Aiming to Compare Apples to Apples: Analysis of Extracellular Vesicles and Other Nanosized Particles by Flow Cytometry

Extracellular vesicles (EVs) are submicrometer-sized biological vesicles released by all cells, and can be found in all body fluids or harvested from cell culture supernatants. It is nowadays widely accepted that EVs can serve as vesicular messengers in various physiological and pathophysiological contexts. Over the last 10–15 years, the EV research community has grown almost exponentially and the field has attracted a lot of attention following numerous studies connecting EVs to therapeutic approaches such as vaccination, antitumor therapy, immunomodulation and drug delivery. However, the exact function or mode of action of EVs in most biological contexts is still poorly understood (1-3). Since the protein composition and cargo of EVs is assumed to resemble the cell releasing them, EVs also have come into focus as potential diagnostic biomarkers. Historically, different EV types have been described mainly based on the cellular pathway they are thought to have originated from. Exosomes, which are derived from the cell's endocytic system and released when multivesicular bodies fuse with the plasma membrane, are one class of EVs that are formed within the cell; however, a lack of specific markers that can be used to resolve exosomes from other EVs of intracellular origin or from similar sized vesicles that can bud directly from the cell surface has called into question many of the compositional and functional characteristics previously assigned to this particular EV subset. Similarly, microvesicles (or ectosomes), which bud from the cell surface, have historically been considered to be larger and bearing distinct cargo, but it is now realized that as for EVs of intracellular origin, these can be produced by several different mechanisms and can have a wide range of sizes that overlap considerably with the so-called small EVs that are formed internally. Apoptotic bodies, a poorly characterized and defined class of particles, may also be found in cell culture supernatants, but, as for the other EV types, there are no specific markers that allow these to be unambiguously defined. The increasing appreciation for the heterogeneity of EVs, and obvious inadequacy of the traditional terminology, has led to the realization that single vesicle analysis of molecular cargo is required to effectively classify and identify EV subsets and the features that are responsible for their functional effects. Due to this complexity and their small diameter, it is challenging to accurately measure individual EVs, and to quantify basic parameters such as diameter, size, and concentration of EVs in a sample of interest. While resistive pulse spectroscopy and nanoparticle tracking analysis can detect single EVs and estimate their size, they cannot effectively measure EV molecular cargo. Flow cytometry is widely used to quantify and distinguish cellular subpopulations in highly heterogeneous samples, but conventional flow cytometry instruments and assay approaches struggle to measure EVs and other submicrometer particles. The amount of scattered light or emitted fluorescence from an EV is orders of magnitudes lower than that from a cell, and most conventional flow cytometers are not sensitive enough to detect EVs smaller than 500 nm in diameter (4). Recently, new flow cytometers with high sensitivity photodetectors have become available that offer significantly improved performance compared with the older generation of PMT-based instruments. However, a more sensitive instrument does not necessarily translate into a more sensitive measurement, and deficiencies in assay design, especially with respect to the necessary controls and calibration, have limited the interpretation and reproducibility of single EV flow cytometry measurements. To address those challenges, an international Working Group of researchers active in the International Society of Extracellular Vesicles (ISEV), International Society for Advancement of Cytometry (ISAC), and/or International Society for Thrombosis and Haemostasis (ISTH) Vascular Biology Scientific Standardization Committee, have developed a framework for reporting EV flow cytometry methods that complements and extends the existing MIFlowCyt (5) and MISEV (6) guidelines. The resulting MIFlowCyt-EV framework, published earlier this year in the Journal of Extracellular Vesicles (7), describes the essential controls and calibrations necessary for interpretable and reproducible results. This Working Group is actively developing educational resources to support the EV flow cytometry field (refer to http://evflowcytometry.org for details and future updates). In addition to such consensus guidelines, it will be essential for the field to develop and qualify standards and reference materials being more suitable for EV research and the dim signals derived from single EVs. While bead standards used in cellular flow cytometry often scatter more light or emit more fluorescence than tens to hundreds of EVs, the field is working on beads and synthetic nanoparticles being more suitable in terms of size and material (8). In this context, a few recent studies have explored the use of biological reference materials by creating EVs or virus-like particles expressing fluorescent reporter proteins through engineering of respective producer cells (9-11). The idea is that such biological reference materials may have the potential to closely resemble signal intensities and other biophysical parameters such as diameter and density obtained from EVs. Especially within the last few years, the EV flow cytometry field also started to implement fluorescence and more recently scatter calibration to improve accuracy and comparability of reported data. This special issue on Flow Cytometry of Extracellular Vesicles provides a snapshot of some of the current work in the field. Clearly, the distinguishing feature of flow cytometry is its ability to make quantitative multiparameter immunofluorescence measurements of individual particles that, if applied to EVs, would enable the identification of distinct compositional and functional subpopulations. However, because small EVs bear few antigens, EV immunofluorescence requires careful optimization and calibration of a sort not commonly encountered in the published literature. Tertel and colleagues (page 602-609) provide a clear example of the optimization and reporting of EV immunofluorescence as applied to an imaging flow cytometer (12). By evaluating the key staining parameters of concentration, temperature, and time, and performing appropriate calibration to report fluorescence in absolute units of MESF (molecules of equivalent soluble fluorochrome), they demonstrate how to produce quantitative and reproducible EV immunofluorescence data. Light scatter is a particularly challenging problem in flow cytometry of EVs. While light scatter signals are commonly used to trigger detection of cells, the small size and low refractive indices of EVs produce very dim signals that can be hard to discriminate from various sources of background. While fluorescence may provide more specific means to detect EVs (13-15), light scatter can, if interpreted correctly, provide additional information on physical properties such as size and refractive index. While most experienced cytometrists understand that polystyrene or silica beads will scatter much more light than a comparably sized vesicle, owing to their higher refractive indices, it is possible to use measurements of beads, in conjunction with Mie theory-based light scattering models, to estimate EV diameter from light scatter. Welsh et al. (page 569-581) present FCMPASS, a software utility for light scatter calibration of EV flow cytometry data postacquisition based upon Mie theory (16). By using a core-shell model of vesicles that accounts for the predicted change in refractive index with EV size, as well as consideration of collection angle and illumination wavelength, the software will provide an instrument-specific calibration that allows estimation of EV diameter from scatter intensity. Because EV refractive index can be different between different vesicle types, the software provides estimates assuming high, medium, and low refractive index values. FCMPASS also performs fluorescence calibration (using commercially available MESF standards) and writes the calibrated data to a new fcs format file, making this software a one-stop solution for EV calibration. In a related publication, Welsh and Tang (page 592-601) illustrate proper fluorescence and light scatter calibration in the context of characterizing an enveloped virus, which can be considered a special class of EVs (17). In this case, the virus has been engineered to express a green fluorescent protein and thus is endogenously fluorescent. In context of a proof-of-concept study carried out as a resource for use at a CYTO2019 workshop, the authors demonstrate its characterization in terms of fluorescence and light scatter calibration and show that calibrated data can be compared between different flow cytometry platforms. While interpreting light scatter data is a key challenge, so is light scatter measurement. Because light scatter signals from EVs are so dim, background from various sources can overwhelm the system and compromise measurements. However, careful optimization of illumination and light collection can reduce this background and improve signal to noise. While conventional wisdom suggests that side scatter is more appropriate for light scatter measurements of small particles, the angular dependence of light scatter may be useful, especially for larger EVs, and forward scatter measurements may provide additional information. Arkesteijn and colleagues (page 610-619) report on the optimization of forward angle light scatter (FALS) for measurement of small particles (18). They evaluated the effects of different blocker bars and pinhole diameters on the relative signal to noise. By improving the rejection of the illuminating laser and out of focus light, they report significant reduction in background signals and improvements in sensitivity. The ability to make effective measurements of both forward and side scatter may improve estimates EV size and refractive index using flow cytometry data. De Rond et al (page 582-591) also address light scatter and describe an approach to characterize and improve scatter sensitivity in both the forward and orthogonal directions on a modified conventional flow cytometer (19). The authors derived quantitative performance metrics based on rigorous characterization of the background and expressed as a separation index. They evaluated obscuration bars and pinhole diameters, as well as sample stream diameter, and report significant improvements in both forward and side scatter performance, pointing the way for instrument designs that might one day be part of a built-for-purpose vesicle flow cytometer. In summary, single EVs still represent an exciting frontier for flow cytometry, with a clear need for more sensitive instruments, improved standards, and validated assays. The development of guidelines to standardize the reporting of methods and results will enable apple-to-apple comparison of results, allowing data-driven instrument and assay performance in absolute terms. This will be a boon to researchers, who will be able to interpret and reproduce data more readily, and to instrument and reagent manufacturers, who will be able to rationally design new products that will have useful and demonstrable performance improvements. The result will be quantitative measurements of EVs that will let us understand their composition, predict their functions, and use these to diagnose, treat, and prevent human disease.