Increased Levels of Extracellular Microvesicle Markers and Decreased Levels of Endocytic/Exocytic Proteins in the Alzheimer’s Disease Brain

Abstract

Background: Alzheimer’s disease (AD) is a chronic neurodegenerative disorder accounting for more than 50% of all dementia cases. AD neuropathology is characterized by the formation of extracellular plaques and intracellular neurofibrillary tangles consisting of aggregated amyloid-β and tau, respectively. The disease mechanism has only been partially elucidated and is believed to also involve many other proteins.

Objective: This study intended to perform a proteomic profiling of post mortem AD brains and compare it with control brains as well as brains from other neurological diseases to gain insight into the disease pathology.

Methods: Here we used label-free shotgun mass spectrometry to analyze temporal neocortex samples from AD, other neurological disorders, and non-demented controls, in order to identify additional proteins that are altered in AD. The mass spectrometry results were verified by antibody suspension bead arrays.

Results: We found 50 proteins with altered levels between AD and control brains. The majority of these proteins were found at lower levels in AD. Pathway analyses revealed that several of the decreased proteins play a role in exocytic and endocytic pathways, whereas several of the increased proteins are related to extracellular vesicles. Using antibody-based analysis, we verified the mass spectrometry results for five representative proteins from this group of proteins (CD9, HSP72, PI42A, TALDO, and VAMP2) and GFAP, a marker for neuroinflammation.

Conclusions: Several proteins involved in exo-endocytic pathways and extracellular vesicle functions display altered levels in the AD brain. We hypothesize that such changes may result in disturbed cellular clearance and a perturbed cell-to-cell communication that may contribute to neuronal dysfunction and cell death in AD.

Keywords

Alzheimer’s disease endocytosis exocytosis exosomes extracellular vesicles immobilized chemistry antibodies mass spectrometry

INTRODUCTION

Alzheimer’s disease (AD) is a chronic and fatal neurodegenerative disorder affecting more than 10% of the population over 65 years [1] and more than 20% of those over 80 years [2]. Importantly, the number of patients will rise dramatically in the future due to the increase in longevity [3]. Symptoms of AD include memory loss, disorientation, and impairment of other cognitive functions. As the disease progresses, the patients also suffer from language impairment and various degrees of dyspraxia [4].

The AD brain is neuropathologically characterized by the formation of extracellular plaques and neurofibrillary tangles, which mainly consist of aggregated amyloid-β (Aβ) and tau, respectively. The formation of plaques and tangles may start 20–30 years prior to the clinical onset of the disease [5]. Whereas 10–15% of all AD cases are caused by dominant mutations in either of three different disease genes (APP, PSEN1, or PSEN2), which all are related to the generation of amyloid-β (Aβ), the vast majority of sporadic cases have a largely unknown etiology. As for diagnostic markers, decreased levels of Aβ₄₂ (the most aggregation prone Aβ species) and increased levels of tau/phospho-tau can be measured in cerebrospinal fluid (CSF). However, although these biomarkers have a fairly good sensitivity and specificity, they do not change during the disease course and are not optimal to measure treatment effects. Hence, there is a great need to identify novel biomarkers that would allow us to monitor disease progress and therapeutic efficacy.

Brain functions such as sensory perception, learning, and memory all emerge as a result of communication between neurons via synapses. Several neurological disorders such as autism, schizophrenia, epilepsy, and AD appear to be related to abnormalities in neural connections, neuronal signaling dysfunctions, or loss of synapses [6 –8]. Previous studies suggest that the loss of synaptic proteins may be an early event in AD [8, 9]. Furthermore, several studies have shown that the loss of presynaptic vesicle proteins is associated with AD pathogenesis by contributing to the learning impairment and cognitive deficits [10, 11]. Synaptic vesicles are uniform organelles of approximately 40 nm in diameter that contain proteins with different roles, including neurotransmitter transport, release, and recycling. Soluble N-ethylmaleinide-sensitive factor attachment receptors (SNARE) proteins mediate the exocytic pathway beginning with the docking of synaptic vesicles at the active zone, which then fuse with the presynaptic membrane. Three neuronal SNAREs: vesicles-associated membrane protein (VAMP2) (v-SNARE), syntaxin 1, and 25-kDa synaptosomal-associated protein (SNAP-25) (t-SNARE) catalyze the synaptic vesicle exocytosis process. The recycling of fused synaptic vesicles is a result of an endocytic process via clathrin-mediated endocytosis (CME), which involves recruitment, budding, fission, and vesicle uncoating [12, 13] or clathrin-independent endocytosis (CIE) [14].

Extracellular vesicles (EVs) are small membrane bound vesicles shed by the cells, e.g., when multivesicular bodies (MVBs) merge with the cell membrane through exocytosis pathway. Once released into the extracellular space, other cells can take up these vesicles or their contents via endocytosis [15]. Since EVs contain substances and cytoplasm from their cells of origin, they represent the intracellular environment of such cells. Studies have reported that EVs and specifically exosomes may be involved in AD pathology as they can harbor and release Aβ and tau [16 –18]. Thus, vesicles displaying neuron or glial specific markers can be isolated from tissues or biofluids and examined for Aβ, tau or other AD-related biomarkers [19, 20]. Along these lines, it was recently suggested that both Aβ and tau from L1CAM positive exosomes, extracted from human plasma, could serve as sensitive and specific biomarkers for AD [21].

The aim of this study was to carry out a comprehensive proteomic profiling of post mortem AD brains and compare the outcome with control brains as well as brains from other neurological diseases. We identified several proteins with altered levels in AD compared to control brains using a label-free shotgun mass spectrometry (MS) approach. SNARE and other synaptic vesicle proteins were found at lower levels, whereas several EV related proteins were found at increased levels in the AD brains. Using antibody-based analysis, six of the proteins [vesicle-associated membrane protein 2 (VAMP2), CD9 antigen (CD9), heat shock-related 70 kDa protein 2 (HSP72), phosphatidylinositol 5-phosphate 4-kinase type-2 alpha (PI42A), transaldolase (TALDO), glial fibrillary acidic protein (GFAP)] were verified to have altered levels in AD compared to control brains.

MATERIALS AND METHODS

Brain specimens

Frozen post mortem human brain tissues from temporal neocortex from eleven ADs, eight non-demented controls, and six other neurological diseases (OND) subjects were included. The AD cases were neuropathologically diagnosed as CERAD C, Braak stages IV–VI [22]. The control tissues did not display any signs of neurodegeneration, whereas the OND tissues were diagnosed according to the respective neuropathological criteria. A summary of the main clinical and neuropathological aspects is shown in Supplementary Table 1.

All brain samples were obtained from the Uppsala Berzelii Technology Centre for Neurodiagnostics biobank at the Uppsala University Hospital. On an average, the samples were collected 27 h (minimum 5 h/ maximum 60 h) post mortem. The samples were put in 1.5 mL Eppendorf tubes and stored at –80°C pending analyses. The project had been approved by the Regional Ethical Review Board in Uppsala, Sweden. The MS experiments were performed twice. The first experiment (D1) contained 17 samples (8 AD, 4 control, and 5 OND cases). The second experiment (D2) contained the same samples as in D1 and four additional human samples, thus in total 10 AD, 5 control, and 6 OND cases.

Chemicals and reagents

Acetonitrile (ACN), methanol, acetic acid (HAc), formic acid, sodium chloride (NaCl), and ammonium bicarbonate (NH₄HCO₃) were obtained from Merck (Darmstadt, Germany). Acetone, ethylenediaminetetraacetic acid tetrasodium salt dihydrate (EDTA), protease inhibitor cocktail, phosphate buffered saline (PBS), trifluoroacetic acid, and n-octyl-β-D-glucopyranoside, and ammonium bicarbonate (NH₄HCO₃) were purchased from Sigma Aldrich (St. Louis, MO, USA). For tryptic digestion, iodoacetamide (IAA), urea, and dithiothreitol (DTT) were obtained from Sigma Aldrich and trypsin (sequencing grade from bovine pancreas 1418475; Roche diagnostic, Basel, Switzerland) were used. Ultrapure water was prepared by Milli-Q water purification system (Millipore, Bedford, MA,USA).

Protein extraction

The temporal neocortex samples were homogenized in liquid nitrogen and the brain powder was stored at –80°C prior to analyses. Aliquots of 30 mg brain powder were homogenized for 60–90 s in a blender (POLYTRON PT 1200, Kinematica) with 1 mL of lysis buffer (10 mM Tris-HCl pH 7.4, 0.15 M NaCl, 1 mM EDTA and PBS containing 1% β-octyl glucopyranoside). After homogenization, the samples were incubated for 90 min at 4°C during mild agitation. The tissue lysates were clarified by centrifugation for 30 min (10000× g at 4°C) using a Sigma 2K15 ultracentrifuge (Sigma Laborzentrifugen GmbH, Osterode, Germany). The supernatant containing extracted proteins was collected and further processed.

Delipidation and protein precipitation

An adapted delipidation protocol according to Mastro et al. was used [23, 24]. Aliquots (200μL) of the protein extracts were mixed with 1.4 mL of ice-cold tri-n-butylphosphate: acetone: methanol mixture (1:12:1) and incubated at 4°C for 90 min. The precipitate was pelleted by centrifugation for 15 min (2800× g at 4°C), washed sequentially with 1 mL of acetone and 1 mL of methanol, and finallyair-dried.

Protein quantification

The total protein concentration of delipidated proteins was determined using the DC Protein Assay Kit (BioRad Laboratories, Hercules, CA, USA), which is based on the modified Lowry method with bovine serum albumin as standard [25]. The protein pellets were redissolved in 200μL of 6% SDS. The DC assay was carried out according to the manufacturer’s instructions using a 96-well microtiter plate reader model 680 (BioRad Laboratories). The measured protein concentration is shown in Supplementary Table 2.

On-filter tryptic digestion of brain proteins

Two hundred μL of the extracts (220–300μg of proteins) were delipidated as described above. Delipidated protein pellets were re-dissolved in 200μL of digestion buffer (8 M urea, β-octyl glucopyranoside in 50% ACN). A volume of 35μL protein solution was used for digestion. An on-filter digestion protocol was used for tryptic digestion of the samples using 3 kDa filters (Pall Life Sciences, Ann Arbor, MI, USA). Centrifugation was carried out at 14,000 × g throughout the protocol. A volume of 10μL of 45 mM aqueous DTT was added to all samples and the mixtures were incubated at 50°C for 15 min to reduce the disulfide bridges. The samples were cooled down to room temperature and 10μL of 100 mM aqueous IAA was added before incubating the mixtures for an additional 15 min at room temperature in darkness to carabamidomethylate the cysteines. The samples were transferred to spin filters that had been pre-washed with 250μL of 50% ACN for 15 min and then 500μL of water for 20 min. Next, the samples were centrifuged for 10 min to remove the added salts, detergents and other interfering substances. An additional volume of 100μL of 50 mM NH₄HCO₃ in 2% ACN was added and the filters were spun for 10 min followed by 100μL of 50 mM NH₄HCO₃ in 50% ACN and 150μL of 50 mM NH₄HCO₃, and centrifugation for another 10 min. Finally, a volume of 100μL of 50 mM NH₄HCO₃ (pH 7.8) and 16μL of trypsin (0.1μg/μL) was added to the samples. The tryptic digestion was performed at 37°C overnight in darkness. The digests were spun through the filter for 20 min to collect the tryptic peptides. An additional volume of 100μL of 50% ACN, 1% HAc was added and the filters were spun for 10 min and pooled with the first tryptic peptide filtrate. The collected filtrates were vacuum centrifuged to dryness using a Speedvac system ISS110 (Thermo Scientific, Waltham, MA,USA).

NanoLC-MS/MS for protein identification

The nanoLC-MS/MS experiments were performed using a 7 T hybrid LTQ FT mass spectrometer (ThermoFisher Scientific, Bremen, Germany) fitted with a nano-electrospray ionization (ESI) ion source. On-line nanoLC separations were performed using an Agilent 1100 nanoflow system (Agilent Technologies, Waldbronn, Germany). The peptide separations were performed on in-house packed 15-cm fused silica emitters (75 μm inner diameter, 375 μm outer diameter). The emitters were packed with a methanol slurry of reversed-phase, fully end-capped Reprosil-Pur C₁₈-AQ 3μm resin (Dr. Maisch GmbH, Ammerbuch-Entringen, Germany) using a PC77 pressure injection cell (Next Advance, Averill Park, NY, USA). The injection volumes were 5μL and corresponded to 2μg of tryptic peptides approximately. The separations were performed at a flow of 200 nL/min with mobile phases A (water with 0.5% acetic acid) and B (89.5% acetonitrile, 10% water, and 0.5% acetic acid). A 100-min gradient from 2% B to 50% B followed by a washing step with 98% B for 5 min was used. Mass spectrometric analyses were performed using unattended data-dependent acquisition mode, in which the mass spectrometer automatically switches between acquiring a high-resolution survey mass spectrum in the FTMS (resolving power 50000 FWHM) and consecutive low-resolution, collision-induced dissociation fragmentation of up to five of the most abundant ions in the ion trap.

Quantification

The raw data from MS was converted to open source format (mzML) by “msconvert” from ProteoWizard [26] and processed using an automated label free pipeline [27] in OpenMS platform [28] (the parameters are present in Supplementary Table 4). Identification was performed using X!Tandem [29] and OMSSA [30] on UniProt/Swiss-Pro human database (release 2014_03) combined with a decoy database (reversed sequences) using the following parameters: Enzyme: Trypsin, missed cleavages: 2 precursor mass tolerance: 10 ppm, fragment mass tolerance: 0.7 Da, minimum charge: 1, maximum charge: 4, fixed modifications: Carbamidomethyl (C), variable modifications: Oxidation (M) and Deamidated (N and Q). The results of identification from the two search engines were combined using “PEPMatrix” voting procedure implemented in “ConsensusID” tool [31]. False discovery rate (FDR) was calculated on the target/decoy database and the peptides with a score lower than 0.01 was chosen as true positive hits. Non-proteotypic peptides (peptides which were assigned to more than one protein) were removed and the data was transformed to log₂ scale and normalized using median normalization [32]. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE [33] partner repository with the dataset identifier PXD004499.

Quality control

In order to analyze clustering of the biological replicates, the median of technical replicates for each peak was calculated and the statistical testing was performed on the resulting data using Limma package in R [34] such that multiple linear models were fitted through the data using “lmFit” function [35] and “ebayes” function [36] was used to compute F-statistic for each peak (including moderated t-statistics for each comparison). Hierarchical clustering (using correlation matrix of the biological replicate) and principal component analysis (PCA) were performed on the peaks with F. p-value lower than 0.05 (missing values were imputed by an iterative stepwise regression algorithm [37] for performing PCA).

Protein quantification

For protein quantification and statistical testing linear models were fitted on all the peaks using “lmFit” and the “duplicateCorrelation” function [38] was used to estimate the correlation between the technical replicates. A contrast matrix was made using “makecontrasts” function, specifying the three comparisons of interest (AD-C, OND-C, AD-OND), which were modeled using the “contrasts.fit” function. Moderated t-statistics for each comparison was calculated using the “ebayes” function. Relative protein expression was estimated such that for each peptide which was identified and quantified in each of the three groups with less than 50% missing values, mean of the intensities in all the replicates were calculated as a measure of peptide abundance. For each protein, the three most abundant peptides [39] based on the calculated mean were selected (all the peptides were selected in case of the proteins identified with less than three peptides) and the minimum of the p-values and the corresponding fold change (individually for each coefficient) and a Fisher’s p-value were calculated as the protein statistical components. The proteins which met the following criteria were selected for further investigation: a protein needs to be identified in both of datasets with absolute log₂ fold change higher than 0.3, it must have consistent direction of fold change in both datasets and it needs be statistically significantly altered in at least one of the datasets (p-value <0.05).

Functional enrichment and protein interaction

Functional Enrichment and protein interaction network analysis were performed using STRING v9.1 (search tool for the retrieval of interacting genes) [40]. The following procedure was performed on the data, using the default settings (confidence level: 0.4; Active Prediction Methods: default): First, the proteins that met the above mentioned selection criteria were set as the target list and the network analysis was performed on the list followed by a GO enrichment analysis, using all the identified proteins as the background set. The target set was then split up into two sets of proteins with increased and decreased levels and the analysis was repeated for each of the new sets, using the same background set as the first step. The terms with p-value <0.05 and at least three proteins were selected as statistically significantly enriched terms.

Antibody suspension bead arrays

To validate the findings from MS, sandwich immunoassays in the form of antibody suspension bead arrays were developed. For the antibody-based analyses we used the same samples as in D2 as well as one additional AD and three additional control samples (Supplementary Table 1). Antibodies were tested as both capture and detection antibodies to find functional pairs. Capture antibodies were coupled to magnetic carboxylated beads (MagPlex, Luminex corp.) at 16 ng/ml according to the manufacturer’s protocol and as described previously [41 –45]. Samples were diluted in 0.5% (w/v) polyvinyl alcohol and 0.8% (w/v) polyvinylpyrrolidone with 0.1% casein in PBS (all Sigma) and heat-treated at 56°C for 30 min followed by 10 min at room temperature before incubated with the capture bead array. Unbound proteins were washed off with PBS-T (1x PBS, 0.05% Tween20). Rabbit detection antibodies were labeled (Zenon R-phycoerythrin rabbit IgG labeling kit, Life Technologies) and applied at 1μg/ml for 30 min. Mouse and goat detection antibodies were biotinylated [46] and applied at 1μg/ml for 1 h, followed by incubation with R-Phycoerythrin conjugated streptavidin (SA1004-4, In vitrogen; diluted 1:500 in PBS-T) for 30 min. Measurements were performed with a FlexMap3D instrument (Luminex Corp.), reporting median fluorescence intensity. For further details, see SupplementaryTable 3.

Western blot analysis

To validate the increased levels of CD9 (including one extra AD and control sample as described in Supplementary Table 1), western blot analysis was performed. Briefly, 25μg of protein from each brain specimen were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) on 4–12% Bis-Tris Criterion XT Precast Gel (BioRad Laboratories), according to the manufacturer’s instructions. The separated proteins were transferred onto PVDF membrane (Amersham Biosciences GE, Little Chalfont, UK). CD9 protein was blocked for 1 h at room temperature with 5% non-fat dry milk dissolved in Tris buffered saline (50 mM Tris, 0.5% Tween-20, pH 8.0), washed briefly with TBS-T twice prior to incubation with primary antibodies in TBS-T overnight at 4°C under gentle agitation. The primary antibodies were diluted as indicated and used for immunoblotting: CD9 (1:1000 dilution; from Sigma Aldrich, St. Louis, MO, USA). The membranes were washed three times, 10 min, with TBS-T and then incubated with secondary IgG anti-rabbit HRP-conjugated (ThermoFisher Scientific, Waltham, MA, USA) for 90 min. After washing three times with TBS-T (20 min each wash step), immunoreactive bands were detected by enhanced chemiluminescence (ECL) detection kit (Amersham Biosciences GE, Little Chalfont, UK) and imaged with ChemiDoc XRS+ (Bio-Rad, Hercules, CA). Equal protein loading was verified by a Ponceau red staining of the membranes. Bands were plotted and quantified using Image Lab (Bio-Rad).

Antibody-based analysis

Kruskal-Wallis test was applied on the data from the antibody suspension bead arrays and western blot data (on all the duplicate measurements) for comparison of the three groups and Mann–Whitney U test was performed for group wise comparisons. Proteins with p-value <0.05 was regarded as statistical significant and relative fold change was calculated as log₂ ratio of “AD/C”, “OND/C”, and “AD/OND”.

RESULTS

Label free shotgun mass spectrometry was used to quantify the relative amount of peptides in the brain samples from AD subjects, non-neurological controls, and OND subjects. In order to replicate and validate the findings, the experiment including tissue dissection, extraction, mass spectrometry, and data processing was performed twice.

Data preparation and quality control

A total of 7,078 and 10,213 protein specific peptides from 1,134 and 1,480 proteins were identified and quantified in the D1 and D2 datasets, respectively. A clear separation of the samples belonging to different disease groups was observed when the clustering analysis was performed on the samples (Fig. 1).

Mass spectrometry protein identification and quantification

In order to translate the peptide information into proteins, the 50% coverage restriction was applied to the peptide list, resulting in 3,768 (689 proteins) in the D1 and 4,282 peptides (724 proteins) in the D2 dataset (549 overlapping proteins), respectively. Comparing AD to non-demented control brains, 163 proteins were found to have statistically significant different levels (p < 0.05) in either the D1 or D2 datasets (Fig. 2; Supplementary Figure 1 for comparison OND-C and AD-OND), of which 50 were significantly altered in both datasets (Supplementary Table 5). Among the 163 proteins, 121 were decreased and 42 were increased in AD compared to controls. Out of these 163 proteins, 14 were found to be significantly altered between AD and ONDs but not between ONDs and controls.

Functional annotation and network analysis

The first step of the functional enrichment analysis comparing AD to non-demented controls was performed on all the proteins with statistical significant altered levels. Proteins enriched were related to nervous system development (p-value = 6.0E-05), neurogenesis (p-value = 1.7E-04), and synaptic transmission (p-value = 1.6E-03) (Table 1). However, when the same analysis was performed only the proteins with decreased levels in AD, the proteins were more specifically related to generation of neurons (p-value = 2.7E-04), regulation of vesicle mediated transport (p-value = 4.6E-04), synaptic transmission (p-value = 3.3E-05), and SNARE binding (p-value = 7.5E-04) (Fig. 3). The enrichment analysis of the proteins with increased levels in AD revealed that they were related to extracellular vesicular exosome (p-value = 1.7E-03) (Fig. 4).

Antibody-based analyses

To validate the results from MS using an orthogonal technique, proteins were targeted for antibody-based verification. Using antibody suspension bead arrays, heat shock-related 70 kDa protein 2 (HSP72, p = 0.020), transaldolase (TALDO, p = 0.001), phosphatidylinositol 5-phosphate 4-kinase type-2 alpha (PI42A, p = 0.005) were found at higher levels in AD compared to controls and vesicle-associated membrane protein 2 (VAMP2, p = 0.007) level was found to be lower in AD than controls (Table 2). Using western blot (Supplementary Figure 2), CD9 antigen (CD9) was found to have statistically significantly (p = 0.020) higher level in AD compared to controls. In addition, the level of the astrocyte marker GFAP (p = 0.01) was higher in OND compared to controls, similarly to the MS data, whereas for AD compared to controls, the antibody result was not statistically significant but the fold change was in agreement between MS and the antibody-based result (Fig. 5A-E).

DISCUSSION

We performed a label free proteomic analysis on neocortical brain tissue samples from subjects with AD, non-demented controls and OND cases. We identified 50 proteins, which were significantly different between AD and control samples in both D1 and D2 datasets. Out of these 50 proteins, seven were significantly altered between AD and OND but not between OND and controls, including Calmodulin (CALM), Catalase (CATA), Coiled-coil domain-containing protein 168 (CC168), Rho GDP-dissociation inhibitor 1 (GDIR1), Isocitrate dehydrogenase [NAD] subunit alpha, mitochondrial (IDH3A), Neutral cholesterol ester hydrolase 1 (NCEH1), Serine/threonine-protein kinase PAK 1 (PAK1). Therefore, the remaining changes might not be solely associated with AD but may be caused by the neurodegeneration per se. In order to validate the mass spectrometry results, we performed antibody-based confirmatory analysis. We successfully verified statistically significant changes of CD9, HSP72, PI42A, VAMP2, and TALDO in AD compared to controls using the antibody-basedtechnique.

Loss of exocytic and endocytic proteins in AD

Neurotransmitter release and vesicle recycling are mediated by membrane trafficking of synaptic vesicle exo-endocytosis process at the nerve terminals. We found lower levels of proteins involved in exo-endocytosis processes in the AD brain compared to controls, indicating a reduced intracellular trafficking in AD brain.

Exocytic proteins

Among the proteins reproduced in the replicated experiment or verified by antibody-based analysis, we found decreased levels of VAMP2, syntaxin-1A (STX1A), and synaptosomal-associated protein 25 (SNAP-25) in the AD brain compared to controls. VAMP2 is localized on the synaptic vesicles and mediates vesicle fusion by forming a tight complex with SNAP-25 and syntaxin-1 present on the pre-synaptic membrane [47]. These proteins are the main components of SNARE complex and involved in the mediation of vesicle docking with presynaptic membrane during exocytosis [48] (Fig. 6). It has been proposed that syntaxin-1 and SNAP-25 proteins might be involved in neuronal survival and that their loss or dysfunction can contribute to neurodegeneration [49]. In the enrichment analysis we found proteins involved in regulation of vesicle mediated transport, synaptic transmission and glutamate secretion. Several of these proteins (e.g., Ras-related protein Rab-3A and septin-5) directly or indirectly interact with SNARE to mediate the vesicle release and transport in neurons [50 –52]. More specifically, these proteins are involved in different stages of the exocytotic process: transporting (Ras-related protein Rab-3A) [53], priming (Syntaxin-binding protein 1) [54], and fusion (Syntaxin-binding protein 1) [11 , 56] many of which have been proposed to be involved in reduced neurotransmission and loss of neuronal activity [11 , 57] (Fig. 6). By profiling human AD hippocampal tissue, Hondius et al. also found protein expression changes related to glutamate signaling and exocytosis. In spite of early disturbance to proteins part of vesicle docking and release, they found no early changes, at least not in hippocampus, to proteins part of the SNARE complex, suggesting that these protein changes occur late in the disease progression [58].

In summary, we found decreased levels of the proteins involved in exocytotic processes and specifically SNARE complex. This indicates lower level of presynaptic proteins in AD compared to controls. However, whether that is a result of the disease progression or in itself promote the neurodegeneration, or both, remain to be elucidated.

Endocytic proteins

In order to sustain synaptic transmission, the fused vesicles should be recaptured and reused to maintain the supply of the synaptic vesicles. In the replicated MS experiments we found statistically significantly decreased levels of clathrin heavy chain 1 (CLH1) and adaptor protein complex AP-2 subunit alpha-1 and alpha-2 (AP2A1 and AP2A2). This result is in agreement with our previous study on AD brains where we found decreased levels of two of these clathrin-mediated endocytic proteins (CLH1 and AP2A1) [59] and in line the decreased levels of AP2A1 and clathrin light chain A (CLTA) already in early stages of the AD pathology in hippocampus [58] and in the later stages of the disease in the olfactory bulb [60]. Clathrin and the adaptor complex participate in budding of vesicles during endocytosis. Clathrin is the major component in the clathrin-coated vesicles that forms a polyhedral lattice, which acts as mechanical scaffold. Microscopic studies showed the presence of CLH1 lines during the invagination of membranes, suggesting its role in vesicle retrieval [61]. In vitro studies have shown that synaptic vesicle recycling is inhibited in the absence of CLH1 [62]. Therefore, decreased level of CLH1 in AD brains could affect the overall clathrin-mediated endocytotic process leading to reduced vesicle recycling (Fig. 6). AP-2 adaptor protein complex is a heterotetramer consisting of four subunits (AP2A1, AP2B1, AP2M1, and AP2S1). The α-subunit (AP2A1) acts as a scaffold for endocytic accessory proteins [63]. The complex binds to both a clathrin lattice and to the lipid and protein components of membrane, enabling clathrin to bind to the membrane components [64]. Clathrin requires the aid of adaptor proteins (as a scaffold) to bind to the membrane components [63, 64]. Therefore, reduced levels of AP2 protein in AD brains might affect the recruitment of clathrin during the endocytosis process (Fig. 6).

Interestingly, we could observe and verify statistically significantly increased levels of PI42A in AD compared to controls. This protein catalyzes the formation of phosphatidylinositol 4,5-bisphosphate (PtdIns(4,5)P2), referred to as PIP2 which has been proposed as a key regulator of clathrin endocytosis [65]. We speculate that since the turnover of PIP2 is essential for endocytosis [66], excessive amount of PI42A in the site of endocytosis might affect this process. However, increased expression of this protein in conjunction with AD remainsunclear.

In summary, the lower level of multiple proteins in AD compared to controls which are involved in the endocytic pathway can lead to reduced fission and invagination processes and to lower activity of endocytic and vesicle recycling pathways.

Increased levels of proteins related to extracellular vesicles in AD compared to control

The functional analyses revealed that the majority of the proteins with higher levels in AD compared to control are related to EVs. We found and verified increased levels of the EVs (e.g., exosome) marker CD9 in AD compared to control brains (Fig. 6). There is increasing evidence that insufficient lysosomal degradation is involved in the pathogenesis of different neurodegenerative diseases, including AD, Parkinson’s disease, and Huntington’s disease [67, 68]. Since lysosomal degradation and secretion of EVs are closely connected, it is likely that Aβ pathology changes the EVs content. We have recently shown that incomplete lysosomal degradation of Aβ₄₂ protofibrils by astrocytes results in EV-induced neuronal cell death in a co-culture system of neurons and glia [69]. EVs and specifically exosomes are present in both extracellular environment and in multivesicular bodies inside the cells. This is also in line with recent studies of human AD hippocampus, where increased levels of proteins part of extracellular matrix and proteins secreted in EVs, e.g., ANXA1 and ANXA5 have been demonstrated. We also observed and confirmed increased levels of the HSP72 protein in AD brain compared to controls. This is in line with the increased levels of HSP72 in the late stages of AD also found in hippocampus [58]. Exosomes containing HSP72 are thought to induce inflammation [70] and could thus be responsible for such reactions in the AD brain [71]. Finally, we verified higher level of TALDO in AD compared to controls. Transaldolase is an enzyme involved in the pentose phosphate pathway [72] and altered levels of this enzyme in AD might be a result of cerebral glucose hypo metabolism previously reported in patients with AD [73, 74]. Interestingly TALDO has also been found in EVs [75] and recently it has been reported as a potential biomarker for Parkinson’s disease in blood [76], suggesting EVs as a potential sources for finding novel biomarkers and pathologically important proteins for neurodegenerativediseases.

Increased levels of the neuroinflammation marker GFAP

In the AD brain, upon injury or disease, reactive astrocytes upregulate GFAP, leading to increased number of GFAP+ cells [77] surrounding and replacing dead or dying neuronal cells [78]. The increased level of GFAP in both OND and AD compared to controls is well in line with the general neuroinflammatory response observed by others [58, 79].

Conclusions

In conclusion, our results indicate that several proteins involved in exo-endocytic pathways and extracellular vesicle functions display altered levels in the AD brain. We hypothesize that such changes may result in disturbed cellular clearance and a perturbed cell-to-cell communication that may contribute to neuronal dysfunction and cell death in AD.

Footnotes

ACKNOWLEDGMENTS

This research was supported by Uppsala Berzelii Technology Centre for Neurodiagnostics, with financing from the Swedish Governmental Agency for Innovation Systems, Alzheimer fonden, Gun och Bertil Stohnes stiftelse, Lars Hiertas Minne stiftelse, stiftelsen för Gamla Tjänarinnor, the Swedish Research Council P29797-1 JB grant (621-2011-4423), ProNova VINN Excellence Centre for Protein Technology (VINNOVA, Swedish Governmental Agency for Innovation Systems), Knut and Alice Wallenberg Foundation, Geriatriska fonden, and the KTH Center for Applied Proteomics funded by the Erling-Persson Family Foundationas well from SciLifeLab. The authors also wish to thank the entire staff of the Human Protein Atlas for their efforts to produce the antibodies included in the study. The authors acknowledge Dr. Levon Manukyan for technical assistance with western blot analysis.

Authors’ disclosures available online ().

Appendix

References

Evans

, Funkenstein

, Albert

, Scherr

, Cook

, Chown

, Hebert

, Hennekens

, Taylor

(1989) Prevalence of Alzheimer’s disease in a community population of older persons. Higher than previously reported. JAMA 262, 2551–2556.

Hye

, Lynham

, Thambisetty

, Causevic

, Campbell

, Byers

, Hooper

, Rijsdijk

, Tabrizi

, Banner

, Shaw

, Foy

, Poppe

, Archer

, Hamilton

, Powell

, Brown

, Sham

, Ward

, Lovestone

(2006) Proteome-based plasma biomarkers for Alzheimer’s disease. Brain 129, 3042–3050.

Fratiglioni

, De Ronchi

, Aguero-Torres

(1999) Worldwide prevalence and incidence of dementia. Drugs Aging 15, 365–375.

Forstl

, Kurz

(1999) Clinical features of Alzheimer’s disease. Eur Arch Psychiatry Clin Neurosci 249, 288–290.

Davies

, Wolska

, Hilbich

, Multhaup

, Martins

, Simms

, Beyreuther

, Masters

(1988) A4 amyloid protein deposition and the diagnosis of Alzheimer’s disease: Prevalence in aged brains determined by immunocytochemistry compared with conventional neuropathologic techniques. Neurology 38, 1688–1693.

Andreasen

(1997) Linking mind and brain in the study of mental illnesses: A project for a scientific psychopathology. Science 275, 1586–1593.

Honer

, Falkai

, Bayer

, Xie

, Hu

, Li

, Arango

, Mann

, Dwork

, Trimble

(2002) Abnormalities of SNARE mechanism proteins in anterior frontal cortex in severe mental illness. Cereb Cortex 12, 349–356.

Reddy

, Mani

, Park

, Jacques

, Murdoch

, Whetsell

Jr , Kaye

, Manczak

(2005) Differential loss of synaptic proteins in Alzheimer’s disease: Implications for synaptic dysfunction. J Alzheimers Dis 7, 103–117; discussion 173-180.

Masliah

, Honer

, Mallory

, Voigt

, Kushner

, Hansen

, Terry

(1994) Topographical distribution of synaptic-associated proteins in the neuritic plaques of Alzheimer’s disease hippocampus. Acta Neuropathol 87, 135–142.

10.

Sze

, Troncoso

, Kawas

, Mouton

, Price

, Martin

(1997) Loss of the presynaptic vesicle protein synaptophysin in hippocampus correlates with cognitive decline in Alzheimer disease. J Neuropathol Exp Neurol 56, 933–944.

11.

Sze

, Bi

, Kleinschmidt-DeMasters

, Filley

, Martin

(2000) Selective regional loss of exocytotic presynaptic vesicle proteins in Alzheimer’s disease brains. J Neurol Sci 175, 81–90.

12.

Saheki

, De Camilli

(2012) Synaptic vesicle endocytosis. Cold Spring Harb Perspect Biol 4, a005645.

13.

Harris

, Schuske

, Jorgensen

(2001) Studies of synaptic vesicle endocytosis in the nematode C. elegans. Traffic 2, 597–605.

14.

, Hamid

, Shin

, Chiang

(2014) Exocytosis and endocytosis: Modes, functions, and coupling mechanisms. Annu Rev Physiol 76, 301–331.

15.

Mulcahy

, Pink

, Carter

(2014) Routes and mechanisms of extracellular vesicle uptake. J Extracell Vesicles 3, 24641.

16.

Rajendran

, Honsho

, Zahn

, Keller

, Geiger

, Verkade

, Simons

(2006) Alzheimer’s disease beta-amyloid peptides are released in association with exosomes. Proc Natl Acad Sci U S A 103, 11172–11177.

17.

Wang

, Dinkins

, He

, Zhu

, Poirier

, Campbell

, Mayer-Proschel

, Bieberich

(2012) Astrocytes secrete exosomes enriched with proapoptotic ceramide and prostate apoptosis response 4 (PAR-4): Potential mechanism of apoptosis induction in Alzheimer disease (AD). J Biol Chem 287, 21384–21395.

18.

Gupta

, Pulliam

(2014) Exosomes as mediators of neuroinflammation. J Neuroinflammation 11, 68.

19.

Lin

, Li

, Huang

, Liu

, Chen

X-M

, Xu

Y-M

, Huang

L-F

, Wang

X-Z

(2015) Exosomes: Novel biomarkers for clinical diagnosis. ScientificWorldJournal 2015, Article ID 657086, 8 pages, 2015. doi:10.1155/2015/657086

20.

Properzi

, Logozzi

, Fais

(2013) Exosomes: The future of biomarkers in medicine. Biomark Med 7, 769–778.

21.

Fiandaca

, Kapogiannis

, Mapstone

, Boxer

, Eitan

, Schwartz

, Abner

, Petersen

, Federoff

, Miller

, Goetzl

(2015) Identification of preclinical Alzheimer’s disease by a profile of pathogenic proteins in neurally derived blood exosomes: A case-control study. Alzheimers Dement 11, 600–607.e601.

22.

Mirra

, Heyman

, McKeel

, Sumi

, Crain

, Brownlee

, Vogel

, Hughes

, Vanbelle

, Berg

(1991) The Consortium to Establish a Registry for Alzheimer’s Disease (CERAD) 2. Standardization of the Neuropathologic Assessment of Alzheimer’s Disease. Neurology 41, 479–486.

23.

Mastro

, Hall

(1999) Protein delipidation and precipitation by tri-n-butylphosphate, acetone, and methanol treatment for isoelectric focusing and two-dimensional gel electrophoresis. Anal Biochem 273, 313–315.

24.

Wetterhall

, Shevchenko

, Artemenko

, Sjodin

MOD

, Bergquist

(2011) Analysis of membrane and hydrophilic proteins simultaneously derived from the mouse brain using cloud-point extraction. Anal Bioanal Chem 400, 2827–2836.

25.

Lowry

, Rosebrough

, Farr

, Randall

(1951) Protein measurement with the folin phenol reagent. J Biol Chem 193, 265–275.

26.

Chambers

, Maclean

, Burke

, Amodei

, Ruderman

, Neumann

, Gatto

, Fischer

, Pratt

, Egertson

, Hoff

, Kessner

, Tasman

, Shulman

, Frewen

, Baker

, Brusniak

, Paulse

, Creasy

, Flashner

, Kani

, Moulding

, Seymour

, Nuwaysir

, Lefebvre

, Kuhlmann

, Roark

, Rainer

, Detlev

, Hemenway

, Huhmer

, Langridge

, Connolly

, Chadick

, Holly

, Eckels

, Deutsch

, Moritz

, Katz

, Agus

, MacCoss

, Tabb

, Mallick

(2012) A cross-platform toolkit for mass spectrometry and proteomics. Nat Biotechnol 30, 918–920.

27.

Weisser

, Nahnsen

, Grossmann

, Nilse

, Quandt

, Brauer

, Sturm

, Kenar

, Kohlbacher

, Aebersold

, Malmstrom

(2013) An automated pipeline for high-throughput label-free quantitative proteomics. J Proteome Res 12, 1628–1644.

28.

Sturm

, Bertsch

, Gropl

, Hildebrandt

, Hussong

, Lange

, Pfeifer

, Schulz-Trieglaff

, Zerck

, Reinert

, Kohlbacher

(2008) OpenMS - an open-source software framework for mass spectrometry. BMC Bioinformatics 9, 163.

29.

Craig

, Beavis

(2004) TANDEM: Matching proteins with tandem mass spectra. Bioinformatics 20, 1466–1467.

30.

Geer

, Markey

, Kowalak

, Wagner

, Xu

, Maynard

, Yang

, Shi

, Bryant

(2004) Open mass spectrometry search algorithm. J Proteome Res 3, 958–964.

31.

Nahnsen

, Bertsch

, Rahnenführer

, Nordheim

, Kohlbacher

(2011) Probabilistic consensus scoring improves tandem mass spectrometry peptide identification. J Proteome Res 10, 3332–3343.

32.

Kultima

, Nilsson

, Scholz

, Rossbach

, Falth

, Andren

(2009) Development and evaluation of normalization methods for label-free relative quantification of endogenous peptides. Mol Cell Proteomics 8, 2285–2295.

33.

Vizcaino

, Csordas

, del-Toro

, Dianes

, Griss

, Lavidas

, Mayer

, Perez-Riverol

, Reisinger

, Ternent

, Xu

, Wang

, Hermjakob

(2016) 2016 update of the PRIDE database and its related tools. Nucleic Acids Res 44, D447–D456.

34.

R Core Team (2013) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL http://www.R-project.org.

35.

Smyth

(2004) Linear models and empirical bayes methods for assessing differential expression in microarray experiments.Article. Stat Appl Genet Mol Biol 3, 3.

36.

Smyth

(2005) Limma: Linear models for microarray data. In Bioinformatics and Computational Biology Solutions using R and Bioconductor, Gentleman

, Carey

, Dudoit

, Irizarry

, Huber

, eds. Springer, New York, pp. 397–420.

37.

Templ

, Kowarik

, Filzmoser

(2011) Iterative stepwise regression imputation using standard and robust methods. Comput Stat Data Anal 55, 2793–2806.

38.

Smyth

, Michaud

, Scott

(2005) Use of within-array replicate spots for assessing differential expression in microarray experiments. Bioinformatics 21, 2067–2075.

39.

Silva

, Gorenstein

, Li

, Vissers

, Geromanos

(2006) Absolute quantification of proteins by LCMSE: A virtue of parallel MS acquisition. Mol Cell Proteomics 5, 144–156.

40.

Franceschini

, Szklarczyk

, Frankild

, Kuhn

, Simonovic

, Roth

, Lin

, Minguez

, Bork

, von Mering

, Jensen

(2013) STRING v9.1: Protein-protein interaction networks, with increased coverage and integration. Nucleic Acids Res 41, D808–D815.

41.

Ayoglu

, Chaouch

, Lochmuller

, Politano

, Bertini

, Spitali

, Hiller

, Niks

, Gualandi

, Ponten

, Bushby

, Aartsma-Rus

, Schwartz

, Le Priol

, Straub

, Uhlen

, Cirak

, t Hoen

, Muntoni

, Ferlini

, Schwenk

, Nilsson

, Al-Khalili Szigyarto

(2014) Affinity proteomics within rare diseases: A BIO-NMD study for blood biomarkers of muscular dystrophies. EMBO Mol Med 6, 918–936.

42.

Bachmann

, Burte

, Pramana

, Conte

, Brown

, Orimadegun

, Ajetunmobi

, Afolabi

, Akinkunmi

, Omokhodion

, Akinbami

, Shokunbi

, Kampf

, Pawitan

, Uhlen

, Sodeinde

, Schwenk

, Wahlgren

, Fernandez-Reyes

, Nilsson

(2014) Affinity proteomics reveals elevated muscle proteins in plasma of children with cerebral malaria. PLoS Pathog 10, e1004038.

43.

Bystrom

, Ayoglu

, Haggmark

, Mitsios

, Hong

, Drobin

, Forsstrom

, Fredolini

, Khademi

, Amor

, Uhlen

, Olsson

, Mulder

, Nilsson

, Schwenk

(2014) Affinity proteomic profiling of plasma, cerebrospinal fluid, and brain tissue within multiple sclerosis. J Proteome Res 13, 4607–4619.

44.

Haggmark

, Mikus

, Mohsenchian

, Hong

, Forsstrom

, Gajewska

, Baranczyk-Kuzma

, Uhlen

, Schwenk

, Kuzma-Kozakiewicz

, Nilsson

(2014) Plasma profiling reveals three proteins associated to amyotrophic lateral sclerosis. Ann Clin Transl Neurol 1, 544–553.

45.

Schwenk

, Gry

, Rimini

, Uhlen

, Nilsson

(2008) Antibody suspension bead arrays within serum proteomics. J Proteome Res 7, 3168–3179.

46.

Dezfouli

, Vickovic

, Iglesias

, Nilsson

, Schwenk

, Ahmadian

(2014) Magnetic bead assisted labeling of antibodies at nanogram scale. Proteomics 14, 14–18.

47.

Schoch

, Deak

, Konigstorfer

, Mozhayeva

, Sara

, Sudhof

, Kavalali

(2001) SNARE function analyzed in synaptobrevin/VAMP knockout mice. Science 294, 1117–1122.

48.

Ungar

, Hughson

(2003) SNARE protein structure and function. Annu Rev Cell Dev Biol 19, 493–517.

49.

Peng

, Liu

, Ruan

, Tepp

, Stoothoff

, Brown

, Johnson

, Yao

, Zhang

, Dong

(2013) Cytotoxicity of botulinum neurotoxins reveals a direct role of syntaxin 1 and SNAP-25 in neuron survival. Nat Commun 4, 1472.

50.

Johannes

, Doussau

, Clabecq

, Henry

, Darchen

, Poulain

(1996) Evidence for a functional link between Rab3 and the SNARE complex. J Cell Sci 109(Pt 12), 2875–2884.

51.

Beites

, Xie

, Bowser

, Trimble

(1999) The septin CDCrel-1 binds syntaxin and inhibits exocytosis. Nat Neurosci 2, 434–439.

52.

Zhou

, Brewer

, Rizo

(2013) Analysis of SNARE complex/synaptotagmin-1 interactions by one-dimensional NMR spectroscopy. Biochemistry 52, 3446–3456.

53.

Leenders

, Lopes da Silva

, Ghijsen

, Verhage

(2001) Rab3a is involved in transport of synaptic vesicles to the active zone in mouse brain nerve terminals. Mol Biol Cell 12, 3095–3102.

54.

Dulubova

, Lou

, Lu

, Huryeva

, Alam

, Schneggenburger

, Sudhof

, Rizo

(2005) A Munc13/RIM/Rab3 tripartite complex: From priming to plasticity? EMBO J 24, 2839–2850.

55.

Verhage

, Maia

, Plomp

, Brussaard

, Heeroma

, Vermeer

, Toonen

, Hammer

, van den Berg

, Missler

, Geuze

, Sudhof

(2000) Synaptic assembly of the brain in the absence of neurotransmitter secretion. Science 287, 864–869.

56.

Wang

, Thiagarajan

, Wang

, Tewolde

, Rich

, Engisch

(2008) Regulation of quantal shape by Rab3A: Evidence for a fusion pore-dependent mechanism. J Physiol 586, 3949–3962.

57.

Toonen

, Wierda

, Sons

, de Wit

, Cornelisse

, Brussaard

, Plomp

, Verhage

(2006) Munc18-1 expression levels control synapse recovery by regulating readily releasable pool size. Proc Natl Acad Sci U S A 103, 18332–18337.

58.

Hondius

, van Nierop

, Li

, Hoozemans

, van der Schors

, van Haastert

, van der Vies

, Rozemuller

, Smit

(2016) Profiling the human hippocampal proteome at all pathologic stages of Alzheimer’s disease. Alzheimers Dement 12, 654–668.

59.

Musunuri

, Wetterhall

, Ingelsson

, Lannfelt

, Artemenko

, Bergquist

, Kultima

, Shevchenko

(2014) Quantification of the brain proteome in Alzheimer’s disease using multiplexed mass spectrometry. J Proteome Res 13, 2056–2068.

60.

Zelaya

, Perez-Valderrama

, de Morentin

, Tunon

, Ferrer

, Luquin

, Fernandez-Irigoyen

, Santamaria

(2015) Olfactory bulb proteome dynamics during the progression of sporadic Alzheimer’s disease: Identification of common and distinct olfactory targets across Alzheimer-related co-pathologies. Oncotarget 6,` 39437–39456.

61.

Heuser

(1980) Three-dimensional visualization of coated vesicle formation in fibroblasts. J Cell Biol 84, 560–583.

62.

Kasprowicz

, Kuenen

, Miskiewicz

, Habets

, Smitz

, Verstreken

(2008) Inactivation of clathrin heavy chain inhibits synaptic recycling but allows bulk membrane uptake. J Cell Biol 182, 1007–1016.

63.

Tian

, Chang

, Fan

, Flajolet

, Greengard

(2013) Adaptor complex AP2/PICALM, through interaction with LC3, targets Alzheimer’s APP-CTF for terminal degradation via autophagy. Proc Natl Acad Sci U S A 110, 17071–17076.

64.

Owen

, Collins

, Evans

(2004) Adaptors for clathrin coats: Structure and function. Annu Rev Cell Dev Biol 20, 153–191.

65.

Antonescu

, Aguet

, Danuser

, Schmid

(2011) Phosphatidylinositol-(4,5)-bisphosphate regulates clathrin-coated pit initiation, stabilization, and size. Mol Biol Cell 22, 2588–2600.

66.

Sun

, Carroll

, Kaksonen

, Toshima

, Drubin

(2007) PtdIns(4,5)P2 turnover is required for multiple stages during clathrin- and actin-dependent endocytic internalization. J Cell Biol 177, 355–367.

67.

Appelqvist

, Waster

, Kagedal

, Ollinger

(2013) The lysosome: From waste bag to potential therapeutic target. J Mol Cell Biol 5, 214–226.

68.

Nixon

, Yang

, Lee

(2008) Neurodegenerative lysosomal disorders: A continuum from development to late age. Autophagy 4, 590–599.

69.

Sollvander

, Nikitidou

, Brolin

, Soderberg

, Sehlin

, Lannfelt

, Erlandsson

(2016) Accumulation of amyloid-beta by astrocytes result in enlarged endosomes and microvesicle-induced apoptosis of neurons. Mol Neurodegener 11, 38.

70.

Anand

, Anand

, Bleck

, Anes

, Griffiths

(2010) Exosomal Hsp70 induces a pro-inflammatory response to foreign particles including mycobacteria. PLoS One 5, e10136.

71.

Heppner

, Ransohoff

, Becher

(2015) Immune attack: The role of inflammation in Alzheimer disease. Nat Rev Neurosci 16, 358–372.

72.

Stincone

, Prigione

, Cramer

, Wamelink

, Campbell

, Cheung

, Olin-Sandoval

, Gruning

, Kruger

, Tauqeer Alam

, Keller

, Breitenbach

, Brindle

, Rabinowitz

, Ralser

(2015) The return of metabolism: Biochemistry and physiology of the pentose phosphate pathway. Biol Rev Camb Philos Soc 90, 927–963.

73.

Palmer

(1999) The activity of the pentose phosphate pathway is increased in response to oxidative stress in Alzheimer’s disease. J Neural Transm (Vienna) 106, 317–328.

74.

Chen

, Zhong

(2013) Decoding Alzheimer’s disease from perturbed cerebral glucose metabolism: Implications for diagnostic and therapeutic strategies. Prog Neurobiol 108, 21–43.

75.

Prunotto

, Farina

, Lane

, Pernin

, Schifferli

, Hochstrasser

, Lescuyer

, Moll

(2013) Proteomic analysis of podocyte exosome-enriched fraction from normal human urine. J Proteomics 82, 193–229.

76.

Alberio

, McMahon

, Cuccurullo

, Gethings

, Lawless

, Zibetti

, Lopiano

, Vissers

, Fasano

(2014) Verification of a Parkinson’s disease protein signature in T-lymphocytes by multiple reaction monitoring. J Proteome Res 13, 3554–3561.

77.

Dimou

, Gotz

(2014) Glial cells as progenitors and stem cells: New roles in the healthy and diseased brain. Physiol Rev 94, 709–737.

78.

Kato

, Gondo

, Hoshii

, Takahashi

, Yamada

, Ishihara

(1998) Confocal observation of senile plaques in Alzheimer’s disease: Senile plaque morphology and relationship between senile plaques and astrocytes. Pathol Int 48, 332–340.

79.

Ben Haim

, Carrillo-de Sauvage

, Ceyzeriat

, Escartin

(2015) Elusive roles for reactive astrocytes in neurodegenerative diseases. Front Cell Neurosci 9, 278.