Cell culture and EVs isolation
BV2 cells, an immortalized murine microglia cell line [26], were continuously maintained, from an original gift of prof. Giulio Levi (Istituto Superiore di Sanità, Rome), at the Dipartimento di Scienze Anatomiche, Istologiche, Medico-Legali e dell’Apparato Locomotore. They were grown in Dulbecco's Modified Eagle Medium (DMEM) supplemented with 10% Fetal Bovine Serum (FBS), 100 U/mL penicillin, 100 μL/mL streptomycin, and 2 mM glutamine at 37°C in 5% CO2/95% humidified air atmosphere. Then they were washed thrice in PBS and posed in a serum free medium to avoid contamination with bovine exosomes.
BV2 cells were left untreated (NT) or treated (Aβ) with the 25–35 fragment (Aβ25–35) of the full lenght Aβ amyloid peptide Aβ1–42 Prior to use, the synthetic form of Aβ25–35 (Bachem) was dissolved to a final concentration of 1 mM in hexafluoroisopropanol (HFIP; Sigma-Aldrich) for its monomerization. HFIP was then removed by evaporation under vacuum, and peptide was solubilised in DMSO for 10 min. After dilution to 100 μM in DMEM, Aβ25–35 was incubated at 4°C for 24 h for polymerization [27]. Then, it was added to the Aβ sample at a final concentration of 25 μM. After 24 h, cell medium was collected from NT and Aβ cells and protease inhibitors were added.
To assess the inflammatory cell state induced by amyloid treatment, nitrite (NO2-) levels were determined by using Griess reagent (1mM sulfanilamide, 1mM naphthylenediamine dihydrochloride and 100 mM HCl) in culture supernatants of NT and Aβ samples at 24 h. Absorbance was measured at 540 nm, and NO2- concentration was determined using sodium nitrite as a standard. Escherichia coli lipopolysaccharide (LPS) (serotype 0127:B8; 0.1 mg/ml) was used as positive control.
Cell viability was assessed by measuring lactate dehydrogenase (LDH) released in the culture medium using a cytotoxicity detection kit (Roche, Mannheim, Germany) according to the manufacturer's protocols. Statistical analyses were conducted using GraphPad Prism version 4.00 software. Data are expressed as means ± SEM. Comparisons were analysed using t-test.
For each biological replicate, a total of 24 ml cell medium both for NT and Aβ samples, was centrifugated at 300g for 10 min and at 3000g for 30 min to eliminate debris and dead cells. The supernatant was filtered on a 0.22 mm filter to remove EVs over 220 nm size. The filtrate was ultra-centrifuged at 100,000g for 1 h; the pellet was resuspended in 24 ml of Phosphate Buffered Saline pH 7.4 (PBS) and ultra-centrifuged again at 100,000g. The pellet was finally resuspended in 100 ml of modified Laemmli buffer containing 8% Sodium Dodecyl Sulfate (SDS) and 1% n-dodecyl-β-D-maltoside.
Proteomic Analyses
Protein quantification of cell lysates from NT and Aβ samples was assayed by Protein Assay Dye Reagent (Bio-Rad). To collect similar amount of EV lysates for the following analytical procedures, aliquots were taken relying on protein content as quantified in cell lysates. These aliquots were fractionated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) on 4-20% Mini-PROTEAN TGX™ gel (Bio-Rad) and stained using a colloidal Coomassie staining.
From each SDS-PAGE lane, twelve slices were excised, and each of them submitted to a carbamidomethylation treatment and trypsin proteolysis according to Brisdelli et al. [28]. Desalting steps were carried out by solid phase extraction (SPE) according to Rappsilber et al. [29]. C18 reverse-phase loaded Empore™ SPE disks were purchased from Sigma-Aldrich. Peptide mixtures were dried, re-suspended in 50 μL of 0.1% TFA, and analysed by nano-liquid chromatography tandem mass spectrometry (nano LC-MS/MS). For this purpose, an Ultimate3000 system (Dionex, Sunnyvale, CA, USA) was equipped with a splitting cartridge for nanoflows and connected on-line via a nano-electrospray ion source (Thermo-Fisher Scientific, Waltham, Massachussetts, USA) to an LTQ Orbitrap XL mass spectrometer (Thermo-Fisher Scientific).
Each sample was automatically loaded from the autosampler module of the Ultimate 3000 system at a flow rate of 20 μL/min onto a trap column (Acclaim®PepMap™ μ-Precolumn, 300 μM x 1 mm, Dionex) in 4% acetonitrile containing 0.1% formic acid. After 4 min, peptides were eluted at 300 nL/min onto a 15 cm column (360 μM OD x 75 μM ID, 15 μM Tip ID; PicoFrit®, New Objective, Woburn, MA, USA), custom packed with a reverse phase (C18, 5 μM particle size, 200 Å pore size; Magic C18 AQ, Michrom), by a two-step gradient of solvent B (from 5% to 40% in 120 min, and from 40% to 85% in 15 min). At the end of each run, eluent was set back to 4% solvent B, and column left to equilibrate for 20 min. Eluted peptides were injected and analysed by LTQ Orbitrap XL as in Correani et al [30]. In particular, tandem mass spectra (MS/MS) were acquired with a data-dependent top-5 method, selecting the five most intense ions with charge states ≥ 2 detected per survey scan by FTMS if they exceeded an intensity of at least 200 counts. To avoid redundant sequencing of the most abundant peptides, dynamic exclusion was enabled with repeat count of 1, repeat duration of 30 s, exclusion list size of 300, and exclusion duration of 90 s.
Raw files from the nano LC-MS/MS analyses were examined by proteomics software package MaxQuant (version 1.6.0.1, Max Planck Institute of Biochemistry, Martinsried) [31]. The Andromeda search engine was configured for the Mus musculus database from UniProtK (release May 2022, 17102 sequences), including the decoy database of reverse peptides as well as a dataset of commonly detected contaminants in proteomics.
The matrix of protein groups identification was filtered out by the Perseus software (version 1.6.0.7. Max Planck Institute of Biochemistry) for reverse identifications, contaminants, peptide “only identified by site”. Protein groups identified with less than one unique peptide and present in only one biological replica were further removed.
Bioinformatic analysis
Gene Ontology (GO) classifications and enrichments were performed using the Database for Annotation, Visualization and Integrated Discovery (DAVID, v6.8; https//david.ncifcrf.gov/) with the entire murine genome as the background. The GO analysis of the identified EV proteins was performed for cellular components. The identified proteomes were compared with Vesiclepedia (ex Exocarta) database (http://www.microvesicles.org) using the Venn diagrams webtool (http://bioinformatics.psb.ugent.be/webtools/Venn/).
Western blot analysis
BV2 cells (∼106) were lysed in 100 μL RIPA buffer containing a suitable cocktail of protease inhibitors. The lysates were incubated on ice for 30 min and centrifuged at 15,600g for 15 min at 4°C. Supernatants were collected and proteins quantification was performed using a Bradford Assay. Aliquots of cellular lysate and of EV samples were separated on 4-12% Bis-Tris Plus Bolt™ (Invitrogen) in MOPS-SDS running buffer and subsequently transferred to nitrocellulose membrane (iBlot®2 NC Mini Stacks) using the iBlot®2 system (Invitrogen). After incubation with 5% ECL blocking agent (GE Healthcare Life Sciences) in TBS-T buffer (50 mM Tris, 150 mM NaCl, 0.1% Tween-20, pH 7.5) for 1 h, membranes were probed with primary rabbit polyclonal antibody anti-CD81 (1:500) (Santa Cruz Biotechnology, sc-9158), rabbit polyclonal antibody anti-CANX/Calnexin (1:500) (BIOSS, bs-1693R), mouse monoclonal anti-Rab11A (1:500) (Santa Cruz Biotechnology, sc-166912) in TBS-T at 4°C o/n with gentle swinging. After three washes with TBS-T, membranes were incubated for 1 h at room temperature with the appropriate horseradish peroxidase-conjugated secondary antibodies: goat anti-mouse IgG-HRP (1:2500) (Santa Cruz Biotechnology, sc-2005) and goat anti-rabbit IgG-HRP (1:2500) (sc-2004, Santa Cruz Biotechnology). Protein signal was visualized by chemiluminescence using ECLTM Prime Western Blotting System (GE Healthcare Life Sciences), detected using a Molecular Imager R_ChemiDoc™, mod. MP System (Bio-Rad Laboratories) and acquired by ImageLab Software (ver. 4.1). Normalization was based on densitometry obtained by Pierce™ Reversible Protein Stain Kit for nitrocellulose membranes (Thermo Scientific). Statistical analysis was performed by applying Student’s T-test; differences in protein expression with p-value of <0.05 were considered statistically significant.