Standardization of extracellular vesicle measurements by flow cytometry through vesicle diameter approximation
Essentials•Platelet extracellular vesicles (EVs) concentrations measured by flow cytometers are incomparable.•A model is applied to convert ambiguous scatter units to EV diameter in nanometer.•Most included flow cytometers lack the sensitivity to detect EVs of 600 nm and smaller.•The model outperforms polystyrene beads for comparability of platelet EV concentrations.Summary: BackgroundDetection of extracellular vesicles (EVs) by flow cytometry has poor interlaboratory comparability, owing to differences in flow cytometer (FCM) sensitivity. Previous workshops distributed polystyrene beads to set a scatter‐based diameter gate in order to improve the comparability of EV concentration measurements. However, polystyrene beads provide limited insights into the diameter of detected EVs.ObjectivesTo evaluate gates based on the estimated diameter of EVs instead of beads.MethodsA calibration bead mixture and platelet EV samples were distributed to 33 participants. Beads and a light scattering model were used to set EV diameter gates in order to measure the concentration of CD61–phycoerythrin‐positive platelet EVs.ResultsOf the 46 evaluated FCMs, 21 FCMs detected the 600–1200‐nm EV diameter gate. The 1200–3000‐nm EV diameter gate was detected by 31 FCMs, with a measured EV concentration interlaboratory variability of 81% as compared with 139% with the bead diameter gate. Part of the variation in both approaches is caused by precipitation in some of the provided platelet EV samples. Flow rate calibration proved essential because systems configured to 60 μL min−1 differed six‐fold in measured flow rates between instruments. ConclusionsEV diameter gates improve the interlaboratory variability as compared with previous approaches. Of the evaluated FCMs, 24% could not detect 400‐nm polystyrene beads, and such instruments have limited utility for EV research. Finally, considerable differences were observed in sensitivity between optically similar instruments, indicating that maintenance and training affect the sensitivity. Essentials•Platelet extracellular vesicles (EVs) concentrations measured by flow cytometers are incomparable.•A model is applied to convert ambiguous scatter units to EV diameter in nanometer.•Most included flow cytometers lack the sensitivity to detect EVs of 600 nm and smaller.•The model outperforms polystyrene beads for comparability of platelet EV concentrations. •Platelet extracellular vesicles (EVs) concentrations measured by flow cytometers are incomparable.•A model is applied to convert ambiguous scatter units to EV diameter in nanometer.•Most included flow cytometers lack the sensitivity to detect EVs of 600 nm and smaller.•The model outperforms polystyrene beads for comparability of platelet EV concentrations. Detection of extracellular vesicles (EVs) by flow cytometry has poor interlaboratory comparability, owing to differences in flow cytometer (FCM) sensitivity. Previous workshops distributed polystyrene beads to set a scatter‐based diameter gate in order to improve the comparability of EV concentration measurements. However, polystyrene beads provide limited insights into the diameter of detected EVs. To evaluate gates based on the estimated diameter of EVs instead of beads. A calibration bead mixture and platelet EV samples were distributed to 33 participants. Beads and a light scattering model were used to set EV diameter gates in order to measure the concentration of CD61–phycoerythrin‐positive platelet EVs. Of the 46 evaluated FCMs, 21 FCMs detected the 600–1200‐nm EV diameter gate. The 1200–3000‐nm EV diameter gate was detected by 31 FCMs, with a measured EV concentration interlaboratory variability of 81% as compared with 139% with the bead diameter gate. Part of the variation in both approaches is caused by precipitation in some of the provided platelet EV samples. Flow rate calibration proved essential because systems configured to 60 μL min−1 differed six‐fold in measured flow rates between instruments. Conclusions EV diameter gates improve the interlaboratory variability as compared with previous approaches. Of the evaluated FCMs, 24% could not detect 400‐nm polystyrene beads, and such instruments have limited utility for EV research. Finally, considerable differences were observed in sensitivity between optically similar instruments, indicating that maintenance and training affect the sensitivity. Extracellular vesicles (EVs) are membrane‐enclosed, cell‐derived particles 1.Brisson A.R. Tan S. Linares R. Gounou C. Arraud N. Extracellular vesicles from activated platelets: a semiquantitative cryo‐electron microscopy and immuno‐gold labeling study.Platelets. 2017; 28: 1-9Crossref Scopus (102) Google Scholar. The pathology of many diseases, including arterial thrombosis, is characterized by elevated concentrations of circulating EVs 2.Gasecka A. Böing A.N. Filipiak K.J. Nieuwland R. Platelet extracellular vesicles as biomarkers for arterial thrombosis.Platelets. 2017; 28: 228-34Crossref PubMed Scopus (41) Google Scholar. As a basis for diagnosis, EV concentrations need to be detectable with sufficient reproducibility between laboratories. EV concentrations can be determined with flow cytometry. A flow cytometer (FCM) can simultaneously detect light scatter and fluorescence signals of single EVs at a rate in excess of 1000 s–1, provided that swarming is avoided 3.Van Der Pol E. van Gemert M. Sturk A. Nieuwland R. Van Leeuwen T. Single vs. swarm detection of microparticles and exosomes by flow cytometry.J Thromb Haemost. 2012; 10: 919-30Crossref PubMed Scopus (288) Google Scholar. However, because of the diameter of EVs, both scatter and fluorescence signals are dim, and the smallest detectable EV diameter differs between FCMs, owing to differences in sensitivity between instruments. Because the smallest EVs are at least an order of magnitude more frequent than the larger EVs 4.van der Pol E. Coumans F.A.W. Gardiner C. Sargent I.L. Harrison P. Sturk A. van Leeuwen T.G. Nieuwland R. Particle size distribution of exosomes and microvesicles by transmission electron microscopy, flow cytometry, nanoparticle tracking analysis, and resistive pulse sensing.J Thromb Haemost. 2014; 12: 1182-92Crossref PubMed Scopus (560) Google Scholar, 5.Robert S. Poncelet P. Lacroix R. Arnaud L. Giraudo L. Hauchard A. Sampol J. Dignat‐George F. Standardization of platelet‐derived microparticle counting using calibrated beads and a Cytomics FC500 routine flow cytometer: a first step towards multicenter studies?.J Thromb Haemost. 2008; 7: 190-7Crossref PubMed Scopus (260) Google Scholar, the minimum EV diameter detectable by each FCM will directly affect the measured EV concentration. A method for measuring EVs in the same diameter range should therefore result in comparable EV concentrations between instruments. Two previous workshops were initiated by the Scientific Standardization Committee on Vascular Biology of the ISTH to standardize a detected diameter range of EVs 6.Lacroix R. Robert S. Poncelet P. Kasthuri R. Key N. Dignat‐George F. Standardization of platelet‐derived microparticle enumeration by flow cytometry with calibrated beads: results of the International Society on Thrombosis and Haemostasis SSC Collaborative workshop.J Thromb Haemost. 2010; 8: 2571-4Crossref PubMed Scopus (292) Google Scholar, 7.Cointe S. Judicone C. Robert S. Mooberry M.J. Poncelet P. Wauben M. Nieuwland R. Key N.S. Dignat‐George F. Lacroix R. Standardization of microparticle enumeration across different flow cytometry platforms: results of a multicenter collaborative workshop.J Thromb Haemost. 2016; 15: 187-93Abstract Full Text Full Text PDF PubMed Scopus (90) Google Scholar. These workshops reduced the interlaboratory variability of platelet EV concentration measurements by setting diameter gates based on measuring polystyrene beads. The second workshop accounted for differences between forward scatter (FSC) and side scatter (SSC) collection angles, by applying different beads for FSC and SSC. However, neither workshop accounted for variations in FSC and SSC collection angles, or for newer, high‐sensitivity, FCM designs 8.van der Vlist E.J. Nolte E.N.M. Stoorvogel W. Arkesteijn G.J.A. Wauben M.H.M. Fluorescent labeling of nano‐sized vesicles released by cells and subsequent quantitative and qualitative analysis by high‐resolution flow cytometry.Nat Protoc. 2012; 7: 1311-26Crossref PubMed Scopus (386) Google Scholar, 9.Chandler W. Yeung W. Tait J. A new microparticle size calibration standard for use in measuring smaller microparticles using a new flow cytometer.J Thromb Haemost. 2011; 9: 1216-24Crossref PubMed Scopus (213) Google Scholar. For both workshops, the diameter range of the selected EVs within the applied bead diameter gate is unknown. Therefore, the aim of this study was to standardize concentration measurements of EVs within the same diameter range by modeling the scatter–diameter relationship for different FCMs. The scatter–diameter relationship depends on the refractive index (RI) of EVs and the FCM configuration, including fluidics, optics, electronics, and settings. In previous studies, the RI of EVs was typically assumed to be constant, with a value near 1.40 10.Konokhova A.I. Yurkin M.A. Moskalensky A.E. Chernyshev A.V. Tsvetovskaya G.A. Chikova E.D. Maltsev V.P. Light‐scattering flow cytometry for identification and characterization of blood microparticles.J Biomed Optics. 2012; 17: 057006Crossref Scopus (43) Google Scholar, 11.Gardiner C. Shaw M. Hole P. Smith J. Tannetta D. Redman C.W. Sargent I.L. Measurement of refractive index by nanoparticle tracking analysis reveals heterogeneity in extracellular vesicles.J Extracell Vesicles. 2014; 3: 25361Crossref PubMed Scopus (103) Google Scholar. All RIs given in the text are for a wavelength of 488 nm. The scatter–diameter relationship can be computed for FCM configurations by the use of matlab scripts from Mätzler 12.Maetzler C. MATLAB functions for Mie scattering and absorption. Institute of Applied Physics, University of Bern, Bern2002Google Scholar, based on Mie theory 13.Bohren C.F. Huffman D.R. Absorption and Scattering by a Sphere Absorption and Scattering of Light by Small Particles. Wiley‐VCH, New York, NY2007: 82-129Google Scholar. In our approach, polystyrene beads are used to establish the scatter–diameter relationship for EVs 3.Van Der Pol E. van Gemert M. Sturk A. Nieuwland R. Van Leeuwen T. Single vs. swarm detection of microparticles and exosomes by flow cytometry.J Thromb Haemost. 2012; 10: 919-30Crossref PubMed Scopus (288) Google Scholar. Figure 1 shows the scatter–diameter relationships for three FCMs that differ in optical configuration, and therefore have different scatter–diameter relationships. Figure 1A shows FSCs and Fig. 1B and Fig. 1C show two different SSC configurations of three different FCMs. The scatter signal of a 400‐nm polystyrene bead (RI = 1.61, black ball in the inset) corresponds to an EV diameter of 700 nm (Fig. 1A), 1790 nm (Fig. 1B), and 1450 nm (Fig. 1C; the orange line and the inset green ball both indicate the EV diameter for RI = 1.40) 3.Van Der Pol E. van Gemert M. Sturk A. Nieuwland R. Van Leeuwen T. Single vs. swarm detection of microparticles and exosomes by flow cytometry.J Thromb Haemost. 2012; 10: 919-30Crossref PubMed Scopus (288) Google Scholar. From this example, it is clear that adjustment for the differences in collection optics between instruments is essential to set comparable EV diameter gates. In the approach tested here, the scatter–diameter relationship for EVs is used to find the EV diameter gates. The choice of EV diameter gate needs consideration. Because some FCMs are insensitive by design, an all‐inclusive strategy will suffer from the least sensitive FCM, and thus ends up gating only large EVs. On the other hand, requiring all EV researchers to work only on FCMs with ‘state‐of‐the‐art’ sensitivity is not feasible. Therefore, we decided to determine the concentrations of EVs in three different diameter gates, whereby EVs within the largest diameter gate are expected to be detectable by most FCMs, and those within the smallest diameter gate by only a few FCMs. These three ranges are 300–600 nm, 600–1200 nm, and 1200–3000 nm. Because the 1200–3000‐nm diameter range of EVs considerably overlaps with platelet diameters, it is impossible to distinguish platelet‐derived EVs from platelets within this diameter range. To evaluate a scatter–diameter model to standardize EV concentration measurements on FCMs. Study participants first calibrated their instruments to determine the EV diameter gates that they could measure. Participants who could measure at least one EV diameter gate proceeded to measure the flow rate and CD61–phycoerythrin (PE)‐stained EV samples. The EV concentration was determined on the basis of immunofluorescence and the detectable EV diameter gates. For comparison, the EV concentration was also determined for a bead diameter gate and by fluorescence alone (i.e. no diameter gate). All participants had a publication track record on detection of EVs by flow cytometry. In addition, two participants without an EV track record but who were very knowledgeable about FCMs or standardization were invited to take part. We distributed frozen aliquots of platelet EVs, also known as ‘platelet microparticles’. See Data S1 for additional details. The concentrations of staining reagents were 5.2 μg mL−1 for lactadherin–fluorescein isothiocyanate (FITC) (Haematologic Technologies, Essex Junction, VT, USA), 0.65 μg mL−1 for CD61–PE (VIPL2; BD Biosciences, San Jose, CA, USA), and 1.8 μg mL−1 for IgG1–PE (mouse BALB/c IgG1, κ; BD Biosciences). Simultaneously, an aliquot of the sample was thawed on melting ice for 1 h and diluted 1 : 7 (v/v) with phosphate‐buffered saline (PBS). The sample was visually inspected for clumps. For staining, 5 μL of lactadherin–FITC and 5 μL of antibody conjugate (either CD61–PE or IgG–PE) was added to 40 μL of sample and incubated for 15 min in the dark. Then, the mixture was diluted with 550 μL of PBS. Besides isotype controls, PBS with and without staining reagent was measured. The Rosetta Calibration system (Exometry, Amsterdam, the Netherlands) was used to determine the diameter gates for each flow cytometer. Rosetta Calibration consists of software and a mixture of seven types of polystyrene bead with traceable mean diameters between 100 nm and 1000 nm. Fluorescent polystyrene beads of 400 nm are included as a marker. The bead mixture is analyzed on the FCM, with the for EV The are analyzed by a software that the bead and a Mie model to the the optical of the used from the for each FCM, optical are to different for each FCM, for differences in both and FCM der Pol E. L. Coumans F.A.W. Böing A.N. Sturk A. Nieuwland R. van Leeuwen T.G. and identification of extracellular vesicles by flow PubMed Scopus Google Scholar, der Pol E. Coumans Sturk A. Nieuwland R. van Leeuwen T.G. index of in using nanoparticle tracking 2014; PubMed Scopus Google for more details. the of an RI of 1.40 for EVs 10.Konokhova A.I. Yurkin M.A. Moskalensky A.E. Chernyshev A.V. Tsvetovskaya G.A. Chikova E.D. Maltsev V.P. Light‐scattering flow cytometry for identification and characterization of blood microparticles.J Biomed Optics. 2012; 17: 057006Crossref Scopus (43) Google Scholar, the scatter–diameter relationship for EVs is The of the software is the Mie model for of including EVs (RI = 1.40) 10.Konokhova A.I. Yurkin M.A. Moskalensky A.E. Chernyshev A.V. Tsvetovskaya G.A. Chikova E.D. Maltsev V.P. Light‐scattering flow cytometry for identification and characterization of blood microparticles.J Biomed Optics. 2012; 17: 057006Crossref Scopus (43) Google Scholar, der Pol E. Coumans Sturk A. Nieuwland R. van Leeuwen T.G. index of in using nanoparticle tracking 2014; PubMed Scopus Google Scholar, polystyrene (RI = and (RI = EV diameter gates are given in scatter units that to three EV diameter 1200–3000 nm, 600–1200 nm, and 300–600 nm. These gates in scatter units can be applied in the analysis software of the The gates were selected to of the scatter–diameter relationship at gate for the included FCMs. measured the Rosetta Calibration bead and the to the The were analyzed with the Rosetta Calibration software to determine the detectable gates. the sensitivity was than expected based on the FCM participants were to their to evaluate the sensitivity could be Participants who could measure the 1200–3000‐nm gate to measurements on the platelet EVs. To the participants measured a bead mixture and beads of the EV as the of EVs the of EVs detected and the sample The measured sample can be from the of flow rate and FCM setting of the flow rate in μL this flow rate is typically and calibration the flow rate is We evaluated two for the flow beads and was as in the a a known of beads is in a known and the of beads measured is directly with the sample that The is determined by a sample and a Participants configured their FCMs to We applying a flow rate of 60 μL measuring for 1 setting the scatter such that no and setting the such that isotype signals are in the first Participants who a flow rate were to measure for to detect a sufficient of EVs. Participants measured the and their analysis, and the determined and EV concentrations with the to the we determined the EV concentration on the basis of immunofluorescence alone to establish the of a no diameter gate. we selected polystyrene bead diameter gates for with approaches 4.van der Pol E. Coumans F.A.W. Gardiner C. Sargent I.L. Harrison P. Sturk A. van Leeuwen T.G. Nieuwland R. Particle size distribution of exosomes and microvesicles by transmission electron microscopy, flow cytometry, nanoparticle tracking analysis, and resistive pulse sensing.J Thromb Haemost. 2014; 12: 1182-92Crossref PubMed Scopus (560) Google Scholar, 5.Robert S. Poncelet P. Lacroix R. Arnaud L. Giraudo L. Hauchard A. Sampol J. Dignat‐George F. Standardization of platelet‐derived microparticle counting using calibrated beads and a Cytomics FC500 routine flow cytometer: a first step towards multicenter studies?.J Thromb Haemost. 2008; 7: 190-7Crossref PubMed Scopus (260) Google Scholar. to the FCM workshop 7.Cointe S. Judicone C. Robert S. Mooberry M.J. Poncelet P. Wauben M. Nieuwland R. Key N.S. Dignat‐George F. Lacroix R. Standardization of microparticle enumeration across different flow cytometry platforms: results of a multicenter collaborative workshop.J Thromb Haemost. 2016; 15: 187-93Abstract Full Text Full Text PDF PubMed Scopus (90) Google Scholar, we selected two bead diameter gates to for differences between SSC and in a bead diameter gate for SSC and a bead diameter gate for both are comparable to a 1200–3000‐nm EV diameter gate the EV concentration in the 1200–3000‐nm EV diameter gate was compared with those in the no diameter gate and the bead diameter gate. we determined EV concentrations in EV diameter gates of 600–1200 nm and 300–600 nm. Figure shows the no diameter bead diameter 1200–3000 nm EV diameter 600–1200 nm EV diameter gate and 300–600 nm EV diameter gate in a selected To the for each gate we determined the of variation of standard of EV concentrations from all FCMs. The bead diameter gates were applied with because the scatter signals for the beads were not in the All other gating was by the participants. All analysis was in matlab Data are The 33 participants a of FCMs. A of FCMs from platelet EV samples. Figure the for Of the FCMs, not 15 because the had to FCMs, but only to measure on one or two FCMs. Two FCMs could not be owing to and one FCM was no The 46 FCMs measured the Rosetta Calibration beads to the sensitivity. Of 46 FCMs, were because the FCMs could not measure scatter from a 400‐nm polystyrene is for the FCM three because the had from a similar FCM, three because they were no and three because participants not All participants measured the Rosetta Calibration beads with the FCM that they applied to study EVs (Fig. FCMs were not sensitive to detect 400‐nm polystyrene beads. Because detection of 400‐nm beads is essential for identification of the different bead the could not determine the optical of the FCM, no scatter–diameter relationship for FCMs could be In to the previous workshops 6.Lacroix R. Robert S. Poncelet P. Kasthuri R. Key N. Dignat‐George F. Standardization of platelet‐derived microparticle enumeration by flow cytometry with calibrated beads: results of the International Society on Thrombosis and Haemostasis SSC Collaborative workshop.J Thromb Haemost. 2010; 8: 2571-4Crossref PubMed Scopus (292) Google Scholar, 7.Cointe S. Judicone C. Robert S. Mooberry M.J. Poncelet P. Wauben M. Nieuwland R. Key N.S. Dignat‐George F. Lacroix R. Standardization of microparticle enumeration across different flow cytometry platforms: results of a multicenter collaborative workshop.J Thromb Haemost. 2016; 15: 187-93Abstract Full Text Full Text PDF PubMed Scopus (90) Google Scholar, we the signals of beads to the to the to find the a scatter signal that is by is not for EV diameter The model was applied to FSC for the and and BD systems 7.Cointe S. Judicone C. Robert S. Mooberry M.J. Poncelet P. Wauben M. Nieuwland R. Key N.S. Dignat‐George F. Lacroix R. Standardization of microparticle enumeration across different flow cytometry platforms: results of a multicenter collaborative workshop.J Thromb Haemost. 2016; 15: 187-93Abstract Full Text Full Text PDF PubMed Scopus (90) Google Scholar. Of the FCMs, the FCM, two of FCMs and one of FCMs could detect the 400‐nm beads. The model was applied to SSC for the FCMs. all BD FCMs and one FCM could not detect the 400‐nm beads. For the FCMs in this the scatter from 400‐nm polystyrene beads corresponds to diameters of nm on FSC and nm on with the value on the of the FCM (Fig. it is for FCMs to detect the 1200–3000‐nm EV diameter but impossible for to detect the 600–1200‐nm EV diameter gate. FCM that detect in the 600–1200‐nm EV diameter gate has limited utility for EV research. For the FCMs that could detect the 400‐nm beads, the model could be applied to determine the EV diameter gates. All participants could the EV diameter gates in their analysis The mean diameters of beads nm and were estimated to be nm of and nm of S1 and Fig. S1 for The largest into an of 100 nm on the EV is a range as compared with the EV diameter to the signal of a 400‐nm polystyrene on the FCM the three FCMs with the largest in bead diameter had for bead that be by optical Of the FCMs that could be had sufficient sensitivity to detect EVs in the 1200–3000‐nm gate. Of FCMs, could measure EVs in the 600–1200‐nm and could measure EVs in the gate. The sample flow rate was set to 60 μL min−1 by of and participants used a flow rate between 5 μL min−1 and μL min−1 because their FCMs had at flow Figure shows the configured rate the rate measured with beads. The FCMs that were configured to 60 μL min−1 had flow rates that between μL min−1 and μL the need to determine the flow rate a measured concentration. The between beads and shows that are in with each other = and flow rates measured by on than flow rates measured with beads (Fig. participants into the or the will result in an of the flow rate measured by In only one rate was the could not be determined for the because this FCM has an to the flow should provide a more flow rate than could be determined with or Because no about the the platelet EV concentrations in this are based on flow rates determined with beads. participants precipitation in the platelet EV samples and the precipitation was also observed of a second the concentration and diameter distribution of platelet EVs. The of platelet EVs was determined within the gates as in (Fig. The ranges of platelet EVs were of for fluorescence of for the bead diameter of for the 1200–3000‐nm EVs, of for the 600–1200‐nm EVs, and of for the EVs and Because of the of we that the observed variation is to and is thus not caused by the FCM and As a the workshop also distributed EV samples For the was for the for the bead diameter and for the 1200–3000‐nm EV diameter concentrations of platelet extracellular vesicles (EVs) on different flow cytometers for 600–1200‐nm EVs and EVs. to the indicate the and of all and indicate and concentrations of three can be at Figure We applied a scatter–diameter model to the interlaboratory variability of platelet EV measurements in and on the most types of with precipitation in the EV we that the from 139% to 81% for a polystyrene gate to an EV that the light scattering model an RI of 1.40 for EVs is a step in the is by the of for the bead diameter gate and for the 1200–3000‐nm EV diameter gate for EV samples a scatter–diameter model scatter units to and thus on the measured For example, an SSC gate on polystyrene beads platelet EVs of 1200–3000 nm. from the it is not to particles within the 1200–3000‐nm gate are platelets platelet relationship of platelet to and PubMed Scopus Google or platelet‐derived EVs. In the diameter of EVs and the of an EV identification is is