Mechanisms of amyloid-β34 generation indicate a pivotal role for BACE1 in amyloid homeostasis

Plasmids and siRNAs

A human BACE1 construct (full length BACE1, isoform A; pcDNA3.1+/Zeo; Invitrogen), APP695 (with an N-terminal Myc-tag; pcDNA3.1+/Zeo; Invitrogen) and APP-C99 (with a C-terminal FLAG-tag; pcDNA3.1+/Zeo; Invitrogen) were used for transient overexpression in HEK293T cells. Point mutations were introduced by site-directed mutagenesis, using PfuUltra II Fusion HS (Stratagene/Agilent) followed by DpnI (NEB) digestion. All constructs were verified by DNA sequencing. For creating stably expressing SH-SY5Y cells, full-length human BACE1 (isoform A) and full-length human APP (isoform APP695, with an N-terminal Myc tag), in the mammalian expression vector pCEP4, Hygro (Invitrogen) were used. Mouse wild-type BACE1 construct with an N-terminal FLAG tag immediately following the propeptide cleavage was generated by overlap extension PCR and cloned in pSport6. LL/AA or D495R variants were then generated by PCR using reverse primers with the mutant sequence. Mock controls for corresponding plasmid backbones were used. For knockdown, siGENOME non-Targeting siRNA Pool #1 (D-001206-13-05), SMARTpool siGENOME Presenilin 1 (M-004998), and Presenilin 2 (M-006018) were used.

Human brain samples

Post-mortem samples were collected from donors with a written informed consent for a brain autopsy and the material, and the use of the material for research purposes and the protocols were approved by the Netherlands Brain Bank by the Implementing Letter regarding project nr. 836 “Diagnostic Potential of AP-clearance Intermediates in Elderly Subjects at Risk for Alzheimer’s Disease”, approved by the Director Netherlands Brain Bank, 26 Aug 2014, Royal Netherlands Academy of Arts and Sciences (KNAVV) acting for and on behalf of The Netherlands Brain Bank (NBB), a department of the Netherlands Institute for Neuroscience (NIN), whose registered office is at Meibergdreef 47, 1105 BA Amsterdam, The Netherlands.

Frozen samples from the temporal cortex from non-demented controls (n = 5) and confirmed Alzheimer disease Braak 4 to Braak 6 (n = 20) were prepared as previously described⁶⁵. In brief, brain samples were thawed on ice, weighed and homogenized in buffer A (100 mM Tris–HCl, 150 mM NaCl, 2 × complete protease inhibitor cocktail (Roche)) using gentleMACS™ M Tubes/Dissociator at 4 °C (Miltenyi Biotech). TritonX-100, final concentration 1%, was added and samples were incubated for 1 h with agitation at 4 °C. Lysates were centrifuged at 10,621 rcf in a microfuge (Eppendorf) at 4 °C for 15 min to remove the nuclear fraction. Samples were measured with bicinchoninic acid assay (BCA assay, Thermo Fisher Scientific Inc., Pierce) and MSD assays.

Mouse brain lysates

We complied with all relevant ethical regulations for animal tissue testing and research. Details on ethics approvals for animal studies are available from our co-authors of the laboratories in Germany that provided the material. Briefly, London APP Transgenic mice and wild-type littermates: All animal studies were conducted in compliance with European and German guidelines for the care and use of laboratory animals and were approved by the Central Animal Facility of the University of Mainz and the ethical committee on animal care and use of Rhineland–Palatinate, Germany.

BACE1 +/+, BACE +/− and BACE1 −/− mice: In agreement with the German animal welfare law all animal handling and care were performed according to the guidelines of the Christian-Albrechts-University of Kiel. The Ministry of Energy, Agri-culture, the Environment and Rural Areas Schleswig–Holstein approved animal experiments under the reference number V242–40,536/2016 (81–6/16).

Cortices of transgenic mice expressing London APP and their wild-type littermates were provided by Dr. Claus Pietrzik’s laboratory at the University of Mainz, Germany. Cortices of BACE1 +/+, BACE +/− and BACE1 −/− mice were provided by Dr. Paul Saftig’s laboratory in University of Kiel, Germany. All mice were on C57BL/6 strain genetic background and were 6-months of age when sacrificed. Frozen mouse brains were thawed on ice, weighed, and homogenized in the homogenization buffer (100 mM Tris–HCl pH: 7.5, 150 mM NaCl and 2 × complete protease inhibitor cocktail (Roche)) using Dounce homogenizer. 10% Triton-X was added to the homogenates (final concentration: 1%). Brain homogenates were lysed at 4 °C for 1 h on a rotator. Lysates were centrifuged at 10,621 rcf in a microfuge (Eppendorf) at 4 °C for an hour to remove the debris. Supernatants were collected and diluted in the appropriate buffers for BCA, Western blot and MSD assays.

Cell culture and transfection

Human Embryonic Kidney (HEK293T) cells (DSMZ No. ACC305; DSMZ, Braunschweig, Germany) cells were grown in DMEM (High glucose (4.5 g/l), 10% fetal bovine serum (FBS), 2 mM glutamine, 1 mM pyruvate) in a humidified incubator at 37 °C 5% CO₂. For transient transfection experiments, cells were seeded on 6-well plates (Fisher) coated with poly-D-lysine (Sigma) and transiently transfected (with the plasmids indicated for each experiment) 20–24 h later by using TransFectin according to the protocol provided by the manufacturer (Biorad). 24 h after transfection, media of the cells were changed, and cells were conditioned for 16 h before sample collection.

Human neuroblastoma (SH-SY5Y) cells (DSMZ No. ACC209; DSMZ, Braunschweig, Germany) stably overexpressing BACE1, APP or APP-C99 were cultured in DMEM/F12 (10% fetal bovine serum (FBS), 2 mM L-glutamine, 1 mM sodium pyruvate) in a humidified incubator at 37 °C 5% CO₂. Stable cell lines were selected with 250 µg/ml Hygromycin B (Milipore). For BACE1 localization, cells were seeded on 6-well plates (Fisher) and transfected with FuGENE HD (Promega) after 24 h. 72 h after the transfection, cells were harvested. For PS1 and/or PS2 knockdown experiments, cells were seeded on 6-well plates (Fisher) and treated with either control, PS1 or PS2 siRNA (concentration of the siRNA depended on the experiment) 24 h later by using RNAiMax according to the protocol provided by the manufacturer (Invitrogen). 72 h after the treatment, cells were harvested. For ICC experiments, cells were seeded on 24-well plates (Fisher) and the same protocols were applied.

Sample preparation

For all experiments performed, cells were harvested on ice. Conditioned media were centrifuged at 2000 rpm at 4 °C for 10 min and Aβ34, Aβ40 and Aβ42 levels were quantified by ELISA. Cells were washed with cold PBS and lysed with Whole Cell Extract Buffer (25 mM HEPES (pH 7.7), 0.3 M NaCl, 1.5 mM MgCl₂, 0.2 mM ethylenediaminetetraacetic acid, 0.1% Triton-X-100, 0.5 mM dithiothreitol, 4 mM NaF, 0.1 mM Na₃VO₄, 1 mM PMSF, Complete Protease Inhibitor Cocktail (Roche)) at 4 °C for 60 min. Cell lysates were cleared from nuclear material by centrifugation at 10,000 rpm at 4 °C for 15 min and protein levels were detected by Western Blot.

Western blot analysis

Samples were prepared by adding LDS loading buffer and 2-Mercaptoethanol to the cell lysates according to the protocol provided by the manufacturer (Invitrogen). The proteins were solubilized and denatured by heating the samples to 70 °C for 10 min. Proteins were separated on 4–12% Bis–Tris gradient gels (Invitrogen) and were transferred to 0.45 µm nitrocellulose (Biorad) or polyvinylidene difluoride (PVDF) (Millipore) membranes at 400 mA at 4 °C for 2.5 h. Proteins were detected by the antibodies indicated in the antibodies section. The primary and secondary antibodies were used in phosphate-buffered saline. Signals were recorded on ImageQuant LAS 500 and LAS 600 (GE Healthcare Life Sciences).

The primary antibodies used for Western Blot analysis were the following: anti-BACE1 1:2,000 dilution (monoclonal D10E5, Cell Signaling), anti-BACE1 1:2,000 dilution (B0681, Sigma-Aldrich), anti-actin 1:5,000 dilution (monoclonal mab1501, Millipore), anti-sAPPβ 1:2,000 dilution (IBL), and anti-APP ectodomain 22C11 1:10,000 dilution (Millipore), anti-flag 1:1,000 dilution (M2, F1804, Sigma-Aldrich), anti-PS2 (ab51249, Abcam), and anti-PS1 1:10,000 dilution (ab76083, Abcam).

The secondary antibodies used for Western Blot analysis were the following: anti-mouse- and anti-rabbit-horseradish peroxidase 1:10,000 dilution (Promega).

Quantification of the Western Blots were performed with ImageJ and all protein levels were normalized to actin.

All gels and blots used in figures are in compliance with the digital image and integrity policies (https://www.nature.com/srep/journal-policies/editorial-policies#digital-image).

Where cropped gels/blots are displayed, respective full-length gels and blots are included in a Supplementary Information file.

Meso scale discovery (MSD) assay

Custom-printed 4-plex plates were used as described previously²⁰. Plates were blocked with 150 µl 5% MSD Blocker A solution for an hour at room temperature with gentle shaking and washed 3 times with 250 µl PBS-T (0.05% tween). Peptide calibrators were diluted in MSD Diluent 35. Plates were loaded with samples and calibrators together with SULFO-TAG™ 4G8 or 6E10 detection antibody diluted in MSD Diluent 100 and incubated overnight at 4 °C with gentle shaking. After three washes with 250 µl PBS-T, 150 µl 2 × MSD read buffer was added to the wells. Plates were read by an MSD QuickPlex SQ 120 Imager and data were analyzed by MSD Workbench® software.

Sandwich-based Enzyme-Linked Immunosorbent Assay (ELISA)

5 μg/ml monoclonal anti-Aβ34 (226), anti-Aβ40 (G2-10) or anti-Aβ42 (G2-13) capture antibody in 100 mM sodium carbonate (pH 9.6) were used to coat the 96-well Nunc™ plates (Thermo Fisher Scientific Inc.). The sealed plates were incubated overnight at 4 °C with gentle shaking. Plates were washed 5 times 10 min with PBS-T washing buffer (1.1 mM NaH₂PO₄, 8.5 mM Na₂HPO₄, 13.7 mM NaCl, (pH 7.4), 0.1% Tween-20). 250 μl Stabil Coat®Immunoassay Stabilizer (SurModics Inc.) was used for blocking and plates were incubated for 2 h at room temperature with gentle shaking. 50 μl of 0.075 μg/ml detection antibody, W02-biotin, in assay buffer (90% 11 mM NaH₂PO₄, 85 mM Na₂HPO₄, 137 mM NaCl, (pH 7.4), 0.5% Tween-20, 1.5% BSA, 0.01% Thimerosal, and 10% SeaBlock blocking buffer (Thermo Fisher Scientific Inc.) was loaded to the wells together with 50 μl sample (cell media) or calibrator (synthetic peptide standards diluted in DMEM or DMEM/F12). After overnight incubation at 4 °C with gentle shaking, plates were washed 5 times for 10 min with PBS-T washing buffer. For Aβ40 ELISA, 100 μl Mono-HRP-conjugated-streptavidin (Pierce) (0.1 μg/ml) in Mono-HRP buffer (11 mM NaH₂PO₄, 85 mM Na₂HPO₄, 137 mM NaCl, (pH 7.4), 0.05% Tween-20, 6% PEG) or for Aβ34 and Aβ42 ELISA (for higher sensitivity), 100 μl Poly-HRP-conjugated-streptavidin (Pierce) (1:20,000 dilution) in Poly-HRP buffer (1.1 mM NaH₂PO₄, 8.5 mM Na₂HPO₄, 13.7 mM NaCl, (pH 7.4), 0.1% Tween-20, 5% BSA)) was added to the wells. Plates were incubated for 1 h at room temperature with gentle shaking and washed 5 times for 10 min with PBS-T washing buffer. For the initiation of enzymatic reaction, 100 μl 1-Step™ Ultra TMB-ELISA Substrate (Thermo Fisher Scientific Inc.) solution was added to the wells and the plates were incubated at room temperature in the dark for up to 30 min. To stop the reaction, 50 μl 1 M H₂SO₄, per well, was added. Using Synergy H1, BioTek Instruments Inc. plate reader, absorbance at 450 nm and 630 nm as a reference was measured. The data analysis was performed with Gen5 BioTek®software. For the fitting of standard curves obtained from the absorbance of calibrators, a non-linear four-parameter logistic fit without weighting was used as follows

where y is signal, x is concentration, b₂ is estimated response at the infinite concentration, b₁ is estimated response at zero concentration, b₃ is mid-range concentration and b₄ is slope factor.

Immunocytochemistry

For all immunofluorescence experiments, 12 mm coverslips were used (Fisherbrand™ Catalog# 12CIR1602811G). SH-SY5Y cell lines were fixed with 4% formaldehyde in phosphate buffered saline. Cells were then permeabilized with 1% Triton X-100 for 10 min and blocked immediately for 30 min with 2% bovine serum albumin (BSA) in phosphate buffered saline solution. After blocking, coverslips were incubated with the primary antibody overnight at 4 °C. The following day, the primary antibody was washed off, the coverslips were washed 3 times in 2% BSA buffer and were then incubated with secondary antibody for 30 min. After incubation, coverslips were washed in PBS and nuclei were stained with NucBlue (ThermoFisher catalog #R37606). In order to visualize the actin network some cells were stained with Phalloidin for 20 min according to the manufacturer’s instructions (ThermoFisher catalog #A22287). Coverslips were then mounted onto microscope slides using the Aqua-Poly/Mount media (Polysciences Catalog #18,606–20).

The primary antibodies used for immunofluorescence were the following: anti-PS1 antibody dilution 1:100 (Invitrogen, MA1-752), anti-PS2 antibody 1:50 dilution (Abcam ab15549), anti-EEA1 antibody 1:200 dilution (Cell Signaling #3288), anti-LAMP1 antibody 1:200 dilution (Cell Signaling #9091), and anti-BACE1 C-term antibody 1:100 dilution (Millipore MAB5308).

The secondary antibodies were acquired from Life Technologies: goat anti-rabbit IgG cross adsorbed Alexa Fluor 647 (ThermoFisher catalog #A-21245; diluted 1:10,000), goat anti-mouse IgG cross absorbed Alexa Fluor 568 (ThermoFisher catalog #A-11031) or goat anti-mouse IgG cross absorbed Alexa Fluor 488 (ThermoFisher catalog #A-11001).

Confocal microscopy and image analysis

Single- or double-immunolabeled (Alexa Fluor-488, -568 or -647) samples were analyzed at the Imaging & Molecular Biology Platform (IMBP; McGill Life Sciences Complex) using a TCS SP8 multi-photon confocal microscope (Leica) with 63x/1.40 oil-immersion objectives (Leica, Wetzlar, Germany). Samples were excited with Coherent Chameleon Vision II multiphoton at 730 nm (2660 mW) for DAPI imaging. For each sample, 12–30 z-stack images were acquired using the same laser intensity settings for quantification. Z-stack images were processed using Image-J (Rasband, W.S., ImageJ, U.S. National Institutes of Health, Bethesda, Maryland, USA, https://imagej.nih.gov/ij/, 1997–2018) and total cell fluorescence was quantified with the analyze tool. To better visualize BACE1 localization, a heatmap was generated using Fire LUT in Image-J. The IMARIS Image Analysis Software (Bitplane (Oxford Instruments), MA, USA) software was used for cross-sectional analysis. BACE1 colocalization with EEA1 and LAMP1 were analyzed using ImageJ plugin JACoP⁶⁶.

Matrix-assisted laser absorption ionization mass spectrometry (MALDI-MS)

Samples were first immunoprecipitated (IP). For each IP (4 °C, 18 h), 0.5 mL of conditioned cell culture supernatant was combined with 5 µg W02 (anti-Ab antibody) and 25 µL protein-G sepharose beads (GE Healthcare) in PBS (1 mL final volume). The samples were sequentially washed with PBS, followed by 10 mM Tris pH 7.5; 150 mM NaCl; 0.2% NP-40; 2 mM EDTA, followed by 10 mM Tris pH 7.5; 500 mM NaCl; 0.2% NP-40; 2 mM EDTA, followed by three-times PBS and finally three-times 100 mM ammonium acetate (pH 7.4). The IPs were eluted twice using 350 µl volumes of 50% acetic acid. The vacuum-dried samples were resuspended in 10 µl of TA30 (33% acetonitrile and ultrasonicated. Samples were mixed 1:1 with α-cyanocinnamic acid matrix (CCA, Bruker Daltonics; 20 mg/mL in TA30) and applied to ground steel MALDI targets using the dried droplet method. Mass spectra were recorded on an UltrafleXtreme MALDI-TOF/TOF system (Bruker Daltonics) using the reflector positive 900–4500 method (ion source 1 = 25 kV; ion source 2 = 22.30 kV; lens = 9.40 kV; reflector = 26.45 kV; reflector 2 = 13.40 kV; pulsed ion extraction = 150 ns) and flexControl v1.4 and flexAnalysis v1.4 software. Ion intensity was evaluated by averaging four measurements of 500 shots each (i.e., 2000 shots total per sample).

Statistical analysis

For all experiments, different conditions were analyzed by one factor ANOVA (between subject design) or two factor ANOVA. Pairwise comparisons were performed either with Dunnet’s or Tukey’s post-hoc tests. The statistical analysis was run by GraphPad Prism 5. For human brain samples, Welch’s t-tests were performed. Data that was not normally distributed (Gaussian normality test) were analyzed with non-parametric tests. Spearman’s correlation was performed to test the correlation between BACE1 and Aβ34 levels.

Ethics approval and consent to participate

Prior to starting the study, ethical approvals have been obtained. The study was conducted in accordance with Helsinki Declaration as revised in 2013 and performed in accordance with respective guidelines. The protocols were approved by The Netherlands Brain Bank (NBB) where the brain post-mortem samples were obtained from, i.e. The Netherlands Institute for Neuroscience, Amsterdam (open access: www.brainbank.nl). All material has been collected from donors having provided written informed consent for a brain autopsy and the use of the material and for whom. Clinical information for research purposes had been obtained by the NBB.