Exosome Isolation by Ultracentrifugation and Precipitation: A Comparison of Techniques for Downstream Analyses

Abstract

Exosomes are 50-150 nm diameter extracellular vesicles secreted by all mammalian cell, except mature red blood cells, and contribute to diverse physiological and pathological functions within the body. Many methods have been used to isolate and analyze exosomes, resulting in inconsistencies across experiments and raising questions about how to compare results obtained using different approaches. Questions have also been raised regarding the purity of the various preparations with regard to the sizes and types of vesicles, and to the presence of lipoproteins. Thus, investigators often find it challenging to identify the optimal exosome isolation protocol for their experimental needs. Our laboratories have compared ultracentrifugation and commercial precipitation and column-based exosome isolation kits for exosome preparation. Here, we present our study of exosomes isolated using two of the most commonly-used methods, ultracentrifugation and precipitation, followed by downstream analyses.. We used NanoSight nanoparticle tracking analysis (NTA) and Flow cytometry (Cytek^® ) to determine exosome concentrations and size. Imaging flow cytometry can be utilized to both size and immunophenotype surface markers on exosomes (ImageStream). High-performance liquid chromatography (HPLC) followed by nano-flow liquid chromatography mass spectrometry (LCMS) of the exosome fractions can be used to determine the presence of lipoproteins, with LCMS able to provide a proteomic profile of the exosome preparations. We found that the precipitation method was six times faster, and resulted in ~2.5 fold higher concentration of exosomes per ml compared to ultracentrifugation. Both methods yielded EVs in the size range of exosomes and both preparations included apoproteins.

INTRODUCTION:

Exosomes are small (50-150 nm diameter) extracellular vesicles secreted by every type of mammalian cell, except mature red blood cells, under both normal physiological and pathological conditions. The term “exosome” has a long and varied history in the biological sciences (Graner, 2019; Witwer & Thery, 2019). A current biological definition is that exosomes are 30-150 nm diameter membrane-enclosed vesicles that are formed in the endosomal system beginning when late endosomes invaginate their limiting membranes to produce intralumenal vesicles (ILVs or intravesicular bodies/IVBs) within the endosomal lumen to form an organelle called a multivesicular body (MVB), which may then fuse with the plasma membrane and release its ILVs outside the cell. When the ILVs are released outside of the cell, they are then called “exosomes” (Graner, 2019; Latifkar, Hur, Sanchez, Cerione, & Antonyak, 2019; Yanez-Mo et al., 2015). Other types of extracellular vesicles include those released directly from the cell surface, which have been given other names (e.g., ectosomes, shed vesicles, microvesicles, microparticles, etc.) (Graner, 2018, 2019). The historical distinction between these different vesicular entities was generally size, e.g., 30-150 nm diameter vesicles were called “exosomes,” whereas vesicles that were greater than 200 nm in diameter were named microvesicles. The size differences between these vesicles resulted in differential sedimentation by centrifugation (e.g., often 10,000 xg vs 100,000 xg) (Gould & Raposo, 2013). However, we now know that in addition to exosomes, other vesicles with a similar smaller size range are likely released directly from the cell surface, and we have no clear, specific, and agreed-upon markers to distinguish the origins of these different vesicles (Kowal et al., 2016).

Guidelines exist regarding nomenclature, separation methods, and characterization of vesicle types (Thery et al., 2018), and the general term “extracellular vesicle” or EV covers vesicles of both types of cellular origin and diameter without specifying origin. However, that same document (Thery et al., 2018) recognizes the vagaries of EV terminology, and suggests that authors may utilize whatever nomenclature they wish as long as it is clearly defined. Thus, vesicle size/diameter is still often used as the defining feature when naming a vesicle class such as exosomes (Pegtel & Gould, 2019), and we will continue to use that terminology here.

Exosomes have been identified in at least 20 different biofluids (http://microvesicles.org/#) including serum/plasma, and of course, cell culture medium. Their presence in the blood means that exosomes can travel anywhere in the body, yet what they carry from cell to cell and from organ to organ, their exact roles in normal physiology versus in disease pathology, and how they may change in response to altered physiological demands, remain largely unknown. The roles of exosomes to date have been debated as facilitating everything from delivery and communication to the clearance and/or seeding of toxic proteins associated with disease. Thus, research that aims to identify exosomal cargo and to delineate the roles of exosomes in healthy individuals versus individuals with a particular disease, is timely and essential. Knowledge of the specific cell types that serve as sources of the exosomes is also of critical importance because it will determine the types of cargo loaded into the exosomes and the potential role(s) of those cell types in normal physiology and/or disease. These cell state-specific markers, in turn, may serve roles as biomarkers of disease and/or as tools for measuring response to treatment(s).

Many methods have been developed for isolating and characterizing exosomes from cells and body fluids. Here, we describe our optimized plasma preparation, our comparison of the use of ultracentrifugation versus a precipitation-based method for the preparation of exosomes, and our downstream analyses to determine the relative lipoprotein contamination of both preparations using high-performance liquid chromatography (HPLC) and liquid chromatography mass spectrometry (LCMS).

BASIC PROTOCOL 1 Pre-Analytic Fluid Collection and Human Plasma Processing Protocol

Introduction:

The University of Colorado Alzheimer’s and Cognition Center (CUACC) Laboratory obtains whole blood samples from participants in several ongoing clinical observational studies and therapeutic trials. Whole blood samples that are collected into anti-coagulation vacutainers (i.e., EDTA or heparin) are routinely processed within two hours of phlebotomy into plasma, buffy coat, and red blood cell components, which are then immediately placed into the CUACC Human Biorepository for storage in our temperature-monitored freezers until needed for various analyses. We aliquot these blood components into cryotubes with 2D barcodes permanently etched on the bottoms, which can be quickly scanned and entered into our freezer management software (FreezerPro^®, Brooks Automation, Inc.), in order to manage each tube’s exact freezer location, and so that retrieval of samples rapidly occurs without threat of thawing and affecting the integrity of subsequent biological analyses. Although the following plasma processing protocol was used herein for the exosome ultracentrifugation and precipitation protocols and downstream analyses, numerous other types of human and animal biological fluids containing exosomes could also have been used, such as heparin-derived plasma, serum, cerebrospinal fluid, urine, cell culture supernatants, and others.

Materials:

BD Vacutainer® EDTA Tubes, Reference #367861

1.5 ml Eppendorf Safe-Lock Tubes, Catalog No. 022363204

Micronic Grey Low Profile Screw Caps, 0.75 ml Internal Thread, MP53320

Micronic 0.75 ml 2D Data-Matrix Coded Screw Cap Tubes V-Bottom Bulk, MP52349

Protocol:

1. In the clinic, collect 4 ml whole blood in lavender top BD Vacutainer^® tubes that contain the anti-coagulant EDTA.

2. Centrifuge the blood for 15 min at 1,500 RCF (xg) at room temperature (22-25°C) in a Beckman Coulter Allegra X-30R centrifuge.

3. For all of the following steps, treat all human blood samples as though they contain blood-borne pathogens, and use biosafety cabinets, closed centrifuges, and proper BSL2 laboratory procedures. In a Class II biosafety cabinet, carefully remove the 1.5 ml of plasma from the top layer of each EDTA Vacutainer^® tube, using sterile filtered polypropylene pipet tips, and place it into 1.5 ml Eppendorf tube. To prevent contamination of the plasma with white blood cells and platelets, be careful not to disrupt the buffy coat by drawing the plasma from each vacutainer in 750 μl increments for a total of 1.5 ml.

4. To reduce possible contamination with white blood cells and platelets, recentrifuge the plasmas for 10 more min at 2,200 RCF (xg) at 4°C in an Eppendorf 5417R centrifuge. This centrifugation step may result in a small pellet of residual cells and platelets at the bottom of the tubes.

5. Distribute the plasma in 6 x 250 μl aliquots into Micronic tubes.

6. Store plasma samples immediately at −80°C.

BASIC PROTOCOL 2 Exosome Isolation by Ultracentrifugation

Introduction:

Ultracentrifugation has been the gold standard for exosome isolation. The major benefit of this state-of-the-art method is that it produces highly enriched EV fractions while also allowing for the collection of additional vesicle fractions (e.g., larger vesicles, which are pelleted first in a lower speed spin at 10,000 xg, and then an EV-free supernatant, which is generated after the high speed spin, as outlined below). Disadvantages of ultracentrifugation include that it is low-throughput (depending on the material used) and that it requires specific infrastructure (i.e., an ultracentrifuge) as well as specific expertise to be performed correctly. The following protocol describes how to isolate EVs from plasma. The protocol is performed in a biosafety cabinet. To generate sterile preparations, the buckets for the SW60 rotor are autoclaved before use, and the bottles and tubes are sterilized with 10% bleach, 70% ethanol, and PBS rinses.

Materials:

Reagents:

Human plasma sample (prepared according to Basic Protocol 1), ice, and 1X PBS (Thermo-Fisher, ref 14190144).

Hardware/Instruments:

Microcentrifuge (such as Sorvall™ Legend™ Micro 21), L-80 Preparative Ultracentrifuge (Beckman Coulter) or equivalent, Type 70.1 rotor (Beckman Coulter, ref 342184, k-factor 42), SW60 rotor (Beckman Coulter, ref 335650, k-factor 31), polycarbonate bottles (Beckman Coulter, ref 355603), polyallomer conical tubes (Beckman Coulter, ref 335650), adapters for SW60 (Beckman Coulter, ref 335650). Pipettes and pipette tips (P1000 and P200).

Procedure:

1. Centrifuge 250 μl of plasma sample (prepared according to Basic Protocol 1) for 10 min at 300 xg, 4°C.

2. Collect the supernatant and transfer it to polycarbonate bottle for use in the Type 70.1 rotor.

3. Fill each bottle with cold 1X PBS and equalize the masses of the bottles. Label the bottles and then place them in the rotor so that the label faces the outside of the rotor and allows for easy visualization of the vesicle pellet.

4. Spin in the preparative ultracentrifuge for 30 min at 10,000 xg, 4°C. Set maximum acceleration and no brake for the deceleration phase. Deceleration should be as slow as possible so as not to disturb the pellet. This setup should be used for all the ultracentrifugation steps

5. Transfer the supernatant to new ultracentrifuge bottles adapted to the Type 70.1 rotor. The pellet (10K-Pellet) can be re-suspended with 1X PBS and kept at -80°C.

6. Equalize the masses of the ultracentrifuge bottles with cold 1X PBS and label them as described in step 3.

7. Spin in the preparative ultracentrifuge for 2 h at 100,000 xg, 4°C .

8. Resuspend the pellet with PBS and transfer the sample to a polyallomer conical tubes adapted for the SW60 rotor. Special adapters need to be in place in order to use conical tubes in the SW60 rotor buckets. Instead of being discarded, the supernatant (EV-free fraction) can be saved and stored at −80°C for later use as a negative control for the EV fraction.

9. Centrifuge 2 h at 100,000 xg, 4°C, to wash the EV pellet.

10. Discard the supernatant and resuspend the pellet (100K Pellet) in 150 μl cold 1X PBS. Transfer the 100K Pellet into 1.5 ml microtubes and store at −80°C. Aliquoting of the samples is recommended to avoid freeze-thaw cycles of the EVs.

ALTERNATE PROTOCOL 1 Exosome Isolation by Precipitation

Introduction:

Although phlebotomy is considered a minimally invasive procedure, blood samples from carefully phenotyped research participants remain a limited resource. Therefore, selecting the type of method used to isolate exosomes should be considered carefully (Macias et al., 2019). Ensuring that the preparation provides a high yield of exosomes from the smallest amount of plasma sample as possible is one of the most important criteria, although the ease and speed of preparation is also paramount when considering the use of exosomes as biomarkers or diagnostics in clinical settings. We previously tested many precipitation kits and focused on their ability to: 1) provide a high yield, and 2) provide vesicles with a size distribution that was exclusively in the range for exosomes. Based on these criteria, we selected the ExoQuick^® ULTRA EV Isolation Kit for Serum and Plasma (Systems Biosciences) precipitation method (Fig. 1, Alternative Protocol 1). In the first step, an exosome pellet is generated. Another deciding factor was that this kit also includes a second step after the initial precipitation and resuspension of the exosome pellet. The second step uses a “purification column” to reduce the levels of two main contaminants — IgG and albumin — in the initial resuspended exosome pellet. Herein, we refer to the initial resuspended exosome pellet as the “PRE” exosome sample, and we refer to the cleaned up exosomes as the “POST” exosome sample. Although the purification column did not remove all of the IgGs or albumin (data not shown), this step significantly reduced the presence of both, in agreement with data shown in the product brochure. The small remaining amounts of albumin and IgG detected may be bound to the exosomes. Our decision to use the ExoQuick^® ULTRA EV Isolation Kit for Serum and Plasma resulted from our testing many kits for extracellular vesicle preparation and comparing their relative enrichment of exosomes based on NTA assessment of size and concentration. This, along with a review of the literature for studies performing similar comparisons, (for example Macias and colleagues, who performed an analysis of six different commercial kits) (Macias et al., 2019) formed the basis of our decision to use the ExoQuick^® ULTRA precipitation approach. This choice was also based on the ExoQuick^® ULTRA kit providing investigators the ability to examine an exosome preparation from which albumin and IgG have been depleted, allowing researchers the ability to answer questions regarding the physiological and pathological roles of exosomes in a biochemically cleaner manner. Our materials and methods essentially follow those in the user manual provided with the kit.

Figure 1. Schematic of the workflow used to isolate and analyze exosomes from plasma.

Schematic of the workflow used to isolate and analyze exosomes from plasma. Initial blood sample were processed to generate plasma samples (Basic Protocol 1), which were used to isolate exosomes using either ultracentrifugation (Basic Protocol 2 ) or precipitation (Alternate Protocol 1). Isolated exosomes were then characterized at a single cell resolution by nanoparticle tracking analysis (Basic Protocol 3) or using flow cytometry and imaging flow cytometry (Alternate Protocol 2). Deep profiling was also carried out by high resolution liquid chromatography (Basic Protocol 4) and nano-flow liquid chromatography mass spectrometry (Basic Protocol 5). Overall expected time is indicated for each described protocol. Abbreviation: SN : supernatant.

Materials:

ExoQuick^® ULTRA EV Isolation Kit for Serum and Plasma (EQULTRA-20A-1, Systems Biosciences). This kit includes the following components:

• ExoQuick

• Purification Columns

• Buffer B

• Buffer A

• 2 ml Eppendorf Tubes

• Collection Tubes

Hardware and Instruments:

EQULTRA-20A-1 kit from Systems Biosciences

Human plasma sample (prepared according to Basic Protocol 1)

Ice bucket

Eppendorf 5417R Centrifuge or equivalent

Pipettes and pipette tips (P200 and P1000)

Protocol:

Exosome Precipitation and Preparation of the PRE Exosome Sample

1. Pipette 250 μl human plasma (prepared according to Basic Protocol 1) into an eppendorf tube.

2. Add 67 μl of ExoQuick precipitation reagent (provided in kit) to each 250 μl of plasma.

3. Mix well by flicking and then, to precipitate the exosomes, incubate the plasma/precipitation reagent mixture on ice for 30 min.

4. To isolate the exosomes, centrifuge at 4°C at 3,000 rpm for 10 min in an Eppendorf 5417R Centrifuge. A beige/white pellet of exosomes and a clear supernatant will be evident post-centrifugation.

5. Remove the supernatant and resuspend the exosome pellet in 200 μl Buffer B (provided with the kit).

6. This is the PRE exosome sample.

Preparation of the POST Exosome Sample

• 7.Add 200 μl Buffer A to the resuspended PRE exosomes described in step 6.
• 8.Prepare the purification columns by snapping off the bottom closures, loosening the caps, and placing the columns into the collection tubes provided with the kit.
• 9.Centrifuge in the Eppendorf 5417R centrifuge at room temperature for 30 sec at 1,000 xg to remove the storage buffer from the columns.
• 10.Discard the storage buffer and place the column back in the collection tube.
• 11.Wash the column by adding 500 μl Buffer B to the column and centrifuging it in the Eppendorf 5417R centrifuge at room temperature for 30 sec at 1,000 xg.
• 12.Wash the column, as described in step 11, one more time.
• 13.Plug the bottom of the column with the bottom closure and add 100 μl Buffer B to the top of the column to prepare it for sample loading.
• 14.Add the exosome mixture prepared in step 7 to the top of the column, recap the tube, and incubate on a rotating shaker for no more than 5 min.

Exosome Sample Elution

• 15.Loosen the cap on the top of the column, remove the bottom closure, and transfer the column to a 2 ml Eppendorf tube provided with the kit.
• 16.Centrifuge at 1,000 xg for 30 sec. The resulting eluate is the POST exosome sample.
• 17.Discard the column.
• 18.Determine the exosome concentration in the PRE exosome sample using NanoSight NTA (Basic Protocol 3).

BASIC PROTOCOL 3 Analysis of Exosomes by Nanoparticle Tracking Analysis (NTA)

Introduction:

NanoSight Nanoparticle Tracking Analysis (NTA) (NanoSight NS300; Malvern Panalytical, Malvern WR14 1XZ, UK) is the most common method utilized in the field for assessing the concentrations and sizes of EVs in solution. This method relies on measuring the characteristic movement of exosomes (and nanoparticles in general) in solution, based on Brownian motion (Bachurski et al., 2019). It then determines information about the particles in solution by capturing scattered light post-illumination with a laser, and uses the Stokes-Einstein equation and cell volume information to determine concentrations and sizes of vesicles. While NTA is deemed reasonably accurate for determining size when particles are greater than 50 nm in diameter, its ability to measure concentrations is less reliable (Bachurski et al., 2019). Hence, researchers are developing alternative approaches to more accurately assess the concentrations of exosomes versus other EVs based not only on size, but also on additional measures, such as surface markers (Alternate Protocol 2).

Materials:

NanoSight instrument (NS300, Software Version: NTA 3.2 Dev Build 3.2.16; Malvern Panalytical, Malvern WR14 1XZ, UK)

ddH₂O (0.1 μm filtered)

1.5 ml Eppendorf tubes

Protocol:

1. Extracellular vesicle sizes and concentrations were determined using the Nanosight nanoparticle tracking analysis (NTA) instrument (NS300, Software Version: NTA 3.2 Dev Build 3.2.16; Malvern Panalytical, Malvern WR14 1XZ, UK)

2. One ml of exosome sample prepared by ultracentrifugation (Basic Protocol 2) or by precipitation (Alternate Protocol 1, PRE exosome sample) was prepared by freshly diluting the exosome sample 1:1,000 and 1:10,000 in ddH₂O (0.1 μm filtered) before reading.

3. Exosome number was captured using the following analytical settings on the NanoSight NS300: camera type: sCMOS, level 11 (NTA 3.0 levels); green laser; slide setting and gain: 600, 300; shutter: 15 ms; histogram lower limit: 0; upper limit: 8235; frame rate: 24.9825 fps; syringe pump speed: 25 arbitrary units; detection threshold: 7; max jump mode: auto; max jump distance: 15.0482; blur and minimum track length: auto; first frame: 0; total frames analyzed: 749; temp: 21.099-22.015 °C; and viscosity: 1.05 cP.

ALTERNATE PROTOCOL 2 Analysis of Exosomes by Flow Cytometry and Imaging Flow Cytometry

Introduction

As discussed in Basic Protocol 3, the ability of NanoSight NTA to accurately measure the sizes and concentrations of exosomes has limitations, inducing researchers to develop alternative approaches to more accurately assess these measures. In this Alternate Protocol 2, we describe one such approach, using a Cytek^® and an Imagestream to quantify the amount of exosomes (particle number per volume) present in the sample and to assess exosomal surface markers, respectively. Once exosomes have been isolated, the following protocol is an option for identification of the surface markers present on the exosomes using exosome fluorescent-based detection by flow cytometry. Additionally, imaging flow cytometry can be utilized to immunophenotype surface markers on exosomes (ImageStream). Quantification by flow cytometry, described below, allows for single-particle quantification as opposed to bulk quantification, such as that typically performed with NanoSight NTA (Basic Protocol 3). Here, we chose SBI Exo-Flow-One for its simple easy-to-use approach. Exo-Flow-ONE targets specific membrane and internal EV components by staining them and achieving near single vesicle resolution with virtually no background binding. The high specificity of Exo-Flow-ONE is far superior to labeling EVs based on the canonical tetraspanins, whose expression is not ubiquitous. Once the quantification is achieved, immunophenotyping of exosomes using imaging flow cytometry allows the user to examine the heterogeneity of tetraspanin expression on the surface of isolated exosomes, which is a gold standard for assessing whether the EVs isolated are in fact exosomes. Additionally, immune cell surface markers or other antibodies of interest could be used to determine the cellular origin of the isolated exosomes.

Materials

1.5 ml Eppendorf tubes

Shaker-incubator set at 37°C

SBI Exo-Flow-ONE Labeling Kit (Cat # EXOF200A-1)

1X PBS pre-filtered with 0.02 μm filter

CD81 PE, clone M38 (Cat # A15781)

CD9 FITC, clone HI9a (Cat # 312103)

CD63 APC, clone H5C6 (Cat # 353007)

Cytek^® Northern Lights spectral cytometer

Amnis^® ImageStream^® XMk II

Note: For quantification, you must always prepare an unstained control in addition to the samples that you will measure.

Quantification Method

1. Resuspend 200-500 μg of exosomes prepared by precipitation (Alternate Protocol 1, PRE exosome sample) in 500 μl of previously filtered 1X PBS.

2. Add 1 μl of Exo-Flow-ONE labeling dye. Keep one sample aside without the labeling dye to function as your unstained control.

3. Vortex and incubate at 37°C with shaking for 20 min.

4. Make 10-fold serial dilutions in previously filtered 1X PBS (1:10 to 1:100,000).

5. Keep samples on ice until analysis.

6. Once at the Cytek^® flow cytometry instrument, first run the unstained control (Fig. 2a), set a threshold for autofluorescence, and then run the labeled/stained samples (Fig. 2b).

7. To generate an accurate dilution/exosome quantity curve, the entire 1 ml sample volume should be run.

8. Plot the total number of events against the sample dilution factor (Fig. 3).

Figure 2. Plasma exosome quantification with the Exo-Flow-ONE staining kit.

A PRE exosome samples (Alternate Protocol 1) were stained with Exo-Flow ONE. a) Unstained exosome sample, showing no spectral signature for the detectors. b) Exosome sample stained with Exo-Flow-ONE garnet far red showing a spectral signature in channel R2 consistent with the excitation/emission 628/643 nm. Data from a representative PRE exosome sample is shown.

Figure 3. Quantification of plasma exosomes using the Exo-Flow-ONE staining kit.

PRE exosome samples (Alternate Protocol 1) were serially diluted as shown above, and each diluted sample was stained with Exo-Flow ONE. Exosome count (read as event count on the flow cytometer) is plotted against the corresponding dilution factor demonstrating a linear correlation between the count and the dilution factor. Data from a representative PRE exosome sample is shown.

To assess accuracy of this quantification method, the quantity/count vs. dilution curve should have an R² value above 0.9. Once this step has been achieved, you can be certain that the quantification is accurate. We recommend doing this procedure in random samples and at random times.

Immunophenotyping Method

1. Pipette the desired amount of exosomes prepared by precipitation (Alternate Protocol 1, PRE exosome sample) into 200 μl of previously filtered 1X PBS.

2. Make sufficient tubes to accommodate the combination of tetraspanin antibodies of your choice (e.g., anti-CD9 and anti-CD81).

3. Pipette 2 μl of antibodies into the tubes containing the exosomes. We recommend using at least one anti-tetraspanin antibody in your staining cocktail, in order to identify the exosomes and, per this optimized protocol, up to three antibodies can be utilized.

4. Incubate for 45 min at room temperature in the dark.

5. Prepare to assess the samples using the Amnis^® ImageStream^® XMk II instrument (Fig. 4).

6. Instrument set-up: All lasers (405, 488, 561, 642, and 785 nm) were set to maximum power (mW). The 60X objective was used, and the speed was set to low with high sensitivity. The diameter core was reduced to 6 μm in order to increase the focus needed to detect exosomes (flow speed tab). To enable detection of the 405 nm SSC, the NF box was unchecked next to the 405 nm laser setting (illumination tab).

Figure 4. Immunophenotyping of plasma exosomes using ImageStream.

PRE exosome samples (Alternate Protocol 1) were stained with fluorescently labeled anti-CD9 (FITC), anit-CD81 (PE), and anti-CD63 (APC) antibodies and analyzed via ImageStream. Exosomes that express CD9 (FITC), CD81 (PE), and CD63 (APC) (left to right) are shown. Each row depicts an individual exosome. Data from a representative PRE exosome sample is shown.

Once the instrument is set up, run the unstained control and gate on the particles based on the side scatter. Afterwards, run your labeled samples, which should appear positive for at least one of your tetraspanin markers, and should also be identifiable on the bright field (Fig. 4).

BASIC PROTOCOL 4 Downstream Analysis of Exosomes Using High-Performance Liquid Chromatography (HPLC)

Introduction

A number of reports suggest that lipoproteins may co-purify with exosomes when using standard methodologies to isolate exosomes from plasma (Huang, Banizs, Shi, Klibanov, & He, 2015; Karimi et al., 2018). Because exosomes are secreted vesicles that are in the same size range as lipoproteins (Table 1), their co-purification is not surprising. However, whether lipoproteins and exosomes overlap in composition and function remains to be determined. Here, we fractionated isolated exosomes by high-resolution size exclusion chromatography (SEC) using standard methodologies optimized for the separation of lipoprotein particles by high-performance liquid chromatography (HPLC). The method described below (using Superpose 6 Increase columns) allows for gentle purification of biomolecules in the molecular weight (M_r) range for globular proteins between 5,000 and 5,000,000 Da. Exosomes were fractioned before and after albumin and IgG removal (PRE and POST Exosome Samples, respectively, from Alternate Protocol 1) and exosomes were also prepared using ultracentrifugation (Basic Protocol 2). The cholesterol and proteomic composition of fractions containing possible lipoproteins were further analyzed by liquid chromatography mass spectrometry (LCMS, Basic Protocol 4). Identification of specific apoproteins that associate with a certain lipoprotein class (see Table 1) will be indicative of co-purification of that class of lipoproteins.

Table 1. Lipoprotein Classes.

Adapted from (Feingold & Grunfeld, 2000).

Lipoprotein	Size (nm)	Major Lipids	Major Apoproteins
Exosomes	50-150	Sphingomyelin, Phosphatidylserine, Phosphatidylcholine, Phosphatidylethanolamine (Skotland, Sagini, Sandvig, & Llorente, 2020)	None reported
Chylomicrons	75-1200	Triglycerides	Apo B-48, Apo C, Apo E, Apo A-I, A-II, A-IV
Chylomicron Remnants	30-80	Triglycerides, Cholesterol	Apo B-48, Apo E
VLDL	30-80	Triglycerides	Apo B-100, Apo E, Apo C
IDL	25-35	Triglycerides, Cholesterol	Apo B-100, Apo E, Apo C
LDL	18-25	Cholesterol	Apo B-100
HDL	5- 12	Cholesterol, Phospholipids	Apo A-I, Apo A-II, Apo C, Apo E
Lp (a)	~30	Cholesterol	Apo B-100, Apo (a)

Materials

0.5 ml Eppendorf tubes

Loading Syringe (Hamilton™ 1700 Series Gastight™ Syringes: N Termination)

HPLC buffer (1 mM EDTA, 0.15 mM NaCl, 0.3 mM NaN₃)

Fluorometric cholesterol assay kit (Cayman Chemical Company, Ann Arbor, MI)

Superose 6 Increase columns (30/1000 GL, Millipore-Sigma)

Hardware and Instruments:

Beckman System Gold HPLC machine

32 Karat software

Protocol

1. Turn on the HPLC machine, fraction collector, and UV lamp to measure the protein concentration as the sample is fractionated.

2. Ensure that respective pumps have sufficient column volumes worth of buffer (120 ml for 2 x column volumes).

3. Increase the flow rate to 0.5 ml. Check pump pressure and column integrity.

4. Inject 200 μl of exosomes isolated by precipitation (Alternate Protocol 1, PRE and POST exosome samples) or ultracentrifugation (Basic Protocol 2) onto the column and turn to the load position.

5. Collect fraction one, which will include any particles larger than 80 nm, including exosomes.

6. Collect fractions 31-33 (VLDL, very low density lipoproteins), 39-51 (LDL, low density lipoproteins), 55-66 (HDL, high density lipoproteins), and 67-70 (HDL) to measure cholesterol concentration and determine lipoprotein content.

7. Freeze fractions at −80°C or continue to downstream analyses by NanoFlow Liquid Chromatography Mass Spectrometry (LCMS).

8. Measure cholesterol in each fraction using a commercially available kit (Cayman Chemical Company, Ann Arbor, MI) following procedures outlined in the package insert. Note, for cholesterol-rich VLDL and LDL lipoprotein fractions, a dilution factor of 5-10 is required. Cholesterol quantification will determine contamination with cholesterol-rich peripheral lipoproteins. Note, exosomes and/or central nervous system-derived lipoproteins are cholesterol-poor. Overall, we observe significantly reduced cholesterol content in lipoprotein fractions from POST exosome samples, suggesting reduced contamination with VLDLs and LDLs after the column step.

9. Determine the protein composition of fraction 1, which contains the exosomes, and the other fractions corresponding to each lipoprotein class. Analysis of apoprotein composition in lipoprotein fractions will validate which lipoproteins are carried over into the exosome preparation (Table 1). Overall, we observe significant apoprotein (particularly Apo-A1) carry-over in fraction 1, which contains the exosomes, and in the lipoprotein fractions from both the PRE and POST exosome samples, even following cholesterol depletion. This suggests that the presence of these apoproteins may be biologically and functionally significant and warrants further study.

BASIC PROTOCOL 5 Downstream Analysis of the Exosome Proteome Using Nano-Flow Liquid Chromatography Mass Spectrometry (LCMS)

Introduction:

Liquid chromatography mass spectrometry (LCMS) based proteomics has become an important tool in many areas of clinical and basic research. A wide range of sample types can be used to investigate proteins, peptides, and modified peptides. Appropriately, there are vast numbers of workflows that can be applied to any given project, and methods will therefore need to be specifically tailored to achieve an investigator’s individual goals. For example, a plasma proteomics workflow often includes the removal of high abundance proteins using affinity chromatography as a first step. Conversely, proteins can be enriched from nuclei or mitochondria using step-wise centrifugation, detergents, and/or sucrose gradients. Finally, whole proteomes can be fractionated using ion exchange or high pH chromatography prior to LCMS analysis to achieve improved proteome coverage. The current protocol is focused on a mass spectrometry-based proteomics method that has been designed specifically for exosome analysis and consists of sample preparation followed by data acquisition and data analysis. The method is based on a similar workflow that was published recently (Reisdorph, Michel, Fritz, & Reisdorph, 2018). Overall, proteomic analyses of exosomes will provide a wealth of information to the researcher, including the identification of biologically important cargo proteins as well as potential contaminants. In this study we examined the exosomes post HPLC purification (Basic Protocol 4), examining the fraction that was enriched for exosomes (Fraction 1) and fractions obtained for each lipoprotein class.

In general, steps involved in a proteomics experiment include the following:

1. Protein solubilization - A variety of methods can be used to extract and solubilize proteins, depending on the sample type. Options include subcellular fractionation, precipitation, addition of various detergents, and centrifugation. In some cases, samples can be immediately denatured and digested. In this experiment, we examined the proteins present in the exosome enriched Fraction 1 and fractions for each lipoprotein class, obtained by HPLC (Basic Protocol 4). An acetone-sodium chloride protein precipitation method was utilized to purify and concentrate proteins from other contaminants in these HPLC fractions.

2. Protein digest - Although quantitative LCMS methods exist for intact proteins, such as molecular imaging and profiling (Spraggins et al., 2016), most often proteins are further solubilized and digested with the protease trypsin prior to analysis by mass spectrometry. This step may include alternative choices of proteolytic enzyme and buffer components.

3. Mass spectrometry - There are several classes of mass spectrometers that are used in proteomics such as quadrupole time-of-flight (QTOF), ion trap, ion cyclotron resonance/orbitrap, and matrix assisted laser desorption ionization time-of-flight (MALDI-TOF). The choice of a QTOF instrument coupled with nano-flow liquid chromatography separations is a good candidate for identifying a large number of proteins and peptides from exosome samples to profile their protein composition. Mass spectrometry parameters and method specifics will vary somewhat depending on the instrument used. The method below is appropriate for an Agilent 6550 QTOF equipped with a nano ESI source.

4. Data analysis- Protein sequence databases are queried using a protein database search program to match spectra to protein sequences in databases resulting in peptide and protein identification within a sample.

Materials:

Water, Burdick and Jackson HPLC grade (VWR, BJ365-4)

Sodium chloride (NaCl) (Fisher, S271-500)

Acetone, HPLC grade (Fisher, A929-1)

Acetonitrile, Optima™ (Fisher, A955-4)

1 ml ampules of formic acid, 99+% (Pierce, 28905)

Low retention 1.5 ml microfuge tubes (Fisher, 02-681-320)

Low retention 0.6 ml microfuge tubes (Fisher, 02-681-311)

1000, 200 and 20 μl pipettes

1000, 200 and 20 μl low-retention pipette tips (USA Scientific: 1182-1730, 1180-8710, 1180-1710)

BCA Protein Assay Kit (Pierce, 23225); Store at Room temperature

Preomics Sample Preparation Kit (8 reactions) (iST 8x kit; Preomics 00001); store everything at room temperature except where noted

• Lysis Buffer (brown cap)

• Protease re-suspension buffer (yellow cap)

• Trypsin/Lys-C mix (red caps); Store at −20°C upon receipt

• Stop Buffer (black cap)

• Wash 1 buffer (green cap)

• Wash 2 buffer (blue cap)

• Elution buffer (purple cap)

• Adapter

• Cartridge

• Waste tube

• Sample collection tube

Autosampler Vials - Polyethylene (Agilent, 5190-3155)

Autosampler Vials - Snap caps (Agilent, 5182-0541)

Acclaim Pepmap 100 75 μm x 2 cm, C18, 3 μm, 100A peptide enrichment column (trapping column; Thermo, 164535)

Acclaim Pepmap RSLC 75 μm x 25 cm, C18, 2 μm, 100A peptide analytical column (analytical column; Thermo, 164941)

20 μm x 550 nanoViper PEEK-shielded fused silica capillaries (nanoViper capillary; Thermo, 6041.5260)

50 μm x 550 mm PEEK-shielded fused silica capillary (loading pump capillary; Agilent, G1375-87310)

0.17 mm x 450 mm pre swaged stainless steel capillary (analytical pump capillary; Agilent, G5067-4658)

SilicaTip Emitter 360 μm OD x 25 μm ID x 10 μm TIP X Precut 5 cm length uncoated (PicoTip) (New Objective, FS3602510N20CT)

Hardware, Instruments, and Software:

−20°C freezer

Timer

Microfuge 18 centrifuge (microfuge, Beckman Coulter) or other centrifuge

Heat Block (water should be in the wells to ensure appropriate heat transfer)

Vortex mixer

NanoDrop 1000 Spectrophotometer (Thermo Scientific) or other appropriate spectrophotometer

Ultrasonic bath sonicator

Corning LSE low speed orbital shaker or other appropriate orbital shaking platform

Savant SPD1010 Integrated SpeedVac System (SpeedVac)or other appropriate SpeedVac

HPLC System (Agilent 1290 Infinity II) or other nano LC system configuration

• high speed pump (Analytical pump, G7120A) equipped with 1260 HIP degasser (G4225A)

• capillary pump (loading pump, G1376A) with degasser (G1379B)

• multisampler (Injector, G7167B)

• multiple column thermostat (MCT, G7116B) with UHPLC switching valve (5067-4117)

• 1290 nanoadapter (G1988A)

6550 Quadrupole Time of Flight (QTOF) mass spectrometer (Agilent, G6550A) equipped with a nano-flow electrospray (ESI) source (Agilent, G1988-6000) Agilent MassHunter Data Acquisition software

Protein Database searching program such as MASCOT (Matrix Sciences) or SpectrumMill (Agilent Technologies). Programs often are packaged with instruments and free versions can be found on-line. Resources can be found at www.expasy.org.

Protocol:

Protein Sample Processing: Keratin is a major contaminant of proteomics experiments and can not only affect your results, but also prevent the identification of low abundant proteins in a mixture. Several on-line resources exist for tips on preventing keratin contamination. Keeping lab coat, eye protection, and rinsing gloves free of keratin throughout the experiment is recommended.

1. Aliquot 250 μl of an exosome-related sample (e.g., exosomes prepared by ultracentrifugation [Basic Protocol 1] or precipitation [Alternate Protocol 1], or their associated HPLC fractions [Basic Protocol 4], including the exosome-containing fraction and fractions for each lipoprotein class) into a 1.5 ml microfuge tube.

2. Add 12.5 μl 1 M aqueous NaCl solution and vortex.

3. Add 1000 μl of ice-cold acetone and vortex for 3 sec and immediately place at -20°C.

4. Store sample at −20°C for 1 h.

5. Remove sample from freezer and pellet protein precipitate at 18,000 xg or max speed at 4°C in a microcentrifuge.

6. Carefully remove the supernatant without agitating the protein pellet using a 1,000 μl pipette. Leave approximately 10-20 μl of liquid.

7. Add 1,000 μl of ice-cold acetone and vortex for 5 sec to wash the pellet.

8. Spin the protein precipitate again at 18,000 xg or max speed at 4°C in a microcentrifuge. Remove supernatant as described in step 6.

9. Open the microfuge tube and let the sample sit on the benchtop until the acetone evaporates and the protein pellet is dry, approximately 10 min.

10. Solubilize the proteins by adding 50 μl lysis buffer to the dried protein pellet and vortex vigorously and/or bath sonicate until the protein pellet is fully resuspended.

11. Perform a BCA assay to determine the protein concentration of the solution. Follow the procedure provided with the Pierce BCA assay protein kit. Briefly:

    1. Aliquot 10 μl of each BSA standard previously prepared using the manufacturer’s protocol (see reagents and solutions section) into a 0.6 ml microfuge tube. Based on the expected sample concentrations use either the standard or enhanced assay standards. Ensure that you mix the standards and quickly spin them down before transferring them to the microfuge tube.
    2. Aliquot 10 μl of each sample resuspended in lysis buffer into a 0.6 ml microfuge tube
    3. Aliquot 200 μl of BCA working reagent into both the standards and samples and mix by vortexing
    4. If using the standard assay, heat the samples in a heat block at 37°C for 30 min.
    5. If using the enhanced assay, heat the samples in a heat block at 60°C for 30 min.
    6. Allow samples and standards to cool to room temperature, mix by vortexing and quickly spin down to remove liquid from the cap.
    7. Measure samples and standards using an appropriate spectrophotometer.
       Based on the concentration results from the protein assay, calculate the estimated total protein content by multiplying the remaining solution volume by the concentration reported. The iST 8x kit digestion reactions can handle between 1-100 μg of starting protein. If the total protein content is within this range, bring the sample back up to 50 μl using lysis buffer and continue with the protein digestion protocol. If the total protein content is greater than 100 μg, aliquot an appropriate amount of solution and add lysis buffer to bring the total volume back up to 50 μl.

12. Digest sample proteins using the Preomics iST 8x kit. Follow the procedure provided with the Preomics iST Sample Preparation Kit (eight reactions). Briefly:

    The iST 8x kit was designed to lyse, reduce, alkylate, and digest proteins and then purify the resulting peptide mixture in a quick and efficient manner. Protease digests are more effective when proteins are reduced and alkylated using chemical reagents. An example of a reducing agent is dithiothreitol, and an example of an alkylating reagent is iodoacetamide. Reducing reagents break cysteine-cysteine disulfide bonds that help form a protein’s tertiary structure (Feige & Hendershot, 2011). Alkylating reagents chemically modify oxidized cysteine residues to prevent disulfide bridges from forming again. The breakdown in tertiary structure of the protein often helps proteases access substrates that would have been more difficult to access and therefore increases digestion efficiency (Suttapitugsakul, Xiao, Smeekens, & Wu, 2017).

    1. Heat samples in a heat block at 95°C for 10 min.
       Heat aids in the reduction of proteins.
    2. Allow samples to cool to room temperature and quickly spin down to remove liquid from caps.
    3. Sonicate in a bath sonicator for 3 min using 30 sec on/off cycles.
    4. Remove trypsin/Lys-C protease tubes (red cap – provided with the preomics iST kit 8x) from the −20°C freezer and add 210 μl protease re-suspension buffer (yellow cap – provided with the Preomics iST 8x kit).
    5. Mix gently by vortexing on the lowest setting for 10 min at room temperature before aliquoting 50 μl trypsin and lys-C protease solution to the 50 μl of lysis buffer containing denatured and alkylated proteins.
       Trypsin will undergo rapid autolysis, especially at room temperature. Always re-suspend the protease solution just before needed, work quickly, and freeze immediately at −20°C after use.
    6. Digest protein into peptides at 37°C using a heat block on an orbital shaker for 2.5 h.
    7. Add 100 μl of stop buffer to inactivate protease enzymes and mix gently for 1 min.
    8. Use the adapter to place the cartridge into the waste tube (all provided in the kit).
    9. Using a 200 μl pipette, pipette digested protein solution up and down a few times and transfer the digested protein solution to the cartridge.
    10. Spin the cartridge in a microcentrifuge at 3,800 xg for 1 min at room temperature.
    11. Add 200 μl of wash 1 to the cartridge (cleans up peptides from hydrophobic contaminants) and repeat step j.
    12. Add 200 μl of wash 2 to the cartridge (cleans up peptides from hydrophilic contaminants) and repeat step j.
    13. Use the adapter to place the cartridge in a fresh collection tube (provided in the kit) and add 100 μl of elution buffer. Repeat step j.
    14. Add another 100 μl of elution buffer to cartridge and repeat step j.
    15. Discard adapter and cartridge and dry peptide elution solution in a speedvac at 45°C until just dry, around 30 min.

13. Re-suspend digested exosome sample in 30 μl of re-suspension buffer by vortexing, let the sample sit for 5 min, and then vortex again.

14. Quickly spin down and transfer sample to an autosampler vial and place in the autosampler prechilled to 4°C.

15. Set up an HPLC connected to your mass spectrometer for nano-flow liquid chromatography.

    1. Connect mobile phase A (MP A) and mobile phase B (MP B) to pump head A and B, respectively, in the analytical pump.
    2. Connect mobile phase C (MP C) to pump head A in the loading pump.
    3. Connect the loading pump capillary from the loading pump outlet to position 1 inlet in the injector valve.
    4. Purge loading pump with MP C for 5 min.
    5. Connect a nanoViper capillary from the position 6 inlet in the injector valve to the position 2 inlet in the multiple column thermostat (MCT) UHPLC switching valve (MCT valve). Set the loading pump to pump 100% MP C at a flow rate of 3.2 μl/min and start pump to purge injector, nanoViper capillary, and seal of the UHPLC switching valve, and stop flow after 20 min.
    6. Connect a nanoViper capillary from the nanoadapter outlet to the position 4 inlet on the MCT valve.
    7. Purge MP A and MP B in the analytical pump. Ensure at least 20 ml of fresh MP is purged in each line (most HPLC systems will not have dead volumes greater than 20 ml from solvent bottle to pump heads).
    8. Close purge valve and pump 50% MP B, 50% MP A at 5.0 ml/min for 4 min to condition the other components of the pump and collect solvent from the standard flow capillary in a waste container.
    9. Stop flow and connect the standard flow capillary from the analytical pump outlet to the nanoadapter inlet. Set the flow rate of the analytical pump to 0.120 ml/min and start pumping 50% MP B, 50% MP A to purge the nanoadapter, nanoViper capillary, and MCT valve seal for approximately 10 min.
    10. Stop flow and attach columns in the proper flow orientation in the MCT valve (see Figure 5). This is the necessary arrangement for nano-trapping columns that have one directional flow such as the Acclaim Pepmap 100 75 μm x 2 cm, C18, 3 μm, 100 A peptide enrichment column used in this experiment.
    11. Pump 50% MP B, 50% MP A through both columns using the analytical pump by setting the MCT valve in the analytical position (Fig. 5, analytical position, valve position 1-2) and the pump flow rate to 0.120 ml/min, which is an effective flow rate of 0.3 μl/min with the analytical column used in this experiment. Flush columns for approximately 30 min.
        The analytical pump used in this experiment is not a true nano-flow pump. The nanoadapter is used to split flow from the analytical pump to produce nano flow rates. If you have a true nano-flow pump, set the flow rate to 0.3 μl/min. If using the nanoadapter, the pump flow rate can be adjusted in 0.05 ml/min increments to achieve an effective flow rate of 0.3 μl/min depending on the specifications of the nano column you are using.
    12. Switch the buffer composition to 3% MP B and allow the columns to equilibrate for approximately 30 min at 0.3 μl/min.
    13. Switch the MCT valve to loading position (Fig. 5, valve position 1-6) and allow system to equilibrate for 10 min.

16. Set up the mass spectrometer with a nano source and connect the analytical column outlet to the inlet of the source.

17. Set up HPLC Method Parameters

    1. High speed pump (analytical pump): The maximum pressure of the system is 1,200 bar. Set the flow rate to 0.120 ml/min for an effective flow rate of 0.3 μl/min when using the nanoadapter. The HPLC gradient is dependent on the sample type, complexity, and goals of the experiment. The gradient listed below is appropriate for deep coverage of the exosome proteome.

       Time	B %	Flow
       0.0	3.0	0.30 μl/min
       2.5	3.0	0.30 μl/min
       4.0	8.0	0.30 μl/min
       52.5	26.0	0.30 μl/min
       60.0	35.0	0.30 μl/min
       62.0	70.0	0.30 μl/min
       67.0	70.0	0.30 μl/min
       70.0	3.0	0.30 μl/min
       92.0	3.0	0.30 μl/min

    2. Capillary Pump (loading pump): Set the loading pump flow rate to pump 100% MP C at 3.2 μl/min. The maximum pressure of the system is 400 bar. An isocratic gradient of MP C will be used to load samples from the multisampler onto the trapping column.
    3. Multisampler (Injector): Set the thermostat to 4°C. Optimal injection range is 0.2 to 0.5 μg of sample on the column. Multisampler default settings can be used.

18. Set mass spectrometer instrument parameters as follows: Capillary voltage=1200-1400 (this is empirically determined and is dependent on the number of hours the Picotip has been in use); drying gas temperature=175°C; flow rate= 11 L/min; Fragmentor = 360 V (Agilent 6550); acquisition mode=MS; TOF spectra range= 290-1700 m/z; acquisition rate= 1.5 spectra/second, 666.7 ms/spectra; ion polarity=positive. The acquisition rate can be increased to sample more spectra per cycle but with a trade-off of reduced response.

19. Set MS/MS parameters as follows: MS = 290-1,700; MS/MS = 50-1,700; MS Acquisition rate = 10 spectra/second, 100 ms/spectra; MS/MS Acquisition rate = 3 spectra/second, 333 spectra/second; Isolation width = narrow; Collision Energy = use formula; charge state 2 = slope 3.1 offset 1; charge state 3 and > 3 = slope 3.6 offset −4.8; Max precursors per cycle = 20 (this will depend on the complexity of the sample); Active Exclusion = exclude after 1 spectra and release after 0.4 min (this depends on the chromatographic peak shape and complexity of sample); Isotope model = Peptides; Only select precursors with a charge state of 2, 3, >3; Sort precursors by abundance only; Use the abundance dependent accumulation, Target = 40,000 counts/spectrum (the lower the value the lower the ion count needed in the MS/MS spectra, which could result in lower quality spectra, but gain faster scan rates and therefore more spectra collected per run), use MS/MS accumulation time limit; Purity stringency = 100%; Purity Cutoff = 30%; or optimize acquisition parameters for your QTOF system based on the manufacturers recommendations or users previous experience.

20. Inject 2 μl of resuspension buffer to serve as a blank. A blank injection will prepare the system for sample analysis.

21. Empirically determine an appropriate injection volume for sample analysis by running 1.0 μl sample and inspecting the chromatogram.

    The most abundant peaks should not exceed an intensity of 107; note that intensity is an arbitrary unit and set by the manufacturer, the value of 107 is appropriate for an Agilent QTOF. Inspect representative extracted ion chromatograms to determine if peaks are saturated. Overloading can result in chromatography artifacts such as retention time shifts and reduced mass accuracy. If necessary reduce the amount of analyte by lowering the injection volume or diluting samples.

22. Collect data on samples of interest using an injection volume that was empirically determined

    Always finish with a blank injection to clean the column and then store it in 50% ACN with no modifiers.

23. Once data for all samples of interest have been collected, move on to data analysis.

24. Search data in Spectrum Mill: Data extraction and database searching.

    1. Extract spectra from the targeted MS/MS raw data files using the following parameters: signal-to-noise = 25:1, maximum charge state allowed = 4, precursor charge assignment = Find.
    2. Search parameters: SwissProt species-specific database (e.g., human), carbamidomethylation as a fixed modification (if IAA used during digest), Digest = trypsin, maximum of one missed cleavage, instrument = Agilent ESI Q-TOF, precursor mass tolerance = 20 ppm, product mass tolerance = 50 ppm, maximum ambiguous precursor charge = 3.
    3. Validate protein identifications using a minimum of 2 peptides per protein, protein score > 10, individual peptide scores of at least 8, and Scored Percent Intensity (SPI) of at least 50%. The SPI provides an indication of the percent of the total ion intensity that matches the peptide’s MS/MS spectrum.
    4. Perform manual inspection of spectra to validate spectrum match to predicted peptide fragmentation pattern as necessary to increase confidence in protein identifications.

Figure 5.

Schematic for high speed pump (analytical pump), capillary pump (loading pump), multisampler (injector), trap column, and analytical column connections and flow paths in both analytical and loading positions in the multiple column thermostat UHPLC switching valve.

Reagents and Solutions

1 M Sodium chloride (NaCl) solution:

Dissolve 0.0584 g of NaCl in 1 ml of water in a 1.5 ml microcentrifuge. Mix well and store at room temperature for up to six months.

BCA Protein Assay Kit (Pierce # 23225):

All necessary solutions (2.0 mg/ml bovine serum albumin standard [23209], Pierce BCA Protein Assay Reagent A, 500 ml x2 [23228], and Reagent B [1859078]) are provided in the kit, except for water. Standard solutions should be made in 1.5 ml microfuge tubes following the protocol provided in the kit. Standards can be stored at 4°C for up to two months. Prepare the BCA Assay working reagent just prior to adding it to standards and samples. Based on estimated protein amounts follow either the Standard Test Tube Protocol (working range = 20-2,000 μg/ml) or the Enhanced Test Tube Protocol (working range = 5-250 μg/ml). Check the compatible substance concentrations table provided in the instructions manual to ensure that there will be no interfering reagents in your sample solution before you start.

Re-suspension Buffer (3% aqueous Acetonitrile [ACN] in 0.1% formic acid):

To 969 ml of water, add 30 ml of ACN and a 1 ml ampule of formic acid, 99+%, in a 1 l wide-mouth bottle. Mix well and store at 4°C for up to 6 months.

Column Storage Solution (50% acetonitrile):

To 500 ml of water, add 500 ml of acetonitrile LCMS grade in a 1 l wide mouth bottle. Expires after one year.

Mobile Phase A (MP A) - Aqueous buffer, 0.1% formic acid:

Add four 1 ml ampules of formic acid, 99+% to a 4 l bottle of water. Expires after six months.

Mobile Phase B (MP B) - Organic buffer, 90% acetonitrile (ACN), 0.1% formic acid:

Combine 900 ml ACN, 100 ml HPLC-grade water, and 1 ml ampule of formic acid, 99+%. Expires after one year.

Mobile Phase C (MP C) - Organic buffer, 3% acetonitrile (ACN), 0.1% formic acid:

Combine 970 ml HPLC-grade water, 30 ml ACN, and 1 ml ampule of formic acid, 99+%. Expires after six months

COMMENTARY

a. Background Information.

While supernatants from cells grown in culture are readily available for exosome isolation, plasma samples from phenotypically characterized human research participants and from animal model experiments are often precious and scarce. Thus, when determining the optimal exosome isolation method, investigators must consider the sample source (Macias et al., 2019). Here, we compared a precipitation based exosome isolation method, the Systems Biosciences ExoQuick^® ULTRA EV Isolation Kit for Serum and Plasma, to ultracentrifugation, which is considered the gold standard method for exosome isolation.

We examined the sizes of the EVs isolated using each method to determine whether they were mostly exosomes or whether they included a mixture of different vesicle sizes and types. Based on NanoSight NTA, both the precipitation and ultracentrifugation methods resulted in the isolation of exosome-sized particles (mean diameter: 160.97 nm for ultracentrifugation versus 148.9 nm for precipitation; mode: 126.6 nm for ultracentrifugation versus 124.8 nm for precipitation). When exosome concentrations were determined by NanoSight NTA, the ultracentrifugation samples had an average concentration of 16 x 10⁹ particles per 250 μls starting plasma, while the PRE exosome samples had an average concentration of 41.7 x 10⁹ particles per 250 μls starting plasma. Based on these criteria, the PRE exosome sample generated by the precipitation method has ~2.5 fold higher exosome concentration. It is necessary to consider that the NanoSight instrument measures particle concentrations and sizes based on Brownian motion. Therefore, alterations in exosomal surface cargo may alter their mobility and thus the measurements of their sizes and concentrations, which may explain some of the differences we observed between the two preparations. Notably, it is difficult to measure the concentration of exosomes in POST exosome samples generated by the precipitation method using NanoSight NTA, either because they are too dilute (2.5 times more dilute than the PRE exosome samples) to allow the minimum of 20 particles per frame necessary to get an accurate read of concentration or because changes in their associated surface molecules may alter their Brownian motion. The former explanation is most likely because in our test runs using less diluted POST samples allowed us to obtain sufficient particles per frame to measure a concentration more accurately, however, a large volume of sample must be used to get robust numbers and the alterations in the surface affecting Brownian motion and measurements must be considered as a confounder to accuracy. Since the sample cannot be recovered and used in downstream analyses post-NTA, we chose not to use NTA to examine the POST exosome samples and to instead rely on the fact that each POST exosome sample is generated from a 200 μl PRE exosome sample that was derived from 250 μl of plasma. It is also noteworthy that the precipitation method required one-third of the starting volume of plasma compared to the ultracentrifugation method.

Exosome concentration is not the only parameter to take into consideration when selecting a method for preparing exosomes. It is also important to consider the downstream utilization of the exosome preparation and the time required to prepare them. For clinical applications using plasma samples, the precipitation method takes only 40 min to generate the PRE exosome sample (described above), and 60 min total to generate the POST exosome sample for which the IgG and albumin contaminants have been significantly reduced by using the purification column. In contrast, ultracentrifugation can take on the order of six hours for six samples, and the throughput of the ultracentrifugation protocol can differ significantly based on the equipment used (i.e., ultracentrifuge rotor). Here, we describe a working set-up to isolate exosomes from six plasma samples at a time using the SW60 rotor. With the precipitation method, it is important to consider that there may be some remaining polymer from the precipitation step that generates the initial exosome pellet and the resulting PRE exosome sample; which is not an issue with the ultracentrifugation method. It is worth mentioning that when comparing treatments with PRE exosome samples versus their buffer-only controls, we have not seen any toxic effects, but including these controls in downstream experiments is important. Another advantage of the ultracentrifugation protocol is that it yields control fractions (i.e., a 10K-pellet with larger EVs and an exosome-free supernatant, see Basic Protocol 2 and Fig. 1). These fractions can be used as negative controls to demonstrate that effects observed are specific to the exosome fraction if these same observations are not present when using one of the control fractions.

Lipoproteins are complex particles that are composed of a central core containing cholesterol esters and triglycerides surrounded by free cholesterol, phospholipids, and apolipoproteins. Plasma lipoproteins can be divided into seven classes based on size, lipid composition, and associated apolipoproteins (Table 1): 1) chylomicrons (75-1,200 nm), 2) chylomicron remnants (30-80 nm), 3) VLDLs (30-80 nm), 4) intermediate density lipoproteins (IDLs, 25-35 nm), 5) LDLs (18-25 nm), 6) HDLs (5-12 nm), and 7) lipoprotein (a) (Lp (a), ~30 nm). Lipoproteins can be isolated from plasma, serum, and other biological samples (e.g., cell culture media) via size exclusion chromatography, such as HPLC.

HPLC is a milder method that is preferred over relatively harsher methods, such as ultracentrifugation, to prevent apoprotein removal during isolation. It is reasonably well-known, albeit under-reported, that lipoproteins and exosomes can co-purify. Therefore, HPLC fractionation of the ultracentrifuged and precipitated exosome preparations allowed us to determine differences in co-purification of canonical lipoproteins between the two purification methods. In this capacity, HPLC can be considered to serve a quality control measure and/or as an alternative purification method. As a quality control measure, HPLC analysis can be used to determine whether different types of lipoprotein contaminants and/or increased levels of lipoprotein contamination are present in a given preparation, which can be important for the downstream interpretation of results. As a gold standard, analysis of both the cholesterol content and the apoprotein composition of the lipoprotein fractions would determine which lipoprotein class has been carried over. One can envision that significant lipoprotein contamination in an exosome preparation may significantly skew downstream applications, such as lipidomics and biomarker identification, and may alter the interpretation of cause and effect in subsequent in vitro or in vivo studies.

b. Critical Parameters.

Biofluid samples, such as plasma samples, from research participants are often available in only limited amounts. In addition, although blood samples are routinely drawn in clinical settings, there is a critical need for fast, reliable tests that can aid in differential diagnoses and/or serve as biomarkers for disease progression. In order to develop and implement tests that will be used in clinical practice, exosome isolation methods should require a relatively small amount of patient sample, should be high-yield, should result in a relatively clean preparation, should be easy to carry out, should not require specialized instrumentation, and should be relatively fast. For example, we include a second centrifugation step during the preparation of our plasma samples in order to ensure that no white blood cells or platelets are present as contaminants in the preparations. The speed, ease of isolation, and yield of exosomes obtained when using the precipitation-based isolation methods have been the main reasons why this field has expanded so rapidly with many companies developing kits that promise all of these facets for the isolation of exosomes (Macias et al., 2019). We chose to compare the ExoQuick^® ULTRA EV Isolation Kit for Serum and Plasma precipitation method to the ultracentrifugation method because ultracentrifugation is the historical gold standard in the field for preparing exosomes.

c. Troubleshooting.

Problem	Solution
NanoSight NS300 may underestimate loss of exosomes due to freeze-thaw and may be less precise, less accurate, and may overestimate concentrations in serially diluted samples based on linear regression modeling. In addition, because surface cargo of exosomes can affect the movement of the exosomes in solution, the assessment of more complex samples can prove problematic (Bachurski et al., 2019).	Consider comparing measurements using multiple methods, such as flow cytometry and imaging flow cytometry (see Alternate Protocol 2) to obtain an accurate assessment of exosome sizes and concentrations in the preparation.
Flow cytometry and imaging flow cytometry for assessing the sizes and concentrations of exosome preparations as well as for confirming the presence of exosomes requires more time, specialized equipment, and expertise.	Consider the downstream analyses and the use of the exosomes with regard to their purity. When particles are greater than 50 nm in diameter, NanoSight NTA can assess size quite accurately, is easier to use, is less costly, is less time-intensive, and requires less specialized training.
Lipoproteins are present in all exosome preparations.	Consider the presence of lipoproteins when drawing conclusions from data obtained from experiments using exosome preparations. For clinical studies, ask research participants to fast prior to obtaining blood samples in order to reduce the amounts of lipids in their plasma.
Exosomes from precipitation kits may still contain traces of polymer from the initial precipitation step.	Compare experimental results to those using exosomes isolated by ultracentrifugation and/or consider the effects that this may have on any downstream applications in which you will use the exosome preparation (e.g., functional assays).
Although cleaning up precipitated exosomes using the purification column reduces the levels of contaminating albumin and IgGs, the clean-up step should be considered in terms of downstream analyses and physiological relevance.	Carefully consider the research question you are addressing and whether it may be affected by the clean-up step. Consider using both the PRE and POST exosome samples to determine whether they show different effects.
NanoSight NTA-based techniques depend on light scatter and motility to measure the sizes and concentrations of exosomes, and the PRE and POST exosome samples produced using the precipitation method may have different associated cargo molecules on their surfaces, which may affect the measurements of their sizes and concentrations.	Use multiple comparative methods to assess the exosome preparations to determine exosome sizes and concentrations and consider the downstream research questions in terms of whether to use the PRE or POST exosome samples. Consider using both types of samples to see if there are differences.

d. Understanding Results.

Exosomes isolated using either precipitation or ultracentrifugation methods contain apoproteins associated with low density lipoproteins (LDLs) and high density lipoproteins (HDLs) in HPLC fractions that were differentially isolated as LDL and HDL lipoproteins, as well as in the HPLC fractions that contained exosomes. This suggests that LDL and HDL lipoproteins are isolated along with and may be adhered to exosomes, making it important to consider the potential roles and contributions of associated lipoproteins when analyzing and interpreting data obtained.

e. Time Considerations.

Preparation of the plasma samples described above takes approximately 60 min per sample. The isolation of exosomes using the precipitation based method requires 40 min to generate the PRE exosome sample (i.e., the first resuspended exosome pellet), including a 30 min incubation and a 10 min centrifugation. This kit also includes a purification column step, which removes albumin and IgG from the initial precipitated exosome sample. This step takes an additional 20 min and generates the POST exosome sample. In contrast, exosome isolation by ultracentrifugation for six samples takes approximately six hours. HPLC fractionation takes 120 min per sample. Mass spectrometry sample preparation (acetone precipitation and protein digest) takes six hours. Setting up the mass spectrometer and HPLC for analysis takes approximately two hours, and sample run time depends on number of samples. Searching spectra on spectrumMill takes approximately 60 min.

f. Summary.

While the field wrestles with developing a deeper understanding of the importance of vesicles of various sizes and origins as well as the significance of their associated molecules, such as apoproteins, addressing the role played by any entity in a biological process is important. The more that entity, a puzzle piece in the process, is understood from the vantage of its physiological role, the more accurately and precisely we can assemble a model, all the while being aware of our limitations in assembling the meaning of a part when measured in isolation, to the whole. Our hope in publishing these techniques as a composite is that we will provide other investigators with tools to address research question(s) of interest so that the field can continue to expand its understanding of the roles of exosomes in cellular physiology and pathology.