Isolation and characterization of exosomes from adipose tissue‐derived mesenchymal stem cells

Abstract

Mesenchymal stem cells (MSCs) are the subject of intense research as they are a potential therapeutic tool for several clinical applications. The new MSCs action models are focused on the use of MSC‐derived secretome which contains several growth factors, cytokines, microRNAs, and extracellular vesicles such as exosomes. Exosomes have recently emerged as a component with great potential involved as mediators in cellular communication. The isolation and identification of exosomes has made it possible for them to be used in cell‐free therapies. The purposes of this study are: (i) to detect exosomes released into adipose‐derived MSC conditioned cell culture medium, (ii) to identify exosome morphology, and (iii) to carry out a complete characterization of said exosomes. Moreover, it is aimed at determining which method for exosome isolation would be best to use. Precipitation has been identified as a highly useful method of exosome isolation since it provides higher efficiency and purity values than other methods. A broad characterization of the exosomes present in the MSC‐conditioned medium was also carried out. This work fills a gap in the existing literature on bioactive molecules which have attracted a great deal of interest due to their potential use in cellular therapies.

Isolation and characterization of exosomes from adipose tissue‐derived mesenchymal stem cells.

1. INTRODUCTION

Mesenchymal stem cells (MSCs), also referred to as multipotent mesenchymal stromal cells, have been the subject of intense scientific research as they are a potential therapeutic tool for several clinical applications. The first adult stem cells identified were the multipotent precursors of bone marrow stroma (BMSCs) (Friedenstein, Piatetzky‐Shapiro & Petrakova, 1966), and today, they still remain the main source of MSCs (Wang et al., 2009). However, MSCs are not exclusive to bone marrow and can be isolated from human adipose tissue, brain, liver, spleen, umbilical cord blood, placenta, lung, dental pulp as well as many other sources (Castro‐Manrreza et al., 2014; da Silva Meirelles, Chagastelles & Nardi, 2006). MSCs derived from adipose tissue (ASCs) are a practical alternative due to the accessibility and abundance of this tissue (Schreml et al., 2009).

Due to species diversity, a variety of tissue sources, and culture conditions, there are no specific MSCs markers (Baglio, Pegtel & Baldini, 2012; Pashoutan Sarvar, Shamsasenjan & Akbarzadehlaleh, 2016). This is why The International Society for Cell Therapy (ISCT) has established minimum criteria for characterizing MSCs: 1) the ability to be plastic‐adherent in standard culture conditions; 2) the expression of markers such as CD105, CD90, or CD73, and a lack of expression of markers such as CD34, CD45, CD14, or CD11b; and 3) the ability to differentiate into osteoblasts, adipocytes, and chondrocytes under in vitro conditions (Dominici et al., 2006).

MSCs have three important biological properties which qualify them as a practical alternative in cell therapy: 1) their capability to differentiate into various cell lineages, 2) their ability to migrate to damaged tissue and promote its repair, and 3) the expression of a broad spectrum of immunoregulatory abilities (Lu et al., 2017; Ma et al., 2014). The initial MSCs action model was based on migration to damaged tissues, insertion and differentiation into functional cells thus regenerating the tissue. However, recent studies have shown that MSCs are generally not grafted in sufficient number or present for sufficient time to explain the results in tissue regeneration (Baglio et al., 2015; Dai et al., 2005; Iso et al., 2007). For this reason, new MSCs regeneration mechanisms have been proposed based on paracrine activity mediated by the secretion of growth factors, cytokines, and hormones, cell‐cell interaction through nanotubes and the secretion of extracellular vesicles (EVs) (Spees, Lee & Gregory, 2016). This new approach opens novel therapeutic perspectives for cell–free therapies involving the use of MSCs secretome which contains several growth factors, cytokines, microRNAs, and EVs involved in intercellular communication for the differentiation of various cell lineages, tissue repair, and decrease in inflammation (Ranganath et al., 2012; Tran & Damaser, 2015; Wang et al., 2015).

The general term “extracellular vesicle” (EV) refers to membrane‐bound vesicles released from various cell types in grown in culture medium (Breakefield, Frederickson & Simpson, 2011; Raposo & Stoorvogel, 2013). These EVs include vesicles of different sizes, biogenesis, composition, and source of origin. EVs are divided into three general categories: 1) exosomes (30‐100 nm), derived from the endolysosomal pathway; 2) microvesicles (100‐1,000 nm), derived from the plasma membrane; and 3) apoptotic bodies (1,000‐5,000 nm), released from apoptotic cells (Kalra et al., 2012; Lai et al., 2016; Schorey et al., 2015).

Exosomes comprise a lipid bilayer surrounding a vesicle with a density of 1.10‐1.20 g/ml. Exosomes are released when multivesicular bodies (MVBs) fuse with the plasma membrane and they carry a variety of membrane‐associated proteins such as tetraspanins (e.g. CD9, CD63, CD81, and CD82), heat‐shock proteins (HSP), MHC‐I and MCH‐II, GTPases, and proteins involved in MVB biogenesis (Alix and TSG101) (Colombo, Raposo & Théry, 2014; Hannafon & Ding, 2013; Lopez‐Verrilli & Court, 2013). The most common exosome surface markers are CD9, CD63, and CD81, which belong to the tetraspanin protein family. These proteins mediate exosome secretion and help arrange the exosomes by assembling tetraspanin‐enriched microdomains (TEMs) (Andreu & Yáñez‐Mó, 2014; van Niel et al., 2011). Apart from proteins, exosomes also contain cytokines, lipids, such as cholesterol, fatty‐acyl chains, ceramide or phosphoglycerides, mRNAs, miRNAs, tRNAs, non‐coding RNA, and to a lesser degree, DNA (Chen et al., 2010; Hannafon & Ding, 2013; Subra et al., 2010; Yoon, Kim & Gho, 2014).

The physiological role of exosomes is not entirely clear. However, they have emerged as a component with great potential in the secretome or conditioned cell culture medium (CM) of different cell types such as MSCs. It was initially thought that the only function of exosomes was in the transport and elimination of excess protein and/or other non‐functional molecules from the cell (Kalluri, 2016; Pan et al., 1985). However, nowadays there is strong evidence that exosomes are involved in cellular communication, the control of physiological processes such as immune response, and the progression of various diseases, above all, cancer (Bellingham et al., 2012; H Rashed et al., 2017; Zhang et al., 2015).

MSC‐derived exosomes were first studied in a mouse model of myocardial ischemia/reperfusion injury (Lai et al., 2010). Various studies have also shown the ability of MSC‐derived exosomes to repair and regenerate tissue, that they appear to have anti‐apoptotic and anti‐inflammatory effects and that they intervene in processes of cardiac regeneration and remodeling (Bruno et al., 2009; Huang et al., 2015; Lee et al., 2012). The isolation and identification of exosomes leads to the possibility of cell‐free therapy which could avoid the disadvantages associated with cell therapy such as immune incompatibility, long waiting times, and the high costs involved in the preparation of biological material (Carlson et al., 2012).

One of the main goals in the isolation of exosomes is to achieve a pure sample to study their mechanisms of action and biomedical applications. Many different techniques have been developed. Several researchers have successfully isolated exosomes using techniques such as ultracentrifugation, ultrafiltration, chromatography, polymer‐based precipitation, and affinity capture (Gurunathan et al., 2019; Peterson et al., 2015). The method chosen to achieve exosome isolation depends on their cellular origin but, the most used method is ultracentrifugation. This method is useful for pelleting lipoproteins, extravesicular protein complexes, aggregates, and other molecules. Nevertheless, it is a lengthy process with numerous steps and it requires costly equipment (Gurunathan et al., 2019; Witwer et al., 2013). In order to avoid these disadvantages, alternative and easier isolation methods have been developed.

Exosomes can also be isolated using precipitation methods followed by capture and collection of exosomes of a certain size (50‐150 nm). These precipitation methods are: (i) are simple and easy to use; (ii) give an efficient enrichment of intact exomes; (iii) require a short and low‐speed centrifugation process and; (iv) do not require any expensive or specialized equipment.

This method allows for exosomes aggregates to be formed, which can be easily isolated using low‐speed centrifugation (Théry et al., 2006).

The aim of this study is to detect the exomes released by ASCs into the CM, to identify their morphology, and to characterize them. Additionally, by detecting and quantifying the number of exosomes produced under different isolation conditions, the hope is to make an analysis of available exosome isolation techniques and determine a method of choice.

2. MATERIALS AND METHODS

2.1. Biological material

The experimental procedures developed in this study were approved by the Medical Committee of the University Hospital of León (Spain). Written consent was obtained from all patients involved following the indications of the Helsinki Declaration of 1975, as revised in 2008, and in adherence with the current Spanish and European laws (RD 53/2013 and EU Directive 2010/63/EU).

Samples of adipose tissue were obtained from healthy donors, of the same gender and similar age, during surgical procedures at the University Hospital of León (Spain). The tissues were obtained using sterile techniques and stored in low‐glucose Dulbecco's Modified Eagle's Medium (DMEM) with 1% (v/v) penicillin/streptomycin.

Nucleos pulposus (NP) and annulus fibrosus (AF) cells were commercially purchased.

2.2. ASCs isolation and culture

In order to isolate the ASCs, approximately 5 g of adipose tissue was digested in DMEM containing 0.075% collagenase type I for 2 hours at 37°C on a shaker. The resulting cell suspension was centrifuged at 1,200× g for 10 minutes. The supernatant was removed after centrifuging and the pellet obtained was resuspended in DMEMc (DMEM) supplemented with 10% (v/v) fetal bovine serum (FBS) and 1% (v/v) penicillin/streptomycin.

Isolated cells were multiplied in a monolayer in T150 flasks using low‐glucose DMEMc at 37°C in a humid atmosphere containing 5% CO₂.

2.3. ASCs characterization

2.3.1. Flow cytometry analysis

We determined the expression of different surface markers: anti‐CD73, anti‐CD90, and anti‐CD105 (1:100) to confirm the identity of the ASCs. Cells were stained with streptavidin‐Alexa 488 antibodies (1:100). About 100 events (minimum) were used for fluorescence capture. Data were acquired and analyzed using a CyAn ADP.

2.3.2. Confocal characterization

The cells were sub‐cultured in an 8‐well Nunc Lab‐Tek Chamber Slide System (2 × 10³ cells/well). Cells were fixed with 2% paraformaldehyde for 15 minutes before incubation with primary mouse anti‐CD73 and anti‐CD105 antibodies (1:100) overnight at 4°C and treated with secondary biotinylated antimouse antibodies (1:100). They were then stained with streptavidin‐Alexa 488 and streptavidin antibodies (1:100). Finally, chamber slides were prepared using Vectashield mounting medium containing DAPI.

2.3.3. ASCs differentiation

Isolated ASCs were cultured under conditions conducive to adipogenesis, osteogenesis, and chondrogenesis in order to assess their multi‐potentiality.

2.3.3.1. In vitro adipogenic differentiation assay

For adipogenic induction, 500 μM isobutyl‐methyl‐xanthine IBMX, 150 μM indomethacin, 1 μM dexamethasone, and 10 μM insulin were added to the culture medium containing 1% antibiotic‐antimycotic solution and 10% fetal bovine serum (FBS). After 15 days, the cultured cells were stained with Oil Red O.

2.3.3.2. In vitro osteogenic differentiation assay

ASCs were cultured under osteogenic culture conditions for 15 days in a solution containing 10 mM β‐glycerophosphate, 10 μM L‐ascorbic acid 2‐phosphate, and 0.1 μM dexamethasone DMEM high glucose with 1% antibiotic‐antimycotic solution. After induction, confirmation of osteogenesis was achieved using Alizarin Red staining.

2.3.4. In vitro chondrogenic differentiation assay

For chondrogenic induction, 1 × 10⁶ micro‐mass cell cultures were incubated under chondrogenic medium containing 6.25 µg/ml ITS liquid media supplement (100×), 50 Nm 2 phospho‐L‐ascorbic acid trisodium salt, and 10 ng/ml transforming growth factor‐beta 1 (TGF‐β1) in DMEM high glucose with 1% antibiotic‐antimycotic solution for 3 weeks. Confirmation of chondrogenesis was achieved through Alcian Blue staining 15 days after induction.

2.4. Supernatant collection

ASCs were kept in DMEMc (DMEM) supplemented with 10% (v/v) fetal bovine serum (FBS) and 1% (v/v) penicillin/streptomycin to approximately 80% confluency (~1 × 10⁶ cells) with two passages. For exosome collection, supernatants were collected from ASC‐conditioned medium 24 hours after the cells were supplemented with serum‐free DMEM with 1% penicillin/streptomycin to avoid possible contamination by exosomes present in the FBS.

2.5. Exosome isolation

One of the most common methods used for exosome purification is differential centrifugation. Firstly, cell culture supernatants were centrifuged at 300× g for 10 minutes at 4°C and 2,000× g for 10 minutes at 4°C to remove cell debris and dead cells. The supernatant was then centrifuged at 10,000× g for 30 minutes at 4°C to remove larger particles. The exosomes were subsequently pelleted by ultracentrifugation at 100,000× g for 70 minutes at 4°C. Finally, the pellet obtained was washed with phosphate‐buffered saline (PBS) and pelleted once more by centrifugation at 100,000× g for 70 minutes at 4°C.

A further exosome precipitation‐isolation method was conducted using a commercial exosome isolation kit (Total Exosome Isolation (from cell culture media) kit, Invitrogen™) following the manufacturer's instructions. In brief, 5 ml of the exosome precipitation solution, provided in the kit, was added to 10 ml of the supernatant and the mixture was incubated overnight at 4°C. The samples were then centrifuged at 10,000× g for 1 hour at 4°C. The pellet, containing the exosomes, was then resuspended in PBS. To determine whether a combined method involving centrifugation and precipitation, using the commercial isolation kit, could provide any improvement of exosome isolation and purification, supernatants that had undergone several different centrifugation regimes were collected and enriched with the isolation reagent.

Exosome isolation was carried out under five different conditions as shown in Table 1: Isolation by serial centrifugation, enrichment using the isolation reagent on raw culture medium without previous centrifugation, and enrichment using the isolation reagent on the supernatants obtained after centrifugation at 2,000 x g, 10,000 x g, and 100,000 x g.

TABLE 1.

Proposed conditions for the isolation of the exosomes released into the conditioned cell culture supernatants of ASCs

Methods for exosomes isolation
1. Without centrifugations plus exosome isolation kit.
2. Centrifugation at 2,000× g plus exosome isolation kit.
3. Centrifugation at 10,000× g plus exosome isolation kit.
4. Centrifugation at 100,000× g plus exosome isolation kit.
5. Centrifugation at 100,000× g without exosome isolation kit.

Characterization and Internalization of Exosomes

To investigate the possible molecular mediators of MSC‐derived

paracrine effects, we isolated exosomes from conditioned me‐

dium as described in Materials and Methods. Tra nsmissi on elec‐

tron microscopy, implemented as described previously [19],

revealed the predicted shape and size of exosomes [20] (Figure 4A).

Immunoblotting also confirmed enrichment of the exosomal

surface marker CD63 in purified exosomal fractions (Figure 4B). To

confirm cellular exosome uptake, exosome fractions from each

conditioned medium were labeled with PKH26 and cocultured

with CMs and HUVECs. After 6 hours, the cells were immuno‐

stained and analyzed by confocal laser microscopy to detect

PKH26 fluorescence. As shown in Figure 4C, both CMs and HUVECs

stained positively for PKH26, confirming exosomal uptake

2.6. Detection and quantification of exosomes in ASC‐conditioned medium

Transmission Electron Microscopy (TEM), protein content assay, and flow cytometry were used to estimate the average number of purified exosomes obtained from a sample of ASCs. Samples obtained under the five conditions described were analyzed considering the number of cultured cells and volume of culture medium as well as analyzing the morphological characteristics of the exosomes.

2.6.1. Transmission electron microscopy

Exosome samples were fixed with 2% paraformaldehyde (PFA). Five µl of each sample was placed onto Formvar‐carbon‐coated EM grids and left to adsorb for 20 minutes at room temperature. Samples were fixed with 1% glutaraldehyde, contrasted in an uranyl oxalate solution, pH 7, and then embedded in a mixture of 4% uranyl acetate and 2% methylcellulose in a ratio of 1:9 on ice. Grids were removed with stainless steel loops and the excess fluid was blotted with filter paper. After drying, the prepared exosomes were visualized using a Transmission Electron Microscope as previously described by other authors (Théry et al., 2006).

2.6.2. Quantification of protein in the isolated exosomes

Pellet obtained under the different isolation conditions (Table 1) was resuspended in PBS. The protein concentration of the isolated exosome fraction was measured using a Micro BCA Protein Assay Kit, using bovine serum albumin (BSA) as a standard as per the manufacturer's instructions. The plate was incubated for 1 hour at 37°C, then its absorbance at 562 nm was quantified in a Multiskan GO multi‐plate spectrophotometer.

2.6.3. Quantification of exosomes in CM using flow cytometry assay

So as to detect and quantify the exosomes present in the conditioned culture medium, flow cytometry assay to identify exosome tetraspanin markers was carried out after attaching exosomes human anti‐CD9 conjugated Phycoerythrin (PE) (1:100). Samples of isolated exosomes from each of the five isolation methods used were resuspended in filtered PBS. Purified exosomes were mixed with the anti‐CD9 and incubated for 1 hour at 4°C while protected from light. After incubation, the exosomes were then washed with PBS and pelleted using ultracentrifugation at 100,000× g for 70 minutes at 4°C. The supernatant was discarded, and the pellet containing the exosomes was resuspended in PBS with 2% paraformaldehyde (PFA) and the samples preserved at 4°C prior to flow cytometry analysis.

A mixture without exosomes was used as a negative control to examine any carry‐over. The samples were analyzed using the MACSQuant Analyzer 10 flow cytometry equipment and counts of 10,000 events per acquisition underwent fluorescence capture using MACSQuantify™ software.

2.7. Mass spectrometry and protein identification

Exosomes were isolated using the Total Exosome Isolation Kit (Invitrogen^®), as high quantities of exosomes had been obtained previously using this method. After centrifugation at 10,000× g for 1 hour at 4°C, the exosome‐rich pellet was suspended in RIPA Lysis and Extraction Buffer with Protease and Phosphatase Inhibitor and incubated at 4°C on a shaker to disrupt the membranes. The supernatant containing the soluble fraction (SF) was then collected and placed in a new tube. Protein concentrations were measured by Micro BCA Protein Assay Kit.

Denaturing SDS‐PAGE was conducted following the method described by Laemmli (Laemmli, 1970). An equivalent amount of proteins extracted from all samples was mixed with loading buffer (1 M Tris [pH 6, 8], 20% w/v SDS, 1% v/v 2‐mercaptoethanol, 50% v/v glycerol, 0.00625% w/v bromophenol blue) in a ratio of 1:4 and boiled at 95°C for 5 minutes. Samples were electrophoresed using 10% Bis‐Tris 1.5 mm mini gels. Gel bands for mass spectrometry were stained with Coomassie blue dye on a shaker for 1 hour at room temperature and then the gels were destained three times on a shaker in 50% methanol and 10% acetic acid for 2 hours.

Protein spots of interest were manually excised from stained gels using biopsy punches. Protein digestion was carried out (Havliš et al., 2003) and processed for further analysis as indicated by Jami and co‐workers (Jami et al., 2010). A 4800 Proteomics Analyzer (MALDI‐TOF/TOF) was used to analyze samples and a 4700 Proteomics analyzer calibration mixture (Cal Mix 5) was used as external calibration. Peptides from the trypsin auto‐digestion were used for internal calibration of all MS spectra. Peptide mass fingerprints (PMF) were produced through the use of MALDI‐TOF/TOF mass spectrometry and the peptides observed (up to 65 peptides per spot) represented a list of monoisotopic molecular weights with a signal to noise (S/N) ratio >20, using the 4000 Series Explorer v3.5.3 software. Later MS/MS analyses excluded all known contaminant ions (trypsin‐ and keratin‐derived peptides). The six most intensive precursors with an S/N > 20 were selected for MS/MS analysis with CID (atmospheric gas was used) in 2‐kV ion reflector mode and precursor mass windows of ±7 Da. For protein identification, Mascot Generic Files combining MS (PMF) and tandem MS (MS/MS) spectra were automatically created and used to query a non‐redundant protein database. Search parameters for tandem MS spectra and peptide mass fingerprints were as follows: 1) Uniprot Mammalia (date 2019.01.10; 2,918,298 sequences, 1,393,207,127 residues); 2) fixed and variable modifications were considered (Cys as S‐carbamidomethyl derivative and Met as oxidized methionine); 3) one missed cleavage site was allowed; 4) precursor tolerance was 100 ppm and MS/MS fragment tolerance was 0.3 Da; 5) peptide charge: 1+ and 6) the algorithm was set to use trypsin as the enzyme. Protein candidates produced by a combined peptide mass fingerprinting/tandem MS search were only considered valid when two criteria were met: 1) global Mascot score >83 (significance level of p < 0.05) and 2) sequence coverage >15%. When identifying two or more fragmentation peptides, this was accepted only if the second criterion was not met.

2.8. Exosome characterization: A multiplex bead‐based platform

In order to carry out a characterisation of the protein composition of the exosomes, a MACSPlex Exosome Kit was used. This kit allows the detection of 37 exosomal surface epitomes plus 2 isotype controls which are identified by staining with different antibodies. The multiplex bead‐based platform contained CD3, CD4, CD19, CD8, HLA‐DRDPDQ, CD56, CD105, CD2, CD1c, CD25, CD49e, ROR1, CD209, CD9, SSEA‐4, HLA‐ABC, CD63, CD40, CD62P, CD11c, CD81, MCSP, CD146, CD41b, CD42a, CD86, CD44, CD326, CD133/1, CD29, CD69, CD142, CD45, CD31, Rea Control, CD20, CD14 and mIgG1 as control.

The ASC‐conditioned medium (CM) was centrifuged at 2,000× g for 10 minutes and the supernatant obtained was collected and centrifuged at 10,000× g for 30 minutes. The supernatant was then filtered through a 0.22‐µm membrane. Pre‐cleared cell culture supernatant was added (120 µl) to a MACSPlex Filter Plate well. The samples were analyzed using the MACSQuant Analyzer 10 flow cytometry equipment, following the protocols of the MACSPlex Exosome Kit protocols supplied by the manufacturer.

2.9. Exosome internationalization assay

In order to estimate the internalization capacity of the exosomes in co‐culture, an ExoGlow^®‐Membrane EV labeling kit was used to label the membranes of isolated exosomes, following the manufacturer's instructions. In brief, 12 μl of Reaction Buffer and 2 μl of Labeling Dye were mixed with 100 μg of exosomes suspended in 1 ml PBS. Excess protein binding was saturated using 1% BSA. PBS 1X was added to the mixture which was centrifuged at 100,000× g for 1 hour at 4°C. The labeled exosomes were suspended in DMEMc supplemented with 10% (v/v) Exosome‐depleted FBS and 1% (v/v) penicillin/streptomycin and co‐cultured with either nucleus pulposus cells (NP) or annulus fibrosus cells (AF) (10,000 cells / 500 μl of DMEM) on an eight‐well Nunc Lab‐Tek Chamber Slide System. After 24 hours of incubation, the cells were fixed with 2% PFA, permeabilized with 0.1% Triton X‐100 and blocked with 3% BSA in PBS. Slides were treated with DAPI (1:1500) and prepared with Vectashield mounting medium. Fluorescence confocal microscopy was conducted using a Zeiss LSM 800 laser scanning confocal microscope. Cells cultured with PBS (exosome free) stained with dye solution were used as a negative control.

Flow cytometry was carried out on cells cultured with labeled exosomes using a MACSQuant Analyzer 10. Cells were harvested using trypsin, washed three times with ice‐cold PBS, and fixed in 4% paraformaldehyde (PFA). Data analysis, based on the fluorescence of the B2‐A channel, was completed using MACSQuantify™ software, with a minimum of 10,000 events per acquisition, excluding doublets and higher aggregates using gating following the manufacturer's indications. As before, cells cultured with PBS (exosome free) and stained with the dye solution were used as the negative control.

2.10. Statistical analysis

The results of this study are expressed at a mean ±standard deviation (SD). Statistical analysis was carried out using IBM SPSS Statistics 17. Significant differences among groups were determined using ANOVA followed by Tukey's post hoc analysis. Results with p ≤ 0.05 were considered statistically significant.

3. RESULTS

3.1. ASC characterization

Flow cytometry analysis showed ASCs to be positive for the expression markers CD73 and CD105. The percentages of positive cell markers and their histograms are given in Figure 1a. Conclusive evidence of MSCs multipotency is found in their capacity to differentiate into different cell types. The ASCs under investigation here, were found to be capable of differentiating in osteocytes, chondrocytes and adipocytes under the appropriate conditions. Adipogenic differentiation was detected by the presence of cytoplasmic lipid vesicles by staining with Oil Red‐O. Osteogenic differentiation was characterized by the mineralization of calcium deposits shown by red staining with Alizarin Red S. Chondrogenic cultures produced deposits of acid mucopolysaccharides, as expected, which were confirmed by Alcian Blue staining (Figure 1b).

FIGURE 1.

Characterization of ASCs. (a) Immunophenotyping analysis using confocal microscopy and flow cytometry. ASCs were positive to CD73 (96.19% expression) and CD105 (97.31% expression). Scale bar: 20 µm. (b) Tri‐lineage differentiation potential of ASCs after 15 days of culture in differentiation medium (magnification 20×). Adypocites stained with Oil Red‐O. Presence of intracytoplasmic lipid‐rich droplets. Osteocytes visualized with Alizarin Red S. Matrix mineralization in induced cultures can be well seen. Chondrocytes stained with Alcian Blue. Scale bar: 100 µm

3.2. Detection of exosomes in ASC‐conditioned medium

Exosomes were successfully isolated, as previously described, and they were detected positively and their purity confirmed using TEM. Samples obtained using the five exosome isolation methods outlined in previous sections (Table 1) were analyzed. Under all conditions, analysis showed the presence, of micro‐vesicles which were morphologically consistent with exosomes in terms of size and shape. The size of these exosomes varied between 30 and 100 nm, and their distribution within the sample was nonhomogeneous; it was observed that some fields were empty while others contained aggregations of exosomes (Figure 2). Some differences were observed between samples produced using different isolation methods. The exosomes in samples obtained by the precipitation method were the highest purity level and tended to be large. These results were similar to those obtained using the precipitation isolation solution in combination with previous centrifugation at 2,000× g (Figure 2a,b). Increasing the number of centrifugations resulted in greater number of smaller exosomes which also had lower membrane integrity. In addition, these samples contained higher levels of residue (Figure 2c,d). The samples of exosomes isolated using the serial centrifugation method also contained large numbers of smaller exosomes but with a lighter preparation background (Figure 2e). These results could suggest that the purified exosomes undergo a partial collapse as they are subjected to more centrifugation steps. Although serial centrifugation is the most common method used to isolate exosomes, it is not possible to rule out the presence of other types of EVs in samples processed using this method. However, the precipitation isolation method clearly causes an enrichment allowing the capture and collection of EVs of a certain size.

FIGURE 2.

Exosomes from each isolation strategy observed using TEM. (a) Isolated exosomes without centrifugation plus isolation kit (magnification 30,000×, scale bar = 50 nm). (b) Isolated exosomes at 2,000× g plus isolation kit (magnification 30,000×, scale bar =50 nm). (c) Isolated exosomes at 10,000× g plus isolation kit (magnification 25,000×, scale bar = 50 nm). (d) Isolated exosomes at 100,000× g plus isolation kit (magnification 30,000×, scale bar = 50 nm). (e) Isolated exosomes by serial centrifugation (magnification 30,000×, scale bar = 50 nm)

3.3. Quantification of protein content in isolated exosomes

To estimate the number of isolated exosomes in a sample its protein concentration was determined. Samples of isolated exosomes produced under the five conditions outlined were analyzed for protein concentration (µg/ml) (Micro BCA protein assay) based on a standard BSA curve.

The amount of protein in samples of purified exosomes per cell number obtained under the five different isolation conditions is shown in Figure 3a. The highest protein concentration was observed in samples produced using the method of precipitation without centrifugation (137.40 ± 13.93). It was observed that as the number of centrifugations used during the purification process increased, the protein concentration decreased. However, the serial centrifugation method resulted in samples with mean protein concentration that was no higher (133.82 ± 2.56) than that obtained with the sample isolated using only the exosome isolation kit (i.e., the precipitation method), although statistical analysis showed no significant differences.

FIGURE 3.

(a) Protein concentration (µg/ml) observed from the proposed conditions of isolation of ASCs‐derived exosomes. Statistical analysis showed no significant differences. (b) Quantification of isolated exosomes determined using flow cytometry. Scatter Plots (SSC‐A vs FSC‐A, PE‐A vs FSC‐A) of isolated exosomes samples stained with PE‐labeled anti‐CD9 antibody. Isolated exosomes samples without labeling were used as negative control. (c) Counts /µl comparative analysis of the proposed isolation conditions of ASCs‐derived exosomes. Data presented as mean and standard error (n = 3) *(p ≤ 0.05), **(p ≤ 0.01)

3.4. Flow cytometry assay for exosome quantification

In order to prove that purified EVs were exosomes, samples were immobilized onto latex beads coated with the CD9 antibody. The presence of this exosomal marker indicated that collected EVs in the culture medium were indeed exosomes. The quantity of isolated exosomes was then determined by flow cytometry. The analysis was carried out with samples obtained under all five isolation conditions outlined. In order to accomplish the analysis, doublets and higher aggregates were excluded by gating. A SSC/FSC dot plot was generated to define the population and a SSC/PE dot plot was used to avoid the error caused by FSC detection limit (Figure 3b). Background noise was quantified using an unstained PBS control. Samples of unstained exosomes were also used as negative control and fluorescent anti‐CD9 marked exosomes were used to identify their location in the SSC/PE dot plot.

The highest percentage of fluorescence was obtained for the samples isolated using the Total Exosome Isolation kit (Invitrogen^®) without centrifugation (32.88% expression). The percentage of fluorescence decreased as the number of centrifugation steps increased for those methods combining pre‐centrifugation with precipitation, the lowest value being observed in the samples obtained with centrifugation at 100,000 x g (1.51% expression). Samples obtained using the serial centrifugation method showed a fluorescence percentage of 2.64% expression. In addition, low fluorescence percentages in the noise were also observed in all the samples analyzed.

The results showed that exosome isolation by precipitation without centrifugation gave significantly higher counts/µl (**p ≤ 0.01) compared to the isolation by precipitation with centrifugation at 10,000× g and 100,000× g or isolation by serial centrifugation, thus indicating its superior efficiency. However, no significant differences were observed between this method and that where precipitation was combined with centrifugation at 2,000× g (Figure 3c).

3.5. Protein profile

The protein expression profile in ASCs‐derived exosomes was evaluated. Coomassie brilliant blue staining of equivalent amounts of proteins, separated under reducing conditions using 10% SDS‐PAGE was carried out. We compared the proteins extracted from conditioned medium (CM), exosomes and the soluble fraction (SF), using DMEM as a negative control. The exosome fraction was pelleted and the soluble fraction was collected. The exosome fraction was obtained using the precipitation isolation kit without centrifugation as this method of isolation has been proved to yield high quantities of exosomes. It is worth noting that the protein pattern observed for the exosome fraction indicated the presence of heavier proteins compared to patterns observed for CM and SF. This is consistent with previous reports (Kogure et al., 2011; Yuan et al., 2009) revealing exosome protein profiles comprising multiple protein bands (Figure 4a). In CM an intense band was observed between 50 and 75 kDa. This pattern was consistent with the presence of albumin or various IgG Chains and considerable contaminating proteins. This band was also observed for exosome samples and SF, although it was not as strong. Both CM and samples of the exosome fraction also contained a protein band at 250 kDa. MALDITOF‐TOF/MS Analysis of three selected bands in all samples identified contaminating serum proteins albumin at 69 kDa. However, Alpha‐2‐macroglobulin at 168 kDa and fibronectin at 250 kDa could not be detected in SF samples (Figure 4b).

FIGURE 4.

Protein assay of ASCs‐derived exosomes. (a) Coomassie blue staining of DMEM (as negative control), conditioned cell culture medium (CM), exosome (EXO) and soluble fraction (SF) samples separated using SDS‐PAGE. (b) Number of unique peptides and corresponding percentage coverage for indicated proteins identified in MALDI TOF/TOF analysis of conditioned cell culture medium (CM), soluble fraction (SF), and exosome (EXO) samples

3.6. Characterization of exosome samples: A multiplex bead‐based platform

The multiplex bead platform allowed us to obtain 39 different bead populations which were detected using flow cytometry. To determine the surface marker profile for ASC‐derived exosomes, pre‐cleared cell culture supernatant was obtained as previously described. The pre‐cleared cell culture supernatant was incubated with the 39 populations and stained with a cocktail of anti‐CD9, anti‐CD81, and anti‐CD63‐APC antibodies. Detection with the commonly used exosome markers anti‐CD9, anti‐CD63 and anti‐CD81‐APC as well as with the detection of the 39 different surface markers allowed us to obtain a broad characterization of the ASC‐derived exosomes.

Flow cytometry allowed the production of the following plots for analysis: a SCC/FSC dot plot to exclude doublets and no bead events, a PE/FITC dot plot to determine the different stained beads populations, and a PE/aCD9/63/81‐APC dot plot to measure signal intensities of single bead populations. Exosome‐depleted CM was incubated with the multiplex beads and a cocktail of detection antibodies to ensure non‐specific binding. The signal intensity obtained was subtracted from the signal emitted by the exosome samples for each population to achieve a background‐corrected signal. As a negative control, we used samples of pre‐cleared CM containing exosomes that had not undergone incubation with the multiplex beads or the cocktail of detection antibodies (Figure 5a).

FIGURE 5.

The multiplex bead‐based platform. (a) Analysis of conditioned cell culture medium samples with exosomes without incubation with multiplex beads or cocktail of detection antibodies used as negative control. (b) Surface markers profile of ASCs‐derived exosomes. Background corrected APC median signal intensities after pre‐cleared cell culture supernatant of ASCs was incubated with the 39 capture antibody bead populations and stained with a cocktail of anti‐CD9, anti‐CD81, and anti‐CD63‐APC antibodies. REA and mIgG1 indicate isotype control beads

The surface expression of several CD molecules was analyzed using flow cytometry to carry out a broad characterization of the exosomes contained in the ASC‐conditioned medium. In order to normalize the data, we used an average of the median signal intensity of the exosome marker beads CD9, CD63, and CD81. The percentage of relative expression of each exosomal surface marker is shown in Figure 5b. Thirty‐one different CD molecules expressed by ASCs‐derived exosomes were detected. CD3⁺, CD45⁺, CD56⁺, HLA‐ABC, and HLA‐DRDPDQ were particularly strongly enhanced in these samples (99.99% ±0.06%, 55.45% ±6.36%, 88.68% ±4.29%, 84.66% ±5.99%, 59.98% ±7.45%, respectively). However, CD42a, CD44, CD62P, CD142, CD136, and MCSP were undetectable.

3.7. Uptake of exosomes by recipient cells

Isolated exosomes from conditioned cell culture supernatants produced by the precipitation isolation method were used to confirm cellular exosome uptake. This method of isolation was used as it has been shown to provide a more optimal quantity and purity of exosomes. Exosome fractions from conditioned cell culture supernatants were labeled with ExoGlow^®‐Membrane EV labeling kit. The labeled exosomes were co‐cultured with either AF or NP cells. These cells are used in our laboratory to study the immunomodulation effect of ASCs and ASCs‐derived in discogenic pain and intervertebral disc degeneration. After 24 hours the cells were immune‐stained and analyzed using confocal microscopy (Zeiss^®) and flow cytometry. As can be seen in Figure 6, both AF and NP cells stained positively, confirming exosomal uptake. Moreover, when a quantification of internal exosomes was made using flow cytometry, significant differences were shown in NP and AF cells in comparison to control cells without exosomes (Figure 6b).

FIGURE 6.

Internalization assay. (a) Cellular internalization of ASCs‐derived exosomes into annulus fibrosus (AF) and nucleus pulposus (NP) cells. Scale bar =10 µm. The cells were also analysed using flow cytometry after the internationalization in order to quantify internalization. Dot plots are shown beside the confocal image. (b) Exosome uptake analysis in AF and NP cells. For all the quantification plots, mean ±SD of three independent experiments is shown. Values significantly different ** (p ≤ 0.01) from control are marked with asterisks

4. DISCUSSION

The main challenge in this study concerning ASCs‐derived exosomes was the determination of an efficient method for exosome isolation and providing an in‐depth immuno‐phenotypic description of these exosomes. For this purpose, an exosome precipitation method used alone or in combination with serial centrifugation was compared with the most commonly used ultracentrifugation method for exosome isolation. In order to do this, cells were cultured under the appropriate conditions and exosomes were isolated from the culture medium using either differential ultracentrifugation or using an exosome precipitation method and the exosome precipitation isolation kit plus serial centrifugation. Exosomes isolated using these different procedures were detected and quantified using a protein assay, transmission electron microscopy (TEM), and flow cytometry. In order to estimate the number of exosomes per cell and conduct a complete characterization, a multiplex bead‐based platform has been developed which can detect up to 39 different surface proteins and allows for exosome sub‐population identification. High‐resolution flow cytometry is currently considered as one of the most promising methods for analyzing surface markers in exosomes (van der Vlist et al., 2012).

ASCs characterization was carried out in advance in order to confirm the mesenchymal stem cells’ phenotype. In culture conditions, ASCs displayed a spindle‐shaped morphology similar to that typical of MSCs and in agreement with the first descriptions of MSCs (extracted from bone marrow) by (Friedenstein et al., 1974; Lopez‐Verrilli et al., 2016) and those of other authors looking at ASC morphology (Zuk et al., 2002). The expression of specific markers further confirmed the characteristics of our ASCs (Locke, Windsor & Dunbar, 2009) (Figure 1a,b). The ability to differentiate into osteoblasts, adipocytes, and chondrocytes constitute the three minimum criteria for characterizing MSCs established by The International Society for Cell Therapy (ISCT) and our positive results for tri‐lineage differentiation agreed with those from other studies of the same cell type (Brown, Squire & Li, 2014; Wetzig et al., 2013).

An accurate exosome isolation method is essential when researching into specific exosome functions and components. The isolation method greatly affects the yield and purity of the exosomes isolated from cell culture media. Importantly, the method of isolation is considered to have a deep impact on the exosomal proteome and transcriptome (Patel et al., 2019; Van Deun et al., 2014). Numerous procedures for isolating exosomes have been described (Momen‐Heravi et al., 2013) and have shown that the composition and properties of exosomes isolated by these different methods vary significantly (Tauro et al., 2012). Obtaining exosomes with serial centrifugation is the most commonly used method to isolate exosomes from CM and there are numerous protocols available. However, this method involves extra working time and changes in the protein and lipoprotein composition of these exosomes are often observed (Li et al., 2017). We have shown that precipitation isolation methods are an excellent way to separate exosomes both quickly and efficiently. In addition, although lower yields are obtained with these techniques, samples are of much higher purity than those obtained using ultracentrifugation (Batrakova & Kim, 2015). TEM images were used to further study the morphology of isolated exosomes. TEM results showed the presence of microvesicles in samples produced under all isolation conditions and, in all cases, the majority of these micro‐vesicles appeared to be morphologically consistent with exosomes (Gurunathan et al., 2019) displaying a micro‐membrane capsule structure and being of a similar size and morphology to the exosomes previously described by other authors (Kim, Tan & Lubman, 2015; Takeda & Xu, 2015). The data obtained showed the effect of different centrifugation protocols on the isolated exosome samples. The increase in centrifuges caused a loss of membrane integrity, meaning that exosome cavities had more electron density components than expected (Griffiths, Heesom & Anstee, 2007). In addition, the increase in the number of centrifugations also caused the loss of larger exosomes, with samples produced under such conditions showing groupings of small‐sized exosomes. These results confirm that the purity and efficiency of exosome isolation depends strongly on the isolation technique used. Using the precipitation isolation method, the purity of samples was higher and the exosomes were larger in size. On the other hand, using the ultracentrifugation method yielded both a lower number of exosomes and these were all of a small size (Figure 2d‐e and Figure 3c). We have also shown that the combination of precipitation and centrifugation did not show any improvement in the isolation and purification of exosomes.

Statistical analysis of results from a BCA protein detection assay did not show any significant differences in the quantity of protein obtained for different isolation methods and, indeed, it was similar to that obtained by other authors, although with some variation depending on the cell line (Brownlee et al., 2014; Wang et al., 2015). Results in this study confirmed the presence of CD9‐positive exosomes in samples obtained under all the five isolation conditions. Nevertheless, there were significant differences between the amount of CD9 observed, specifically, for exosomes purified using the precipitation kit without previous centrifugation CD9⁺ was significantly higher than for samples isolated using the other methods. It must be borne in mind, however, that this method of isolation also causes the precipitation of other types of micro‐vesicle which also test positive for the CD9 marker, thus the results do not reflect levels produced by exosomes alone (Kanninen et al., 2016; Mead & Tomarev, 2017; Pospichalova et al., 2015; Villatoro et al., 2019).

In order to determine the protein profiles of CM, exosomes, and SF, samples were stained with Coomassie Blue. All samples showed a different protein profile. The presence of an intense band between 50 and 75 kDa is seen to be especially prominent in CM and SF samples. This band was also observed in exosome samples. MALDITOF‐TOF/MS analysis of three selected bands identified contamination by the serum proteins albumin, alpha‐2‐macroglobulin, and fibronectin. This contamination is common and increases with prolonged centrifugation (Cvjetkovic, Lötvall and Lässer, 2014). Thus further studies are needed focusing on the reduction of contamination in samples of isolated exosomes.

Although in recent years several articles have appeared related to exosome characterization (Batrakova & Kim, 2015; Brownlee et al., 2014; Wetzig et al., 2013), only a few work‐studies have carried out a comprehensive exosome characterization (Mead & Tomarev, 2017; Villatoro et al., 2019). It has been reported that tetraspanins, such as CD9, CD63, and CD81, are enriched in exosome membranes in general as they are involved in the regulatory role played by exosomes, thus they are frequently used as exosome markers (Simons and Raposo, 2009; Kooijmans et al., 2012). Our results showed that CD9, CD63, and CD81, were detected in all samples of ASC‐conditioned CM (10.19% ±2.21%, 4.33% ±1.98%, 43.7% ±4.42%, respectively). The use of a large spectrum of antibodies also showed that exosomes expressed both HLA‐ABC (MHC‐I) and HLA‐DPDQDR (MHC‐II) antigens. This antigen expression is characteristic of exosomes (Buschow et al., 2010). Previous studies have shown that the proteins present inside exosomes reflect their cellular origin (Pegtel et al., 2010). Due to the endosomal origin of the exosomes, they all include membrane‐associated proteins, such as tetraspanins (e.g., CD9, CD63, and CD81), MHC‐I and MHC‐II, and CDs related to the inflammatory response therefore explaining why exosomes are capable of exerting an immunomodulatory effect (Hannafon & Ding, 2013; Pashoutan Sarvar, Shamsasenjan & Akbarzadehlaleh, 2016). CD45⁺ exosomes have been studied due to their implication in inflammatory response (Transfeldt, Raaii, Mehbod, 2010; Alexander et al., 2015). CD3⁺ exosomes also appear to be involved in the immune response as T cell co‐receptors helping to activate T cells (both CD8⁺ and CD4⁺) (Wahlgren et al., 2012; Domenis et al., 2017). Still further studies have reported the expression of the natural killer cell marker, mainly present in immune cells, CD56⁺ (also known as neural cell adhesion molecule) in exosomes as a membrane receptor (Gross et al., 2003).

We also studied the degree to which ASC‐derived exosomes could be internalized into NP and AF cells. Recently, it has been shown that exosomes are key performers in intercellular communication and are released from different cell types to act as mediators of cell‐cell communication (Boriachek et al., 2018). We confirmed that exosomes isolated and purified from the supernatant of ASC‐conditioned medium were capable of being internalized by NP and AF cells. When the number of internalized exosomes was analyzed, no significant differences were observed between cell types, although significant differences were seen compared to controls without exosomes. It should be noted that, in other studies completed in our laboratory, we have observed immunomodulation mediated by exosomes in NP and AF cells (results not shown). Thus, we suggest that exosome internalization could explain the paracrine effects exerted by MSCs on different cell types (Pashoutan Sarvar, Shamsasenjan & Akbarzadehlaleh, 2016).

5. CONCLUSIONS

The growing interest in the study of exosomes has led to an increase in the efficiency of exosome isolation methods. In this study, a method of isolating exosomes by precipitation emerged as the method of choice since it can enrich the exosome markers as well as the associated protein content of samples compared to differential centrifugation methods. Where precipitation was combined with centrifugation, little or no improvements were seen, even for the greatest maximum centrifugation forces. Effective internalization of exosomes within NP and AF cells was also achieved. Our data provide a useful reference for researchers interested in studying the immunological features of these cells which are increasingly attracting a wide range of interest for their potential use in cellular therapies, regenerative medicine, and tissue engineering. Further research into different methods for exosome isolation and characterization is required to enhance the quality of the isolated exosomes and so improve their effectiveness.

CONFLICT OF INTEREST

We confirm that none of the authors have any competing interests in the manuscript.

AUTHOR'S CONTRIBUTION

All the authors were involved in drafting the article or revising it critically for major intellectual content, and all the authors approved the final version to be published. Villar‐Suárez had full access to all the data in the study and was responsible for the integrity of the data and the accuracy of the data analysis.

Conception and design of the study was by Villar‐Suárez, González‐Cubero, and González‐Fernández. Data acquisition was by González‐Cubero, González‐Fernández, Gutiérrez‐Velasco, and Navarro‐Ramírez. Analysis and interpretation of the data was carried out by all the authors.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available on request from the corresponding author.