EVs isolation from whole fruit

Apple-derived NVs (ADNVs) were isolated from DOP Golden variety apple cultivar fruit (Malus domestica sp.), cultivated in Val Di Non (Trentino, Italy). Apples were selected from two distinct ‘Golden Delicious’ diploid cultivar conditions, organic (GDBio) and integrated (GD). The isolation protocol is described by [29]. Briefly, four apples of approximately 250 g each were washed thoroughly. Juice was extracted through cold press mechanism. The released pulp, comprised of fibrous and liquid fractions, was subjected to a series of three centrifugations at increasing speed (650 rcf, 3.000 rcf and 10.000 rcf), to gradually remove denser and denser debris from the solution. At the end of each centrifugation step, the EVs-enriched supernatant was transferred in a novel tube, while the fibrous fraction was discharged. After centrifugation at 10.000 rcf, the EVs-enriched fraction was filtered at 0,22 µm with vacuum filtering systems for 10 min (Sartorius, Germany). The solution was centrifuged at 15.000 rcf with Ultracentrifuge Optima L-70 (Beckman Coulter Inc., USA), type 70 Ti rotor, to remove macrovesicles and cell-wall debris. The supernatant was further centrifuged at 110.000 rcf, to deposit low-density EVs. The pellet, resulting from this last centrifugation, was resuspended in 1 mL of PBS (Thermo Fisher Scientific, MA, USA), and was used as ADNVs-enriched fraction in all following experiments. All ADNVs-enriched fractions were conserved at – 80 °C until use.

ADNVs quantification and size characterisation

Following isolation, ADNVs were quantified through Tuneable Resistive Pulse Sensing (TRPS) (qNANO Gold, Izon Science Ltd, Cambridge, MA, USA). The analysis also provided data on size distribution of particles in the ADNVs-enriched fraction. The nanopore (NP100, Izon Science Ltd, Cambridge, MA, USA) was stretched 49 mm wide and measured with a digital calliper. Pore installation, wetting and cleaning was performed with reagents provided by the manufacturer, following manufacturer’s instructions. As standard procedure, each sample was measured both at 10 atm and 20 atm of pressure. During each measurement, particle rate was maintained above 200 particles/minutes and the total particle count surpassed 500 particles. Calibration particles (CPC100, Izon Science Ltd, Cambridge, MA, USA) were used for calibrating each sample measurement during data analysis, at 10 atm and 20 atm, ensuring a maximum current divergence of 5%. Each measurement was performed in triplicate.

Cell culture

Fibroblasts from the derma were cultured in Dulbecco’s Modified Medium (DMEM) completed with 10% Foetal Bovine Serum. For both Ca2+ signalling analysis and mitochondrial superoxide assay, cells were seeded onto coverslips or plates respectively, and cultured for 24 h before the beginning of the experiment. For the Mitochondrial Superoxide Assay and ABTS Assay, fibroblasts were treated with TNFα (100 ng/mL) for 6 h. Cells were then treated with 0, 100, 200 and 300 μg of ADNVs (GD), dissolved in 1 mL of DMEM complete medium for 12 h.

Ca 2+signalling analysis

The cytosolic Ca2 + response was evaluated using the fluorescent Ca2+ indicator Fura-2 (Life Technologies, Invitrogen). Fibroblasts, grown onto 24-mm coverlids, were incubated for 30 min at 37 °C in 1 mM Ca2 + /KRB supplemented with 2,5 mM Fura-2 AM, 0.02% Pluronic F-68 (SigmaAldrich), and 0.1 mM sulfinpyrazone (Sigma-Aldrich). For samples necessitating extracellular Ca2+ removal, the above-descripted medium was spiked with EGTA (5 μM). At the end of incubation, coverslips were washed and supplemented with either 1 mM Ca2 + /KRB or 1 mM Ca2 + /KRB + 5 μM EGTA. The live cells were placed in an open Leyden chamber on a 37 °C thermostat-controlled stage. Acquisition was performed by exposing the cells to 340 nm/380 nm wavelength light, using the Olympus xCellence multiple-wavelength fluorescence microscopy system, equipped with an ORCA ER CCD camera (Hamamatsu Photonics) and an Uplan FLN 40 × oil objective (Olympus). The data, expressed as emission ratio, was then plotted as a curve, and the area underneath (AUC) was calculated to compare conditions.

Mitochondrial superoxide assay

After treatment, cells were assayed with MitoSOX™ Mitochondrial Superoxide Indicators (Thermo Fisher Scientific), as for manufacturer’s instructions. Cells were then detached from the culture plate and resuspended in 25 μL of PBS. The MitoSOX dye permeates cellular membrane of live cells, is oxidized by superoxide, and produces red fluorescence. Fluorescence was measured with Tali Image-Based Cytometer (Thermo Fisher Scientific) at absorption/emission 396/610 nm. Un-stained cells were used for background reference.

DPHH assay

Radical scavenging assay was performed with the radical DPPH (1,1-diphenyl-2-picrylhydrazyl), as described by Yu et al. [30]. A solution of 400 μg/mL DPPH in methanol was prepared, filtered with syringe filters at 0,22 μm and kept away from light. To test ADNVs scavenger ability, ADNVs at different concentrations (100, 200 and 300 μg/mL) were added to the DPPH solution at 1:1 ratio. The solutions were mixed and incubated at RT for 30 min. Samples were placed in a 96 well, two wells per-sample, and absorbance was measured at 516 nm with a plate reader (Victor3, PerkinElmer, MA, USA). Results are shown as percentage relative to the control sample (DPPH), after normalization over the blank sample (methanol). The experiment was performed in triplicate.

TEM imaging of ADNVs

Both GD and GDBio ADNVs fractions were prepared for Transmission Electron Microscopy (TEM), imaging following a procedure for the staining of extracellular vesicles described by Corona et al., 2023 [31]. Briefly, EVs were deposited at a concentration of 109 particles/mL and a TEM Grids 200 Mesh Cu/Pd. Sample and grid were fixed with 2% paraformaldehyde (PFA) 1% glutaraldehyde in 100 mM PBS, pH 7,4 solution. For staining, a contrasting solution of Methylcellulose/Uranyl Acetate was applied for 10 min. The grid was dried, and pictures of the grid were taken with TEM Zeiss EM 910 instrument (Carl Zeiss Microscopy, Oberkochen, German). Images were recorded by a CCD digital camera (Ultrascan 1000, Gatan, Munich, Germany).

Proteomics sample preparation and nLC-MS/MS analysis

Total protein content was assessed though Bradford assay with Pierce™ BCA Protein Assay kit (Thermo Fisher Scientific, Waltham, MA, USA) following manufacturer’s instructions. Absorbance was measured at 570 nm using multilabel plate reader Victor 3 (Perkin Elmer, Milano, Italy). ADNVs of both GD and GDBio conditions were digested with trypsin as previously reported [32]: After digestion and clean-up, MS analysis was performed by nanoflow ultra-high performance liquid chromatography-high resolution mass spectrometry using an Ultimate 3.000 nanoLC (Thermo Fisher Scientific, Bremen, Germany) coupled to an Orbitrap Lumos tribrid mass spectrometer (Thermo Fisher Scientific) with an Easy nano electrospray ion source (Thermo Fisher Scientific). Peptides were trapped in a PepMap trap column (Thermo Fisher), and then loaded and separated onto a C18-reversed phase column (250 mm × 75 μm I.D, 2.6 µm, 100Å, Biozen). Mobile phases were A): 0,1% HCOOH in water v/v; B): 0,1% HCOOH in ACN/Water v/v 80/20, a linear 60 min gradient was performed. HRMS analysis was performed in data dependent acquisition (DDA), with MS1 range 375–1.500 m/z, HCD fragmentation was used with normalized collision energy setting 27. Resolution was set at 120.000 for MS1 and 15.000 for MS/MS. Single and unassigned charge were excluded. Quadrupole isolation was set to 3Da. Each sample was analysed by LC–MS/MS in duplicate. MS data were acquired using a data-dependent method, dynamically choosing the most abundant precursor ions from the survey scan (375–1.500 m/z) using HCD fragmentation. Maximum ion injection times for MS (OT) and the MS/MS (OT) scans were set to auto and 50 ms respectively, and ACG values were set to standard. Dynamic exclusion: 30 s. For data processing, raw MS data were analysed using Proteome Discoverer v 2.5 (Thermo Fisher) MS/MS was matched against Malus Domestica proteome (Uniprot 03/2022 version). The following parameters were used: enzyme trypsin, missed cleavages max 1, mass accuracy tolerance 10 ppm and 0,6 Da for precursors and fragments respectively. Sequest search and Percolator algorithm were used. Carbamidomethylcysteine was used as fixed modification while methionine oxidation as variable. Proteins were considered identified with at least one unique peptide, using a false discovery rate (FDR) threshold of < 0,1. Each analysis was performed in triplicate.

Lipid sample preparation and UHPLC-TIMS analysis

Lipid extraction and mass spectrometry‐based lipid detection were performed on ADNVs derived from GD and GDBio samples, in triplicate. 225 µL of cold MeOH (methanol) containing a mix of deuterated standards (EquiSPLASH® LIPIDOMIX®, Avanti Polar Lipids, Alabaster, AL, U.S.A) were added to ADNVs, vortexed and then incubated for 1 min at -30°C. The samples were shacked and sonicated for 10 min (× 2). Subsequently, 800 µL of cold MTBE (methyl tert-butyl ether) were transferred to the tube and the solution was incubated in a thermomixer (Eppendorf, Hamburg, Germany) for 1 h, 500 rpm at 4 °C. To induce phase separation, 188 µL of H2O were added and samples were centrifuged at 14680 rpm, for 10 min at 4 °C. The upper layer was collected and evaporated using a SpeedVac (Savant, Thermo Scientific, Milan, Italy). The dried samples were dissolved in 100 µL of CHCl3/MeOH/IPA 1/2/4 (v/v %) before the UHPLC-TIMS analysis. Unless otherwise described, all solvents and additives were LC–MS grade and purchased by Merck (Darmstadt, Germany).

UHPLC-TIMS analyses were performed on a Thermo Ultimate RS 3000 coupled online to a TIMS-TOF-Pro quadrupole Time of flight (Q-TOF) (Bruker Daltonics, Bremen, Germany) equipped with an Apollo II electrospray ionization (ESI) probe, as previously reported [32]. The separation was performed with an Acquity UPLC CSH C18 column (100 × 2,1 mm; 1,7 μm) protected with a VanGuard CSH precolumn (5,0 × 2,1 mm; 1,7 μm, 130 Å) (Waters, Milford, MA, U.S.A). The column temperature was set at 55 °C, a flow rate of 0,4 mL/min was used, mobile phase consisted of (A) ACN/H2O containing HCOONH4 (10 mM) and 0,1% HCOOH 60:40 (v/v %) and (B) IPA/ACN containing HCOONH4 (10 mM) and 0,1% HCOOH 90:10 (v/v %). The following gradient has been used: 0 min, 40% B; 2 min, 43% B; 2.10 min, 50% B; 12 min, 54% B; 12.10 min, 80% B; 15 min, 99% B; 17 min, 99% B, 17.10 min, 40% B and then 2.9 min for column re-equilibration. The analyses were performed in data-dependent parallel accumulation serial fragmentation (DDA-PASEF) with both positive and negative ionization, in separate runs. The oven temperature was set to 55°C and the injection volume was 2 µL for the positive mode and 5 µL for the negative mode.

Source parameters: Nebulizer gas (N2) pressure: 3,0 Bar, Dry gas (N2): 10 L/min, Dry temperature: 250°C. Mass spectra were recorded in the range m/z 50–1500, with an accumulation and ramp time to 100 ms each. The ion mobility was scanned from 0,55 to 1,80 Vs/cm2. Precursors for data-dependent acquisition were isolated within ± 2 m/z and fragmented with an ion mobility-dependent collision energy ranging from to 20–40 eV in positive mode while a fixed collision energy (40 eV) was employed in negative mode. The total acquisition cycle was of 0.32 s and comprised one full TIMS-MS scan and two PASEF MS/MS scans. Exclusion time was set to 0.1 min, Ion charge control (ICC) was set to 7.5 Mio. The instrument was calibrated for both mass and mobility using the ESI-L Low Concentration Tuning Mix with the following composition: [m/z, 1/K0: (322,0; 0,7 Vs cm−2), (622,0; 1,0 Vs cm−2), (922,0; 1,2 Vs cm−2), (1222,0; 1,4 Vs cm−2)] in positive mode and [m/z, 1/K0: (302,0; 0,7 Vs cm−2), (602,0; 0,9 Vs cm−2), (1034,0; 1,2 Vs cm−2), (1334,0; 1,4 Vs cm−2)] in negative mode.

4D data alignment, filtering and annotation was performed with MetaboScape 2021 (Bruker) employing a feature finding algorithm (T-Rex 4D) that automatically extracts buckets from raw files. Feature detection was set to 500 and 250 counts for positive and negative modes. The minimum number of data points in the 4D TIMS space was set to 100, and recursive feature extraction was used. The spectra were processed in positive mode using [M + H]+, [M + Na]+, [M + K]+, [M + H-H2O]+ and [M + NH4]+ ions in positive mode, while [M–H]−, [M + Cl]−, [M + HCOO]− and [M + H-H2O]+ in negative mode, the assignment of the molecular formula was performed for the detected features using Smart Formula™ (SF). Lipid annotation was performed first with a rule-based annotation and subsequently using the LipidBlast spectral library of MS DIAL (http://prime.psc.riken.jp/compms/msdial/main.html) with the following parameters: tolerance: narrow 2 ppm, wide 10 ppm; mSigma: narrow 30, wide 250, MS/MS score: narrow 800, wide 150. CCS%: narrow 2, wide 3. For the assessment of repeatability and instrument stability over time was used a mixture of lipid standards [LightSPLASH®, Avanti Polar Lipids], blank samples were used to assess and exclude background signals. At the end all lipids missing more than 75% of samples or influenced by carry-over effects were deleted. Lipids were quantified using the corresponding deuterated internal standard as reported previously [33]. Each analysis was performed in triplicate.

Small RNA sequencing

Small RNAs were isolated from ADNVs samples of GD and GDBio derivation with Exosome Purification and Exosomal RNA Isolation kit (Norgen Biotek Corp., Thorold, ON, Canada) following manufacturer instructions. The extracted RNA quality and concentration was verified with NanoDrop One (Thermo Fisher Scientific, Waltham, MA, USA). RNA was then stored at – 80 °C until use.

Sequencing of all miRNAs was performed with Illumina sequencing technology. Briefly, 250ng of RNAs were processed using QIAseq miRNA Library Kit (QIAGEN; Hilden, GE). Sequencing was performed on a Novaseq 6000 Sequencing System (Illumina; San Diego, CA, USA) in 2 × 150 paired-end mode. Identification of miRNAs in the samples was performed using the QIAseq miRNA-NGS data analysis software considering Single Read as read type and Read 1 Cycles 75 as read cycles. Each analysis was performed in triplicate.

Enrichment and statistical analysis

Pathway enrichment analysis was performed on proteomics results, following “Pathway enrichment analysis and visualization of omics data using g:Profiler, GSEA, Cytoscape and EnrichmentMap” nature protocol [34]. Briefly, normalized output from GD and GDBio proteomics were enriched using g:Profiler open-access online enrichment software (https://biit.cs.ut.ee/gprofiler/gost). Parameters were chosen as follows: Malus domestica organism, significance was calculated by the software with FDR-adjusted p-value (p-valueadj) < 0,05, and multiple testing correction was performed with g:GOSt tailor-made algorithm g:SCS for reducing significance score. Enrichment was performed on three different data sources, biological process (BP), cellular compartment (CC) and molecular functions (MF) on gene set database Gene Ontology (GO). “Inferred from electronic annotations” (IEA) matches were not removed, since the specie Malus domestica is less-well-studied compared to the human genome and model organisms. For the same reason, gene set size was not reduced during following steps of enrichment analysis. Therefore, larger and smaller pathways are included. Enrichment data was uploaded on network visualization software Cytoscape, where EnrichmentMap application was used to create the networks of pathways. A hierarchical layout was chosen for the visualization, with operator manipulation. For data interpretation, AutoAnnotate application was used to cluster pathways into groups with a summarized name, to provide an overview of the enrichment result themes.

miRNA sequences were aligned with Malus domestica transcriptome though Qiagen analysis software Ingenuity Pathway Analysis (IPA). Screening of miRNAs was performed by selecting miRNA present in three or more sample replicates with an average mean READ count abundance over 100. Abundance figures display the READs mean of all replicates with ± standard error (SE), they were obtained with GraphPad Prism 8 software version 8.0.0 for Windows (GraphPad Software, San Diego, California USA, www.graphpad.com). Comparison between GD and GDBio samples was analysed statistically by one-way ANOVA, followed by the Bonferroni post-hoc multiple comparison test (GraphPad Software). Differences between groups were considered significant given p-values < 0,05.

Lipid profiling was conducted with LipidSig (http://chenglab.cmu.edu.tw/lipidsig/). Univariate statistics and pathway analysis was performed with open-access software MetaboAnalyst 5.0 (https://www.metaboanalyst.ca/). Pathway analysis was performed by selecting Arabidopsis thaliana KEGG pathway library, using all compounds of the pathway library, though Hypergeometric test type enrichment. A scatter plot was produced for the visualization of significant features.

Principal Component Analysis (PCA) plots, Volcano plots and Heatmaps of significant proteins, miRNAs or lipids were produced with SRplot (https://www.bioinformatics.com.cn/en) online tool (cluster orientation: bidirectional; cluster method: complete; distance method: Euclidean).