Biochimica et Biophysica Acta 1843 (2014) 1150–1161

Contents lists available at ScienceDirect

Biochimica et Biophysica Acta journal homepage: www.elsevier.com/locate/bbamcr

Loss of proteostasis induced by amyloid beta peptide in brain endothelial cells Ana Catarina Fonseca a,b, Catarina R. Oliveira a,c, Cláudia F. Pereira a,c, Sandra M. Cardoso a,c,⁎ a b c

Center for Neuroscience and Cell Biology, University of Coimbra, Largo Marquês de Pombal, 3004-517 Coimbra, Portugal Department of Life Sciences, Faculty of Sciences and Technology, University of Coimbra, Apartado 3046, 3001-401 Coimbra, Portugal Faculty of Medicine, University of Coimbra, Rua Larga, 3004-504 Coimbra, Portugal

a r t i c l e

i n f o

Article history: Received 27 September 2013 Received in revised form 20 February 2014 Accepted 23 February 2014 Available online 28 February 2014 Keywords: Alzheimer's disease Amyloid-beta Autophagy Endoplasmic reticulum stress Ubiquitin–proteasome system Apoptosis

a b s t r a c t Abnormal accumulation of amyloid-β (Aβ) peptide in the brain is a pathological hallmark of Alzheimer's disease (AD). In addition to neurotoxic effects, Aβ also damages brain endothelial cells (ECs) and may thus contribute to the degeneration of cerebral vasculature, which has been proposed as an early pathogenic event in the course of AD and is able to trigger and/or potentiate the neurodegenerative process and cognitive decline. However, the mechanisms underlying Aβ-induced endothelial dysfunction are not completely understood. Here we hypothesized that Aβ impairs protein quality control mechanisms both in the secretory pathway and in the cytosol in brain ECs, leading cells to death. In rat brain RBE4 cells, we demonstrated that Aβ1–40 induces the failure of the ER stress-adaptive unfolded protein response (UPR), deregulates the ubiquitin–proteasome system (UPS) decreasing overall proteasome activity with accumulation of ubiquitinated proteins and impairs the autophagic protein degradation pathway due to failure in the autophagic flux, which culminates in cell demise. In conclusion, Aβ deregulates proteostasis in brain ECs and, as a consequence, these cells die by apoptosis. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Alzheimer's disease (AD) is the most frequent cause of dementia in the elderly. One neuropathological hallmark of AD is the extracellular accumulation of amyloid-beta (Aβ) peptide in brain parenchyma and, according to the ‘Amyloid Cascade Hypothesis’, Aβ triggers a cascade of toxic events leading to selective synaptic and neuronal loss [1,2]. In addition, the majority of patients with AD have cerebral amyloid angiopathy (CAA), thus showing vascular deposition of Aβ [3] as a consequence of impaired transport of Aβ across the blood–brain barrier (BBB) [4,5]. Moreover, Aβ contributes to BBB leakage in humans with CAA and in transgenic mice [6]. A growing body of evidences indicate that neurovascular dysfunction occurs before the first symptoms of the disease and plays an important role in the neurodegenerative process and consequent cognitive decline in AD [4,7]. Recent findings in mice modeling AD demonstrated that microvascular impairment directly correlates with Aβ accumulation [8]. Indeed, results obtained in human and animal cultured cells and animal isolated vessels suggest that the age-dependent degeneration of brain vasculature and dysfunction of brain capillary endothelium is due to the toxic effects of Aβ peptide on endothelial cells (ECs) [9–12]. In human ECs, Aβ causes ⁎ Corresponding author at: Center for Neuroscience and Cell Biology, Faculty of Medicine, University of Coimbra, Largo Marquês de Pombal, 3004-517 Coimbra, Portugal. Tel.: +351 239820190; fax: +351 239822776. E-mail address: [email protected] (S.M. Cardoso).

http://dx.doi.org/10.1016/j.bbamcr.2014.02.016 0167-4889/© 2014 Elsevier B.V. All rights reserved.

irreversible morphological and functional changes that result in inhibition of their proliferative activity through accumulation of autophagic vesicles, reduces survival and induces apoptosis [13–15]. However, the molecular mechanisms underlying Aβ-induced endothelial dysfunction have not been fully elucidated yet. The endoplasmic reticulum (ER) is an organelle involved in folding and processing of proteins in the secretory pathway and in Ca2 + homeostasis. When the amount of proteins with abnormal conformation exceeds the repair capacity of the ER, the signaling pathways of the unfolded protein response (UPR), initiated by the ER stress sensors PERK (protein kinase RNA-like ER kinase), IRE1 (inositol-requiring protein1) and ATF6 (activating transcription factor 6), are activated in order to re-establish ER homeostasis and avoid cell damage. The UPR mediators lead to the transcription of genes encoding ER-resident chaperones, such as GRP78, genes involved in ER and Golgi biogenesis, and also genes implicated in ER-associated degradation (ERAD), attenuating general translation to decrease the demand of ER and activate the ERAD to degrade the aberrant proteins [16,17]. During ERAD, an ubiquitin chain is added to misfolded proteins that are then retrotranslocated to the cytosol for degradation in the proteasome [18]. However, if UPR and ERAD fail, ER stress can also trigger the lysosome-mediated autophagic protein degradation pathway to preserve cell survival [19]. In the event of chronic or unmitigated ER stress, the UPR activates apoptotic cascades [20]. The aim of this work was to explore the hypothesis that Aβ damages microvascular brain ECs through the impairment of protein quality

A.C. Fonseca et al. / Biochimica et Biophysica Acta 1843 (2014) 1150–1161

control mechanisms in the secretory pathway and in the cytosol leading to activation of apoptotic cell death pathways. Our results show that the Aβ1–40 isoform, which accumulates in brain vasculature [21], activates ER stress, promotes the accumulation of ubiquitinated proteins and reduces the proteasome activity, inhibits macroautophagy and, in the absence of compensatory mechanisms, activates apoptosis in ECs derived from rat brain microvessels. This knowledge could contribute to the development of novel therapeutic strategies to prevent or delay the progression of AD. 2. Materials and methods 2.1. Materials Collagen was obtained from Roche Applied Science (Mannheim, Germany). Mem-Alpha medium with Glutamax-1, Nut Mix F-10 W/ GLUTAMAX-1, fetal bovine serum (FBS), and geneticin were acquired from Invitrogen Life Science (Paisley, UK). The synthetic Aβ1–40 peptide and the Suc-Leu-Leu-Val-Tyr-AMC substrate were from Bachem (Bubendorf, Switzerland). Boc-Leu-Arg-Arg-AMC and Z-Leu-Leu-GluAMC substrates were acquired from Peptide Institute (Osaka, Japan). Hoechst 33342 and Alexa Fluor 594 goat anti-rabbit IgG conjugate were obtained from Molecular Probes (Leiden, The Netherlands). Glycergel Mounting Medium was obtained from DakoCytomation Inc. (Carpinteria, CA, USA). Colorimetric substrates for caspase-2, -3, -9, and -12 Ac-VDAVD-pNA, Ac-DEVD-pNA, Ac-LEHD-pNA, and Ac-LEVD-pNA were purchased from Calbiochem (Darmstadt, Germany). Polyvinylidene difluoride (PVDF) membrane, goat alkaline phosphatase-linked antirabbit and anti-mouse secondary antibodies, and the enhanced chemifluorescence (ECF) reagent were acquired from Amersham Pharmacia Biotech (Buckinghamshire, UK). Mouse monoclonal antibodies reactive against GRP78 and Beclin-1 were from BD Biosciences (Heidelberg, Germany). Rabbit polyclonal antibodies reactive against HDAC6 and XBP-1 (spliced and unspliced form) were from Abcam plc (Cambridge, UK). Rabbit monoclonal antibodies reactive against Bcl-2, LC3B-XP, Aβ and rabbit polyclonal antibodies reactive against LC3B and p62 were acquired from Cell Signaling Technology, Inc. (Danvers, MA, USA). Mouse monoclonal antibody against glyceraldehyde-3phosphate dehydrogenase (GAPDH) was from Chemicon International Inc. (Temecula, CA, USA). Mouse monoclonal anti-lysosomal-associated membrane protein (LAMP)-1 (clone H4A3) was from the Developmental Studies Hybridoma Bank (University of Iowa, Iowa City, IA, USA). Bio-Rad protein dye assay reagent, acrylamide and the pre-stained Precision Plus Protein All Blue Standard were purchased from Bio-Rad (Hercules, CA, USA). Anti-α-tubulin mouse monoclonal antibody, Trypsin-EDTA solution, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), protease inhibitors (leupeptin, pepstatin A, chymostatin, and aprotinin), phenylmethylsulfonyl fluoride (PMSF), 4-(2hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), Tris, sodium dodecyl sulfate (SDS), dimethyl sulfoxide (DMSO), 1,4-dithiothreitol (DTT), ethylene glycol tetraacetic acid (EGTA), glycerol, bromophenol blue, NaH2PO4, ethylenediamine tetraacetic acid (EDTA), recombinant human basic fibroblast growth factor (bFGF), Tween 20, glucose, CHAPS (3-[(3-Cholamidopropyl)-dimethylammonio]-1-propane sulfonate), paraformaldehyde, phosphate-buffered saline (PBS), Coomassie G-250, thapsigargin, NH4Cl, 3-methyladenine (3-MA), rapamycin and lactacystin were obtained from Sigma Chemical Co. (St. Louis, MO, USA). Bovine serum albumin (BSA), NaCl, KCl, MgCl 2 , CaCl 2 , 2propanol, HCl, Triton X-100, and methanol were purchased from Merck (Darmstadt, Germany). 2.2. Cell culture and treatments The RBE4 cell line from rat brain microvasculature was provided by Dr. Jon Holy (University of Minnesota, Duluth, USA). It is a continuous, immortalized cell line that retains a stable phenotype reminiscent of

1151

BBB endothelium in vitro [22]. RBE4 cells at passages 10–35 were grown on 75 cm2 tissue culture flasks coated with 4.15 μg/cm2 collagen in MEM-Alpha medium with Glutamax-1 and Nut Mix F-10 W/ GLUTAMAX-1 (1:1 vol/vol), supplemented with 10% heat inactivated FBS, 1 ng/ml bFGF and 0.3 mg/ml geneticin. The cultures were maintained at 37 °C in a humidified atmosphere of 5% CO2/95% air. The cells were plated on collagen-coated (4.15 μg/cm2) multiwells for cell viability, western blotting (WB) analysis or measurement of proteasome, and caspase activities or coverslips coated with poly-L-lysine (0.1 mg/ml) plus collagen (4.15 μg/cm2) for LC3B sub-cellular localization or analysis of necrotic/apoptotic cell death markers, at a density of approximately 22,000 cells/cm2, and treated with 20 mM NH4Cl, 10 mM 3-MA, 0.1 μM rapamycin, 2.5 μM Aβ1–40, 2 μM thapsigargin, or 2 μM lactacystin. Synthetic Aβ1–40-HCl was dissolved in sterile water at a concentration of 6 mg/ml and then diluted to 1 mg/ml (231 μM) in PBS and stored at −20 °C. Thapsigargin (non-competitive inhibitor of ER Ca2+-ATPase that depletes Ca2 + in the ER), rapamycin (macroautophagy inducer) and lactacystin (inhibitor of 20S proteasome) were diluted in DMSO and stored at − 20 °C until use. 3-MA (macroautophagy inhibitor) was prepared in culture medium and NH4Cl (inhibitor of lysosomal function by increasing the intra-lysosomal pH) was dissolved in sterile water both immediately before use. 2.3. Measurement of protein levels by western blotting analysis After treatments, RBE4 cells were washed 2 times with PBS (pH 7.4), lysed and scrapped in ice cold lysis buffer (in mM): 25 HEPES-Na, 2 MgCl 2, 1 EDTA, 1 EGTA, supplemented with 0.1% Triton X-100, 100 μM PMSF, 2 mM DTT, and 1:1000 of a protease inhibitor cocktail (1 μg/ml leupeptin, pepstatin A, chymostatin, and antipain). The cellular extracts were then rapidly frozen in liquid N2 and thawed three times and centrifuged for 1 min at 106 ×g at 4 °C. The supernatant was collected and the protein content was measured using the BioRad protein dye assay reagent. Total cellular extracts containing 15 μg (for the ER stress markers GRP78 and XBP-1) or 50 μg (for LC3B, p62, Beclin-1, Bcl-2, LAMP-1, HDAC6 and ubiquitin) of protein were separated by electrophoresis on 15% (for LC3B) or 10% (for other proteins) (w/v) SDS-polyacrylamide gel (SDS-PAGE) after dilution (1:5) and denaturation at 95 °C for 5 min in sample buffer (in mM): 100 Tris, 100 DTT, 4% (v/v) SDS, 0.2% (w/v) bromophenol blue, and 20% (v/v) glycerol. For Aβ aggregation analysis, 40 μg of extracted cellular proteins or an 8.66 μg aliquot of the Aβ1–40 stock was separated by electrophoresis on 4–16% Tris–Tricine SDS-PAGE [23] after dilution (1:2) in sample buffer: 40% (w/v) glycerol, 2% (w/v) SDS, 0.2 M Tris–HCl, pH 6.8, and 0.005% (w/v) Coomassie G-250. To facilitate the identification of proteins of interest, the pre-stained Precision Plus Protein All Blue Standards was used. Proteins were then transferred to PVDF membranes, which were further blocked for 1 h at room temperature (RT) with 5% (w/v) BSA in Tris-buffered saline (150 mM NaCl, 50 mM Tris, pH 7.6) with 0.1% (w/v) Tween 20 (TBS-T). The membranes were next incubated overnight at 4 °C with primary mouse monoclonal antibodies against GRP78 (1:1000 dilution in TBS-T), Beclin-1 (1:1000 dilution in TBS-T), LAMP-1 (1:1000 dilution in TBS-T), with a primary rabbit monoclonal antibody against Bcl-2 (1:1000 dilution in TBS-T) or βAmyloid (D54D2) XP® (1:200 dilution in TBS-T); or with primary rabbit polyclonal antibodies against LC3B (1:1000 dilution in TBS-T), p62 (1:1000 dilution in TBS-T), HDAC6 (1:1000 dilution in TBS-T), or ubiquitin (1:500 dilution in TBS-T). Control of protein loading was performed using primary mouse monoclonal antibodies against α-tubulin (1:20,000 dilution in TBS-T) or GAPDH (1:10,000 dilution in TBS-T). After washing, membranes were incubated for 1 h at RT with a goat alkaline phosphatase conjugated secondary anti-mouse or anti-rabbit antibodies (1:20,000 dilution in TBS-T). Bands of immunoreactive proteins were visualized after membrane incubation with ECF reagent for about 5 min, on a Versa Doc 3000 Imaging System (Bio-Rad, Hercules,

1152

A.C. Fonseca et al. / Biochimica et Biophysica Acta 1843 (2014) 1150–1161

A.C. Fonseca et al. / Biochimica et Biophysica Acta 1843 (2014) 1150–1161

CA, USA) and densities of protein bands were calculated using the WCIF ImageJ program (Wayne Rasband, Research Services Branch, National Institute of Mental Health, Bethesda, Maryland, USA). The ratios between Aβ, p62, Beclin-1, Bcl-2, LAMP-1, HDAC6, or ubiquitin and GAPDH; or between GRP78, or LC3B and α-tubulin were calculated and results were normalized to control values. 2.4. Evaluation of LC3B localization by immunocytochemistry In order to analyze the subcellular localization of LC3B in the autophagic vacuoles (AVs), treated or untreated RBE4 cells grown in glass coverslips were washed three times with PBS (pH 7.4) and then fixed with 4% (w/v) paraformaldehyde at RT for 20 min. After that, cells were rinsed three times with PBS and permeabilized with 100% methanol at −20 °C for 10 min. After washing three times with PBS the cells were incubated for 30 min at RT in blocking solution containing 3% (w/v) BSA in PBS to prevent non-specific binding. Subsequently, cells were incubated over night at 4 °C with a primary rabbit monoclonal antibody reactive against LC3B-XP (1:200 dilution in blocking solution). Thereafter, coverslips were rinsed in PBS and incubated for 1 h at RT with the secondary antibody anti-rabbit IgG labeled with Alexa Fluor 594 (1:200 dilution) in PBS with 1% (w/v) BSA. Following additional rinses in PBS, cells were stained in the dark for 5 min at RT with Hoechst 33342 (15 μg/ml) prepared in PBS. Finally, the preparations were mounted using Glycergel Mounting Medium. Images were collected using a Zeiss LSM 510 Meta confocal microscope (Carl Zeiss, New York, USA) with a 63 × objective. The experiments were performed three times in duplicate, and a minimum of 150 cells was scored for each coverslip. 2.5. Analysis of proteasomal activity Proteasome activity was measured as previously with some modifications [24]. Briefly, after treatments, RBE4 cells were washed twice with PBS, lysed and scrapped in ice cold lysis buffer (1 mM EDTA, 10 mM Tris–HCl pH 7.5, 20% glycerol, and 4 mM DTT), in the presence or absence of 2 mM ATP to measure the activity of 26S or 20S proteasome, respectively. Cells were then rapidly frozen in liquid N2 and thawed three times and centrifuged at 13,000 ×g for 10 min, at 4 °C. Supernatants were collected and protein concentrations were assayed using the BioRad protein dye assay reagent. Cellular extracts containing 50 μg protein were incubated in proteasome activity buffer (0.5 mM EDTA and 50 mM Tris–HCl, pH 8.0) in the absence or presence of 400 μM ZLeu-Leu-Glu-AMC, 100 μM Boc-Leu-Arg-Arg-AMC, or 50 μM SucLeu-Leu-Val-Tyr-AMC that are substrates of PHGH, trypsin-like, and chymotrypsin-like proteolytic activities, respectively. The fluorescence signals obtained with the fluorogenic substrates were monitored for 1 h at 37 °C in a microplate reader (SpectraMax Gemini EM fluorocytometer, Molecular Devices, California, USA) at 380 nm excitation and 460 nm emission and the results were expressed relatively to the control. 2.6. Analysis of caspase-2, -3, -9 and -12-like activity Caspase-like activities were measured in total cellular extracts prepared after treatment of ECs. Cells were washed two times with PBS at 4 °C and scrapped in reaction buffer (25 mM HEPES-Na, 10% sucrose, 10 mM DTT, and 0.1% CHAPS, pH 7.4) at 4 °C. The cellular extracts

1153

were rapidly frozen and thawed three times in liquid N2, and centrifuged for 10 min at 20,800 ×g at 4 °C. The supernatants were collected and assayed for protein content using the Bio-Rad protein dye assay reagent. Aliquots of cellular extracts containing 30 μg protein were reacted for 2 h, at 37 °C, with 0.1 mM Ac-VDAVD-pNA, Ac-DEVD-pNA, Ac-LEHD-pNA, or Ac-LEVD-pNA, chromogenic substrates for caspase2, -3, -9, and -12, respectively, in reaction buffer. Caspase-2, -3, -9, and -12-like activities were determined by measuring substrate cleavage at 405 nm in a microplate reader (SpectraMax Plus 384) and results were expressed relatively to the control. 2.7. Analysis of cell viability by the MTT assay Cell viability was evaluated using the MTT assay, which measures the ability of metabolic active cells to form formazan through cleavage of the tetrazolium ring of MTT. Briefly, cells were washed in sodium medium (in mM: NaCl 132, KCl 4, NaH2PO4 1.2, MgCl2 1.4, glucose 6, HEPES 10, and CaCl2 1, pH 7.4) and incubated with 0.5 mg/ml MTT for 2 h at 37 °ºC. The blue formazan crystals formed were dissolved in an equal volume of 0.04 M HCl in isopropanol and quantified spectrophotometrically by measuring the absorbance at 570 nm using a microplater reader (SpectraMax Plus 384, Molecular Devices, California, USA). Results were expressed as the percentage of the absorbance determined in control cells. 2.8. Data analysis Data were expressed as means ± SEM of determinations performed in duplicate, from at least three independent experiments. Statistical significance analysis was determined using one-way ANOVA followed by Dunnett's post hoc tests or using Student's t-test. The differences were considered significant for P values b 0.05. 3. Results 3.1. Levels of intracellular Aβ In RBE4 cells incubated with extracellularly added Aβ1–40 for 6 h (Fig. 1A–D) or 24 h (Fig. 1E–H) an increase in the levels of intracellular Aβ was found, suggesting that Aβ is internalized and can potentiate its own synthesis and/or impair its degradation in ECs. Thapsigargininduced ER stress, lactacystin-induced proteasomal inhibition and macroautophagy inhibition with 3-MA enhanced the intracellular levels of Aβ (Fig. 1). Moreover, a higher increase in Aβ levels, especially in monomers and aggregates, occurred in the simultaneous presence of these compounds and Aβ1–40 (Fig. 1A–C, E–G), in particular Aβ1–40 and thapsigargin (Fig. 1E–G). These results suggest a close relationship between the accumulation of Aβ and the compromise of protein quality control mechanisms in brain ECs. 3.2. ER stress in brain endothelial RBE4 cells The activation of UPR in RBE4 cells was analyzed by measuring the protein levels of GRP78 (Fig. 2A and B), spliced (active) and unspliced XBP-1 (Fig. 2C–F) by WB after incubation for 6 h with 2.5 μM Aβ1–40, 2 μM thapsigargin (ER stress inducer), 2 μM lactacystin (proteasome inhibitor), or 10 mM 3-MA (macroautophagy inhibitor). Co-treatments were performed simultaneously, in the absence of pre-incubation.

Fig. 1. Intracellular Aβ in brain endothelial cells. The aggregation state of Aβ peptide, in the concentrated Aβ1–40 stock and inside either in untreated cells or in cells treated for 6 h (A–D) or 24 h (E–H) with 2.5 μM Aβ1–40, 2 μM thapsigargin, 2 μM lactacystin, or 10 mM 3-metyladenine, alone or in combination, was analyzed by western blot (A). The levels of aggregated Aβ with more than 50 kDa (B and F), Aβ monomers (C and G), as well as total Aβ levels (D and H) were quantified. Proteins from cell lysates (40 μg) or the Aβ stock (8.66 μg) were separated by SDS-PAGE and immunoblotted with a monoclonal antibody against Aβ. An anti-GAPDH antibody was applied as a protein loading control and used to normalize Aβ levels. The results were calculated relatively to control values and represent the means ± SEM of at least four independent experiments. *p b 0.05, **p b 0.01 and ***p b 0.001 significantly different from untreated cells (control).

1154

A.C. Fonseca et al. / Biochimica et Biophysica Acta 1843 (2014) 1150–1161

Fig. 2. ER stress induction in brain endothelial cells. ER stress was analyzed in treated and untreated cells by western blot. Proteins (15 μg) from lysates obtained from cells treated with 2.5 μM Aβ1–40, 2 μM thapsigargin, 2 μM lactacystin, or 10 mM 3-methyladenine alone or in combination for 6 h (in the absence of pre-treatment) were separated by SDS-PAGE and immunoblotted with antibodies against GRP78 (A and B), unspliced (uXBP-1) and spliced (sXBP-1) forms of XBP-1 (C–F). An anti-α-tubulin antibody was applied as a protein loading control and used to normalize protein levels of ER stress markers. The results were calculated relatively to control values and represent the means ± SEM of at least eight independent experiments. *p b 0.05, **p b 0.01 and ***p b 0.001 significantly different from control. #p b 0.05 and ##p b 0.01 significantly different from Aβ1–40 treatment.

Under these conditions, the levels of the ER stress marker GRP78 were significantly increased. Enhanced GRP78 levels were detected when cells were incubated simultaneously with Aβ1–40 and lactacystin or 3MA showing that Aβ1–40-induced ER stress is potentiated when accumulated misfolded proteins cannot be removed by the proteasome or macroautophagy (Fig. 2A and B). In the same way, unspliced XBP-1 (uXBP-1), which increases upon ER stress, and spliced XBP-1 (sXBP-1) that results from XBP-1 mRNA cleavage by the ER-resident ER stress sensor IRE1 were also enhanced by Aβ1–40 and 3-MA (Fig. 2C–F). Taken together, these results show that Aβ1–40 triggers an ER stress response that is potentiated by the impairment of proteostasis in brain ECs due to inhibition of the proteasome and of lysosome-mediated macroautophagy.

3.3. UPS impairment in brain endothelial RBE4 cells In RBE4 cells treated with 2.5 μM Aβ1–40, 2 μM thapsigargin, 2 μM lactacystin, or 10 mM 3-MA for 6 h, ubiquitin proteasomal system (UPS) activity was investigated through analysis of the levels of ubiquitinated proteins by WB and measurement of the activity of the proteasome β1, β2 and β5 catalytic subunits of the 20S core particle in the presence (26S proteasome) or absence (20S proteasome) of ATP. Aβ1–40, thapsigargin, lactacystin, and 3-MA increased the levels of ubiquitinated proteins (Fig. 3A and B). Accordingly, Aβ1–40 and thapsigargin significantly decreased β1 (PGPH-like), β2 (trypsin-like) and β5 (chymotrypsin-like) activities (Fig. 3C–E). 3-MA significantly decreased β1 and β5 activities without affecting β2 activity (Fig. 3C–E).

A.C. Fonseca et al. / Biochimica et Biophysica Acta 1843 (2014) 1150–1161

1155

Fig. 3. Proteasomal dysfunction in brain endothelial cells. (A and B) The levels of ubiquitinated proteins in RBE4 cells treated with 2.5 μM Aβ1–40, 2 μM thapsigargin and 2 μM lactacystin for 6 h were analyzed by western blot. Proteins (50 μg) from cell lysates were separated by SDS-PAGE and immunoblotted with a polyclonal antibody against ubiquitin. An anti-GAPDH antibody was applied as a protein loading control and used to normalize levels of ubiquitinated proteins. (C–E) Moreover, the activity of β1 (C), β2 (D) and β5 (E) catalytic subunits of proteasome were measured with (26S proteasome) or without (20S proteasome) ATP after 6 h incubation of RBE4 cells with 2.5 μM Aβ1–40, 2 μM thapsigargin, 2 μM lactacystin, or 10 mM 3-metyladenine. The results were calculated relatively to control values and represent the means ± SEM of at least four independent experiments. *p b 0.05, **p b 0.01 and ***p b 0.001 significantly different from control.

These data show that Aβ1–40 induces the accumulation of ubiquitinated proteins, indicating a direct role in the alteration of proteostasis mechanisms. Furthermore, our data show that macroautophagy inhibition is not compensated by activation of proteasome-mediated degradation. 3.4. Autophagy in brain endothelial RBE4 cells To evaluate whether Aβ1–40, ER stress and proteasome inhibition upregulated macroautophagy in RBE4 cells, as a compensatory mechanism, we determined the LC3-II/α-tubulin ratio and the LC3 punctuate, which directly correlates with the amount of autophagosomes [25], by WB and immunocytochemistry, respectively, after treatment with Aβ1–40 (2.5 μM), thapsigargin (2 μM) or lactacystin (2 μM) for 6 h (Fig. 4A, B and D). Only lactacystin increased significantly the basal LC3-II/α-tubulin ratio (Fig. 4A and B). NH4Cl, which changes the pH inside the lysosomes leading to inhibition of lysosomeresident hydrolases, was used to evaluate the lysosomal degradation of autophagosomes' cargo (autophagic flux). As expected, a significant increase in the LC3-II/α-tubulin ratio was determined in cells

stimulated with Aβ1–40, thapsigargin or lactacystin in the presence of NH4Cl (Fig. 4A and B) and the calculated ratio between levels measured in the presence and absence of NH4Cl, which corresponds to the autophagic flux, decreased approximately 32%, 47%, and 55%, respectively, in Aβ1–40-, thapsigargin-, or lactacystin-treated cells compared to untreated cells (Fig. 4C). These data indicates that macroautophagy does not compensate for the loss of proteostasis under ER stressful conditions or compromised proteasome-mediated protein degradation. As the p62 protein is able to bind LC3 in autophagosomes, its amount is frequently used as a marker of autophagic flux. p62 protein levels significantly increased with NH4Cl treatment but Aβ1–40 or thapsigargin didn't change significantly the levels of p62. However, proteasomal inhibition with lactacystin significantly increased p62 levels (Fig. 4E and F). Beclin-1 is involved in the formation of autophagic vesicles and its activity is regulated by interaction with Bcl-2. Therefore, the protein levels of Beclin-1 and Bcl-2 were analyzed by WB in total and cytosolic cellular extracts in the experimental conditions described above. Although thapsigargin slightly decreased Beclin-1, but not Bcl-2 levels, Aβ1–40, thapsigargin or lactacystin was not able to significantly change

1156

A.C. Fonseca et al. / Biochimica et Biophysica Acta 1843 (2014) 1150–1161

A.C. Fonseca et al. / Biochimica et Biophysica Acta 1843 (2014) 1150–1161

the levels of these proteins in total (Fig. 4E, G and H) and in cytosolic (data not shown) extracts, which means that autophagic induction is not compromised in this cells. HDAC6 is involved in recruitment of cargo autophagic proteins to be degraded and is also implicated in the lysosomal dynamics [26]. Therefore, HDAC6 levels were analyzed by WB but significant changes were not found between controls and treated cells (Fig. 4E and I). Lamp1 protein levels, analyzed by WB, were used as lysosomal markers after incubation of RBE4 cells with Aβ1–40, thapsigargin or lactacystin, in the absence or presence of NH4Cl. As expected, the inhibition of lysosomes with NH4Cl increased the amount of Lamp1, a measure to compensate for lysosomal loss of function. Aβ1–40 did not affect Lamp1 levels while thapsigargin and lactacystin led to a significant decrease in Lamp1 (Fig. 4E and J). A decrease in the levels of Lamp1 measured in the presence of NH4Cl in Aβ-, thapsigargin- or lactacystin-treated cells was observed demonstrating that the compensatory lysosomal biogenesis triggered by NH4Cl is compromised. All together, these findings indicate that macroautophagy is compromised in brain ECs treated with Aβ1–40 or exposed to ER stress conditions or proteasome dysfunction, mainly due to inhibition of autophagic flux. 3.5. Moderate ER stress stimulates macroautophagy in brain endothelial RBE4 cells In order to analyze the effect of moderate ER stress on macroautophagy, RBE4 cells were treated for 6 h with low doses of thapsigargin, 25 and 50 nM, and co-treated with 20 mM NH4Cl in the last 4 h of incubation. Both concentrations of thapsigargin increased the protein levels of the ER stress marker GRP78 (Fig. 5A and B) showing that low thapsigargin levels were able to induce ER stress and activate the UPR. At a concentration of 50 nM, thapsigargin enhanced the levels of the macroautophagy marker LC3-II (Fig. 5A and C) but the autophagic flux was not affected compared to untreated ECs in the presence of both tested thapsigargin concentrations (Fig. 5D). In conclusion, a moderate induction of ER stress can stimulate macroautophagy, which might represent an attempt of ECs to restore homeostasis. 3.6. Apoptosis in brain endothelial RBE4 cells After 12 h of treatment with Aβ1–40 (2.5 μM), thapsigargin (2 μM), lactacystin (2 μM), or 3-MA (10 mM), the activity of caspase-2, -3, -9, and -12 was quantified in cellular extracts prepared from RBE4 cells. 3MA and Aβ1–40 significantly activated all caspases analyzed, thapsigargin activated all caspases except caspase-2 and lactacystin only significantly increased caspase-2 activity (Fig. 6A–D). Similar results were obtained after 6 and 24 h of treatment (data not shown). After incubation of RBE4 cells during 12 h with Aβ1–40 (2.5 μM), thapsigargin (2 μM), or 3-MA (10 mM), a significant decrease in cell survival, as measured by the MTT test, was detected. Moreover, the co-incubation with Aβ1–40 and 3-MA further decreased cell survival, as compared with 3-MA or Aβ1–40 alone (Fig. 6E). Additionally, the co-incubation of ECs with 0.1 μM rapamycin and Aβ1–40 (2.5 μM), thapsigargin (2 μM) or lactacystin (2 μM) for 12 h partially reverted the decrease in cell viability induced by Aβ1–40 or thapsigargin alone, but not by lactacystin (Fig. 6F). In conclusion, the impairment of protein quality control mechanisms leads to caspase-dependent apoptotic cell death in brain ECs.

1157

4. Discussion RBE4 cells are commonly used as an in vitro model of the ECs present at the BBB [27], which protect the central nervous system, acting as filters for molecules and microorganisms and also synthesizing neurotrophic factors. Therefore, alterations in the survival of these cells play a major role in several neurological disorders. In particular, neurovascular dysfunction arising from the toxic effects of Aβ on brain ECs has been implicated in the cognitive decline occurring in AD [4]. Previously, we have shown that Aβ 1–40 induces ER stress and consequently activates the UPR and deregulates ER Ca2+ homeostasis, both in rat brain cortical neurons and in ECs, leading to activation of mitochondrial-dependent and -independent apoptotic cell death pathways [28,29]. Here we further investigated the toxic effects of Aβ on brain ECs, using RBE4 cells as an in vitro model, through the analysis of protein quality control mechanisms in the secretory pathway (ER stress-induced UPR) and in the cytosol (UPS, and macroautophagy). In parallel, we tested the effect of well-known modulators of the cellular mechanisms involved in proteostasis, namely thapsigargin (ER stress inducer), lactacystin (proteasome inhibitor), and 3-MA (macroautophagy inhibitor). The ubiquitin–proteasome and the autophagy–lysosome systems are two major protein degradation pathways that can be activated in eukaryotic cells when the ER fails to repair unfolded or misfolded proteins, preventing cell death. However, apoptotic cell death can be triggered when these mechanisms are not able to cope with the excessive accumulation of abnormal proteins. In this work, exposure to extracellular Aβ1–40, as well as perturbation of protein quality control mechanisms (ER stress, inhibition of UPS and inhibition of macroautophagy) were shown to increase the levels of Aβ, mainly monomers and aggregates in the brain ECs. These results, obtained in RBE4 cells co-treated with Aβ1–40 and quality control inhibitors indicate that exogenous Aβ can be internalized and potentiate its own accumulation, through inhibition of proteostatic pathways. In fact, Aβ can be internalized by ECs in the BBB regulating the efflux and influx of Aβ to the brain [30,31]. Furthermore, the accumulation of AVs and ER stress, which delays the transport of proteins along the secretory pathway, can increase Aβ production by enhancing the exposure of amyloid precursor protein (APP) to β- and γ-secretases, which are more abundant in the ER, Golgi and endosomes than in the plasma membrane [32–34]. ER stress was previously shown to increase β- and γ-secretase activities enhancing Aβ production [35,36]. Hence, the significant increase of Aβ monomers in ECs co-treated with Aβ1–40 and thapsigargin, compared to Aβ alone, suggest that Aβ production is increased under ER stress. Since aggregated Aβ is degraded by macroautophagy in physiological conditions, the impairment in macroautophagy reduces the degradation of Aβ [32]. Additionally, an increase in total Aβ levels can potentiate its aggregation [37]. Accordingly, ECs treated with Aβ1–40 alone or in combination with an ER stress inducer, proteasome or autophagy inhibitor, increased the levels of Aβ aggregates with a molecular weight higher than 50 kDa. As expected, Aβ 1–40 and the inhibition of proteasome and macroautophagy activated the UPR since accumulation of abnormal proteins can induce ER stress. These proteins can also be degraded in the proteasome. However, Aβ1–40, as well as thapsigargin, increased the amount of polyubiquitinated proteins suggesting that damaged proteins, targeted for degradation in the proteasome, are not efficiently removed. Accordingly, the activity of β1, β2 and β5 catalytic subunits of

Fig. 4. Macroautophagy deregulation in brain endothelial cells. RBE4 cells were incubated for 6 h with 2.5 μM Aβ1–40, 2 μM thapsigargin or 2 μM lactacystin and co-treated for 4 h with 20 mM NH4Cl. The protein levels of LC3-I (A), LC3-II (A and B), p62 (E and F), Beclin-1 (E and G), Bcl-2 (E and H), HDAC6 (E and I) and Lamp1 (E and J) were analyzed by western blot. Proteins (50 μg) from cell lysates were separated by SDS-PAGE and immunoblotted with the specific antibodies. Anti-α-tubulin or anti-GAPDH antibodies was applied as protein loading controls and used to normalize the levels of the analyzed proteins. Furthermore, the ratio between LC3-II levels measured in the presence and absence of NH4Cl, for each experimental condition was calculated in order to analyze the autophagic flux (C). The results were calculated relatively to control values and represent the means ± SEM of at least four independent experiments. *p b 0.05, **p b 0.01 and ***p b 0.001 significantly different from control. (D) In addition, the subcellular distribution of LC3B protein in RBE4 cells was followed by immunocytochemistry using a specific antibody against LC3B and Hoechst 33342 to label nuclei and the fluorescence images of control and treated cells were visualized using a confocal microscope. Selected fields in merged images (white squares) were amplified images.

1158

A.C. Fonseca et al. / Biochimica et Biophysica Acta 1843 (2014) 1150–1161

Fig. 5. Autophagy under moderate ER stress conditions. Proteins in cellular lysates from RBE4 cells treated with 25 or 50 nM thapsigargin during 6 h and co-treated with NH4Cl for 4 h were separated by SDS-PAGE and immunoblotted with antibodies against GRP78 (A and B) or LC3B (A and C). In order to analyze the autophagic flux, the ratio between LC3-II levels calculated in cells treated with 25 or 50 nM thapsigargin in the presence and absence of NH4Cl was determined (D). An anti-α-tubulin antibody was applied as a protein loading control and used to normalize protein levels of ER stress and macroautophagy markers. The results were calculated relatively to control values and represent the means ± SEM of at least eight independent experiments. **p b 0.01 and ***p b 0.001 significantly different from control.

the 26S proteasome was inhibited by Aβ1–40 and thapsigargin in ECs. These results are in agreement with results obtained by Nijholt and collaborators that found diminished activity of β2 and β5 subunits of proteasome in neuronal cells in the presence of the ER stress inducer tunicamycin [38]. This inhibition of proteasome can be due to the excessive accumulation of aggregated misfolded proteins resulting from ER stress that block the proteasome. Other authors reported that soluble Aβ oligomers inhibit the activity of the proteasome [39] and fibrillar Aβ1–40 was shown to impair proteasomal activity through the increase expression of the E2 ubiquitin-conjugating enzyme E2-25K/Hip2 leading to a proteolytic activation of ER-resident caspase-12 in the brains of Tg2576 and APPswe/PS1dE9 mice [40]. Furthermore, impairment in proteasome function but not expression of proteasome subunits has been described in the brain of AD patients [41,42]. Under conditions of proteasome inhibition, cytotoxic proteins such as tau and Aβ accumulate, promoting the formation of inclusion bodies [39,43]. In mouse neuroblastoma N2a cells, the accumulation of cytoplasmic prion aggregates inside the cells leads to ER stress, impairment of UPS and induction of autophagy and apoptosis [44], supporting a close relationship between impaired protein quality control mechanisms and activation of cell death. More recently, protein aggregates resulting from

proteasome dysfunction were shown to upregulate LC3-II in rat brain oligodendrocytes [45]. This increase in LC3-II, also observed in human ECs [15], can be due to an increase in macroautophagy induction or to an impairment of autophagosome degradation. Macroautophagy plays an important role in removing protein aggregates and organelles that are not degraded by the UPS [46]. Dysfunctional macroautophagy results in the accumulation of ubiquitinated and aggregated proteins and damaged organelles [47]. Autophagosomes were shown to accumulate in the brain of AD patients [47]. Our data suggest that, in the brain endothelium, this accumulation can be due to a decrease in autophagic flux induced by Aβ due to excessive or sustained ER stress or proteasome inhibition. A functional autophagy is needed for the clearance of large polyubiquitinated proteins [48]. A link between ubiquitinated proteins and the autophagic machinery to enable degradation in the lysosome is p62 which binds to polyubiquitinated proteins co-localizes with ubiquitinated protein aggregates in numerous neurodegenerative diseases and also binds to LC3 [49,50]. Recent evidences demonstrate that p62 has an important role in AD since it is involved in the degradation of tau [51]. p62 is used as an autophagic flux marker because it could accumulate when autophagic flux is inhibited [52]. However, it is also involved in proteasome-mediated degradation and accumulates when proteasome is inhibited [53], which makes the use of p62 as a general marker of autophagy a questionable issue. Accordingly, p62 protein levels were not significantly altered after incubation of RBE4 cells with Aβ1–40 or thapsigargin, but increased when proteasome activity was inhibited showing that proteasome is important for p62 degradation. Previously, we found that Aβ1–40 and thapsigargin increase ROS production in RBE4 cells (data not shown). Indeed, oxidative damage to the p62 promotor reduces p62 expression that is in accordance with the decreased levels of p62 in the brain of AD patients [54,55]. This decrease can compensate any increase in p62 protein levels due to the inhibition of autophagy flux induced by Aβ1–40 and thapsigargin. The increase in ROS production, as well as the rise in cytosolic Ca2+ concentration and UPR activation, which were demonstrated to be induced by Aβ1–40 in RBE4 cells [29], are able to boost macroautophagy in neurons and neuron-like cells [56,57]. However, the levels of Beclin-1 and Bcl-2 were not significantly altered by Aβ1–40, thapsigargin, or lactacystin suggesting that the activation step in the macroautophagic pathway is not affected or up-regulated by Aβ1–40, ER stress, or proteasome inhibition in this type of cells. When the formation of inclusion bodies overwhelms its clearance, the protein aggregates are transported to microtubule-organizing center (MTOC)-associated aggresomes. This process requires HDAC6 that [58,59] binds to dynein to transport polyubiquitinated misfolded proteins along microtubules to the aggresome that are then recognized by the autophagosome [58]. HDAC6 is also involved in lysosomal dynamics [26]. Thus, a decrease in HDAC6 protein levels can reduce autophagic flux. Nevertheless, its levels were not altered by Aβ1–40, thapsigargin, or lactacystin, and thus the observed reduction of autophagic flux with these conditions must be independent of HDAC6. Previously we showed that Aβ inhibits autophagic flow in a human neuroblastoma SH-SY5Y cell line [60]. These alterations were due to Aβ-induced mitochondria dysfunction, which causes microtubule disruption and consequent inhibition of the transport of autophagosomes and lysosomes. Consequently, autophagosome degradation was decreased. Interestingly, the use of the microtubule stabilizer taxol reduced mitochondrial and cytosolic Aβ levels [60] showing that Aβ can be degraded by macroautophagy. Since ER stress triggered by Aβ1–40 or thapsigargin induces mitochondria dysfunction through ER-to-mitochondria Ca2 + transfer [61–64], we can hypothesize that ER stress can cause microtubule disruption and subsequently affects autophagy flux. The 26S proteasome is involved in the formation and restructuration of microtubules [65]. Hence, the observed inhibition of proteasome by Aβ1–40 or thapsigargin or by lactacystin can impair the functioning of microtubules and consequently reduce the transport of AVs and lysosomes. On the other hand, dysfunctional microtubules

A.C. Fonseca et al. / Biochimica et Biophysica Acta 1843 (2014) 1150–1161

1159

Fig. 6. Apoptosis in brain endothelial cells. RBE4 cells were treated with 10 mM 3-MA, 2.5 μM Aβ1–40, 2 μM thapsigargin or 2 μM lactacystin for 12 h and caspase-2 (A), caspase-3 (B), caspase-9 (C) and caspase-12 (D) -like activities were determined in cell lysates (30 μg of protein) using the colorimetric substrates Ac-VDAVD-pNA, Ac-DEVD-pNA, Ac-LEHD-pNA and Ac-LEVD-pNA, respectively. Moreover, cell viability in ECs treated with 2.5 μM Aβ1–40, 2 μM thapsigargin or 2 μM lactacystin, with or without 0.1 μM rapamycin or 10 mM 3-MA, for 12 h was analyzed by measuring the capacity of the cells to reduce MTT (E). The presented results were normalized to control and represent the means ± SEM of at least three independent experiments performed in duplicate. *p b 0.05, **p b 0.01 and ***p b 0.001 significantly different from control, ##p b 0.01 and ###p b 0.001 significantly different from Aβ1–40 treatment, $$p b 0.01 significantly different from 3-MA treatment and &&p b 0.01 significantly different from thapsigargin treatment.

can also decrease the biogenesis of lysosomes [66]. Moreover, the activation of the UPR decreases general protein translation in order to reduce ER stress [67]. A strong or prolonged activation of the UPR can significantly diminish the amount of crucial proteins. Consequently, the function or the number of specific organelles, especially those dependent on the secretory pathway, can be diminished. In accordance, the treatment of RBE4 cells with thapsigargin or lactacystin reduced the levels of lysosomes, which impair the capacity of the cells to degrade autophagosomes, as demonstrated by the reduction in autophagic flux. Contrarily to severe ER stress induced by high thapsigargin levels that impaired macroautophagy, moderate ER stress induced by lower doses of thapsigargin was shown to stimulate macroautophagy without affecting the autophagic flux, possibly as a compensatory mechanism to degrade abnormal proteins, as demonstrated previously [68]. Under normal conditions, autophagy and UPS have an anti-apoptotic effect because the clearance of soluble and aggregated misfolded proteins is enhanced [57]. Since Aβ1–40 inhibited macroautophagy and the proteasome in RBE4 cells, damaged proteins and organelles resulting from Aβ1–40-induced ER stress were not repaired and accumulated, further enhancing ER stress. In addition, excessive ER stress triggers apoptosis through several pathways such as the elevation of ROS and cytosolic Ca2+ levels, activation of the ER-resident caspase-12, upregulation of the pro-apoptotic CHOP, release of cytochrome c from mitochondria

and consequent activation of caspase-9 and -3 as observed in the present study [29,69]. Moreover, caspase-2 was activated by Aβ 1–40 and lactacystin but not by thapsigargin. This caspase is localized at the Golgi complex and also promotes the translocation of Bax to mitochondria and consequent permeabilization of mitochondria and cytochrome c release upon genotoxic stress or JNK activation [70–72]. 3-MA, a well-known inhibitor of macroautophagy, increased Aβ levels inside RBE4 cells, induced ER stress, decreased proteasomal activity and robustly activated caspases leading to cell death by apoptosis. This was probably due to an increased accumulation of damaged proteins and organelles that were not efficiently degraded by macroautophagy. Aβ can be degraded by macroautophagy but this peptide can also be generated within endosomes and in autophagosomes when lysosomal degradation is compromised [34,60,73]. Despite the fact that Aβ 1–40, thapsigargin and lactacystin did not significantly induce macroautophagy, autophagosome accumulation due to a decrease in their turnover can potentiate the formation of Aβ. In addition, macroautophagy activation by 1 μM rapamycin was not able to revert the effects of Aβ1–40, thapsigargin and lactacystin (data not shown), but 0.1 μM rapamycin partially reverted the loss of viability induced by Aβ1–40 and thapsigargin. These findings demonstrate that severe induction of macroautophagy is toxic to ECs, probably through an excessive accumulation of autophagosomes that the cells cannot eliminate, but a slight/moderate activation of macroautophagy can compensate

1160

A.C. Fonseca et al. / Biochimica et Biophysica Acta 1843 (2014) 1150–1161

Fig. 7. Apoptosis triggered by Aβ1–40-through deregulation of proteostasis in brain endothelial cells. ER stress caused by Aβ leads to the accumulation of misfolded proteins that can be targeted for degradation in the proteasome or in the lysosome by macroautophagy when become aggregated. However, oligomeric and large protein aggregates block the proteasome and further induce ER stress. Under conditions of severe ER stress, general protein translation is inhibited and impairs ER functioning, compromising the biogenesis of organelles, such as the lysosome. As a consequence, the degradation of autophagosomes' cargo decreases and protein aggregates accumulate. Therefore, ER stress and proteasome blockage are exacerbated, leading to cell death by apoptosis. Similarly, the inhibition of autophagosome formation by 3-methyladenine (3-MA), the inhibition of the proteasome by lactacystin, and the induction of ER stress by thapsigargin increase the accumulation of aggregated proteins and consequently impair protein quality control and induce apoptosis.

the toxic effects arising from deregulation of proteostasis by degrading damaged proteins and organelles. 5. Conclusion Our findings support the idea that Aβ1–40 induces ER stress and impairs proteasomal activity in brain ECs. The inhibition of the proteasome in turn potentiates ER stress and the accumulation of abnormal proteins can aggregate and be targeted for degradation by macroautophagy within the lysosome. However, sustained and excessive ER stress increases the amount of dysfunctional abnormal proteins, reduces de novo protein synthesis that can impair the biogenesis of lysosomes, activates ER stress-dependent apoptotic pathways and also deregulates Ca2 + homeostasis. Due to the close crosstalk between the ER and mitochondria, this later organelle is also affected, and ATP production decreases, which may deplete vesicular transport, subsequently compromising autolysosome formation, and mitochondria-dependent apoptotic pathways are activated. As a result, the ECs die by apoptosis (Fig. 7), which can contribute to the dysfunction of the BBB found in AD. Disclosure statement The authors declare that this work was conducted without financial or commercial relationships that could be considered as a potential conflict of interest. Acknowledgements The authors are grateful to Dr. Jon Holy (University of Minnesota, Duluth, USA) for the generous gift of RBE4 cells, to Luísa Cortes for microscopy assistance, and to Daniela M Arduíno for the scientific discussion of some results. This work was supported by ‘Fundação para a Ciência e a Tecnologia’, Portugal (the project PEst-C/SAU/LA0001/ 2011, PhD fellowship attributed to AC Fonseca: SFRH/BD/47573/2008). References [1] D.J. Selkoe, D. Schenk, Alzheimer's disease: molecular understanding predicts amyloid-based therapeutics, Annu. Rev. Pharmacol. Toxicol. 43 (2003) 545–584. [2] J. Hardy, G. Higgins, Alzheimer's disease: the amyloid cascade hypothesis, Science 256 (1992) 184–185. [3] D.J. Selkoe, Alzheimer's disease: genes, proteins and therapy, Physiol. Rev. 81 (2001) 741–766. [4] B.V. Zlokovic, Neurovascular pathways to neurodegeneration in Alzheimer's disease and other disorders, Nat. Rev. Neurosci. 12 (2011) 723–738.

[5] A.P. Sagare, R.D. Bell, B.V. Zlokovic, Neurovascular dysfunction and faulty amyloid beta-peptide clearance in Alzheimer disease, Cold Spring Harb. Perspect. Med. 2 (2012). [6] A.M. Hartz, B. Bauer, E.L. Soldner, A. Wolf, S. Boy, R. Backhaus, I. Mihaljevic, U. Bogdahn, H.H. Klunemann, G. Schuierer, F. Schlachetzki, Amyloid-beta contributes to blood–brain barrier leakage in transgenic human amyloid precursor protein mice and in humans with cerebral amyloid angiopathy, Stroke 43 (2012) 514–523. [7] N.L. Marchant, B.R. Reed, N. Sanossian, C.M. Madison, S. Kriger, R. Dhada, W.J. Mack, C. DeCarli, M.W. Weiner, D.M. Mungas, H.C. Chui, W.J. Jagust, The aging brain and cognition: contribution of vascular injury and abeta to mild cognitive dysfunction, JAMA Neurol. 70 (2013) 488–495. [8] A. Dorr, B. Sahota, L.V. Chinta, M.E. Brown, A.Y. Lai, K. Ma, C.A. Hawkes, J. McLaurin, B. Stefanovic, Amyloid-beta-dependent compromise of microvascular structure and function in a model of Alzheimer's disease, Brain 135 (2012) 3039–3050. [9] M. Chisari, S. Merlo, M. Sortino, S. Salomone, Long-term incubation with betaamyloid peptides impairs endothelium-dependent vasodilatation in isolated rat basilar artery, Pharmacol. Res. 61 (2010) 157–161. [10] H. Du, P. Li, J. Wang, X. Qing, W. Li, The interaction of amyloid beta and the receptor for advanced glycation endproducts induces matrix metalloproteinase-2 expression in brain endothelial cells, Cell. Mol. Neurobiol. 32 (2012) 141–147. [11] E. Kuznetsova, R. Schliebs, Beta-amyloid, cholinergic transmission, and cerebrovascular system — a developmental study in a mouse model of Alzheimer's disease, Curr. Pharm. Des. 19 (2013) 6749–6765. [12] S.Y. Kook, H.S. Hong, M. Moon, C.M. Ha, S. Chang, I. Mook-Jung, Abeta1–42-RAGE interaction disrupts tight junctions of the blood–brain barrier via Ca2+-calcineurin signaling, J. Neurosci. 32 (2012) 8845–8854. [13] J. Rajadas, W. Sun, H. Li, M. Inayathullah, D. Cereghetti, A. Tan, V. de Mello Coelho, F.J. Chrest, J.W. Kusiak, W.W. Smith, D. Taub, J.C. Wu, J.M. Rifkind, Enhanced Abeta1–40 production in endothelial cells stimulated with fibrillar Abeta1–42, PLoS One 8 (2013) e58194. [14] S. Fossati, J. Ghiso, A. Rostagno, Insights into caspase-mediated apoptotic pathways induced by amyloid-beta in cerebral microvascular endothelial cells, Neurodegener. Dis. 10 (2012) 324–328. [15] S. Hayashi, N. Sato, A. Yamamoto, Y. Ikegame, S. Nakashima, T. Ogihara, R. Morishita, Alzheimer disease-associated peptide, amyloid beta40, inhibits vascular regeneration with induction of endothelial autophagy, Arterioscler. Thromb. Vasc. Biol. 29 (2009) 1909–1915. [16] P. Walter, D. Ron, The unfolded protein response: from stress pathway to homeostatic regulation, Science 334 (2011) 1081–1086. [17] A.M. Gorman, S.J. Healy, R. Jager, A. Samali, Stress management at the ER: regulators of ER stress-induced apoptosis, Pharmacol. Ther. 134 (2012) 306–316. [18] J.L. Goeckeler, J.L. Brodsky, Molecular chaperones and substrate ubiquitination control the efficiency of endoplasmic reticulum-associated degradation, Diabetes Obes Metab 12 (Suppl. 2) (2010) 32–38. [19] S. Deegan, S. Saveljeva, A.M. Gorman, A. Samali, Stress-induced self-cannibalism: on the regulation of autophagy by endoplasmic reticulum stress, Cell. Mol. Life Sci. 70 (2012) 2425–2441. [20] R. Jäger, M.J. Bertrand, A.M. Gorman, P. Vandenabeele, A. Samali, The unfolded protein response at the crossroads of cellular life and death during endoplasmic reticulum stress, Biol. Cell 104 (2012) 259–270. [21] J. Miao, F. Xu, J. Davis, I. Otte-Holler, M.M. Verbeek, W.E. Van Nostrand, Cerebral microvascular amyloid beta protein deposition induces vascular degeneration and neuroinflammation in transgenic mice expressing human vasculotropic mutant amyloid beta precursor protein, Am. J. Pathol. 167 (2005) 505–515. [22] P.O. Couraud, J. Greenwood, F. Roux, P. Adamson, Development and characterization of immortalized cerebral endothelial cell lines, Methods Mol. Med. 89 (2003) 349–364. [23] H.W. Klafki, J. Wiltfang, M. Staufenbiel, Electrophoretic separation of betaA4 peptides (1–40) and (1–42), Anal. Biochem. 237 (1996) 24–29.

A.C. Fonseca et al. / Biochimica et Biophysica Acta 1843 (2014) 1150–1161 [24] B. Martin-Clemente, B. Alvarez-Castelao, I. Mayo, A.B. Sierra, V. Diaz, M. Milan, I. Farinas, T. Gomez-Isla, I. Ferrer, J.G. Castano, Alpha-synuclein expression levels do not significantly affect proteasome function and expression in mice and stably transfected PC12 cell lines, J. Biol. Chem. 279 (2004) 52984–52990. [25] N. Mizushima, T. Yoshimori, How to interpret LC3 immunoblotting, Autophagy 3 (2007) 542–545. [26] A. Iwata, B.E. Riley, J.A. Johnston, R.R. Kopito, HDAC6 and microtubules are required for autophagic degradation of aggregated huntingtin, J. Biol. Chem. 280 (2005) 40282–40292. [27] F. Roux, P.O. Couraud, Rat brain endothelial cell lines for the study of blood–brain barrier permeability and transport functions, Cell. Mol. Neurobiol. 25 (2005) 41–58. [28] E. Ferreiro, C.R. Oliveira, C.M.F. Pereira, The release of calcium from the endoplasmic reticulum induced by amyloid-beta and prion peptides activates the mitochondrial apoptotic pathway, Neurobiol. Dis. 30 (2008) 331–342. [29] A.C.R.G. Fonseca, E. Ferreiro, C.R. Oliveira, S.M. Cardoso, C.F. Pereira, Activation of the endoplasmic reticulum stress response by the amyloid-beta 1–40 peptide in brain endothelial cells, Biochim. Biophys. Acta 1832 (2013) 2191–2203. [30] P. Candela, F. Gosselet, J. Saint-Pol, E. Sevin, M.C. Boucau, E. Boulanger, R. Cecchelli, L. Fenart, Apical-to-basolateral transport of amyloid-beta peptides through blood– brain barrier cells is mediated by the receptor for advanced glycation end-products and is restricted by P-glycoprotein, J. Alzheimers Dis. 22 (2010) 849–859. [31] T. Pflanzner, M.C. Janko, B. Andre-Dohmen, S. Reuss, S. Weggen, A.J. Roebroek, C.R. Kuhlmann, C.U. Pietrzik, LRP1 mediates bidirectional transcytosis of amyloid-beta across the blood–brain barrier, Neurobiol. Aging 32 (2011) (2323 e2321-2323 e2311). [32] Z. Cai, L.J. Yan, Rapamycin, autophagy, and Alzheimer's disease, J. Biochem. Pharmacol. Res. 1 (2013) 84–90. [33] Y. Urano, S. Ochiai, N. Noguchi, Suppression of amyloid-beta production by 24S-hydroxycholesterol via inhibition of intracellular amyloid precursor protein trafficking, FASEB J. 27 (2013) 4305–4315. [34] A.C.R.G. Fonseca, R. Resende, C.R. Oliveira, C.M.F. Pereira, Cholesterol and statins in Alzheimer's disease: current controversies, Exp. Neurol. 223 (2010) 282–293. [35] X. Sai, Y. Kawamura, K. Kokame, H. Yamaguchi, H. Shiraishi, R. Suzuki, T. Suzuki, M. Kawaichi, T. Miyata, T. Kitamura, B. De Strooper, K. Yanagisawa, H. Komano, Endoplasmic reticulum stress-inducible protein, Herp, enhances presenilin-mediated generation of amyloid beta-protein, J. Biol. Chem. 277 (2002) 12915–12920. [36] A. Nogalska, W.K. Engel, V. Askanas, Increased BACE1 mRNA and noncoding BACE1-antisense transcript in sporadic inclusion-body myositis muscle fibers—possibly caused by endoplasmic reticulum stress, Neurosci. Lett. 474 (2010) 140–143. [37] J.H. Walton, R.S. Berry, F. Despa, Amyloid oligomer formation probed by water proton magnetic resonance spectroscopy, Biophys. J. 100 (2011) 2302–2308. [38] D.A. Nijholt, T.R. de Graaf, E.S. van Haastert, A.O. Oliveira, C.R. Berkers, R. Zwart, H. Ovaa, F. Baas, J.J. Hoozemans, W. Scheper, Endoplasmic reticulum stress activates autophagy but not the proteasome in neuronal cells: implications for Alzheimer's disease, Cell Death Differ. 18 (2011) 1071–1081. [39] B.P. Tseng, K.N. Green, J.L. Chan, M. Blurton-Jones, F.M. LaFerla, Abeta inhibits the proteasome and enhances amyloid and tau accumulation, Neurobiol. Aging 29 (2008) 1607–1618. [40] S. Song, H. Lee, T. Kam, M. Tai, J. Lee, J. Noh, S. Shim, S. Seo, Y. Kong, T. Nakagawa, C. Chung, D. Choi, H. Oubrahim, Y. Jung, E2-25K/Hip-2 regulates caspase-12 in ER stress-mediated Abeta neurotoxicity, J. Cell Biol. 182 (2008) 675–684. [41] V. Cecarini, Q. Ding, J.N. Keller, Oxidative inactivation of the proteasome in Alzheimer's disease, Free Radic. Res. 41 (2007) 673–680. [42] J.N. Keller, K.B. Hanni, W.R. Markesbery, Impaired proteasome function in Alzheimer's disease, J. Neurochem. 75 (2000) 436–439. [43] E.J. Bennett, N.F. Bence, R. Jayakumar, R.R. Kopito, Global impairment of the ubiquitin– proteasome system by nuclear or cytoplasmic protein aggregates precedes inclusion body formation, Mol. Cell 17 (2005) 351–365. [44] R. Chen, R. Jin, L. Wu, X. Ye, Y. Yang, K. Luo, W. Wang, D. Wu, X. Ye, L. Huang, T. Huang, G. Xiao, Reticulon 3 attenuates the clearance of cytosolic prion aggregates via inhibiting autophagy, Autophagy 7 (2011) 205–216. [45] L. Schwarz, O. Goldbaum, M. Bergmann, S. Probst-Cousin, C. Richter-Landsberg, Involvement of macroautophagy in multiple system atrophy and protein aggregate formation in oligodendrocytes, J. Mol. Neurosci. 47 (2012) 256–266. [46] B. Boland, A. Kumar, S. Lee, F.M. Platt, J. Wegiel, W.H. Yu, R.A. Nixon, Autophagy induction and autophagosome clearance in neurons: relationship to autophagic pathology in Alzheimer's disease, J. Neurosci. 28 (2008) 6926–6937. [47] B. Levine, G. Kroemer, Autophagy in the pathogenesis of disease, Cell 132 (2008) 27–42. [48] M. Komatsu, S. Waguri, T. Ueno, J. Iwata, S. Murata, I. Tanida, J. Ezaki, N. Mizushima, Y. Ohsumi, Y. Uchiyama, E. Kominami, K. Tanaka, T. Chiba, Impairment of starvation-induced and constitutive autophagy in Atg7-deficient mice, J. Cell Biol. 169 (2005) 425–434.

1161

[49] S. Pankiv, T.H. Clausen, T. Lamark, A. Brech, J.A. Bruun, H. Outzen, A. Overvatn, G. Bjorkoy, T. Johansen, p62/SQSTM1 binds directly to Atg8/LC3 to facilitate degradation of ubiquitinated protein aggregates by autophagy, J. Biol. Chem. 282 (2007) 24131–24145. [50] G. Bjorkoy, T. Lamark, S. Pankiv, A. Overvatn, A. Brech, T. Johansen, Monitoring autophagic degradation of p62/SQSTM1, Methods Enzymol. 452 (2009) 181–197. [51] A. Salminen, K. Kaarniranta, A. Haapasalo, M. Hiltunen, H. Soininen, I. Alafuzoff, Emerging role of p62/sequestosome-1 in the pathogenesis of Alzheimer's disease, Prog. Neurobiol. 96 (2012) 87–95. [52] M. Komatsu, Q.J. Wang, G.R. Holstein, V.L. Friedrich Jr., J. Iwata, E. Kominami, B.T. Chait, K. Tanaka, Z. Yue, Essential role for autophagy protein Atg7 in the maintenance of axonal homeostasis and the prevention of axonal degeneration, Proc. Natl. Acad. Sci. U. S. A. 104 (2007) 14489–14494. [53] I.B. Wilde, M. Brack, J.M. Winget, T. Mayor, Proteomic characterization of aggregating proteins after the inhibition of the ubiquitin proteasome system, J. Proteome Res. 10 (2011) 1062–1072. [54] Y. Du, M.C. Wooten, M. Gearing, M.W. Wooten, Age-associated oxidative damage to the p62 promoter: implications for Alzheimer disease, Free Radic. Biol. Med. 46 (2009) 492–501. [55] Y. Du, M.C. Wooten, M.W. Wooten, Oxidative damage to the promoter region of SQSTM1/p62 is common to neurodegenerative disease, Neurobiol. Dis. 35 (2009) 302–310. [56] R. Scherz-Shouval, E. Shvets, E. Fass, H. Shorer, L. Gil, Z. Elazar, Reactive oxygen species are essential for autophagy and specifically regulate the activity of Atg4, EMBO J. 26 (2007) 1749–1760. [57] K.M. Doyle, D. Kennedy, A.M. Gorman, S. Gupta, S.J. Healy, A. Samali, Unfolded proteins and endoplasmic reticulum stress in neurodegenerative disorders, J. Cell. Mol. Med. 15 (2011) 2025–2039. [58] Y. Kawaguchi, J.J. Kovacs, A. McLaurin, J.M. Vance, A. Ito, T.P. Yao, The deacetylase HDAC6 regulates aggresome formation and cell viability in response to misfolded protein stress, Cell 115 (2003) 727–738. [59] H. Ouyang, Y.O. Ali, M. Ravichandran, A. Dong, W. Qiu, F. MacKenzie, S. Dhe-Paganon, C.H. Arrowsmith, R.G. Zhai, Protein aggregates are recruited to aggresome by histone deacetylase 6 via unanchored ubiquitin C termini, J. Biol. Chem. 287 (2012) 2317–2327. [60] D.F. Silva, A.R. Esteves, D.M. Arduino, C.R. Oliveira, S.M. Cardoso, Amyloidbeta-induced mitochondrial dysfunction impairs the autophagic lysosomal pathway in a tubulin dependent pathway, J. Alzheimers Dis. 26 (2011) 565–581. [61] C. Giorgi, D. De Stefani, A. Bononi, R. Rizzuto, P. Pinton, Structural and functional link between the mitochondrial network and the endoplasmic reticulum, Int. J. Biochem. Cell Biol. 41 (2009) 1817–1827. [62] P. Pizzo, T. Pozzan, Mitochondria-endoplasmic reticulum choreography: structure and signaling dynamics, Trends Cell Biol. 17 (2007) 511–517. [63] R.O. Costa, E. Ferreiro, I. Martins, I. Santana, S.M. Cardoso, C.R. Oliveira, C.M. Pereira, Amyloid beta-induced ER stress is enhanced under mitochondrial dysfunction conditions, Neurobiol. Aging 33 (2012) (824 e825-816). [64] R.O. Costa, E. Ferreiro, S.M. Cardoso, C.R. Oliveira, C.M. Pereira, ER stress-mediated apoptotic pathway induced by Abeta peptide requires the presence of functional mitochondria, J. Alzheimers Dis. 20 (2010) 625–636. [65] J. Kurepa, S. Wang, J. Smalle, The role of 26S proteasome-dependent proteolysis in the formation and restructuring of microtubule networks, Plant Signal. Behav. 7 (2012) 1289–1295. [66] J. Lacombe, G. Karsenty, M. Ferron, Regulation of lysosome biogenesis and functions in osteoclasts, Cell Cycle 12 (2013). [67] J.J. Hoozemans, W. Scheper, Endoplasmic reticulum: the unfolded protein response is tangled in neurodegeneration, Int. J. Biochem. Cell Biol. 44 (2012) 1295–1298. [68] M. Ogata, S. Hino, A. Saito, K. Morikawa, S. Kondo, S. Kanemoto, T. Murakami, M. Taniguchi, I. Tanii, K. Yoshinaga, S. Shiosaka, J.A. Hammarback, F. Urano, K. Imaizumi, Autophagy is activated for cell survival after endoplasmic reticulum stress, Mol. Cell. Biol. 26 (2006) 9220–9231. [69] E. Lai, T. Teodoro, A. Volchuk, Endoplasmic reticulum stress: signaling the unfolded protein response, Physiology (Bethesda) 22 (2007) 193–201. [70] H.H. Park, Structural features of caspase-activating complexes, Int. J. Mol. Sci. 13 (2012) 4807–4818. [71] B. Zhivotovsky, S. Orrenius, Caspase-2 function in response to DNA damage, Biochem. Biophys. Res. Commun. 331 (2005) 859–867. [72] M. Enoksson, J.D. Robertson, V. Gogvadze, P. Bu, A. Kropotov, B. Zhivotovsky, S. Orrenius, Caspase-2 permeabilizes the outer mitochondrial membrane and disrupts the binding of cytochrome c to anionic phospholipids, J. Biol. Chem. 279 (2004) 49575–49578. [73] S.M. Son, H. Song, J. Byun, K.S. Park, H.C. Jang, Y.J. Park, I. Mook-Jung, Accumulation of autophagosomes contributes to enhanced amyloidogenic APP processing under insulin-resistant conditions, Autophagy 8 (2012) 1842–1844.