers, fibrils and plaques [5,20], whereas the Takano et al. [19] assessed the APPE693D model, which generates Ab oligomers in the absence of plaques [19]. Given that we localized increased PEA-15 immunodetection exclusively to plaque-associated reactive astrocytes, we propose that the generation of Ab-containing plaques is critical to PEA-15 expression in surrounding astrocytes. In the absence of Ab-containing plaques, as seen in the APPE693D mouse model, a similar localization of PEA-15 to reactive astrocytes is unlikely to occur. Further, the altered PEA15 expression observed in the two studies was found in two different brain regions: Takano and colleagues reported decreased expression of PEA-15 in the hippocampus of APPE693D mice [19], whereas we detected increased PEA-15 expression in the neocortices of TgCRND8 mice. Finally, our proteomic discovery of increased levels of PEA-15 in the neocortices of TgCRND8 mice was validated by immunostaining of brain tissues and immunoblotting of brain proteins in these mice. Most importantly, we provide translational evidence that increased PEA-15 immunostaining of plaque-associated reactive astrocytes in transgenic mice is also seen in human AD tissue, where Ab plaques are surrounded by PEA-15-immunopositive reactive astrocytes.
2453
Collectively, these data show that 2D gel electrophoresisfractionation of neocortices from 6-month old TgCRND8 and nontransgenic mice yields consistent patterns of differential expression for several proteins, in particular increased levels of Ab and PEA-15. Using mass spectrometry, Ab was only found in the proteomes of TgCRND8 mice, confirming its accumulation in the brains of these transgenic animals. The other protein of interest, PEA-15, was detected at high levels in the proteomes of TgCRND8 mice; this increase in PEA-15 was confirmed by both immunoblotting and tissue immunostaining. In the neocortices of the TgCRND8 mice, PEA-15 expression was localized to plaque-associated reactive astrocytes, a pattern that was also observed in post-mortem human AD tissue. As a biomarker of AD, increased levels of astrocytic PEA-15 may be used, along with other pathological hallmarks, to assess severity of the disease progress at the time of death. Future studies will assess whether PEA-15 is detectable in human cerebrospinal fluid and/or serum, and if such detection can be linked to cognitive impairment associated with AD prior to death. Acknowledgements We would like to thank Ms. Mary E. Brown for assisting with harvesting the mouse tissues, Mr. Casey N. Petrie for tissue homogenizations, Ms. Verna Norkum for the preparation of the tissues for immunostaining and examination at the transmission electron microscope level, and Mr. David McLeod for his assistance with the mass spectrometry (Queen’s Protein Function Discovery Facility). Ms. Thomason is supported by an Ontario Graduate Scholarship and the Heart and Stroke Foundation Centre for Stroke Recovery Trainee Stimulus Fellowship, and Ms. Smithson is supported by a Banting and Best Scholarship from the Canadian Institutes of Health Research. This work was supported by CIHR (J.M.: MOP-10246) and The Botterell Foundation of Queen’s University (M.D.K.). References [1] Götz, J., Ittner, L.M., Schonrock, N. and Cappai, R. (2008) An update on the toxicity of A in Alzheimer’s disease. Neuropsychiatr Dis. Treat. 4, 1033–1042. [2] World Health Organization (WHO) (2012), World Alzheimer Report 2012, a public health priority. [3] Association, Alzheimer.’s. (2010) Alzheimer’s disease facts and figures. Alzheimers Dement. 6, 158–194. [4] Bird, T. D. (2010) Alzheimer’s disease overview. Gene Rev. http:// www.ncbi.nlm.nih.gov/books/NBK1161/. Last updated December 2012, (accessed 21.02.2012). [5] Chishti, M.A., Yang, D.S., Janus, C., Phinney, A.L., Horne, P., Pearson, J., Strome, R., Zuker, N., Loukides, J., French, J., Turner, S., Lozza, G., Grilli, M., Kunicki, S., Morissette, C., Paquette, J., Gervais, F., Bergeron, C., Fraser, P.E., Carlson, G.A., St. George-Hyslop, P. and Westaway, D. (2001) Early-onset amyloid deposition and cognitive deficits in transgenic mice expressing a double mutant form of amyloid precursor protein 695. J. Biol. Chem. 276, 21562–21570. [6] Graham, D.R., Garnham, C.P., Fu, Q., Robbins, J. and Van Eyk, J.E. (2005) Improvements in two-dimensional gel electrophoresis by utilizing a low cost ‘‘in-house’’ neutral pH sodium dodecyl sulfate-polyacylamide gel electrophoresis system. Proteomics 5, 2309–2314. [7] Shevchenko, A., Jensen, O., Podtelejnikov, A., Sagliocco, F., Wilm, M., Vorm, O., Mortensen, P., Shevchenko, A., Boucherie, H. and Mann, M. (1996) Linking genome and proteome by mass spectrometry: large-scale identification of yeast proteins from two dimensional gels. Proc. Natl. Acad. Sci. USA 93, 14440–14445. [8] McDonald, T.G., Scott, S.A., Kane, K.M. and Kawaja, M.D. (2009) Proteomic assessment of sympathetic ganglia from adult mice that possess null mutations of ExonIII or ExonIV in the p75 neurotrophin receptor gene. Brain Res. 1253, 1–14. [9] Eckert, A., Böck, C., Tagscherer, K.E., Haas, T.L., Grund, K., Sykora, J., HeroldMende, C., Ehemann, V., Hollstein, M., Chneiweiss, H., Wiestler, O.D., Walczak, H. and Roth, W. (2008) The PEA-15/PED protein protects glioblastoma cells from glucose deprivation-induced apoptosis via the ERK/MAP kinase pathway. Oncogene 27, 1155–1166. [10] Kitsberg, D., Formstecher, E., Fauquet, M., Kubes, M., Cordier, J., Canton, B., Pan, G., Gloinski, J. and Chneiweiss, H. (1999) Knock-out of the neural death effector domain protein PEA-15 demonstrates that its expression protects astrocytes from TNF-induced apoptosis. J. Neurosci. 19, 8244–8251.
2454
L.A.M. Thomason et al. / FEBS Letters 587 (2013) 2448–2454
[11] Estellés, A., Yokoyama, M., Nothias, F., Vincent, J.-D., Glowinski, J., Vernier, P. and Chneiweiss, H. (1996) The major astrocytic phosphoprotein PEA-15 is encoded by two mRNAs conserved on their full length in mouse and human. J. Biol. Chem. 271, 14800–14806. [12] Sharif, A., Renault, F., Beuvon, R., Castellanos, B., Canton, L., Barbeito, M., Junier, P. and Chneiweiss, H. (2004) The expression of PEA-15 (phosphoprotein enriched in astrocytes of 15 kDa) defines subpopulations of astrocytes and neurons throughout the adult mouse brain. Neuroscience 126, 263–275. [13] Böck, B.C., Tagscherer, K.E., Fassl, A., Krämer, A., Oehme, I., Zentgraf, H.-W., Keith, M. and Roth, W. (2010) The PEA-15 protein regulates autophagy via activation of JNK. J. Biol. Chem. 285, 21644–21654. [14] Formisano, P., Ragno, P., Pesapane, A., Alfano, D., Alberobello, A.T., Rea, V.E.A., Giusto, R., Rossi, F.W., Beguinot, F., Rossi, G. and Montuori, N. (2012) PED/PEA-15 interacts with the 67 kDa laminin receptor and regulates cell adhesion, migration, proliferation and apoptosis. J. Cell Mol. Med. 16, 1435–1446. [15] McGeer, P.L. and McGeer, E.G. (2002) Local neuroinflammation and the progression of Alzheimer’s disease. J. Neurovirol. 8, 529–538.
[16] Fillit, H., Ding, W.H., Buee, L., Kalman, J., Altstiel, L., Lawlor, B. and Wolf-Klein, G. (1991) Elevated circulating tumor necrosis factor levels in Alzheimer’s disease. Neurosci. Lett. 129, 318–320. [17] Alvarez, A., Cacabelos, R., Sanpedro, C., García-Fantini, M. and Aleixandre, M. (2007) Serum TNF-alpha levels are increased and correlate negatively with free IGF-I in Alzheimer disease. Neurobiol. Aging 28, 533–536. [18] Luccarini, I., Grossi, C., Traini, C., Fiorentini, A., Dami, T.E. and Casamenti, F. (2012) A plaque-associated glial reaction as a determinant of apoptotic neuronal death and cortical gliogenesis: a study in APP mutant mice. Neurosci. Lett. 506, 94–99. [19] Takano, M., Maekura, K., Otani, M., Sano, K., Nakamura-Hirota, T., Tokuyama, S., Min, K.S., Tomiyama, T., Mori, H. and Matsuyama, S. (2012) Proteomic analysis of the brain tissues from a transgenic mouse model of amyloid oligomers. Neurochem. Int. 61, 347–355. [20] McLaurin, J., Kierstead, M.E., Brown, M.E., Hawkes, C.A., Lambermon, M.H., Phinney, A.L., Darabie, A.A., Cousins, J.E., French, J.E., Lan, M.F., Chen, F., Wong, S.S., Mount, H.T., Fraser, P.E., Westaway, D. and St. George-Hyslop, P. (2006) Cyclohexanehexol inhibitors of Abeta aggregation prevent and reverse Alzheimer phenotype in a mouse model. Nat. Med. 12, 801–808.
journal homepage: www.FEBSLetters.org
Reactive astrocytes associated with plaques in TgCRND8 mouse brain and in human Alzheimer brain express phosphoprotein enriched in astrocytes (PEA-15) Lynsie A.M. Thomason a,1, Laura J. Smithson b,1, Lili-Naz Hazrati c, JoAnne McLaurin a, Michael D. Kawaja b,d,⇑ a
Department of Laboratory Medicine and Pathobiology, University of Toronto, Toronto, Ontario, Canada Centre for Neuroscience Studies, Queen’s University, Kingston, Ontario, Canada Tanz Centre for Research in Neurodegenerative Diseases, University of Toronto, Toronto, Ontario, Canada d Department of Biomedical and Molecular Sciences, Queen’s University, Kingston, Ontario, Canada b c
a r t i c l e
i n f o
Article history: Received 22 March 2013 Revised 30 April 2013 Accepted 6 June 2013 Available online 19 June 2013
Edited by Jesus Avila
a b s t r a c t To identify potential biomarkers associated with Alzheimer’s disease (AD)-like neuropathologies in the murine brain, we conducted proteomic analyses of neocortices from TgCRND8 mice. Here we found that phosphoprotein enriched in astrocytes 15 kDa (PEA-15) is expressed at higher levels in the neocortical proteomes from 6-month old TgCRND8 mice, as compared to non-transgenic mice. Immunostaining for PEA-15 revealed reactive astrocytes associated with the neocortical amyloid plaques in TgCRND8 mice and in post-mortem human AD brains. This is the first report of increased PEA-15 expression in reactive astrocytes of an AD mouse model and human AD brains. Ó 2013 Published by Elsevier B.V. on behalf of the Federation of European Biochemical Societies.
Keywords: Alzheimer’s Astrocyte PEA-15 TgCRND8 Human
1. Introduction Alzheimer’s disease (AD), a neurodegenerative disease characterized by the presence of amyloid deposits, neurofibrillary tangles and neuronal loss [1], is currently the leading cause of dementia, estimated to affect approximately 35.6 million people worldwide [2]. While no treatment exists for AD, studies have shown that active medical management of individuals diagnosed with AD can improve their quality of life [3], suggesting that early diagnosis of AD would allow for a potentially better outcome for patients. There remains, however, a lack of reliable and early diagnostic biomarkers for AD, with definitive diagnosis still relying on postmortem neuropathological findings of amyloid plaques and neurofibrillary tangles [4].
Abbreviations: AD, Alzheimer’s disease; DED, death effector domain; PEA-15, phosphoprotein enriched in astrocytes ⇑ Corresponding author at: Department of Biomedical and Molecular Sciences, Queen’s University, Kingston, ON, Canada K7L 3N6. Fax: +1 613 533 2022. E-mail address: [email protected] (M.D. Kawaja). 1 These two authors contributed equally to this project.
The TgCRND8 mouse model shows both amyloid deposition and cognitive deficits by 3 months of age, as well as amyloid deposition patterns similar to that seen in human AD patients [5]. These mice express a mutated form of the human amyloid precursor protein (APP) under the control of the hamster prion promoter. This APP contains 2 common mutations seen in individuals with familial AD, the Swedish (K670N, M617L) and Indiana (V717F) mutations [5]. These mutations near the b-secretase and c-secretase cleavage sites on APP are believed to enhance its cleavage, and thus lead to an accelerated generation and accumulation of the amyloid-b peptide (Ab). The TgCRND8 mouse model is useful as an AD research tool due to its robust and early onset Ab pathology [5]. The aim of the present study was to conduct an unbiased broad-spectrum proteomic analysis of the neocortices from TgCRND8 mice, with the expectation of revealing potential biomarkers or therapeutic targets for AD. We have identified and validated that PEA-15 (phosphorylated protein enriched in astrocytes, 15 kDa) is found at higher levels in the neocortices of 6-month old TgCRND8 mice, as compared to age-matched non-transgenic littermates. Importantly, we show that PEA-15-immunopositive reactive astrocytes surround those plaques in the brains of TgCRND8 mice and those amyloid plaques in the postmortem brains of AD patients.
0014-5793/$36.00 Ó 2013 Published by Elsevier B.V. on behalf of the Federation of European Biochemical Societies. http://dx.doi.org/10.1016/j.febslet.2013.06.015
L.A.M. Thomason et al. / FEBS Letters 587 (2013) 2448–2454
2. Methods 2.1. Animals and tissue collection TgCRND8 mice and non-transgenic littermates (both males and females) were maintained on a C3H/C57Bl6 background [5] with food and water ad libitum. All methods were approved by the University of Toronto’s Animal Care Committee, following the guidelines set forth by the Canadian Council on Animal Care. Under deep anesthesia with sodium pentobarbital, these animals were perfused transcardially with either (1) 0.1 M phosphate buffered saline (PBS; pH 7.4, 50 ml volume) to allow for harvest of fresh tissues for proteomic analyses; (2) 4% paraformaldehyde in 0.1 M PBS (pH 7.4) to obtain fixed tissues for immunolocalization at the light microscope level; and (3) 2% paraformaldehyde and 2% glutaraldehyde in 0.1 M PBS (pH 7.4) to obtain fixed tissues for examination at the transmission electron microscope level. The fresh or fixed neocortices were removed from each animal. The fresh tissues were placed in labeled tubes and snap frozen in liquid nitrogen, and the fixed tissues were stored in fixative. For fresh tissue collection, 6-month old transgenic and non-transgenic mice (n = 10 mice per genotype) were sacrificed, and for fixed tissue collection, both 3- and 6-month old transgenic and non-transgenic mice (n = 14 per age and genotype) were sacrificed for fixation. 2.2. Isoelectric focusing and 2 dimensional (2D) gel electrophoresis Individual samples of frozen neocortices were homogenized in whole tissue homogenization buffer containing 8 M urea, 2 M thiourea, 4% w/v CHAPS in 15 mM Tris (pH 7.0). Samples were centrifuged and supernatants were stored at 80 °C. Protein concentration was determined using the 2D-Quantification Kit (Amersham Biosciences, Baie d’Urfe, QC, Canada). 200 lg of sample protein was reduced using 1% (w/v) DTT for 15 min and diluted to 450 ll in rehydration buffer containing 8 M urea, 2 M thiourea, 4% (w/v) CHAPS, 2% (v/v) HED and 0.5% ampholytes. 24-cm (pH range 4–7) immobilized pH gradient (IPG) strip gels (Amersham) were rehydrated overnight at room temperature in the buffer containing the sample protein. Isoelectric focusing was carried out on a Protein IEF cell machine (BioRad; Mississauga, ON, Canada, Queen’s University Protein Function Discovery Facility) according to the following program: 8 h at 250 V; 1 h at 500 V; 1 h at 1000 V; 3 h linear ramping to 10 000 V; 65 kV h at 10 000 V. Focused IPG strips were stored at 80 °C to prevent protein diffusion. A minimum of four IPG (pI 4–7) strips were prepared for each sample (three pools of cortices per phenotype). 2D SDS–PAGE was carried out according to Graham and colleagues [6]. Focused IPG strips were treated for 20 min in equilibration buffer containing 8 M urea, 30% (v/v) glycerol and 2% SDS in 50 mM Bis–Tris (pH 8.8), supplemented with 1% (w/v) DTT and again for 20 min in equilibration buffer supplemented with 2.5% (w/v) iodoacetamide. Equilibrated strips were placed on top of 1 mm thick 10% acrylamide gels (with Bis–Tris (pH 7.4)) and large format 2D SDS–PAGE electrophoresis was carried out in a DALT6 apparatus (Amersham) in 1 MOPS electrophoresis buffer and run at 80 V at 4 °C overnight. Gels were stained with a standard silver stain [7] and scanned using a UMAX Powerlook2100XL. Gels were dried between porous cellophane sheets for spot excision and long-term storage. 2.3. Spot identification and mass spectrometry Protein samples were prepared for mass spectroscopy according to McDonald and colleagues [8]. Scanned images of gels were ana-
2449
lyzed manually, and overlaid images were compared visually to identify differences in location or intensity of corresponding spots (relative to background and surrounding spots). Spots were selected as candidates for further processing if the difference in abundance was present in every gel from transgenic compared to non-transgenic proteome (i.e., three pools of cortices per phenotype, four gels per phenotype, and four spots picked from every phenotype). Candidate spots were then manually excised from the gels, rehydrated and de-stained with aqueous 15 mM potassium ferricyanide and 50 mM sodium thiosulfate. Gel plugs digested with 6 ng/ml trypsin (sequencing Grade-modified trypsin; Promega, Napean, ON, Canada) for 5 h at 37 °C. Peptides were extracted using 1% formic acid/2% acetonitrile followed by 50% acetonitrile. Extraction solutions were combined and evaporated down to 10 ll. Peptides were spotted onto a MALDI target plate on top of a re-crystallized spot of alpha-cyanno-4-hydroxy-cinnamic acid matrix and spectra were collected using a Voyager DE-Pro matrixassisted laser desorption ionization time of flight massspectrometer (MALDI-TOF MS; PerSeptive Biosystems, Framingham, MA, USA). Peaks were analyzed using Data Explorer software; spectra were adjusted to baseline, noise filtered, externally calibrated against a spectrum of a known 4-peptide mix and internally calibrated against Trypsin I and Trypsin III peaks, when present. Search parameters allowed for one missed cleavage as well as peptide oxidation and carbamidomethylation. Peptide masses were manually selected and the resulting peak mass lists were searched against the NCBI database using the peptide mass fingerprinting database search tool MASCOT PMF search engine (http://www.matrixscience.com). Spectra which did not contain adequate peaks to get a statistically significant hit (significance as determined by MASCOT) were ZIP-TIPPED (Millipore Corporation, Bedford MA, USA) and eluted into 70% acetonitrile in 0.1% TFA in ddH2O to purify and concentrate the peptides and new MS spectra were collected. Only protein identifications with statistically significant MOWSE scores and which approximately matched the predicted molecular weight and pI values were included. 2.4. Immunoblotting For 2D immunoblotting, 200 lg of protein per sample was diluted to 450 ll in rehydration buffer and run out on 24-cm IPG strips (pH 4–7; Amersham) and equilibrated as described above. Equilibrated strips were placed on top of 1 mm thick 10% acrylamide gels (with Bis–Tris (pH 7.4)) and large format 2D SDS–PAGE electrophoresis was carried out as also described above. A 7.0 cm section of the 10% acylamide gel which contained the spot identified as phosphoprotein enriched in astrocytes 15 kDa (PEA-15) was cut out and transferred (for 1 h at 0.4 Å) onto nitrocellulose membranes. Membranes were blocked and then incubated in primary antibody against rabbit anti-PEA-15 IgG (dilution 1:500; Abcam, Cambridge, MA) at room temperature for 2 h. After rinsing, the membranes were incubated in goat anti-rabbit IgG (1:10,000; HRP conjugated; Jackson Immunoresearch, West Grove, PA, USA). Membranes were then treated with Chemi-Illuminescence (Amersham) according to the manufacturer’s protocol. 2.5. Electron microscopy Neocortical mouse tissues fixed with 2% paraformadehyde and 2% gluataraldehyde were cut at 100 lm thickness using a Vibratome. The sections were trimmed into 1 1 cm2 pieces, which were subsequently fixed overnight in 1% phosphate-buffered osmium tetroxide. After rinsing and dehydrating through a graded series of methanols, the sections were cleared with propylene oxide and embedded Epon. Both semithin and ultrathin sections were cut using a LKB Ultramicrotome, the latter of which were col-
2450
L.A.M. Thomason et al. / FEBS Letters 587 (2013) 2448–2454
lected onto copper grids and stained with uranyl acetate and lead citrate. Stained ultrathin sections were viewed and photographed with a Hitachi 7000 transmission electron microscope. 2.6. Post-mortem human neocortical tissues Human specimens were obtained from a collection of classified paraffin-embedded frontal and temporal neocortical tissues. NonAD tissues included those from the following three adults: (1) 72-year old female who died of coronary artery disease; (2) 76year old male who died of coronary artery disease; and (3) 88-year old male who died of pneumonia. None of these tissues had ADassociated changes. AD tissues included those from the following four adults: (1) 71-year old male with an AD Braak score of 6/6; (2) 73-year old male with an AD Braak score of 4/6; (3) 78-year old female with an AD Braak score 5/6; and, 88-year old female with an AD Braak score of 5/6. 2.7. Immunohistochemistry Fixed mouse brains were sectioned at 40 lm thickness using a freezing microtome. The sections were collected in 0.1 M Tris buffered saline (TBS; pH 7.4) and processed for the immunohistochemical detection of PEA-15, as well as glial fibrillary acidic protein (GFAP; astrocyte marker; DakoCytomation), Ionized calcium binding adaptor molecule 1 (Iba1, microglial marker), and neuronal nuclei (NeuN; neuronal marker; Millipore). Sections were first incubated in 0.3% hydrogen peroxide in 0.1 M TBS for 1 h and then 10% bovine serum albumin (BSA) in 0.1 M Tris and 0.25% TritonX-100 (TBX) for 1 h at room temperature. Sections were then incubated in avidin and biotin for 1 h each at room temperature. Next, sections were incubated in rabbit anti-GFAP IgG (1:1000), rabbit anti-Iba1 IgG (1:1000), or mouse anti-NeuN IgG (1:1000) (each diluted in 3% BSA in TBX) for two days at room temperature. After rinsing in 0.1 M TBS for 15–30 min, the sections were incubated with biotinylated goat anti-rabbit IgG (1:100; Vector Laboratories) for 2 h at room temperature. The sections were then incubated with avidin–biotin complex (EliteÒ ABC Peroxidase kit; Vector Laboratories) for 2 h at room temperature. The sections were then reacted in a solution containing 0.05% diaminobenzidine (DAB) tetrahydrochloride, 0.04% nickel chloride, and 0.015% hydrogen peroxide in 0.1 M TBS until the brown reaction product was clearly visible, as viewed with a stereomicroscope. The sections were then immediately transferred into buffer to stop the reaction. The sections were rinsed thoroughly (3) in 0.1 M TBS for no less than 30 min. For the benzidine dihydrochloride (BDHC) blue reaction, sections were first rinsed in 0.01 M phosphate buffer (pH 6.6) for 30 min, and then incubated in 10% BSA in 0.25% TBX for 1 h at room temperature. Next the sections were incubated in rabbit anti-PEA-15 IgG (1:500) or GFAP (1:1000) diluted in 3% BSA in 0.25% TBX for 2 days at room temperature. After rinsing in 0.1 M TBS, the sections were incubated with biotinylated goat anti-rabbit IgG (1:100; Vector Laboratories) for 2 h at room temperature. The sections were then incubated with avidin–biotin complex (EliteÒ ABC Peroxidase kit; Vector Laboratories) for 2 h at room temperature. The sections were then reacted in a solution containing 0.01% BDHC, 0.025% sodium nitroferricyanide, and 0.03% hydrogen peroxide in 0.01 M phosphate buffer (pH 6.6). Once the blue crystalline reaction product was clearly visible, the reaction was stopped by rinsing the sections in 0.01 M PBS (pH 6.6). Sections were then mounted onto slides, dried overnight at room temperature, dehydrated through a graded series of ethanols, cleared in HistoClearÒ, and then coverslipped in Permount mounting media. Sections were viewed under bright field optics, and images were acquired using a Zeiss microscope and Axiovision software.
Procedures for immunostaining of human neocortical tissues were the same as that described above, with the only exception being that the paraffin-embedded tissues (already mounted on slides) were cleared of paraffin and hydrated to water before incubating in 0.3% hydrogen peroxide.
3. Results Neocortical proteins were fractionated by 2D gel electrophoresis. No fewer than three individual samples of neocortices isolated from 6-month old TgCRND8 and non-transgenic littermates were resolved and silver-stained to reveal protein spot differences. Four proteins were found to be consistently present at more abundant levels in the TgCRND8 proteomes, as compared to age-matched non-transgenic proteomes (Fig. 1). Using mass spectrometry, two of these proteins in the TgCRND8 proteomes, labeled 3 and 4 (Fig. 1A, B) were identified as PEA-15 and Ab, respectively. The spot for PEA-15 on the silver-stained gels was consistently more intense in the TgCRND8 proteomes, as compared to the non-transgenic proteomes. The spot for Ab on the silver-stained gels was consistently found only in the TgCRND8 proteomes, providing direct evidence of the ectopic expression of Ab in these transgenic mice. To validate the presence of PEA15 in the neocortical proteomes of 6-month old TgCRND8 (as compared to the neocortical proteomes of non-transgenic mice), protein samples were again fractioned by 2D gel electrophoresis (with no subsequent silver staining) and then transferred to membranes for immunoblotting. Membranes were incubated with PEA-15 those having TgCRND8 neocortical tissues yielded two immunopositive spots, where as the non-transgenic neocortical tissues did not (insets of Fig. 1A and B). These spots had molecular weights of 15 kDa and isoelectric points of pH 5, and were consistent with previously reported 2D gelmigration patterns for PEA-15 [9]. Since PEA-15 can exist in non-phosphorylated, single-phosphorylated, or double-phosphorylated forms [9], the identification of two immunopositive spots may represent shifts in the phosphorylation status of PEA-15 in the brains of 6month old TgCRND8 mice. Regarding the described phenotype of 3- and 6-month old TgCRND8 mice, the single most conspicuous difference is the increased incidence of Ab plaques throughout the neocortex of the older aged mice. Non-transgenic littermates at both ages lacked such Ab plaques (Fig. 1C; double immunostaining for NeuN (pan neuronal marker) using the chromogen DAB to reveal positive immunoreactivity – brown reaction product, and for PEA-15 using the chromogen BDHC to reveal positive immunoreactivity – blue reaction product). In 6-month old TgCRND8 mice, these plaques found in the grey matter (Fig. 1D) and white matter (Fig. 1E) of the neocortices displayed prominent immunostaining for PEA-15 (blue reaction product). The intensity of this plaque-associated immunostaining was greater in the 6-month old TgCRND8 mice as compared to their 3-month old transgenic counterparts (data not shown). This immunostaining was deemed to be positive because no staining was seen in either the absence of the primary anti-PEA-15 IgG or the presence of the incorrect secondary IgG (e.g., using biotinylated rabbit-anti-goat IgG). Positive immunostaining for PEA-15 (blue reaction product) was also seen among capillaries of both transgenic and non-transgenic mice, likely confined to the end-foot processes of astrocytes. Double immunostaining for GFAP (brown reaction product) and PEA-15 (blue reaction product) yielded GFAP-immunopositive reactive astrocytes associated with plaques in the neocortices of 6-month old TgCRND8 mice; importantly, these GFAP-immunopositive astrocytes displayed positive immunoreactivity for PEA-15 (Fig. 1F). Double immunostaining for Iba1 (using the chromogen DAB) and GFAP
L.A.M. Thomason et al. / FEBS Letters 587 (2013) 2448–2454
2451
Fig. 1. Silver-stained 2D gels of neocortical proteins from 6 month old non-transgenic mice (A) and age-matched TgCRND8 mice (B) revealed 4 protein differences between these two proteomes. Mass spectrometry identified these proteins as small nuclear RNA-activating complex polypeptide 1 (spot 1), IgG-dependent histamine-releasing factor (spot 2), phosphoprotein enriched in astrocytes (PEA-15, spot 3), and amyloid-b (spot 4). Immunostaining for PEA-15 on 2D gels of neocortical tissues from 6 month old nontransgenic and TgCRND8 mice (insets in A and B) revealed two protein spots only in the proteomes of transgenic mice. Neocortical tissues of 6 month old non-transgenic mice (C) displayed PEA-15-immunopositive blood vessels among NeuN-immunopositive neurons. Neocortical tissues of 6 month old TgCRND8 mice (D, E) displayed PEA-15immunopositive blood vessels among NeuN-immunopositive neurons, as well as PEA-15-immunopositive plaques (D, box; E, circle) in both the grey and deep white matter. These neocortical plaques (⁄ in F) were surrounded by reactive astrocytes displaying positive immunostaining for GFAP (brown reaction product) and PEA-15 (blue reaction product). These neocortical plaques (⁄ in G) were composed of Iba1-immunopositive microglia (brown reaction product) and were surrounded by GFAP-immunopositive reactive astrocytes (blue reaction product). Scale bars = 100 lm (C, D) and 20 lm (F, G).
(using the chromogen BDHC) yielded Iba1-immunopositive reactive microglial cells in the core of the neocortical plaques of 6month old TgCRND8 mice; in contrast to the reactive astrocytes, these Iba1-immunopositive microglial cells displayed no positive immunoreactivity for GFAP; double immunostaining with Iba1 and PEA-15 failed to label PEA-15-positive astrocytes in the neocortices of 6-month old TgCRND8 mice (Fig. 1G).
The presence of PEA-15 immunostaining in those reactive astrocytes associated with neocortical plaques of 6-month old TgCRND8 mice warranted further examination of these glial cells at the ultrastructural level. Reactive astrocytes in immediate proximity to dystrophic neurites in the plaques had fragmented rough endoplasmic reticulum, pieces of which were found in smaller astrocytic processes throughout the plaques (Fig. 2A and B).
2452
L.A.M. Thomason et al. / FEBS Letters 587 (2013) 2448–2454
Fig. 2. Transmission electron microscopy of neocortical plaques from 6 month old TgCRND8 mice revealed astrocytic processes (shaded blue) having fragmented pieces of rough endoplasmic reticulum (arrowheads) (A, B); these astrocytic processes were in close proximity to dystrophic neurites (⁄). Neocortical tissues (frontal lobe) from an AD patient having a Braak score of 4/6 had plaques surrounded by reactive astrocytes immunopositive for both GFAP (brown reactive product) and PEA-15 (blue reaction product) (C, D). Some of these reactive astrocytes were also seen extending processes to adjacent blood vessels (BV). Though control neocortical tissues (frontal lobe) from a non-AD patient lacked plaques, astrocytes in grey and white matter displayed positive immunostaining for GFAP (brown reactive product) and PEA-15 (blue reaction product) (E, F). Scale bars = 1 lm (A), 0.8 lm (B), 20 lm (C, D), and 40 lm (E, F).
Postmortem human tissues revealed GFAP-immunopositive reactive astrocytes associated with neocortical (frontal and temporal) plaques in specimens diagnosed with AD (Braak scores ranging from 4/6 to 6/6). These plaque-associated reactive astrocytes displayed positive immunostaining for PEA-15 (Fig. 2C and D). These same PEA-15-immunopositive reactive astrocytes could be seen extending processes in and around the plaques, as well as toward adjacent capillaries. It is important to note that PEA-15 immunoreactivity was found through the neocortices of both AD and non-AD human specimens, including both neuropil and adjacent white
matter tracts (Fig. 2C–F). This pattern was not seen in either transgenic or non-transgenic mice, in which PEA-15 immunoreactivity was restricted to those reactive astrocytes associated with the plaques of TgCRND8 mice. 4. Discussion PEA-15 is a death effector domain (DED) containing protein found predominantly in the central nervous system, with expression previously observed in astrocytes and neurons alike [10]. This
L.A.M. Thomason et al. / FEBS Letters 587 (2013) 2448–2454
protein is highly conserved among mammals, with 96% similarity between the mouse and human forms [11]. PEA-15 was originally described as a protein related to diabetes mellitus and is now known to be involved in regulating numerous other cell processes [9]. The N-terminal DED in PEA-15 has led to investigations into its role in regulating apoptosis [9,10,12]. More recently it has been discovered that PEA-15 has a diverse set of regulatory functions, including roles in proliferation, glucose metabolism, cell cycle control, cell adhesion, migration and autophagy and its expression has been shown to be altered in various cancers [9,10,12–14]. Given the diversity of functions attributed to PEA-15, its enhanced detection among plaque-associated reactive astrocytes in TgCRND8 mice might have beneficial effects, exacerbate glial pathology, and/or merely serve as a by-product of the disease state. Though the scope of our study did not include a functional analysis of increased PEA-15 expression in the neocortices of TgCRND8 mice, it is tempting to speculate that PEA-15 expression in plaque-associated reactive astrocytes may be playing an anti-apoptotic role in this glial sub-population. It has been previously shown that PEA-15 expression decreases the susceptibility of astrocytes to tumor necrosis factor alpha (TNFa)-induced apoptosis, providing a possible mechanism as to why astrocytes are often resistant to apoptosis during inflammation or disease states [10]. Neuroinflammation plays a large role in the pathogenesis of AD [15] and increased TNFa levels have been found in patients with AD [16,17]. Further, in the TgCRND8 mouse model, intense TNFa immunostaining has been observed in reactive microglia and astrocytes surrounding Ab plaques in the neocortex and hippocampus [18]. Therefore, it is possible that increased PEA-15 expression in plaque-associated reactive astrocytes may protect the astrocytes from apoptosis induced by TNF in the vicinity of the Ab plaques. It is also possible that a more general relationship may exist between astrocyte PEA-15 expression and the diseased central nervous system, given that ‘‘marked‘‘ PEA-15 expression was found in reactive astrocytes located in the grey matter of the spinal cord of an ALS mouse model (SOD1 G93A mutant) [12]. As previously mentioned, our report provides direct and validated evidence that detection of PEA-15 is increased in the brains of an AD mouse model. Contrary to our study, Takano et al. [19] reported a decrease in PEA-15 expression in the hippocampus of APPE693D mice (as determined solely by quantitative comparison between wild type and transgenic hippocampal proteomes). This discrepancy may be explained by certain differences between the two studies, including the different mouse models and brain regions examined. Our experiments used the TgCRND8 model of AD, which shows the presence of Ab oligomers, fibrils and plaques [5,20], whereas the Takano et al. [19] assessed the APPE693D model, which generates Ab oligomers in the absence of plaques [19]. Given that we localized increased PEA-15 immunodetection exclusively to plaque-associated reactive astrocytes, we propose that the generation of Ab-containing plaques is critical to PEA-15 expression in surrounding astrocytes. In the absence of Ab-containing plaques, as seen in the APPE693D mouse model, a similar localization of PEA-15 to reactive astrocytes is unlikely to occur. Further, the altered PEA15 expression observed in the two studies was found in two different brain regions: Takano and colleagues reported decreased expression of PEA-15 in the hippocampus of APPE693D mice [19], whereas we detected increased PEA-15 expression in the neocortices of TgCRND8 mice. Finally, our proteomic discovery of increased levels of PEA-15 in the neocortices of TgCRND8 mice was validated by immunostaining of brain tissues and immunoblotting of brain proteins in these mice. Most importantly, we provide translational evidence that increased PEA-15 immunostaining of plaque-associated reactive astrocytes in transgenic mice is also seen in human AD tissue, where Ab plaques are surrounded by PEA-15-immunopositive reactive astrocytes.
2453
Collectively, these data show that 2D gel electrophoresisfractionation of neocortices from 6-month old TgCRND8 and nontransgenic mice yields consistent patterns of differential expression for several proteins, in particular increased levels of Ab and PEA-15. Using mass spectrometry, Ab was only found in the proteomes of TgCRND8 mice, confirming its accumulation in the brains of these transgenic animals. The other protein of interest, PEA-15, was detected at high levels in the proteomes of TgCRND8 mice; this increase in PEA-15 was confirmed by both immunoblotting and tissue immunostaining. In the neocortices of the TgCRND8 mice, PEA-15 expression was localized to plaque-associated reactive astrocytes, a pattern that was also observed in post-mortem human AD tissue. As a biomarker of AD, increased levels of astrocytic PEA-15 may be used, along with other pathological hallmarks, to assess severity of the disease progress at the time of death. Future studies will assess whether PEA-15 is detectable in human cerebrospinal fluid and/or serum, and if such detection can be linked to cognitive impairment associated with AD prior to death. Acknowledgements We would like to thank Ms. Mary E. Brown for assisting with harvesting the mouse tissues, Mr. Casey N. Petrie for tissue homogenizations, Ms. Verna Norkum for the preparation of the tissues for immunostaining and examination at the transmission electron microscope level, and Mr. David McLeod for his assistance with the mass spectrometry (Queen’s Protein Function Discovery Facility). Ms. Thomason is supported by an Ontario Graduate Scholarship and the Heart and Stroke Foundation Centre for Stroke Recovery Trainee Stimulus Fellowship, and Ms. Smithson is supported by a Banting and Best Scholarship from the Canadian Institutes of Health Research. This work was supported by CIHR (J.M.: MOP-10246) and The Botterell Foundation of Queen’s University (M.D.K.). References [1] Götz, J., Ittner, L.M., Schonrock, N. and Cappai, R. (2008) An update on the toxicity of A in Alzheimer’s disease. Neuropsychiatr Dis. Treat. 4, 1033–1042. [2] World Health Organization (WHO) (2012), World Alzheimer Report 2012, a public health priority. [3] Association, Alzheimer.’s. (2010) Alzheimer’s disease facts and figures. Alzheimers Dement. 6, 158–194. [4] Bird, T. D. (2010) Alzheimer’s disease overview. Gene Rev. http:// www.ncbi.nlm.nih.gov/books/NBK1161/. Last updated December 2012, (accessed 21.02.2012). [5] Chishti, M.A., Yang, D.S., Janus, C., Phinney, A.L., Horne, P., Pearson, J., Strome, R., Zuker, N., Loukides, J., French, J., Turner, S., Lozza, G., Grilli, M., Kunicki, S., Morissette, C., Paquette, J., Gervais, F., Bergeron, C., Fraser, P.E., Carlson, G.A., St. George-Hyslop, P. and Westaway, D. (2001) Early-onset amyloid deposition and cognitive deficits in transgenic mice expressing a double mutant form of amyloid precursor protein 695. J. Biol. Chem. 276, 21562–21570. [6] Graham, D.R., Garnham, C.P., Fu, Q., Robbins, J. and Van Eyk, J.E. (2005) Improvements in two-dimensional gel electrophoresis by utilizing a low cost ‘‘in-house’’ neutral pH sodium dodecyl sulfate-polyacylamide gel electrophoresis system. Proteomics 5, 2309–2314. [7] Shevchenko, A., Jensen, O., Podtelejnikov, A., Sagliocco, F., Wilm, M., Vorm, O., Mortensen, P., Shevchenko, A., Boucherie, H. and Mann, M. (1996) Linking genome and proteome by mass spectrometry: large-scale identification of yeast proteins from two dimensional gels. Proc. Natl. Acad. Sci. USA 93, 14440–14445. [8] McDonald, T.G., Scott, S.A., Kane, K.M. and Kawaja, M.D. (2009) Proteomic assessment of sympathetic ganglia from adult mice that possess null mutations of ExonIII or ExonIV in the p75 neurotrophin receptor gene. Brain Res. 1253, 1–14. [9] Eckert, A., Böck, C., Tagscherer, K.E., Haas, T.L., Grund, K., Sykora, J., HeroldMende, C., Ehemann, V., Hollstein, M., Chneiweiss, H., Wiestler, O.D., Walczak, H. and Roth, W. (2008) The PEA-15/PED protein protects glioblastoma cells from glucose deprivation-induced apoptosis via the ERK/MAP kinase pathway. Oncogene 27, 1155–1166. [10] Kitsberg, D., Formstecher, E., Fauquet, M., Kubes, M., Cordier, J., Canton, B., Pan, G., Gloinski, J. and Chneiweiss, H. (1999) Knock-out of the neural death effector domain protein PEA-15 demonstrates that its expression protects astrocytes from TNF-induced apoptosis. J. Neurosci. 19, 8244–8251.
2454
L.A.M. Thomason et al. / FEBS Letters 587 (2013) 2448–2454
[11] Estellés, A., Yokoyama, M., Nothias, F., Vincent, J.-D., Glowinski, J., Vernier, P. and Chneiweiss, H. (1996) The major astrocytic phosphoprotein PEA-15 is encoded by two mRNAs conserved on their full length in mouse and human. J. Biol. Chem. 271, 14800–14806. [12] Sharif, A., Renault, F., Beuvon, R., Castellanos, B., Canton, L., Barbeito, M., Junier, P. and Chneiweiss, H. (2004) The expression of PEA-15 (phosphoprotein enriched in astrocytes of 15 kDa) defines subpopulations of astrocytes and neurons throughout the adult mouse brain. Neuroscience 126, 263–275. [13] Böck, B.C., Tagscherer, K.E., Fassl, A., Krämer, A., Oehme, I., Zentgraf, H.-W., Keith, M. and Roth, W. (2010) The PEA-15 protein regulates autophagy via activation of JNK. J. Biol. Chem. 285, 21644–21654. [14] Formisano, P., Ragno, P., Pesapane, A., Alfano, D., Alberobello, A.T., Rea, V.E.A., Giusto, R., Rossi, F.W., Beguinot, F., Rossi, G. and Montuori, N. (2012) PED/PEA-15 interacts with the 67 kDa laminin receptor and regulates cell adhesion, migration, proliferation and apoptosis. J. Cell Mol. Med. 16, 1435–1446. [15] McGeer, P.L. and McGeer, E.G. (2002) Local neuroinflammation and the progression of Alzheimer’s disease. J. Neurovirol. 8, 529–538.
[16] Fillit, H., Ding, W.H., Buee, L., Kalman, J., Altstiel, L., Lawlor, B. and Wolf-Klein, G. (1991) Elevated circulating tumor necrosis factor levels in Alzheimer’s disease. Neurosci. Lett. 129, 318–320. [17] Alvarez, A., Cacabelos, R., Sanpedro, C., García-Fantini, M. and Aleixandre, M. (2007) Serum TNF-alpha levels are increased and correlate negatively with free IGF-I in Alzheimer disease. Neurobiol. Aging 28, 533–536. [18] Luccarini, I., Grossi, C., Traini, C., Fiorentini, A., Dami, T.E. and Casamenti, F. (2012) A plaque-associated glial reaction as a determinant of apoptotic neuronal death and cortical gliogenesis: a study in APP mutant mice. Neurosci. Lett. 506, 94–99. [19] Takano, M., Maekura, K., Otani, M., Sano, K., Nakamura-Hirota, T., Tokuyama, S., Min, K.S., Tomiyama, T., Mori, H. and Matsuyama, S. (2012) Proteomic analysis of the brain tissues from a transgenic mouse model of amyloid oligomers. Neurochem. Int. 61, 347–355. [20] McLaurin, J., Kierstead, M.E., Brown, M.E., Hawkes, C.A., Lambermon, M.H., Phinney, A.L., Darabie, A.A., Cousins, J.E., French, J.E., Lan, M.F., Chen, F., Wong, S.S., Mount, H.T., Fraser, P.E., Westaway, D. and St. George-Hyslop, P. (2006) Cyclohexanehexol inhibitors of Abeta aggregation prevent and reverse Alzheimer phenotype in a mouse model. Nat. Med. 12, 801–808.