Authors: Saba Zanganeh, Gabriel F. Arias, Allaura S. Cone, Runjie Yuan, Dirk P. Dittmer
Categories: Protocol, Biotechnology and bioengineering, Material sciences, Molecular biology, Protein biochemistry, Protein expression and purification
Source: STAR Protocols
Authors: Saba Zanganeh, Gabriel F. Arias, Allaura S. Cone, Runjie Yuan, Dirk P. Dittmer
Extracellular vesicles (EVs) hold promise as biomarkers and drug delivery vehicles; however, their broader use is limited by the lack of effective and scalable purification methods. We present a protocol for purifying EVs from human plasma without ultracentrifugation. The workflow incorporates tangential flow filtration (TFF) for volume reduction, polyethylene glycol (PEG)-based precipitation for enrichment, nuclease treatment to remove extravesicular nucleic acids, and multimodal and affinity chromatography for contaminant depletion. We further describe procedures for biochemical, biophysical, and functional EV characterization.
Extracellular vesicles (EVs) have attracted significant attention due to their remarkable potential as biomarkers, drug-delivery carriers for therapeutic cargo, and mediators of intracellular communication, providing valuable insights into physiological processes and disease mechanisms.^1^ However, the absence of effective and scalable purification protocols continues to hinder the advancement of EV research and the translation of their promising applications.^2^^,^^3^^,^^4^ We designed a high-yield purification pipeline for human plasma that circumvents ultracentrifugation. This protocol provides detailed steps on EV isolation from human plasma, using a sequential purification workflow incorporating tangential flow filtration (TFF), polyethylene glycol (PEG) precipitation, and multimodal and affinity chromatography. The EVs are characterized according to Minimal Information For Studies of Extracellular Vesicles (MISEV2023) guidelines,^5^ including characterizing EV size, concentration, charge, protein content, and RNA content using a variety of approaches, such as Nanoparticle Tracking Analysis (NTA), Western blotting, TapeStation RNA analysis, and direct stochastic optical reconstruction microscopy (dSTORM) to ensure purity and consistency across multiple batches. In addition, an assay for EV uptake and fusion with cells is provided to ascertain biological activity.
This elaborate purification pipeline can be utilized for laboratory and large-scale purification of EVs. It has been used specifically to isolate EVs from a large volume of human plasma, which contains substantial quantities of albumin. The proposed approach below can also be utilized for EV isolation and purification from the cell culture media of different cell lines, milk, or other body fluids.^2^^,^^3^^,^^4^
Rigorous, robust, and reproducible experiments require pure, well-validated biomaterials. EV or exosomes can only be attained by purification from natural sources. This protocol applies to any fluid, including human plasma, which can be obtained as clinical-grade material. We provide workflow and QC assays as STAR methods that translate discovery-grade EV purification into GMP-compliant purification. This method does not use ultracentrifugation, previously used as the gold standard experimental procedure for EV purification, as it is now known to malform EVs through partial EV aggregation and degradation and impair their functionality, structural, and biological integrity under the external force of high-speed rotation.^6^^,^^7^^,^^8^ Our protocol includes two multimodal resin-based steps and completely and demonstrably removes all contaminating DNA and RNA. This method is BSL-3 compatible, allowing for the purification of EVs from infectious material. The EVs isolated by this method are biochemically purer and at higher concentrations than was previously possible. They are fusion-competent, biologically functional, and dSTORM analysis-ready, as validated by our assays.
Any work with human material requires approval by the institutional review board or certification that the work is considered non-human subjects research (NHSR). All human specimen collections must adhere strictly to applicable institutional and national ethical guidelines and regulations.
Any work with biological material requires approval from the institutional biosafety board (IRB). The procedures outlined here can be performed under BSL-1, BSL-2, and BSL-3. Since this procedure does not involve ultracentrifugation, it reduces the risk of aerosols and is suitable for BSL-3 and work with infectious pathogens.
REAGENT or RESOURCESOURCEIDENTIFIERAntibodiesCD63 Mouse mAb, clone TS63 (1:3000 dilution)AbcamAb59479Anti-Hu CD9, clone MM2-57 (1:500 dilution)EMD Millipore CorpCBL162Alix Rabbit mAB, clone E6P9B (1:2000 dilution)Cell Signaling92880SAnti-Syntenin mAB, clone EPR8102 (1:2000 dilution)Abcamab133267Flotillin-2 Rabbit mAb, clone C42A3 (1:2000 dilution)Cell Signaling3436STSG101 Rabbit mAb, clone EPR7130B (1:2000 dilution)Abcamab125011Goat anti-HRS/HGS Antibody, polyclonal (1:2000 dilution)MyBioSource.comMBS420850Goat Anti-Rabbit IgG, Peroxidase (1:4000 dilution)Vector LaboratoriesPI-1000-1Horse Anti-Mouse IgG, Peroxidase (1:4000 dilution)Vector LaboratoriesPI-2000-1Bovine Anti-Goat IgG-HRP (1:4000 dilution)Santa Cruz BiotechnologySc-2378CD9 (used for dSTORM only)BD PharmingenCat # 555370Biological samplesDefibrinated Normal Human plasma, SeraCon™ ISeraCare1800–0009Fetal Bovine Serum (FBS)Corning35-015-CVChemicals, peptides, and recombinant proteinsEndothelial Cell Basal Medium 2 kit EBM-2)PromoCellC-22211Penicillin Streptomycin (Pen Strep)Gibco15140–122Sodium Hydroxide (NaOH)Fisher ChemicalS318-1Polyethylene Glycol 8000 (PEG)Fisher BioreagentsBP233-1RNase A, DNase and Protease-Free (10 mg/mL)Thermo ScientificEN0531RQ1 RNase-Free DNasePromegaREF: M610APageRuler Plus Prestained Protein LadderThermo Scientific266192-Propanol (Isopropanol)Fisher ChemicalA416-1Sodium Chloride (NaCl)Fisher BioreagentsBP358-212Potassium Chloride (KCl)Sigma-AldrichP3911-500GSodium Phosphate Dibasic Anhydrous (Na2HPO4)Fisher ScientificS374-1Sodium Phosphate Monobasic Monohydrate (NaH2PO4)Fisher ScientificBP330-500Tris-HydrochlorideFisher BioreagentsBP153-1Magnesium Chloride (MgCl2)EMD Millipore Corp442611-500GMCalcium Chloride (CaCl2)Acros Organics349610250Sodium Dodecyl SulfateFisher ScientificBP166-500Tween 20Fisher BioreagentsBP337-500MethanolFisher ChemicalA411-4Absolute ethanolFisher BioreagentsBP2818-500ChloroformFisher ScientificBP1145-1Acetic Acid, Glacial (Certified ACS)Fisher ScientificA38-500Coomassie Brilliant Blue R-250Fisher ScientificBP101-25DithiothreitolMillipore SigmaD0632Sodium Dodecyl Sulfate(SDS)Fisher ScientificBP166-100Bromophenol blueMillipore SigmaB0126GlycerolFisher BioreagentsBP229-1β-Mercaptoethanol (BME)Millipore SigmaM6250-10MLGlucoseMillipore SigmaRDD016-500GGlucose OxidaseMillipore SigmaG7141-10KUCatalaseMillipore SigmaC3515-25MGPoly-L-Lysine hydrobromideMillipore SigmaP2636-25MGCellMask™ Plasma Membrane Stains Deep Red (CMDR)InvitrogenC10046Paraformaldehyde (PFA)Fisher Scientific15714-SDAPI (4′,6-diamidino-2-phenylindole, dihydrochloride)InvitrogenCat# D1306Dil Stain (1,1′-Dioctadecyl-3,3,3′,3′-Tetramethylindocarbocyanine Perchlorate (‘DiI’; DiIC18(3)))InvitrogenCat# D3911Bovine Serum Albumin (BSA)Fisher BioreagentsBP9703-100Critical commercial assaysPierce BCA Protein Quantitative KitThermo ScientificCat# 23225MiRNeasy Kit (50)Qiagen217084Alexa Fluor 568 Antibody Labeling KitThermo FisherA20184Alexa Fluor 488 Antibody Labeling KitThermo FisherA20181Experimental cell lineshTERT-HUVEC (Human Telomerase Reverse Transcriptase (hTERT), Human umbilical vein endothelial cells (HUVEC))ATCCCRL-4053 ™Software and algorithmsUnicorn 1.0Cytiva1.0ZetaViewZetaViewV 8.06.01 SP1NanometrixNanometrixV2.7.7.7OtherCentrifuge 5910REppendorf5910RCentrifuge rotor S-4xUniversalEppendorfEP5895200001Centrifuge rotor FA-6x250EppendorfEP5895175007Vacuum-Driven Filter 1000mL PES membrane 0.45 μmGenCloneCat# 25-230Vacuum-Driven Filter 1000mL PES membrane 0.22 μmGenCloneCat# 25-229MidGee hoop ultrafiltration hollow fiber cartridgesCytivaUFP-750-C-H24LAÄKTA™ flux tangential flow filtration systemCytiva29038437ÄKTA start™ chromatography systemCytiva29022094ZetaView® Multi-laser TWIN Nanoparticle Tracking AnalyzerParticle MetrixPMX-230-Z-488/6403000 Series Nanosphere™ Size StandardsThermo Scientific3100ApH test strips 0-14VWR ChemicalsBDH35309.606Capto Core 700 multimodal resinCytiva17548102HiScreen Capto Core 700 4.7 mL ColumnCytiva17548115HiTrap Heparin High-Performance 1 mL ColumnCytiva17040601Whatman® UNIFLO® 0.2 μm PES filtersCytivaWHA9915-25028-well glass bottom chamber slidesIbidiCat# 80827SurePAGE™, Bis-Tris, 10x8, 4-12%, 10 wellsGenScriptCat# M00652Tris-MOPS-SDS Running Buffer PowderGenScriptCat# M00138Mini Trans-Blot Electrophoretic Transfer CellBiorad1703930PowerPac™ Basic Power SupplyBiorad1645050Nitrocellulose Membrane, 0.45 μmBiorad1620115Whatman GB003 gel blotting filter paperCytiva10426892iBright™ CL1500 Imaging SystemInvitrogen™A441144200 TapeStation SystemAgilentG2991BAHigh Sensitivity RNA ScreenTapeAgilent5067–5579High Sensitivity RNA ScreenTape LadderAgilent5067–5581High Sensitivity RNA ScreenTape Sample BufferAgilent5067–5580
1X DNase bufferReagentFinal concentrationAmountTris-HCl10 mM788 mgMgCl22.5 mM254 mgCaCl20.5 mM36.8 mgDI waterN/A500 mLTotalN/A500 mL
CRITICAL: Filter the DNase buffer through a 0.22 μm vacuum filter. Note: 1X DNase buffer is prepared in deionized water with a final pH of 7.6 and stored at 25°C for up to 1 year. 10X Calcium-Magnesium Free Phosphate-Buffered Saline (CMF-PBS buffer)ReagentFinal concentrationAmountNaCl1.37 M80 gKCl26.8 mM2 gNa2HPO4101.4 mM14.4 gNaH2PO420 mM2.4 gDI waterN/A1 LTotalN/A1 L
CRITICAL: Filter the 10X CMF-PBS buffer through a 0.22 μm vacuum filter. Note: 10X CMF-PBS buffer is prepared in deionized water with a final pH of 7.4 and stored at 25°C for up to 1 year. CRITICAL: It is important to use Calcium-Magnesium Free PBS, considering that Ca^2+^ and Mg^2+^ promote bridging between negatively charged surfaces such as EV membranes and proteins, leading to aggregation and clumping of EVs and other contaminants, which can hinder the purification process. 1M NaOH + 30% isopropanolReagentFinal concentrationAmountNaOH 5M1 M20 mLIsopropanol30%30 mLDI waterN/A50 mLTotalN/A100 mL
Note: 1M NaOH + 30% isopropanol is typically prepared in deionized water and stored at 25°C for up to 1 month.
5X Non-reducing Laemmli BufferReagentFinal concentrationAmountTris-HCl250 mM3.75 mL of 1 M Tris-HClSDS10% (w/v)1.5 g SDSBromophenol blue1 mg/mL0.015 gGlycerol50% (v/v)7.5 mlddH2ON/ATo 15 mL total volumeTotalN/A15 mL
Note: 5X Non-reducing Laemmli Buffer is prepared in deionized water with a final pH of 6.8 and stored at −20°C for up to 1 year.
5X Reducing Laemmli BufferReagentFinal concentrationAmountTris-HCl250 mM3.75 mL of 1 M Tris-HClSDS10% (w/v)1.5 g SDSBromophenol blue1 mg/mL0.015 gGlycerol50% (v/v)7.5 mlβ-Mercaptoethanol (BME)5%0.75 mLDithiothreitol (DTT)0.5 M1.16 gddH2ON/ATo 15 mL total volumeTotalN/A15 mL
Note: 5X Reducing Laemmli Buffer is prepared in deionized water with a final pH of 6.8 and stored at −20°C for up to 1 year.
Tris-buffered saline + 0.1% Tween 20 (1X TBST)ReagentFinal concentrationAmountTris-HCl100 mM12.11 g Tris base, adjust to pH 7.5 with HClNaCl0.9% (w/v)9 gTween-200.1% (v/v)1 mLddH2ON/ATo 1 L total volumeTotalN/A1 L
Note: 1X TBST is prepared in deionized water and stored at 25°C for up to 1 month.
10X Transfer Buffer stock solution for western blottingReagentFinal concentration (10X)AmountTris base0.48 M58 gGlycine0.39 M29 gSDS12.8 mM3.7 gddH2ON/ATo 1 L total volumeTotalN/A1 L
CRITICAL: The 10X Transfer Buffer stock solution is prepared in deionized water and stored at 4°C for up to 3 months. The working solution for western blotting is 1X Transfer Buffer, which is made from diluting the 10X Transfer Buffer stock in deionized water and adding 20% (v/v) Methanol. The 1X Transfer Buffer working solution can be stored at 4°C and used for up to 1 week.
Coomassie stain solutionReagentFinal concentrationAmountCoomassie Brilliant Blue R-2500.2% (w/v)2 g95% Ethanol50% (v/v)500 mLGlacial Acetic Acid10% (v/v)100 mLddH2ON/ATo 1 L total volumeTotalN/A1 L
Note: Coomassie stain is prepared in deionized water and stored at 25°C for up to 1 year.
Destain solutionReagentFinal concentrationAmount95% Ethanol25% (v/v)250 mLGlacial Acetic Acid10% (v/v)100 mLddH2ON/ATo 1 L total volumeTotalN/A1 L
Note: Destain solution is prepared in deionized water and stored at 25°C for up to 1 year.
TN Buffer^9^ for dSTORMReagentFinal concentrationAmountTris–HCl, pH 8.050 mM60.6 mgNaCl10 mM5.85 mgddH2ON/ATo 10 mL total volumeTotalN/A10 mL
Note: TN Buffer is prepared in deionized water and stored at 25°C for up to 1 year. Oxygen Scavenging Buffer (OSB)^9^ for dSTORM imagingReagentFinal concentrationAmountGlucose10% (W/v)200 mgBME (1M)143 mM20.58 μL of 1M stockGlucose Oxidase (5 mg/mL)0.5 mg/mL200 μL of 5 mg/mL stockCatalase (14 mg/mL)40 μg/mL6.15 μL of 14 mg/mL stockTotal volume diluted in TN BufferN/A2000 ul
CRITICAL: First add the glucose and BME to the OSB, this reagent is stable for up to 24 h at 4°C. When ready to image, add the catalase and glucose oxidase. Once the enzymes are added, the OSB will only be usable for up to 3 h.
Timing: ∼1 h
Our EV purification protocol begins with slow centrifugation and filtration of the human plasma sample. The centrifugation of the sample at 1,000 x g removes intact cells, dead cells, and large cellular debris. The centrifugation at 20,000 x g helps separate platelets and large microvesicles from plasma EVs. Physical size exclusion through vacuum filters eliminates remaining large particles and aggregates. Note that this will also eliminate aggregates of multiple EVs and of viruses (if present).1.Centrifuge the human plasma sample at 1,000 x g for 10 min at 4°C. After centrifugation, a large pellet should be visible at the bottom of the tubes, while the plasma appears clear. Transfer the supernatant to new tubes.2.Using a 5910R (Eppendorf) centrifuge with a FA-6x250 rotor, centrifuge the tubes at 20,000 x g for 30 min at 4°C. After the centrifugation is completed, transfer the supernatant to new tubes.CRITICAL: For all centrifugation steps, ensure that all sets of conical tubes containing medium are balanced adequately by weight.Optional: In this step, you can keep the 20K pellets for further analysis of large EVs by dislodging and resuspending the pellets in 1 mL of PBS.3.After completing these centrifugation steps, use a 0.45 μm vacuum filter to remove aggregates larger than 450 nm.4.Dilute 50 mL of human plasma by adding 450 mL of 1X-CMF-PBS.Optional: If you are primarily interested in enriching small EVs in your purification process, use a 0.22 μm vacuum filter to further filter the plasma to remove aggregates larger than 220 nm. The filtered plasma is now ready for Tangential Flow Filtration.Note: See Troubleshooting Problem 1 for how to enrich your output for large EV subpopulations.Note: This step is optimized for a 500 mL maximum capacity TFF (Table 1) often found in laboratory settings. It can be scaled to fit 5000 mL TFF reservoirs, which would process 500 mL plasma.
Note: Plasma is the liquid component of blood that still contains clotting factors such as fibrinogen. In contrast, serum is the liquid portion remaining after blood has clotted. It lacks fibrinogen and other clotting proteins. Both serum and plasma can be used on this purification pipeline. **Pause ** The filtered plasma can be stored at 4°C for up to 24 h before proceeding to the next step. Table 1TFF Cartridge detailsFilterModel numberPore size (NMWC)Flow path (cm)Membrane area (cm^2^)Fiber i.d. (mm)Cartridge o.d. (cm)Material (fiber)MidGee hoop ultrafiltration hollow fiber cartridgesUFP-750-C-H42LA750 000110410.50.3Polysulfone (PS)
Timing: ∼5 h
TFF device using a 750 kDa molecular weight cut off (M.W.C.O) filter concentrates the plasma sample by 10-fold and removes low-molecular-weight soluble contaminants. TFF enables the processing of a large volume of samples without ultracentrifugation. TFF is expected to remove free proteins, small peptides, and salts, while EVs, lipoproteins, and large protein complexes are retained in the sample. Hence, TFF results in a highly concentrated sample of plasma EVs.^2^^,^^10^5.Preparation of the ÄKTA™ Flux TFF system.a.Turn on the TFF instrument using the power switch. After the machine fully boots up, hook the feed hose into the peristaltic pump and aim the retentate collection valve to waste. Follow the manufacturer’s instructions for installing the UF hollow fiber cartridge (Table 1) to the ÄKTA™ Flux TFF system.b.If applicable, first remove the storage solution (0.1 N NaOH) from both reservoirs (feed reservoir and transfer reservoir).i.Start the pump at a feed flow of 28 RPM and open the retentate collection valve to empty the system of NaOH storage solution until air starts pumping.ii.Close the retentate collection valve and open the feed collection valve to empty the system from there for a few seconds until air starts pumping.iii.Close all collection valves and turn off the feed flow to stop the pump.c.Water Flush: Fill both feed and transfer reservoirs with 500 mL of DI water.i.Place a stir bar in the feed reservoir and set the Mixer to 100 RPM.ii.Start the pump with a feed flow of 28 RPM and allow the machine to recirculate the water for about 5 min until the whole system is filled with water.iii.Aim the retentate collection valve at the waste container and open it. After draining 100 mL of water from the retentate collection valve, close it.iv.Aim the feed collection valve at the waste container and open it to drain 10 mL from there. Close the feed collection valve as well.v.Aim the permeate to waste and open it. Start tightening the pressure control knob slowly until the Transmembrane Pressure (TMP) is 5 psi. Allow the permeate to drain for at least 10 min (about 100 mL). At the same time, turn on the transfer pump at a flow rate of 5 mL/min to supply more water from the transfer reservoir to the feed reservoir. Ensure the feed reservoir does not overfill.vi.Once the 10-min time is up and enough water has drained from the permeate, initiate the Normal Water Permeability test.d.Normal Water Permeability test (NWP): For this test, set the feed flow to 28 RPM and ensure the pressure control knob is set at 5 psi TMP.i.Retrieve a 15 mL conical tube and direct the permeate into it while simultaneously setting a 1-min timer.ii.After collecting the permeate for 1 min, direct the permeate to waste. Record the volume of permeate collected.Note: NWP is a measure of the cleanness of the Hollow Fiber (HF). The NWP for a fresh HF is usually around 12 mL and can be used for multiple runs. If the NWP is less than 4 mL, the filter must be re-sanitized or replaced.e.After all valves are flushed with water, release all pressure from the control knob and close the permeate.i.Empty the system of DI Water by manually inverting both reservoirs into the waste container.ii.Drain both collection valves (retentate and feed) at 28 RPM feed flow until only air remains in the valves (similar to step 5b).iii.Close all collection valves.6.Equilibration of the Hollow Fiber.a.To equilibrate HF, add 500 mL of 1X CMF-PBS to each reservoir (feed and transfer reservoirs).i.Turn on the Mixer to 100 RPM and begin the flow of the buffer through the system at a feed flow of 28 RPM.ii.Allow the machine to recirculate PBS for about 5 min until the whole system is filled with PBS.iii.As done previously in Water Flush (step 5c), aim the retentate collection valve to waste and open it. After draining 100 mL PBS from the retentate collection valve, close it.iv.Afterward, aim the feed collection valve toward the waste and open it to drain 10 mL of PBS. Close the feed collection valve as well.v.Aim the permeate to waste and open it. Start tightening the pressure control knob slowly until 5 psi TMP is obtained. Allow the permeate to drain for at least 10 min (about 100 mL). At the same time, turn on the transfer pump to a 5 mL/min flow rate to supply more PBS from the transfer reservoir to the feed reservoir. Ensure the feed reservoir does not overfill.b.After all valves are flushed with PBS, test each collection valve and permeate with a pH test strip. The measured pH in all valves should be close to 7.4.Note: If the pH is too high and alkaline, it indicates that there is still NaOH remaining in the valves. To resolve this, continue flushing the valves with more 1X CMF-PBS buffer.c.When the system is at the appropriate pH, stop all pumps.i.Release all pressure from the control knob and close the permeate.Note: Ensure all collection valves are closed as well.ii.Manually discard the buffer from the feed reservoir only. The system is ready for the sample now.CRITICAL: Do not discard the CMF-PBS in the transfer reservoir, since it will be necessary for the Chase step.7.Sample Equilibration.a.The filtered plasma should be equilibrated to room temperature before adding to the feed reservoir. You can collect a small sample of the filtered plasma if it is necessary for further analysis and comparison (store at 4°C for up to 2 weeks).Note: A temperature change can affect HF filtration efficiency and alter fluid dynamics and shear stress for the EVs; therefore, ensure that the samples reach room temperature prior to filtration.b.Fill the feed reservoir with the filtered plasma sample and turn on the Mixer to 100 RPM.c.Start the feed flow of the sample at 28 RPM and allow the machine to recirculate it for 5 min until the entire system is filled with the sample. Figure 1 shows the general setup of our TFF during the sample run.Figure 1The laboratory TFF setup during the sample run8.Ultrafiltration and Sample Collection.a.Place a collection bottle that can hold your target permeate volume underneath the HF. Open the bottle and aim the permeate at the bottle, then open the permeate (Figure 2B).Figure 2The two outputs produced during the TFF step(A) The collected sample through TFF Retentate, (B) The TFF Permeate.b.Start tightening the pressure control knob slowly until a target TMP of 5 psi is obtained. The target TMP range is not to exceed 7 psi or fall below 3 psi.Note: Getting exactly 5 psi for TMP is not necessary as long as you stay within the suggested range of 3–7 psi. As you concentrate your sample, its viscosity and concentration keep changing, which increases the pressure inside the system.c.Allow the sample to concentrate, making sure to regularly check and adjust the pressure control knob to ensure the target TMP range is not exceeded.Note: Many human plasma stocks are highly concentrated and enriched in a wide array of proteins, particles, and impurities. If necessary, you can repeat TFF ultrafiltration of your sample multiple times by bringing the sample volume back up to 500 mL again using 1X-CMF-PBS when it reaches 100 mL.d.Once your sample volume gets below 100 mL, reduce the Mixer to 60 RPM since mixing too fast can cause foaming of the sample.e.When your sample volume reaches 20 mL, release all pressure from the pressure control knob, close the permeate, and turn off the Mixer.f.Retrieve a 50 mL conical tube and direct the retentate collection valve to the tube.g.Open the retentate collection valve to collect the entire sample from the system until only air remains in the valve.h.Turn off the feed pump and close the retentate collection valve. Close the conical tube containing your sample and place it aside for now.9.Chase.a.Set the transfer pump to 10 mL/min. Allow 1X CMF-PBS to flow from the transfer reservoir into the feed reservoir for 75 s, then turn off the transfer pump.b.Set feed flow to 28 RPM and allow the buffer to recirculate in the system for at least 3 min.c.Once the 3-min timer is up, open the conical tube containing the samples and aim the retentate collection valve at the tube again. Open the retentate collection valve and collect all remaining sample in the system until only air/bubbles remain in the valves (Figure 2A).CRITICAL: As the retentate collection valve was previously drained to air, it may be necessary to flick the tubing vigorously to get the flow going.d.Close the retentate collection valve and turn off the feed flow. Close the conical tube containing the sample and mix it by inversion. This is the final TFF product, and the sample is now ready for the next steps.e.Discard the remaining PBS from the transfer reservoir into the waste container.10.Sanitization of Hollow Fiber.a.To perform sanitization, prepare 0.5 N NaOH. Add 500 mL of 0.5 N NaOH only to the feed reservoir.i.Turn on the Mixer to 100 RPM and begin the flow of NaOH through the system with a feed flow of 28 RPM.ii.Allow the machine to recirculate NaOH for about 5 min until the whole system is filled with NaOH.iii.Aim the retentate collection valve at the waste and open it. After draining 10–15 mL NaOH from the retentate collection valve, close it.iv.Afterward, aim the feed collection valve to the waste and open it to drain 10–15 mL NaOH from there. Close the feed collection valve as well.v.Aim the permeate to the waste and open it. Start tightening the pressure control knob slowly until 5 psi TMP is obtained. Allow the permeate to drain 10–15 mL NaOH.b.Stop the feed pump and release all pressure from the control knob. Remove the 0.2 μm filter attached to the top of the feed reservoir and replace it with the permeate to recirculate from the top of the feed reservoir.CRITICAL: Do not allow liquid to enter the filter. If this filter gets wet, it impedes the flow of fluid out of the retentate collection valve. If this problem occurs, replace the filter with a new 0.2 μm syringe filter.i.Resume the flow of 28 RPM from the feed pump and use the pressure control knob to set TMP to 5 psi. Allow NaOH to recirculate for 1 h.ii.Stop the feed pump and release all pressure from the control knob. Empty the system by manually discarding the feed reservoir into the waste container. Start the feed flow of 28 RPM and empty both collection valves one by one until only air/bubbles remain (similar to step 5b).iii.Stop the feed pump and close all collection valves. Remove the permeate from the top of the feed reservoir and close the permeate. Put the syringe filter back on the feed reservoir.c.Add 500 mL of DI water to each of the feed and transfer reservoirs. As previously done in the water flush (step 5c), using a 28 RPM feed flow, fill the system with DI water, flush the collection valves, permeate, and transfer feed with DI water for the same volumes specified in the water flush section.i.Empty the system and collection valves as previously done in step 5b. Stop all pumps and empty the transfer reservoir by manually discarding the remaining solution into the waste container.11.Storage of the Tangential Flow Filtration device.a.To place the system in storage, prepare 1 L of 0.1 N NaOH and add 500 mL to both the transfer and feed reservoirs.Note: 0.1 N NaOH is the storage solution for this system.b.Flush the entire system with the storage solution using 28 RPM feed flow. Flush the collection valves, transfer feed pump, and permeate with storage solution.i.Aim the retentate collection valve at the waste and open it. After draining 10–20 mL NaOH from the retentate collection valve, close it.ii.Aim the feed collection valve at the waste and open it to drain 10–20 mL NaOH. Close the feed collection valve as well.iii.Aim the permeate at the waste and open it. Start tightening the pressure control knob slowly until 5 psi TMP is obtained. Allow the permeate to drain 10–20 mL NaOH. At the same time, turn on the transfer pump to a 5 mL/min flow rate.c.Check the pH of the solution from all valves and ensure it is highly basic (pH>10).d.Stop all pumps and release the pressure from the control knob. Close the permeate and ensure both collection valves are closed.e.Turn off the system using the shutdown button on the screen and the power switch at the bottom of the instrument. Unhook the feed hose from the peristaltic pump.
Timing: ∼1 h
Capto™ Core 700 (CC700) multimodal chromatography resin can be used to capture impurities in the TFF retentate prior to PEG addition, thereby optimizing the precipitation of EVs and proteins of interest and yielding a cleaner pellet. This step is important for high-albumin-containing samples, such as human plasma, but less so for cell culture media.^2^12.Add 0.25 mL of CC700 resin to a new tube for every 1 mL of collected TFF retentate sample (1:4 v/v).Optional: We suggest using 0.2–0.5 mL CC700 resin per 1 mL sample amount for this step. You can increase the ratio of CC700 resin to your TFF retentate if necessary. We recommend increasing the ratio of CC700 resin to TFF retentate if elevated system pressure, capacity-limited flowthrough, or active leaking is observed in subsequent chromatography steps. We also recommend increasing the ratio of CC700 resin if cellular debris or high levels of unwanted impurities are detected in the final output.CRITICAL: The CC700 resin is stored in 20% ethanol as a preservative, which should be removed prior to adding TFF retentate, since ethanol can damage EV membranes.13.To wash ethanol out of the CC700 solution, use 1X-CMF-PBS for equilibration of the beads (1:5 v/v). Add cold 1X-CMF-PBS to your beads and mix well by pipetting. Centrifuge the resin at 500 x g for 3 min at 4°C. Aspirate the supernatant close to the resin bed. Repeat this wash step 5 times with 1X-CMF-PBS.14.After carefully removing the supernatant on top of the CC700 bead bed, add the collected TFF retentate sample to the equilibrated CC700 resin and mix well by pipetting and inversion. Incubate this sample at 4°C on a rotator for 30 min.15.After the incubation, centrifuge this sample at 500 x g for 5 min at 4°C and collect the supernatant in a new tube. If any beads are still visible floating in your sample, centrifuge it at 1000 x g for 5 min at 4°C.***Note:***Figure 3A shows the sample and CC700 resin bed after 500 x g centrifugation, which separates them into two layers.
Optional: You can also repeat the CC700 resin incubation for multiple rounds based on the extent of albumin and impurity present in your plasma sample. Figure 3The output of CC700 resin batch incubation and PEG percipitation steps(A) The TFF retentate sample incubated with CC700 resin formed two layers after centrifugation, (B) PEG precipitated pellet after centrifugation.
Timing: ∼10 min for preparation and 24 h for incubation
PEG precipitates EVs and some protein complexes by reducing their solubility, resulting in a 50-fold concentration of the plasma sample. The PEG precipitated sample is slowly centrifuged to remove PEG-solubilized supernatant proteins and contaminants and obtain a highly concentrated pellet of EVs, while preserving EV integrity.16.Prepare polyethylene glycol (PEG-8000) at a stock concentration of 400 mg/mL in PBS (equal to 40% w/v) and filter through 0.22 μm vacuum filters, then store at 4°C. Considering the total volume of the TFF retentate sample you have collected after the CC700 resin step, add 400 mg/mL PEG solution to the collected sample to achieve a final concentration of 40 mg/mL PEG (4% w/v).17.Once PEG is added, close the tube and mix it well by inversion. Incubate the PEGylated sample at 4°C for at least 18 h on rotation.**Pause ** The PEG overnight incubation provides a pause point for up to 24 h.18.The following day, centrifuge the PEGylated sample at 1000 x g for 1 h at 4°C. Figure 3B shows the PEG pellet after centrifugation is completed.
Timing: ∼20 min
This step uses nuclease treatment to remove free nucleic acids not encapsulated within EVs, while retaining EV-encapsulated nucleic acids and intact EVs in the sample.19.Prepare 1x DNase I buffer per instructions in the “materials and equipment” section and filter through a 0.22 μm vacuum filter.20.After the centrifugation of the PEGylated sample at 1000 x g for 1 h at 4°C is completed, aspirate out all the visible liquid on top of the PEG pellet, making sure not to touch the pellet at the bottom of the tube.21.Carefully resuspend the PEGylated EV pellet in 1X DNase buffer using gentle pipetting.Note: See Troubleshooting Problem 2 for how to handle difficulties with resuspending a rigid PEG pellet.22.Add RNase A and DNase I to the resuspended PEG pellet sample at a 100 dilution to a final concentration of 100 μg/mL and 10 U/mL in the sample, respectively. Mix by gentle rotation. Incubate the sample on a 37°C heat block for 10 min.23.After 10 min, remove the sample from the heat block and allow it to equilibrate to room temperature for 5 min. The Nuclease-treated sample is now ready to proceed to the next step.Note: It is advised to incubate one sample at a time in nucleases. If you have multiple aliquots, keep the remainder at a 4°C fridge while you work with one sample at a time.CRITICAL: This step does not use heat or chemicals to inactivate nucleases, as this would damage the EVs. The nucleases are removed during CC700 column chromatography.
Timing: ∼2.5 h
CC700 multimodal chromatography combines size exclusion and internal ligand binding to remove contaminants such as smaller proteins, albumin, and the nuclease enzymes digestion products in the sample from the nuclease treatment step. The CC700 flow through (FT) contains clean EVs and highly purified vesicles excluded from the resin pores.^2^^,^^4^ It avoids heat-inactivation, which may cause aggregation and denaturation of proteins on the EVs.24.Preparation for ÄKTA start™ Chromatography.a.Turn on the ÄKTA start™ instrument and hook the pressure sensor valves into the pump.b.Open the Unicorn 1.0 application on the computer connected to ÄKTA start™. Click on the “System administration” tab and “Connect” to allow the system to load. When connected, the screen will show a chromatogram section, a run log, and a schematic of the whole system.Note: The chromatogram section displays all traces in real time as the instrument runs. The schematic of the whole system serves as a guide to where the fluid is coming from and going. When a run is active, the schematic will change to show a green line for any active flows. If there is only a gray line, then material is not actively flowing through there.c.On the instrument, make sure that Buffer A container has a sufficient amount of 1x CMF-PBS, and Buffer B container has a sufficient amount of 1x CMF-PBS + 1 M NaCl.CRITICAL: During equilibration, sanitization, water flush, or storage for the CC700 column, the only difference between these steps is the buffer in the sample line. Equilibration buffer is 1x CMF-PBS. Sanitization buffer is 1 M NaOH + 30% isopropanol (w/v in water). Water refers to DI water. Storage buffer is 20% ethanol.25.Preparation and CC700 column connection.a.Preparation: Move the sample input line at the sample valve to a bottle of 1x CMF-PBS. Aim the green fractionator output line at waste.i.Start a “manual run” with a flow rate of 1 mL/min. Allow the machine to flow for 2 min.ii.In “execute manual controls”, set the “flow path” in “sample valve” to “sample”. Allow this to flow to waste for 2 min.iii.On “flow path”, switch “wash valve” to “column”. This allows the buffer to flow through the column path to waste.b.Connecting a Install the HiScreen CC700 5 mL column while buffer is flowing and tighten it. Check the column for any leaks.Note: See Troubleshooting Problem 3 for guidance on addressing leaks.c.Once the column is properly connected, switch the “flow path” in “outlet valve” to “collection.” Liquid should be seen flowing from the green collection tubing. Allow to flush for 3 min.d.Switch the “flow path” in “outlet valve” to “waste” and allow to flush for 3 min.e.Once the system is fully flushed, press “End”. The manual run has been completed.26.Prime, Equilibration, and Sample Run on CC700 column.a.To start a run with your sample, redirect the green fraction collector tubing to the fraction collector, making sure to place the tubing slightly above an open 1.7 mL tube in the fraction collector. Figure 4A shows the general setup of the ÄKTA start™ Chromatography instrument using the CC700 column during a sample run.Figure 4Multimodal chromatography setup and output(A) CC700 Chromatography setup during sample run, (B) CC700 flow through fractions collected containing EVs (peak 1).b.Redirect the sample line from the 1X CMF-PBS to the tube of nuclease-treated sample. Ensure the line reaches the bottom of the tube.CRITICAL: It is crucial that no air enters the sample tubing during transfer to the sample; therefore, transfer gently.Note: See Troubleshooting Problem 4 on how to deal with air entering the sample tubing.c.Design the method for HiScreen Capto Core 700 5 mL column based on the parameters shown in Table 2.Note: This CC700 method for EVs is run using a flow rate of 1 mL/min. The method starts with sequential Prime steps using 5 mL of Buffer B and 5 mL of Buffer A. The method continues with Equilibration of the column with 3 Column volumes (CV) of Buffer A. This method proceeds to fractionate the sample by pumping from sample line (for a total volume of 10 mL) through the column and out of the green fraction collector tubing for 20 fractions (0.5 mL set as volume per fraction). The method ends with a Wash-out Unbound step using Buffer A for 1 CV, and a fraction volume of 1 mL.Table 2Method used for EV purification using CC700 column chromatographyPhaseVariableValueMethod SettingsColumnHiScreen Capto Core 700Method SettingsColumn Volume (CV)4.657 CVMethod SettingsFlow Rate (mL/min)1.00 mL/minFraction SettingsSample Fractionation Volume (mL)0.5 mLPrimeBuffer B volume (mL)5 mLPrimeBuffer A Volume (mL)5 mLEquilibrationEquilibration Volume (CV) using Buffer A3.00 CVSample runSample Volume (CV)2.00 CVWash-out unboundWash Volume (CV)1.00 CVSanitization (Post-method run)1M NaOH + 30% isopropanol volume (mL)25 mLWater flush (Post-method run)Water volume (mL)25 mLStorage (Post-method run)20% ethanol volume (mL)15 mLd.Select the method of interest in “method run” and start the run.e.The method begins with Prime steps, followed by Equilibration for 3 CVs. After Equilibration, a new run mark will appear on the chromatogram, and the system schematic will indicate that the flow is now coming from the sample line.f.Prepare a bottle of 1x CMF-PBS, a 1 mL pipette, and clean 1 mL tips.Note: You must stay at the instrument and monitor vigilantly. There is approximately 1 mL of sample, but the process will take 10 mL (2 CV) from the sample line. You must avoid any air from entering the sample line or column.i.To avoid air, monitor the sample tube.ii.Allow the sample to first reach near the bottom, then add 0.5 mL of 1x CMF-PBS.iii.Once that reaches near the bottom of the tube again, add another 0.5 mL of 1x CMF-PBS.iv.Once that reaches near the bottom of the tube again, fill the tube up to the top with 1x CMF-PBS.v.Continue to add buffer until the end of the sample run phase when the instrument switches to the wash-out phase.g.After the sample run, the instrument switches back to buffer A, and it has now entered the “wash-out” step. Allow this to continue until the instrument naturally stops and the run completes.h.Looking at the obtained chromatogram, determine which fractions are in peak 1, which is the “flow through” containing the EVs. Pool peak 1 fractions together and store at 4°C for now (Figure 4B).Optional: You might also observe a small second peak, corresponding to the non-EV content, including soluble proteins, nucleic acids, metabolites, and other contaminants trapped or delayed by CC700 resin. If the peak 2 sample is desired to be kept for further analysis, pool those fractions together and store at 4°C; otherwise, discard.Note: Plasma samples are highly enriched in proteins, leading to a high UV absorbance value shown as a sharp peak 1. Given that the UV absorbance of peak 2 is much lower than that of plasma peak 1, the second peak might not be easily visible or distinguishable on the final chromatogram.Note: See Troubleshooting Problem 5 for how to resolve overloaded columns.27.Sanitization, Water Flush, and Storage of the CC700 column.a.Sanitization: Run another manual run on the computer, setting a flow rate of 1 mL/min. Place a bottle containing 1M NaOH + 30% isopropanol into the sample line. Direct the green fraction collector tubing to waste.b.In “flow path”, set “sample valve” to “sample”, “wash valve” to “column”, “outlet valve” to “collection”, and execute these commands. Allow this manual run to proceed for 5 CVs (25 ml), which takes approximately 25 min. Then, “Pause”.c.Water Flush: Place the sample line in a bottle of DI water. Press “continue” and allow water to flow for 5 CVs (25 mL), which takes approximately 25 min. Then, press “Pause”.d.Storage: For this step, the buffer to be put onto the sample line depends on whether there is another run of CC700 or if it needs to be stored for future reuse.i.If another run is still needed, place 1x CMF-PBS into the sample line. After a 3 CV wash with 1x CMF-PBS, you can restart from step 26 to run your next sample.ii.If storage is the goal, place a bottle of 20% ethanol into the sample line.iii.For either buffer, continue the flow for 3 CV (15 ml), which takes approximately 15 min. Then, press “End.”e.To remove the column, disconnect all connections to the column and replace them with proper stoppers on both sides of the column.Note: The column can be kept at 25°C or at 4°C for long-term storage. If the column is stored at 4°C, allow the column to reach room temperature before using it again.
Timing: ∼2 h
This step performs a selective depletion of heparin-binding soluble proteins and contaminants from the CC700 FT sample, which produces highly pure human plasma EVs, referred to as Non-heparin-binding EVs.^4^28.Prime and Equilibration of Heparin column.a.Attach the tubing adapter to the column tubing port. Place a bottle of 1x CMF-PBS into the sample line, once again ensuring no air has entered the tubing. Direct the green fraction collector tubing to waste.b.Start a manual run on the computer, with a flow rate of 0.5 mL/min. In “flow path”, set “sample valve” to “sample”, “wash valve” to “column”, “outlet valve” to “collection”, and execute these commands.c.As done previously with the CC700 column, connect the tubing coming from the injection valve to the red top of the HiTrap Heparin 1 mL column. Ensure flow is coming from the bottom of the column, then attach the other end of the column to the adapter port. Tighten and check for leaks.Note: See Troubleshooting Problem 3 on how to prevent leaks.d.Allow the flow of 1x CMF-PBS to continue for 10 CVs (10 mL), which approximately takes 20 min. Once equilibrated, press “End”.29.Running a sample on the Heparin column.a.Place the sample line tubing into the pooled CC700 peak 1 tube, ensuring no air enters the sample tubing, and the tubing reaches the bottom of the sample.Note: See Troubleshooting Problem 4 on how to deal with air entering the sample tubing.b.Design the method for the HiTrap Heparin High-Performance 1 mL Column based on the parameters listed in Table 3.Note: This Heparin method for EVs operates at a flow rate of 0.5 mL/min. The method begins with column equilibration using Buffer A for 3 CV (equal to 3 mL). Then, it fractionates by pumping the sample through the column from the sample line for 5 CV (total volume of 5 mL) and collects the fractions into the green fraction collector tubing, with 0.5 mL designated as the volume per fraction. Next, perform a Wash-out Unbound step using Buffer A for 1 CV, with a fraction volume of 0.5 ml. Finish the method with Step Elution using an increasing gradient of NaCl up to 1 M (10%–50% of Buffer B), and cleanse the column with a final Wash-out Unbound step using Buffer A for 3 CV (equal to 3 mL).Table 3Method used for EV purification using Heparin column chromatographyPhaseVariableValueMethod SettingsColumnHiTrap Heparin High-PerformanceMethod SettingsColumn Volume (CV)0.962 CVMethod SettingsFlow Rate (mL/min)0.5 ml/minFraction SettingsSample Fractionation Volume (mL)0.5 mLPrime (Pre-method run)1X CMF-PBS Volume (CV)10.00 CVEquilibrationEquilibration Volume (CV) using Buffer A3.00 CVSample runSample Volume (CV)5.00 CVWash-out unboundWash Volume (CV)1.00 CVElution step 1Target concentration of Buffer B: 10 %1.00 CVElution step 2Target concentration of Buffer B: 20 %1.00 CVElution step 3Target concentration of Buffer B: 30 %1.00 CVElution step 4Target concentration of Buffer B: 40 %1.00 CVElution step 5Target concentration of Buffer B: 50 %3.00 CVWash-out unboundWash Volume (CV)3.00 CVWater flush (Post-method run)Water volume (mL)25 mLStorage (Post-method run)20% ethanol volume (mL)15 mLc.Select the method of interest in “method run” and start the run. Figure 5 shows the general setup of the ÄKTA start™ Chromatography instrument using the Heparin column during a sample run.Figure 5The setup of Affinity-based Chromatography using Heparin column during sample run stepsd.The method begins with column equilibration for 3 CVs. After equilibration, a new run mark will appear on the chromatogram, and the system’s schematic will indicate that the flow is now coming from the sample line.e.As before with CC700, prepare a bottle of 1x CMF-PBS, a 1 mL pipette, and clean 1 mL tips. You must stay at the instrument and monitor vigilantly.Note: There is approximately 3 mL of sample, but the process will take 5 mL from the sample line. You must avoid any air from entering the sample line or column.i.To avoid air, monitor the sample tube.ii.Allow the sample to first reach near the bottom, then add 0.5 mL of 1x CMF-PBS.iii.Once it reaches near the bottom of the tube again, add another 0.5 mL of 1x CMF-PBS, ensuring the sample never reaches air.iv.Once that reaches near the bottom of the tube again, fill the tube up to the top with 1x CMF-PBS.v.Continue adding buffer until the end of the sample run step, when the machine switches to Buffer A.f.Allow the method to run, resuming with the Wash-out Unbound step, and then a step elution, and finally, the cleansing of the column. Allow the method to resume running until completion.g.Once the method is complete, you will observe 2 peaks on the obtained chromatogram. The first peak is the Non-Heparin Binding (NHB) peak. Pool those fractions together. The second peak, which came during the step elution, is the Heparin-Bound (HB) peak. Pool those fractions together.Note: You will be left with two samples, NHB and HB. These are your final products, and EVs are found in the NHB fraction. The HB fraction contains extracellular particles. Place all products in 4°C for up to 2 weeks, or aliquot and freeze at −80°C for long-term storage.Note: See Troubleshooting Problem 5 for how to resolve overloaded columns.30.Water Flush and Storage of the Heparin column.a.Water Flush: Place the sample tubing into a bottle of DI water. Direct the green fractionation tubing to waste. Start a “manual run” with a flow rate of 0.5 mL/min.Note: The Heparin method stated above already includes a “sanitization” of the column using Buffer A and Buffer B; therefore, further sanitization is not required. Keep in mind that the Heparin column will be destroyed if any NaOH is used to sanitize it.b.As before with equilibration of the columns, set the flow path to go from the sample, through the column, and through the fraction collector. In “flow path”, set “sample valve” to “sample”, “wash valve” to “column”, “outlet valve” to “collection”, and execute these commands. Allow 10 CV (10 mL) to flow through the column, which takes approximately 20 min. Then, “Pause”.c.Storage: Place the sample tubing into a bottle of 20% ethanol. Press “continue” and allow ethanol to flow for 10 CV (10 mL), which takes approximately 20 min.d.In “flow path”, set “outlet valve” to “waste”. Allow to flow for 3 CV (3 mL), which takes about 6 min. Afterward, press “End”.e.To remove the column, simply remove all connections to the column and replace them with proper plugs on both sides of the column. Reconnect the ends of the instrument together to ensure a closed system. Store the Heparin column at 25°C for short-term storage, or 4°C for long-term storage. If kept at 4°C, allow to reach room temperature before using again.f.The instrument and computer can be turned off now.
Purified EVs were comprehensively characterized based on particle size, concentration, charge, protein content, RNA content, and biological function.31.EV Characterization using Nanoparticle Tracking Analysis (NTA).Note: This step describes a method for determining the size distribution, concentration, and charge of particles in a sample using Brownian motion. Here, we used ZetaView® PMX-230 TWIN (Particle Metrix, Germany) Nanoparticle Tracking Analyzer for characterization of our purified EVs isolated from human plasma.Timing: ∼15 min per samplea.Turn on the instrument and follow the initialization prompts on ZetaView Software (version 8.06.01 SP1). Using 100 nm Nanosphere beads, standardize the instrument at the Sensitivity of 65 and Shutter of 100.b.Dilute the samples in ddH2O to stay within a range of 50–200 particles per view window. Inject the diluted sample into the NTA instrument.c.For particle concentration and size distribution measurements, acquire data with the Sensitivity at 85, Shutter at 200, Max Area at 1000, Min Area at 10, Min Brightness at 20, and Frame Rate at 30 fps with High video resolution, using a Laser Wavelength of 488 nm in Scatter mode of the filter, evaluated in 3 cycles and throughout 11 positions.d.For Zeta potential measurements, acquire data with the Sensitivity at 85, Shutter at 200, Max Area at 1000, Min Area at 10, Min Brightness at 20, and Frame Rate at 30 fps with High video resolution, using a Laser Wavelength of 488 nm in Scatter mode of the filter evaluated in 2 continuous stationary positions.32.EV Characterization using gel electrophoresis and Coomassie stain.Note: This step uses a qualitative approach to demonstrate protein content at different stages of the EV purification pipeline and to confirm the removal of albumin.Timing: 2 daysa.Sample Based on the NTA results on the particle concentration of all samples of interest, calculate the volume containing the highest amount of equal particles that can fit in the wells of your gel. Here, we have loaded 1.5 × 10^11^ particles per well for each sample. Add 5X Reducing Laemmli Buffer (with 3% β-mercaptoethanol added as the reducing agent) to the samples that will be targeted in reducing conditions.i.All samples need to be loaded at equal volume; thus, bring the sample volume up to reach the same value using PBS.ii.Vortex all samples gently for a few seconds to mix.iii.Heat all samples at 95°C for 15 min to ensure lysis and linearization of all the proteins in your samples.b.Gel Prepare the GenScript Tris-MOPS-SDS Running Buffer for gel electrophoresis based on the manufacturer’s instructions.i.Fill the gel electrophoresis tank with Running Buffer and carefully load the samples and protein ladder onto the GenScript SurePage gel or similar gels.ii.Using Biorad PowerPac™ Basic Power Supply, run the gel at a constant voltage of 80–120V for about 2 h, or until the dye at the front approaches the bottom of the gel.c.After the dye reaches the bottom of the gel, stop the current, and wash the gel with DI water 3 times.d.Incubate the gel in Coomassie stain for 2 h. Afterward, wash the gel with DI water at least 3 times.e.Incubate the gel in Destain solution for 10 min (Destain stock solution made per instructions in “materials and equipment” section).f.Wash the gel with DI water at least 3 times.g.Incubate the gel in DI water for 24 h and image the gel on the Invitrogen™ iBright™ CL1500 Imaging System the next day.33.EV Characterization using Western Blotting.Note: This step implements the well-known method of western blotting for the identification of proteins in samples and evaluates the presence of EV protein markers in our purified plasma.^11^^,^^12^^,^^13^Timing: 2 daysa.Sample Based on the NTA results on the particle concentration of all samples of interest, calculate the volume containing the highest amount of equal particles that can fit in the wells of your gel.Note: Here, we have loaded 1.5 × 10^11^ particles per well for each sample.i.Add 5X Non-reducing Laemmli Buffer to the samples that will be targeted in non-reducing conditions. On the other hand, add 5X Reducing Laemmli Buffer to the samples that will be targeted in reducing conditions (Table 4).Table 4The antibodies used for western blotting, their clones, preparation conditions, the dilutions for the primary antibody, and the secondary antibody type and dilution for each target proteinAntibodyCloneConditionPrimary antibody dilutionSecondary antibody (1:4000 dilution)CD63TS63Non-reducing1:3000Anti-MouseCD9MM2-57Non-reducing1:500Anti-MouseAlixE6P9BReducing1:2000Anti-RabbitSynteninEPR8102Reducing1:2000Anti-RabbitFlotillin-2C42A3Reducing1:2000Anti-RabbitTSG101EPR7130(B)Reducing1:2000Anti-RabbitHRS/HGSPolyclonalReducing1:2000Anti-Goatii.All samples need to be loaded at equal volume; thus, bring the sample volume up to reach the same value using PBS.iii.Vortex all samples gently for a few seconds to mix.iv.Heat all samples at 95°C for 15 min to ensure lysis and linearization of all the proteins in your samples.Note: The protocol for making Reducing and Non-reducing Laemmli Buffer has been provided in the “materials and equipment” section.b.Run the gel electrophoresis by following the same protocol as reported previously (in step 32b).c.Using Biorad Mini Trans-Blot Electrophoretic Transfer Cell system (includes two gel holder cassettes, foam pads, electrodes, tank, blue cooling unit, lid with cables) and a Biorad PowerPac™ Basic Power Supply, perform wet transfer of the gel onto a Nitrocellulose membrane at a constant voltage of 30V at 4°C for 24 h (or 120V for 2 h) in a tank filled with the 1X Transfer Buffer working solution, which is freshly made by mixing 700 mL DI water and 200 mL methanol with 100 mL of 10X Transfer Buffer stock solution (refer to “materials and equipment” for composition).d.Block the transferred membrane with 5% non-fat milk or 5% BSA in 1X TBST for 1 h on rotation.e.Discard the blocking solution and incubate the membranes with primary antibody for 24 h at 4°C.f.The following day, wash the membranes with 1X TBST 3 times for 5 min each. The instructions for making 1X TBST have been provided in the “materials and equipment” section.g.Incubate the membranes in the secondary antibody for 1 h. The antibodies used in this study are listed in key resources table and Table 4.h.Wash the membrane with 1X TBST 3 times for 5 min each, and image the blots on Invitrogen™ iBright™ CL1500 Imaging System. For more details on how to perform a western blot, you can refer to the ref.^4^^,^^13^34.Validation of external RNA removal – RNA Extraction.Note: To perform RNA extraction, we used miRNeasy Micro Kit (QIAGEN), which is used to purify total RNA and small RNAs. We followed every step of the kit’s protocol on 3 PEG pellet resuspended in 1X DNase buffer prior to adding Nuclease, Nuclease-treated sample after 37° C incubation prior to CC700 Chromatography, and the CC700 FT pooled fractions.Timing: ∼1 ha.RNA Extraction using miRNeasy Micro Kit: Follow the detailed steps of the miRNeasy Micro Handbook, which can be easily found on the QIAGEN website.b.After the RNA extraction is completed, store your RNA samples at 4°C for up to 2 weeks or preferably at −80°C for long-term storage.35.Validation of external RNA removal – Tape Station.Note: This step is conducted to assess the efficacy of our purification pipeline in purifying the free-floating RNA from the human plasma EV sample.Timing: ∼1 ha.Follow the manufacturer’s manual for operating Agilent 4200 TapeStation instrument, which can be found as “4200 TapeStation System Manual” on the Agilent website.b.To conduct the High Sensitivity RNA Screen Tape assay, follow the detailed instructions of the kit “High Sensitivity RNA ScreenTape Assay for TapeStation Systems”, which can be found on Agilent website, Document No: SD-UF0000089 Rev. D.c.Start this assay by adding 1 μL High Sensitivity RNA ScreenTape Sample Buffer to each tube.d.Add 2 μL of High Sensitivity RNA ScreenTape Ladder to the first tube containing the High Sensitivity RNA ScreenTape Sample Buffer. Mix by vortexing for 1 min. Spin the tube down for 1 min.e.Add 2 μL of your sample to the tube containing the High Sensitivity RNA ScreenTape Sample Buffer. Mix by vortexing for 1 min. Spin the tube down for 1 min.f.Heat the tubes on 72°C for 3 min.g.Cool down the samples on ice for 2 min.h.Spin the tubes down for 1 min. Load the tubes on the Agilent 4200 TapeStation instrument and run the assay.36.EV Characterization using Fusion Assay.Note: The EV fusion assay is used to functionally validate the biological integrity of purified EVs. This uptake assay demonstrates whether EVs retain the ability to interact with and fuse into recipient cell membranes and remain biologically active after extensive purification.Timing: ∼3 daysa.On day 1 of the fusion assay, seed 100,000 hTERT-HUVEC per well on a coverslip in a 6-well plate in 1 mL EBM-2 medium supplemented with 2% EV-depleted FBS and 1% Penicillin Streptomycin and incubate the plate for 24 h at 37°C.i.To create EV-depleted FBS, PEG is added to FBS to a final concentration of 4%. The mixture is then rotated for 24 h at 4°C. The following day, centrifuge the PEGylated sample at 1000 x g for 1 h at 4°C to pellet and remove the serum EVs. The supernatant, which is now EV-depleted FBS, is stored at −20°C for future use.b.The samples of interest include CC700 EV, NHB EV, HB peak, and No EV control. Prepare EV-DiI using equal EV particle number (based on the NTA results) and DiI dye to a final concentration of 10mM DiI in each EV sample. Incubate the EV-DiI samples on rotation at 4°C for 24 h.c.On day 2 of the fusion assay, prepare CC700 resin with 5 PBS washes as reported in “Capto Core 700 resin batch incubation” major step. Add the EV-DiI samples to the CC700 resin (1:2) and incubate at 4°C on rotation for 1 h to filter out the free dye that wasn’t uptaken by EVs. Then centrifuge the solutions for 3 min at 500 x g and transfer the top aqueous layer to new tubes.d.Add the EV-DiI samples on top of the hTERT-HUVEC cells in different wells and incubate at 37°C for 8 h.e.After the incubation, gently wash the wells with PBS 3 times to remove the EVs that were not uptaken by the cells. Fix the cells by incubating in 4% paraformaldehyde (PFA) for 10 min at 25°C. Discard PFA and wash the wells with PBS another 3 times.f.Prepare DAPI stain stock solution by dissolving 10 mg DAPI in 2 mL of DI water as per the manufacturer’s instructions. Add 1 mL DAPI stain to a final concentration of 0.5 μg/mL (equal to 10000 dilution of the DAPI stock solution in PBS) on each coverslip in all wells and incubate for 10 min at 25°C, followed by 3 PBS washes.g.Mount the coverslips on a microscope glass slide using mounting oil and let them dry for 24 h at 25°C.h.On day 3, image the slides on a widefield microscope.37.EV imaging using dSTORM.Note: dSTORM is used to directly visualize individual EVs at nanoscale resolution, validate EV purity and identity by confirming EV size and morphology at the single-particle level, and demonstrate spatial localization and co-expression of canonical EV surface markers including CD9 and CD63 as definitive structural evidence of EVs.Timing: ∼3 daysa.Stain CD9 antibody with 488 labeling kit in accordance with Alexa Fluor® Antibody Labeling Kits’ instructions.b.Stain CD63 antibody with 568 labeling kit in accordance with Alexa Fluor® Antibody Labeling Kits’ instructions.c.Antibody conjugation manufacturer’s i.Prepare a 1 M sodium bicarbonate solution by resuspending Component B in 1 mL of deionized water. These components are provided in the kit.ii.Adjust the antibody concentration to 1.0 mg/mL, then add 1/10th volume of the 1 M sodium bicarbonate solution.iii.Add 100 μL of the antibody to the vial of Alexa Fluor® dye. Invert to mix and incubate for 1 h at 25°C.iv.Assemble the Purification Column. Prepare 1.5 mL of resin bed and centrifuge the column at 1100 x g for 3 min.v.Add the reaction mix and centrifuge the column at 1100 x g for 5 min to collect the labeled antibody.d.dSTORM staining i.Prepare a solution of PBS+ 0.1% Tween-20 (PBS-T) for washes and a solution of 2% BSA in PBS-T as the block buffer.ii.Based on the NTA results, calculate the volume containing 1.5 × 10^9^ EVs in 200 μL of PBS, and stain with 1 μl of CMDR on ice for 30 min.Optional: You can also stain at 4°C for 16 h.iii.Add 200 μL of Poly-Lysine solution to ibidi 8-well glass bottom chamber slides for 15 min at 25°C.Optional: You can also incubate at 4°C for 16 h.iv.Remove Poly-Lysine and wash the slide 3 times with PBS.v.Add the EVs to slides and incubate at 4°C for 16 h.vi.Carefully remove PBS from the EVs in the slide and add 200 μL of 0.5% PFA to wells. Incubate at 25°C for 30 min.vii.Remove PFA and wash the slide 3 times with PBS.viii.Add 200 μL of blocking buffer (2% BSA in PBS-T) to the slide. Incubate at 25°C for 30 min.ix.Dilute conjugated antibodies in block buffer by a 200 ratio.x.Remove block buffer from wells and add 200 μL of the antibody solution. Incubate at 25°C for 2 h.Optional: You can also incubate at 4°C for 16 h.xi.Remove antibody solution and wash wells 3 times with PBS-T.xii.Remove PBS-T and add fresh PBS until ready to image. Slides can be kept at 4°C for up to 1 week for imaging.e.dSTORM Imaging Instructions:i.Make OSB based on the protocol in the “materials and equipment” section. First, dilute glucose and BME in TN buffer. Once the glucose has been fully dissolved, add the glucose oxidase and catalase.Note: The enzymes are only good for a maximum of 3 h at 25°C after being added to the buffer.ii.Remove PBS from wells and add 200 μL of OSB.iii.Add oil to the microscope objective, then carefully add the slide as to not spill any liquid.iv.Focus your sample using the 647 laser (on the CMDR stain).v.Set laser power by seeing when the fluorophores start to blink. Every microscope will have different settings. This paper used an Abbelight system.CRITICAL: The laser power will have to be high for the fluorophores to blink, or photoswitch.vi.Image your samples, starting from the longest wavelength (i.e., 647) to the shortest wavelength (i.e., 488). Take between 5,000 and 10,000 frames per color to build an accurate reconstruction.vii.After images were taken, the data was uploaded to Nanometrix image analysis software for processing.
In this article, we have proposed a detailed experimental protocol for the purification and characterization of EVs from human plasma. Figure 6 shows the Capto Core 700 curve for human plasma samples, demonstrating the replicability of the EV peak throughout 3 trials. The collected fractions from this peak were further purified using Affinity-based chromatography, and Figure 7 shows the Heparin curve for human plasma samples, which reveals the NHB peak containing high-purity human plasma EVs. We have characterized purified EVs based on their particle size, concentration, charge, protein content, RNA content, and fusion capability.Figure 6Capto Core 700 curve of human plasma samples demonstrating the UV absorbance vs volume pumped in the system and the designed method stepsThis peak contains our plasma EVs, and these fractions were collected.Figure 7Heparin curve of human plasma samples demonstrating the UV absorbance and Conductivity vs volume pumped in the system and the designed method stepsThe NHB peak contains our plasma EVs and the fractions for both peaks were collected for further analysis.
The particle concentration, size distribution, and zeta potential of our human plasma samples were measured by NTA and are shown in Figure 8. Our purification pipeline results in a high yield of pure human plasma EVs in the NHB fraction, which falls within a size range of 80–120 nm and a charge of −27.86 mV + 1.74 mV, in line with the values reported for exosomes and small EVs isolated from human plasma in the literature.^14^^,^^15^^,^^16^ Table 5 presents the expected volume, total particle number, step recovery rate, and cumulative yield rate for each purification stage.Figure 8Nanoparticle tracking analysis of human plasma EV samples at different steps of purificationNTA results for the human plasma and EV samples showing (A) the size distribution of particles, (B) normalized frequency distribution of the size of NHB EVs, (C) distribution of particle concentration, and (D) zeta potential distribution.Table 5Expected volume and yield for each purification stageSampleVolume (mL)Median diameter ± SDParticle concentration (Particle/mL)Total particle count (particles)Step recovery (%)Cumulative yield (%)Filtered plasma50086.0 ± 58.17.8 × 10^11^3.9 × 10^14^100%100%Post-TFF5075.2 ± 42.73.1 × 10^12^1.6 × 10^14^41%41%Post-CC7003.5103.1 ± 52.01.3 × 10^13^4.6 × 10^13^28%12%NHB3110.6 ± 57.31.2 × 10^13^3.6 × 10^13^78%9%HB2.5114.2 ± 64.23.6 × 10^12^9 × 10^12^25%2%Note that this differs from a traditional purification yield table as EVs are a heterogenous mixture that is progressively enriched across sequential purification steps. Hence, the particles counted in the early stages are not identical to the particles in the later stages. For instance, the initial filtered plasma sample contains not only EVs but also other nanoscopic contaminants and protein aggregates that are not EVs, while the NHB fraction is highly enriched for pure EVs. The step recovery and cumulative yield refer to the EV class enriched by the particular purification step. The total particle count was calculated by multiplying particle concentrations with volume. The step recovery rate was calculated by dividing particle count throughout each stage by that of the previous stage. The cumulative yield rate was calculated by dividing particle count throughout each stage by that of the filtered plasma.
The Coomassie stain in Figure 9 indicates that our purification process effectively removes albumin, major contaminants, and non-EV proteins from the human plasma sample, yielding a high-purity NHB fraction that is highly enriched in EV proteins.Figure 9Coomassie stain illustrating the protein content in different stages of the purification process
As shown by the western blot in Figure 10, the CC700 peak and non-Heparin binding fractions are highly enriched for EV markers, including CD63, CD9, Alix, TSG101, Syntenin, Flotillin, and HRS, while all these markers are absent in the heparin-binding fractions.Figure 10The bold presence of EV markers in our purified EV fractions as shown by western blotting
To assess the biological function of the purified EVs, we conducted an EV uptake and fusion assay using hTERT-HUVEC cells and DiI dye, imaged on a widefield microscope. Exposure time was kept constant across all images, and only the signal-to-noise ratio was slightly adjusted. Figure 11 indicates the successful uptake of our EVs by hTERT-HUVEC cells and their fusion capability.Figure 11Fusion assay demonstrating the uptake of EVs from different fractions by cellsEV populations were stained with DiI dye, then incubated with hTERT-HUVEC cells for 16 h. Cells were then fixed and DNA was stained with DAPI. Images taken at 20X and 100X magnification. Scale bar=10 μm
Furthermore, to evaluate the removal of free-floating RNA from the human plasma sample, we used TapeStation to analyze 3 the pre-nuclease sample (PEGylated EV pellet resuspended in 1X DNase buffer), the nuclease-treated sample, and the CC700 FT fractions (peak 1). As shown in TapeStation results in Figure 12, the nuclease step removed the majority of free-floating RNA in the EV sample, and the CC700 chromatography successfully removed any excess nuclease and RNA smears remained in the EV sample.Figure 12Validation of external RNA removal using TapeStation
For visualization and structural validation, dSTORM was used to image NHB EVs stained with antibodies against CD63 (568-green) and CD9 (488-blue). Figure 13 represents the dSTORM images of the EVs in NHB fraction as the final pure product of this protocol, which contains individual EVs with correct size, morphology, and marker distribution, free from aggregation and major contaminants.Figure 13Representative image of dSTORM analysis of NHB fractionAfter purification of EVs by heparin chromatography, EVs were stained with CellMask DeepRed (647-magenta). EVs were then seeded, fixed, and stained with antibodies against CD63 (568-green) and CD9 (488-blue). EVs were then imaged on an Abbelight system, with 10,000 frames taken per laser color. After image acquisition, the data was then analyzed using the Nanometrix Pipeline (v2.7.7.5) and single particle data was obtained. Scale bar = 1 μm in large square, and scale bar = 200 nm in small squares.
Plasma has an abundance of proteins and physiological protein aggregates that are not considered EVs, such as high-density and low-density lipoprotein aggregates.^17^ These can diminish the purification efficiency by overwhelming the capacity at multiple steps. Platelets are one of the major contaminants for plasma-derived EVs due to size overlap, but our purification pipeline effectively removes cell components and cell fragments through the size-exclusion filter steps. Additionally, other free soluble proteins, such as serum albumin, immunoglobulins, and fibrinogen, are major contaminants of plasma-derived EVs, but our elaborate protocol can successfully remove these proteins from the NHB EV sample.
Although the proposed purification process results in a high yield of clean EVs, it can be challenging to find the proper dilution factor for the plasma input into the TFF, the amount of resin needed for the pre-PEG batch purification, and the CC700 and Heparin column bed volume. If need be, the listed quantities can be doubled or quadrupled.
The rich protein content of plasma EVs can exceed the maximum threshold of UV absorbance in CC700 and Heparin columns’ limit of detection on FPLC; therefore, it is important to find the proper dilution factor to stay within the range of detection of these columns on your FPLC system, as the accuracy of these FPLC peaks directly impacts your EV fractions.
The purification final output is mostly enriched in small EVs (NHB Peak).
If you’re more interested in large EV subpopulations, skip the 0.22μm filtration (Step 4) to allow the larger EVs through the purification pipeline.
The PEG pellet is too hard and rigid to resuspend by pipetting after centrifugation. (Steps 19–21).
Plasma samples being highly enriched in proteins can make it hard to resuspend the PEG pellet by pipetting. In this case, you can consider letting the pellet soak in the DNase buffer for a few minutes on gentle agitation at 4 °C until the pellet loosens, then continue to mix by pipetting with a wide-bore tip. If the pellet is still very thick and impossible to mix, consider adding more cleaning rounds using CC700 resin in the previous steps or at higher ratio of the resin (Steps 12–15). If none of the above solutions resolved your problem, consider using a higher dilution factor for the initial TFF input when you start your purification process in the next trials (Steps 4, or 7–8).
Leakage from chromatography columns (step 25b and 28c).
As the instrument’s tubing isn’t the best size for the connectors, there may be some leaks. To amend these issues, use small amounts of parafilm to create seals on the tubing to prevent leaking.
Air entered the sample tubing during transfer to the sample (Step 26b).
If air enters the sample tubing, run a manual method and flush the sample line with 1x CMF-PBS through the column valve to waste at a flow rate of 1–2 mL/min. Then, redirect the sample line from the 1X CMF-PBS to the tube of nuclease-treated sample. Ensure the line reaches the bottom of the tube and no air enters the system.
Overloaded CC700 and/or Heparin columns (peaks exceeding the range of detection of your column and chromatography instrument) (Steps 24–30).
Diluting the initial input before TFF step can be the most effective solution (Step 3–4, or 7–8). If your sample is too concentrated, it can decrease the efficiency of the TFF filter, CC700 beads, PEG precipitation, CC700 and Heparin FPLC columns. It might take some trial and error to find the proper dilution factor for your sample. For our samples, diluting the plasma sample in 1X CMF-PBS (1:10) prior to 0.22 μm filtration as the TFF initial input starting with a particle count around 1 × 10^14^ (based on NTA results) yielded the best results. The necessary dilution factor and particle count can be different in plasma samples based on the plasma source.
Dr. Dirk P. Dittmer, dirkdittmer@me.com.
Saba Zanganeh, sabaz@unc.edu.
This study did not generate new unique reagents.
This study did not generate datasets and code.
We thank Linda J. Pluta and Kyle W. Shifflet for technical support and helpful discussion. The project is supported by an NCI Cancer Center Core Support Grant (CA016086) and public health service grants CA163217 and DE034179 to DPD. Graphical abstract was created using Biorender.com.
D.D., S.Z., G.F.A., and A.C. conceived the project. D.D. supervised the project. S.Z. and A.C. conducted the experiments. S.Z., A.C., and R.Y. analyzed and visualized the data. S.Z. wrote the manuscript, which was reviewed and edited by all authors.
Neither the funders nor the University of North Carolina had any role in the study design, data collection, or interpretation, and they were not involved in the opinions expressed here. Dittmer and Cone declare competing interests regarding the possible commercialization of some of the information presented. The University of North Carolina manages these. Dittmer serves as CEO of Vironomics, LLC.