Cutting Edge Approaches for Rapid Characterization of Airway Exosomes

Abstract

Exosomes have been described as promising biomarkers for understanding disease progression and prognosis. These lipid membrane nanoparticles derived from airway cells have been shown to have immunomodulatory effects, such as driving inflammatory responses in asthma. These emerging evidences demonstrating an important pathophysiological role of exosomes warrants the development of novel approaches for isolation and rapid characterization of exosomes, which would be applicable for both translational and clinical studies. In this review article, we describe two methods of rapid exosomes characterization: (1) imaging flow cytometry using ImageStream; and (2) conventional flow cytometry using the BD Symphony A5 platform. We also explore sorting of exosomes using the BD Aria.

Graphical Abstract

Overview of exosome characterization from either conditioned media or biological fluid, such as Broncho Alveolar Lavage (BAL) fluid, using both imaging and conventional flow cytometry.

1. Introduction

Exosomes are appreciated as important cellular couriers with diverse biological effects. They are endosomally-derived vesicles of size ranging from 50–150nm, which nucleotides, enzymes, and metabolites [1]. These vesicles are secreted into the extracellular space where adjacent cells can internalize them or travel to distal sites through the vasculature [2, 3]. Extracellular vesicles, such as exosomes, have been shown to play a role in regular homeostasis as well as in disease pathogenesis [4, 5].

In the field of pulmonary immunology and allergy, exosomes have been shown to regulate immune responses in the lung and contribute to allergic diseases such as asthma [6]. Exosomes from the airway of asthmatics have been shown to be enriched with class-II antigen presentation molecules (HLA-DR) as compared to healthy controls [7, 8]. Exosomes are derived from various cell types that reside in the airways, including myeloid-lineage cells, epithelial cells and fibroblasts [9–11]. We have isolated and characterized exosomes produced by purified airway cells sorted from the airways, in particular, those produced by myeloid derived regulatory cells (MDRCs). These exosomes express HLA-DR on the surface, and are internalized by CD4⁺ T cells [9]. Furthermore we have evidence to show that the MDRC-derived exosomes transfer mitochondria to autologous T cells [9]. Exosomes from a mouse model of chronic obstructive pulmonary disease (COPD) have been shown to contain neutrophil elastase (NE), a potent mediator of COPD [12]. These exosomes were shown to protect NE from α1-antitrypsin, a potent inhibitor of NE [12]. The protected NE in neutrophil-derived exosomes from COPD mice is capable of “transferring” the disease to healthy mice demonstrating the pathogenic potential of exosomes in disease.

In addition to the pathogenic roles of exosomes, they have translational value as characteristic differences in these particles have been shown to be correlated with disease stages, for example in cancer [13–15]. Micro-RNA-1290 and miR-375 contained in plasma exosomes of castration-resistant prostate cancer has been shown to be promising prognostic markers and have been associated with poor overall survival [15]. These reports indicate that exosomes may be used as novel biomarkers and can easily be obtained through liquid biopsy, which is generally less invasive.

The translational value and biological implications of exosomes warrants the development of rapid techniques to characterize these nanoparticles. Different isolation methods have been developed, such as differential centrifugation, precipitation, cushioned sedimentation, size-exclusion chromatography, and gradient density ultracentrifugation [16–18]. Each of these techniques have their pros and cons, ranging from quality and purity of isolated EVs to the requirement of specific equipment, such as an ultracentrifuge. High quality and pure exosomes may be required for functional experiments, however, for the purpose of rapid characterization, we describe the differential centrifugation and precipitation methods, which are rapid and provide sufficient yields for characterization by flow cytometry or ImageStream. Methods for multiparametric analysis of circulating exosomes have been described previously using imaging flow cytometry [19]. Here, we describe in depth the use of flow cytometry and ImageStream imaging cytometry for the detection and characterization of exosomes as a rapid means to obtain population statistics and characteristics of exosomes. Furthermore, we explore the novel approach of sorting exosomes using the BD Aria.

2. Exosome Isolation for flow cytometry and ImageStream analysis

2.1. Isolation of exosomes from Broncho Alveolar Lavage (BAL) fluid

The ultracentrifugation method is commonly used to isolate exosomes from large volumes of fluid, such as bronchoalveolar lavage (BAL) fluid [8, 17]. This method has been adapted from an exosome isolation protocol described by Théry, et al [17]. Although ultracentrifugation may pellet other types of extracellular vesicles, the isolation procedure can be accomplished in one day and is sufficient for the purpose of rapid exosome characterization by flow cytometry or ImageStream. Any vesicles devoid of exosome markers are excluded in the analysis by only gating for vesicles with positive fluorescence signal.

BAL fluid is first centrifuged at 300xg for 10 minutes at 4°C to remove any cells. The resulting supernatant is transferred to a new tube and samples are kept on ice or centrifuged in a refrigerated centrifuge at 4°C. The cell pellet can be preserved and used for other assays. The transferred supernatant is then centrifuged further at 2,000xg for 15 minutes at 4°C to remove any cellular debris, such as apoptotic bodies. Following the 2,000xg spin, the supernatant is centrifuged in an ultracentrifuge at 10,000xg for 30 minutes at 4°C in a Ti-70 rotor. The resulting supernatant is transferred to a new tube and filtered through a 0.2μm filter to exclude particles larger than 200nm. When transferring the supernatant, it is important to take care not to disturb the pellet at the bottom of the centrifugation tube. After filtration, the supernatant is centrifuged at 100,000xg at 4°C for 70 minutes in a Ti-70 rotor. It is important to balance all tubes prior to the ultracentrifugation step as an unbalanced rotor can result in injury or even death. Marking the side of the centrifugation tube that faces outwards is also suggested as it will help locate where the pellet should be in case the pellet is small or invisible. The resulting pellet, if visible, may appear as a clear oil-droplet like pellet or a slightly hazy pellet. The supernatant is carefully removed and the pellet washed with sterile PBS. Often times the pellet is not visible and therefore a large are of the tube should be washed with PBS to ensure the pellet is resuspended. The sample is then centrifuged again at 100,000xg at 4°C for 70 minutes in a Ti-70 rotor, and the final pellet resuspended in 100μl of sterile phosphate buffered saline (PBS). Same care should be taken when reuspending the pellet by pipetting a large are of the tube to ensure that the pellet is fully reconstituted in the PBS solution. Exosome samples can be stored at −80°C until use.

2.2. Isolation of exosomes from cell culture supernatant

When isolating exosomes from smaller volumes of conditioned media, the precipitation method using an exosome isolation kit, such as the Total Exosome Isolation from Cell Culture Media (Invitrogen, Cat# 4478359) used in this paper, can yield sufficient exosomes for rapid characterization by flow cytometry or ImageStream. Similar to the differential centrifugation method, you may precipitate protein aggregates or vesicles devoid of exosome markers, which can be excluded from your analysis by gating for vesicles with positive fluorescence signals for the markers of interest.

The procedures for isolating exosomes using the precipitation method requires less steps compared to the differential centrifugation method, however, takes 2 days to complete. First, conditioned media is harvested and centrifuged at 2,000xg at 4°C for 30min to remove cells and cell debris. If cells are required for a different assay, then a 300xg centrifugation step can be performed prior to the 2,000xg centrifugation. The supernatant is transferred to a new sterile conical or microtube (choose appropriate container based on volume). Half volume (0.5) of Isolation Media from the Total Exosome Isolation kit is added to the supernatant. For example, for 1ml of conditioned media, 0.5ml of Isolation Media is added. The Isolation Media is viscous and thus slow pipetting will allow for accurate and reliable dispensing of the Isolation Media into the samples. Once the Isolation Media has been added, the samples are vortexed and incubated at 4°C overnight. The incubated samples are then centrifuged at 10,000xg at 4°C for 60 minutes. After centrifugation, a white pellet may be visible. The supernatant is carefully removed and the resulting pellet then resuspended with 100μl of PBS. Gentle and persistent pipetting is recommended when resuspending the pellet as it may stick to the pipet tip resulting in lower yields. Exosome samples can be stored at −80°C until use.

3. Exosome characterization by ImageStream

3.1. Sample preparation for ImageStream analysis

Quantitation of the exosome preparations prior to staining with antibodies is recommended in order to preserve limiting sample. Quantitation can be done with any nanoparticle tracking platform, such as the NanoSight [20]. Initial trial-and-error is required to determine the optimal dilution for your exosome source, as yield may vary based on exosome source (serum, BAL, conditioned media, etc.) and the isolation method. For exosomes purified from approximately 40–80ml of BAL, a dilution factor of 1000-fold is sufficient. Dilutions should be performed in PBS. The time required to quantitate exosomes on the NanoSight is approximately 10min per sample but may vary based on your quantitation platform.

Once samples have been quantitated, a minimum of 1×10⁶ exosomes is recommended for ImageStream analysis. Preparation of samples for the ImageStream takes approximately 1 hour, but may vary depending on panel complexity and sample number. First, transfer approximately 1×10⁶ exosomes to a new 1.5ml centrifugation tube and adjust the volume to 30μl with PBS that does not contain Mg²⁺ or Ca²⁺. Aliquot exosomes for unstained control in the same manner as the sample. If multiple exosome samples are being analyzed on the ImageStream, the sample exosomes can be pooled for the unstained control to limit use of exosome samples. Single color controls are prepared using the Invitrogen UltraComp eBeads Compensation Beads (Cat# 01–2222-41). The Invitrogen UltraComp eBeads are provided in a dropper and therefore cannot be pipetted. One drop of the UltraComp eBeads is sufficient for preparing each single color control. Prior to dispensing UltraComp eBeads, vortex the tube to ensure homogenous bead content. Add one drop of UltraComp eBeads to a new centrifugation tube and add 1μl of a single fluorescently conjugated antibody for the single color control. Repeat the steps for creating single color controls for each antibody used in the analysis panel. The amount of antibodies to add should be pre-determined by titration, however, 1–2μl can be used safely. A final antibody only control is prepared by adding 1μl of each antibody used in the panel to 30μl of PBS to determine if free floating antibodies are interfering with the acquisition of sample on the ImageStream. Lastly, the exosome samples are labeled with the antibodies used in the analysis panel. For the purpose of this paper, the panel used for the BAL exosomes on the ImageStream is: CD63 eFluor450 (eBioscience, Lot: 1946493, Clone: H5C6, Cat# 48–0639-41), CD81 PE-Cy7 (BioLegend, Lot: B276513, Clone: 5A6, Cat# 349511), TSG101 FITC (Novus, Lot: 43353–040919-F, Clone: 4A10, Cat# NB200–112F), HLA-DR APC (eBioscience, Lot: 1994140, Clone: LN3, Cat# 17–9956-42), CD16 PE (eBioscience, Lot: 4303330, Clone: eBioCB16, Cat# 12–0168-42). Samples and single color controls are labeled with antibodies in the dark for 30 minutes on ice. No washing or re-isolation step is required as the antibody alone control will be used to confirm that free floating antibodies do not interfere with the data acquisition. Keep stained samples, unstained control and single color controls in the dark on ice.

3.2. ImageStream sample acquisition

Acquisition of data on the ImageStream usually takes 5–10min per sample but may vary depending on exosome concentration and sample volume. The ImageStream MkII used in this methods paper is equipped with the following lasers: 405nm, 488nm, 561nm, 642nm. These lasers were set to maximum power to ensure that even low signals can be captured at acquisition. The laser powers used for the ImageStream MkII are as follows: 120mW (405nm), 200mW (488nm), 200mW (561nm), and 150mW (642nm).

To acquire exosome data on the ImageStream, it is important to determine localize exosomes on the forward-scatter (size) vs side-scatter (granularity) plot. ApogeeMix beads of known size and fluorescence will help determine the initial gate for exosomes. Transfer 100ul of ApogeeMix beads to a clean centrifugation tube for acquisition on the ImageStream. ApogeeMix beads have fluorescent 110nm latex beads (FITC) which can be used to determine the initial acquisition gate. This is especially important on the ImageStream as the machine runs calibration beads called Speed Beads which may appear as exosomes if not carefully excluded. A dot plot of FITC vs side-scatter or FITC vs size (Figure 1A–C) can be used to determine where on the size vs side-scatter plot the exosomes will be present. Once the location of the 110nm beads have been determined, create a region around it which will be used as the acquisition region. Acquire 5,000 events on all events for the ApogeeMix beads as this will be used to determine the analysis region later on. Although we use ApogeeMix beads which contain both silica and latex beads, the fluorescent beads that are in the size range of exosomes are latex. Recent reports indicate that silica beads have closer refractive indices to exosomes and thus are recommended over latex beads [21, 22].

Figure 1 –

Exosome analysis on ImageStream. (A) Size (Area_M09) versus SSC (Ch12) of ApogeeMix beads and calibration speed beads. (B) Size versus FITC (Intensity_Ch02) illustrating the location on the side scatter that is positive for the fluorescent ApogeeMix beads of 110nm size. (C) Visualizing the ~100nm region from (B) on a size versus SSC plot demonstrating its location when acquiring samples near the ~100nm size range. (D) Representative size versus SSC plot of exosomes purified from human BAL. (E) Representative size versus SSC plot of MDRC-derived exosomes. (F) Representative size versus SSC plot of antibodies alone in PBS as a negative control. (G) Representative histogram of CD63⁺ BAL exosomes from the ~100nm region. (H) Representative histogram of TSG101⁺ BAL exosomes from the ~100nm region. (I) Representative histogram of CD81⁺ BAL exosomes from the ~100nm region. (J) Quantitation of percent of BAL exosomes positive for each of the 5 markers used in the sample acquisition panel (TSG101, CD81, CD63, HLA-DR, CD16). (K) Multiparameter analysis showing percent of BAL exosomes positive for a combination of the 5 markers used in the sample acquisition panel.

After creating an acquisition region for the exosomes, unstained control should be loaded to determine negative signal for each of the channels being acquired. Create histograms for each channel that will be acquired and note the negative peaks. Next, load the single color controls to verify that adequate signal and resolution can be acquired. Acquire a minimum of 5,000 events of data for the unstained control and each of the single colors. The unstained and single color controls will be used to setup the analysis of the sampled data. Once the controls have been acquired, acquire a minimum of 10,000 events for each sample from the acquisition region that was created for exosomes. For BAL exosome samples tend to have a broader spread of particles with differing granularities as seen in Figure 1D compared to the MDRC-derived exosomes that were isolated using a precipitation method (Figure 1E). Following acquisition of samples, load the antibody alone control and acquire the data to confirm that free floating antibodies do not interfere with data acquisition. No events should be seen where the 100nm particles were present during sample or ApogeeMix bead acquisition (Figure 1F). It is important to acquire only the events that occur inside the acquisition region to avoid dilution of data by Speed Beads.

3.3. Analysis of ImageStream data on IDEAS software

Analysis of ImageStream data is performed on the IDEAS software made available by the manufacturers. Analysis time may take approximately 30–60min per sample, but may vary depending on panel and analysis complexity. The raw data files that come off of the ImageStream are called Raw Image Files (RIF) and have an extension “.rif”. When a RIF file is analyzed on IDEAS, two additional files are generated: 1) Data Analysis File (.daf) which stores the analysis plots and statistics; and 2) Compensated Image File (.cif) which stores the fluorescence compensation.

The first step in analysis is generating the fluorescence compensation matrix by either using the automated tools available in IDEAS or manually. Once the compensation matrix has been generated, the matrix can be applied to each samples through IDEAS. For each sample and unstained control, create an analysis region for particles around the 100nm size range using the ApogeeMix beads (Figure 1D and E). You may see differences in FSC versus SSC between exosomes from different sources, as is seen between BAL exosomes and MDRC-derived exosomes (Figure 1D and 1E). The differences may result from the purity and quality of the exosome isolation. Each marker that was included in the panel will be analyzed from the exosome analysis region. For example, the percentages of CD63⁺ exosomes (Figure 1G), TSG101⁺ exosomes (Figure 1H), and CD81⁺ exosomes (Figure 1I) can be gleaned from the representative histograms. Lastly batch analysis and multi-marker analyses can be done at the sample level, yielding useful characteristic information regarding the exosomes (Figure 1J and 1K).

4. Exosome characterization by flow cytometry

4.1. Sample preparation for flow cytometry (Symphony A5 or LSR II)

Similar to the sample preparation outlined for ImageStream analysis, quantitation of the exosome preparations prior to staining with antibodies is recommended to preserve limited samples. Quantitation can be done with any nanoparticle tracking platform, such as the NanoSight (details in Section 3.1). Sample preparation for flow cytometry takes approximately 1 hour, but may vary depending on panel complexity and sample number. A minimum of 1×10⁶ exosomes is recommended for flow cytometry analysis. First, transfer approximately 1×10⁶ exosomes to a new polystyrene tube and adjust the volume to 100μl with PBS that does not contain Mg²⁺ or Ca²⁺. Aliquot exosomes for unstained controls in the same manner as the sample. If multiple exosome samples are being analyzed, the sample exosomes can be pooled for the unstained control to preserve limiting exosome samples.

Single color controls are prepared using the Invitrogen UltraComp eBeads Compensation Beads (Cat# 01–2222-41). The Invitrogen UltraComp eBeads are provided in a dropper and therefore cannot be pipetted. One drop of the UltraComp eBeads is sufficient for preparing each single color control. Prior to dispensing UltraComp eBeads, vortex the tube to ensure homogenous bead content. Add one drop of UltraComp eBeads to a new polystyrene tube and add 1μl of a single fluorescently conjugated antibody for the single color control. Repeat the steps for making single color controls for each antibody used in the analysis panel. The amount of antibodies to add should be pre-determined by titration, however, 1–2μl can be used. A final antibody only control is prepared by adding 1μl of each antibody used in the panel to 100μl of PBS to determine if free floating antibodies are interfering with the acquisition of sample on the flow cytometer.

Lastly, the exosome samples are labeled with the antibodies used in the analysis panel. For the purpose of this paper, the panel used for the BAL exosomes on the Symphony A5 is: CD63 eFluor450 (eBioscience, Lot: 1946493, Clone: H5C6, Cat# 48–0639-41), CD81 PE-Cy7 (BioLegend, Lot: B276513, Clone: 5A6, Cat# 349511), TSG101 FITC (Novus, Lot: 43353–040919-F, Clone: 4A10, Cat# NB200–112F), HLA-DR APC (eBioscience, Lot: 1994140, Clone: LN3, Cat# 17–9956-42), CD11b APC-Cy7 (BD Biosciences, Lot: 8277993, Clone: ICRF44, Cat# 557754), CD16 PE (eBioscience, Lot: 4303330, Clone: eBioCB16, Cat# 12–0168-42), CD66b PerCP-Cy5.5 (BioLegend, Lot: B260775, Clone: G10F5, Cat# 305108), and EpCAM Alexa Fluor 549 (BioLegend, Lot: B261247, Clone: 9C4, Cat# 324228). Samples and single color controls are labeled with antibodies in the dark for 30 minutes on ice. No washing or re-isolation step is required as the antibody alone control will be used to confirm that free floating antibodies do not interfere with the data acquisition. Keep stained samples, unstained control, and single color controls in the dark on ice.

4.2. Symphony A5 sample acquisition

Acquisition of exosome data on the Symphony A5 or similar flow cytometer is similar to the ImageStream. To acquire exosome data on the Symphony A5, it is important to determine where on the forward-scatter (FSC-A) vs side-scatter (SSC-A) plot the exosomes will be present. To do this, use beads of known size and fluorescence, such as the ApogeeMix beads, to determine the initial gate for exosomes. Transfer 500ul of ApogeeMix beads to a clean polystyrene tube for acquisition on the Symphony A5. ApogeeMix beads have fluorescent 110nm latex beads (FITC) which can be used to determine the initial acquisition gate. Load the ApogeeMix beads onto the flow cytometer and adjust the photomultiplier tube (PMT) voltages for FSC and SSC until you obtain optimal resolution of the different bead sizes present in the ApogeeMix beads. A dot plot of SSC-A vs FSC-A should yield discrete populations representing each of the different sized particles contained in the ApogeeMix beads (Figure 2A). Similar results can be obtained on a BD LSR II however not to the same resolution as seen in the Symphony (Figure 2A and 2B). Additionally, adjust the thresholding on both the SSC and FSC to exclude as much of the electronic noise and background. Although we use ApogeeMix beads, which contains both latex and silica beads, the beads that are in the size range of exosomes are latex. As mentioned in the earlier section (3.2), recent studies suggest that silica beads have refractive indices that closely resemble vesicles, and are recommended over latex [21, 22].

Figure 2 –

Exosome analysis on the Symphony A5. (A) FSC-A versus SSC-A of ApogeeMix beads on the Symphony A5 illustrating the resolution of the different bead sizes contained in the ApogeeMix. (B) FSC-A versus SSC-A of ApogeeMix beads on the BD LSR II illustrating that smaller beads can be resolved but not to the same degree as seen with the Symphony A5 (A). (C) Representative FSC-A versus SSC-A plot of exosomes purified from human BAL. (D) Representative FSC-A versus SSC-A plot of MDRC-derived exosomes. (E) Representative FSC-A versus SSC-A plot of antibodies alone in PBS as a negative control. (F-G, I-J) Representative plots of BAL exosomes demonstrating the type of gating and analysis that can be performed on multiple parameters. (H) Quantitation of percent of BAL exosomes positive for each of the 8 markers used in the sample acquisition panel (TSG101, CD63, CD81, HLA-DR, CD16, CD66b, CD11b, EpCAM). (K) Multiparameter analysis showing percent of BAL exosomes positive for a combination of the 8 markers used in the sample acquisition panel.

The 110nm beads can be determined by plotting FITC vs side-scatter and observing where the fluorescent signal falls on the SSC-A vs FSC-A plot. Once the location of the 110nm beads have been determined, create a gate around it which will be used as the acquisition gate. Acquire 10,000 events using the acquisition gate as a stopping gate for the ApogeeMix beads. The ApogeeMix beads data will be used to determine the analysis gates later on.

After creating an acquisition gate for the exosomes, load the unstained control to adjust the laser power to optimal levels. Create a “n × n” matrix for each PMT being acquired and adjust the PMT voltages such that the signal falls between the origin and the second decade (10²). After adjusting the laser, load the single color controls to verify that adequate signal and resolution can be acquired and do the same for the samples. After completing the adjustment of the PMTs, perform an automated compensation using the DIVA software and the single color controls.

Once the compensation has completed, acquire a minimum of 100,000 events for each sample using the exosome gate created earlier as the stopping gate for acquisition. Both BAL exosomes and MDRC-derived exosomes have a “wispy” characteristic to the forward- and side-scatter properties (Figure 2C and 2D). Lastly, an antibody only control is also acquired to verify that free floating antibodies do not interfere with sample acquisition (Figure 2E).

4.3. Analysis of flow cytometry data on FlowJo

Analysis of flow cytometry data is performed on FlowJo or other analysis programs that can read FSC files. Time required for analysis is approximately 30–60min per sample but may vary depending on analysis or panel complexity. Once the sample data files are loaded on FlowJo or a similar platform, confirm that the acquisition defined compensation is correct and adjust for fluorescent spill over as needed. Manual adjustments can be made on FlowJo.

The first step in analysis is determining the 100nm-500nm bead range using the ApogeeMix bead data that was acquired. This initial gate will include 100–500nm particles as the resolution between these sizes on the A5 is more challenging as compared to the ImageStream. This data can be used to create a gate that will encompass the exosomes that will be characterized (Figure 2C and 2D). Once the exosome gate has been created, unstained samples are used to determine where the negative signal ends (Figure 2I) in order to aid in proper gating for positive signals (Figure 2J). Single markers can be analyzed using histograms, or two markers can be analyzed simultaneously using a dot plot as seen in Figure 2F–G and 2I–J. The percentages for each marker can be calculated and graphed as seen in Figure 2H and 2K, yielding useful characteristic.

5. Sorting of Exosomes on the BD Aria

Sorting exosomes can open new avenues to understand exosome biology, such as observing functional differences and studying their cargo. Sorting of exosomes can be very challenging and validation experiments are required to determine successful sorting. However, when successful, it provides a rapid way to categorize exosomes based on fluorescence markers. The labeling of exosomes for sorting is the same as for flow, with the only difference being the sample tubes should contain as many exosomes as possible. Sorting was conducted in 1.5ml tubes with the caps removed. The 1.5ml tubes were placed in the 5ml polystyrene FACS tube holders for the Aria. The FACS data obtained was comparable to that of the LSRII or A5 (Figure 3) but a major challenge was obtaining sufficient sort events to conduct meaningful downstream experiments. Our initial experiments prove that sorting exosomes is feasible albeit further optimization and validation is warranted.

Figure 3 –

Sorting of exosomes on the BD Aria. BAL exosomes were labeled with HLA-DR APC and a mitochondrial dye MitoTracker Green. Data acquired on BD Aria using a 70-micron flow cell at a flow rate of 1.0. Acquisition thresholding was set at 1,000 on both SSC and FSC. Exosomes were sorted into 1.5ml tubes placed in the 5ml polystyrene tube holders.

6. Conclusions

Exosomes are natural couriers of biomolecules containing cargo from the originating cell. Robust characterization of exosomes can provide insight on the state of the exosome-generating cell, and the contribution of exosome-generating cell to disease state. The use of rapid methods of characterization opens up translational opportunities to quickly assess disease progression and prognosis in a minimally invasive manner (such as liquid biopsy and phlebotomy) while obtaining results quickly.

The two methods of characterization described in this paper offer both advantages and disadvantages. Both methods allow for rapid characterization of surface markers of exosome and proteins using fluorescently conjugated antibodies. We describe the use of ApogeeMix beads and antibody alone controls to accurately acquire exosomes on both platforms. The use of ApogeeMix beads helps identify the region of the forwards/size versus side-scatter where exosome-sized particles should be. However, recent reports indicate that silica beads have refractive indices that better represent vesicles such as exosomes, as compared to latex beads that were used in our methods. Additionally, the use of antibodies along in PBS helps confirm that free floating antibodies do not interfere with the acquisition process.

Furthermore, the use of non-exosome markers such as Grp94 and Arf6 in the analysis panel is strongly recommended, especially for exosomes that are isolated using the ultracentrifugation method [7]. The ultracentrifugation method may pellet extracellular vesicles that are not endosomally-derived and large protein aggregates. Grp94 is a marker for the endoplasmic reticulum [23] and Arf6 is a plasma membrane protein found on microvesicles [24], which are neither found on exosomes and can therefore be reliable used to exclude non-exosomes from the characterization process. When characterizing exosomes using flow cytometry or ImageStream, the use of exosomal markers is required to exclude vesicles devoid of exosomal markers and when combined with non-exosomal markers, such as Grp94 and Arf6, and provide additional control measures. Additionally, fluorescent dyes such as lipophilic membrane dyes and cationic dyes can be used in addition to antibodies to obtain additional phenotypic information. Although differential centrifugation and precipitation methods [18] for exosome isolation are sufficient for characterizing exosomes by flow cytometry and ImageStream, if purer and higher quality preparations of exosomes are desired, alternative methods such as cushioned-density gradient centrifugation methods are available [16]. Each method has its pros and cons such as quality, purity, concentration, resource availability, time and cost, thus these factors should be evaluated prior to an experiment to determine the optimal isolation method.

The ImageStream strategy described in this paper has been extensively validated by other investigators demonstrating feasibility and reliability of imaging flow cytometry-based characterization methods [19, 25, 26]. The ImageStream does have a limitation on the number of fluorochromes that can be used compared to the Symphony A5. We have used 5 colors representing exosome and cellular markers on the ImageStream without major complications with fluorescent spill over. The BD Symphony platform is capable of acquiring 23–32 colors depending on the configuration of the cytometer. The Symphony A5 used in this paper is capable of acquiring up to 31 colors enabling a large panel of cellular and exosome markers, as well as lipophilic and cationic dyes to characterize other phenotypic details, such as mitochondrial membrane potential [9].

In this paper we have used a total of 8 markers including exosome markers (CD63, CD81, TSG101) and cellular markers (HLA-DR, CD11b, CD16, CD66b). HLA-DR is an antigen presentation marker important in CD4⁺ T lymphocyte activation and has been found significantly enriched in exosomes from asthmatics [6–9]. CD11b is a pan-macrophage, monocyte, and neutrophil marker [27, 28], CD16 is a marker that helps further classify monocytes into their phenotypic subsets [29], and CD66b is an adhesion marker that is expressed exclusively on granulocytes, such as eosinophils [30].

Characterization of exosomes and other extracellular vesicles by conventional flow cytometry has been generally described by using capturing beads [31, 32]. However, in this paper, we describe the direct analysis of exosomes on the Symphony A5 or similar flow cytometer without the need for a dedicated cytometer for nanoparticle analysis or the use of capture beads. A major strength of the Symphony platform, is the ability to design large panels for rapid characterization exosomes that include an extensive set of both clinically and biologically relevant markers. For example, the 31 channel Symphony A5 is able to characterize exosome markers (eg., CD63, CD81, TSG101), non-exosome markers (eg., Grp94 and Arf6), and cellular markers (eg., CD11b, CD16, CD66b, HLA-DR, etc.) to determine origin of the BAL exosomes while probing for mitochondria using cationic dyes, such as MitoTracker or TMRE. Such an analysis can provide a wealth of phenotypic information.

Both methods pose challenges as the resolution limit of the underlying optics is being pushed to the limit. In particular, it is possible to acquire false signal that may appear as exosomes. Furthermore, with conventional flow cytometry, it may be challenging to determine if the events are single exosome events or an aggregate due to swarm effect, as discussed by Van Der Pol, et al [21]. Another challenge is the possibility of steric hindrance to prevent antibody binding to the target protein on the exosome. Although we have been able to successfully capture signal for multiple different protein targets on the exosome using fluorescently conjugated antibodies, because of the limited surface availability on exosomes, steric hindrance may prevent the detection of target proteins that are in lower abundance due to binding of antibodies to more abundant targets. These limitations must be considered when performing characterization experiments of exosomes using flow cytometers.

Lastly, designing of an antibody/fluorochrome panel that is compatible with the cytometer and performing fluorescence compensation may be challenging at first for users who are not familiar with flow cytometry. However, despite these challenges, both cytometry platforms prove to be robust methods for multiparameter characterization of exosomes and other extracellular vesicles in a rapid manner.

Highlights.

• Rapid multiparameter analysis of airway exosomes by imaging flow cytometry and conventional flow cytometry methods using ImageStream Amnis and Symphony A5.

• Standardization of exosome analyses using ApogeeMix beads and evaluation of percent exosomes contributed by various airway cell populations by multiparameter analyses.

• Exosome isolation from bronchoalveolar lavage fluid using differential centrifugation.

• Airway myeloid-derived regulatory cell (MDRC)-derived exosomes were isolated form conditioned media by precipitation using exosome isolation kit.