Extracellular vesicles and nanoparticles at a glance

ABSTRACT

Cells can communicate with neighboring and more distant cells by secretion of extracellular vesicles (EVs). EVs are lipid bilayer membrane-bound structures that can be packaged with proteins, nucleic acids and lipids that mediate cell–cell signaling. EVs are increasingly recognized to play numerous important roles in both normal physiological processes and pathological conditions. Steady progress in the field has uncovered a great diversity and heterogeneity of distinct vesicle types that appear to be secreted from most, if not all, cell types. Recently, it has become apparent that cells also release non-vesicular extracellular nanoparticles (NVEPs), including the newly discovered exomeres and supermeres. In this Cell Science at a Glance article and the accompanying poster, we provide an overview of the diversity of EVs and nanoparticles that are released from cells into the extracellular space, highlighting recent advances in the field.

Summary: Cells communicate over distances by releasing extracellular vesicles and non-vesicular nanoparticles. We review the diversity and heterogeneity of these structures, including several recently identified examples.

Introduction

Cells can release extracellular vesicles (EVs), lipid bilayer membrane-enclosed structures, as a means of intercellular signaling through cell surface interactions or by delivery and internalization of their cargo in recipient cells. The general secretion of EVs appears to occur from all cells across the evolutionary spectrum from bacteria to plants and animals (Cai et al., 2018; Jeppesen et al., 2023; Toyofuku et al., 2023). More specialized EVs are also secreted from specific cell types, such as synaptic vesicles released from neuronal synapses (Holm et al., 2018). Long assumed to be primarily a way for cells to jettison unwanted material (Johnstone et al., 1987), it is now well established that EVs actively function in cell–cell communication of multicellular organisms by interacting with cells both locally and at distant sites (Jeppesen et al., 2023; Mathieu et al., 2019; van Niel et al., 2022). Due to the great diversity of EVs (see poster); their general heterogeneity of size, composition and biogenesis; and the technical difficulty in obtaining pure samples of EV subtypes, the generic term ‘extracellular vesicles’ has been promoted and adopted to encompass all categories (Théry et al., 2018; Welsh et al., 2024) (see Box 1 for further details on EV nomenclature). Although the pace of EV research proceeds at breakneck speed, the field has come to recognize that not everything secreted into the extracellular space is carried by a membrane vesicle. Non-vesicular extracellular nanoparticles (NVEPs) are a diverse group of amembranous particles released from cells (see poster) (Debnath et al., 2023; Jeppesen et al., 2023; Zhang et al., 2019, 2021b). Although NVEPs lack a lipid bilayer membrane, many of them are of sizes similar to EVs, and as a result they are commonly found to co-purify with EVs, leading to their incorrect assignment as membrane vesicles rather than NVEPs (Jeppesen et al., 2019; Tosar et al., 2021; Zhang et al., 2023, 2021b). In this Cell Science at a Glance article and the accompanying poster, we review the current understanding of EVs and NVEPs, focusing primarily on knowledge from mammalian cells.

See Supplementary information for a high-resolution version of the poster.

Box 1. EV nomenclature and a word of caution.

The variable criteria and imprecise nomenclature used to describe the multitude of EV types reported has presented challenges for the EV field (Buzas, 2023; Clancy and D'Souza-Schorey, 2023; Jeppesen et al., 2023; Mathieu et al., 2019; Thery et al., 2018). The generic term ‘EV’ has become the preferred name for all lipid bilayer membrane-enclosed vesicles released from cells to the extracellular environment (Welsh et al., 2024). The terms ‘small EV’ (sEV) and ‘large EV’ are used to describe EVs with sizes of less than 200 nm and more than 200 nm, respectively; these categories have gained acceptance due to both their (relative) ease of separation and the fact that many studies have found these two EV populations to be functionally distinct. There are two main classes of EVs, exosomes and ectosomes, which are distinguished by their mode of biogenesis (Buzas, 2023; Jeppesen et al., 2023; van Niel et al., 2022). The term ‘exosomes’ refers specifically to EVs of less than 150 nm that are released from multivesicular endosomes (MVEs) or amphisomes upon fusion of these organelles with the plasma membrane. The term ‘ectosomes’ refers to EVs that are released directly from the plasma membrane. Ectosomes are a diverse class of small to very large EVs that range in size from 30 nm to 10 µm. When a mode of biogenesis cannot be ascribed to a specific subpopulation of EVs, it is prudent to use the terms ‘small EVs’, ‘large EVs’ or even just ‘EVs’. Although there is not a full consensus in the field on all aspects of nomenclature, standardization efforts are ongoing (Welsh et al., 2024). Confusion regarding nomenclature stems from the use of now obsolete terms from the older literature or from too liberal use of the term ‘exosomes’ to describe a mix of smaller and larger vesicles. In studies in which EVs have been isolated, there is often insufficient characterization of biogenesis or demonstration of purity to allow identification of the vesicle population under study as ‘exosomes’, even though some exosomes are likely present and detectable using accepted markers in the preparations. Furthermore, there is a scarcity of protein markers that reliably mark only one subpopulation of EV.

EVs in homeostasis and cancer

Although EVs were previously thought to be a type of cellular debris, they are now known to play many roles in both normal physiology and pathophysiology. EVs can affect functions in distant tissues by transfer of proteins, nucleic acids and lipids in order to maintain normal tissue homeostasis and repair (Roefs et al., 2020). However, transfer of cargo by EVs is also a mechanism exploited by cancer cells to promote tumor growth (Skog et al., 2008). EVs can attach or ‘dock’ to cells through interaction of specific EV surface proteins with cell surface proteins to trigger signaling in the target cell. Indeed, it has been proposed that specific integrin proteins expressed on the surface of tumor EVs determine which organ takes up the EV, which in turn functions in the establishment of a pre-metastatic niche (Hoshino et al., 2015). Furthermore, by selectively jettisoning deleterious material from a cell, EVs can support maintenance of normal cellular homeostasis; examples include the removal of damaged mitochondria and protein aggregates from cells by EV secretion (Jiao et al., 2021; Melentijevic et al., 2017). In cancer, aberrant removal of cellular components from a cell by EVs can function to sustain tumorigenesis or enhance metastatic potential. An example of this is the selective removal of specific tumor-suppressor microRNAs (miRNAs) from bladder cancer cells by EVs, which increases the invasive capacity of the cells (Ostenfeld et al., 2014). The heterogenous nature of EVs likely reflects the myriad ways that cells have adapted to use EV secretion as a means to maintain homeostasis.

Types of EVs

Exosomes

Exosomes are defined as small extracellular vesicles (sEVs) produced from endosomal compartments, as opposed to ectosomes, which are EVs generated from the plasma membrane (see Box 1). In the endocytic trafficking pathway, endocytosis starts with inward budding of the plasma membrane to form early endosomes. Later, intraluminal vesicles (ILVs) are generated by inward budding of the limiting membrane of late endosomes, and accumulation of ILVs in the late endocytic compartment generates a multivesicular endosome (MVE) (Geuze et al., 1983; van Niel et al., 2011). Importantly, the inward budding of late endosomes results in reorientation of endocytosed membrane proteins on ILVs such that their topology is similar to that at the cell surface, thereby rendering them signaling competent upon release from the cell. The MVE with its contents of ILVs can be targeted to a lysosome for degradation of the cargo or can move towards the cell surface (Geuze et al., 1983; Pan et al., 1985). Alternatively, the MVE can fuse with an autophagosome to form an amphisome, which is then targeted either to a lysosome for degradation or towards the cell surface (Jeppesen et al., 2019). After trafficking to the cell surface, docking and fusion of the MVE or secretory amphisome with the plasma membrane results in release of ILVs to the extracellular space as 30–150 nm exosomes (see poster) (Jeppesen et al., 2019; Ostrowski et al., 2010; Pan et al., 1985).

The presence of one or more of the tetraspanin proteins CD63, CD81 and CD9 has long been considered a defining feature of exosomes (Jeppesen et al., 2019; Mathieu et al., 2019; van Niel et al., 2018), although it is becoming clear that these proteins can also be present on ectosomes (Fordjour et al., 2022; Mathieu et al., 2021). CD63 is highly expressed in MVEs and ILVs, and appears be the tetraspanin most specifically associated with exosomes (Edgar et al., 2014; Jeppesen et al., 2019; van Niel et al., 2011). The proteins of the endosomal sorting complex required for transport (ESCRT) machinery, as well as ESCRT accessory proteins, regulate sorting of cargo and formation of ILVs in the MVE. Many of these proteins are often highly expressed in exosomes, including TSG101, ALIX (also known as PDCD6IP) and VPS4 (herein referring to VPS4A and VPS4B); however, these proteins also function in outward budding of the plasma membrane and may therefore also be present on some ectosomes (Choi et al., 2018; Choudhuri et al., 2014; Nabhan et al., 2012). Syntenin-1, which is typically highly abundant in exosomes (Jeppesen et al., 2019; Kugeratski et al., 2021), recruits ALIX to cause inward budding of ILVs and is another protein involved in exosome biogenesis and loading cargo into ILVs (Baietti et al., 2012; Ferreira et al., 2022). Formation of ILVs, and therefore production of exosomes, can also occur through an ESCRT-independent but ceramide-dependent pathway (Trajkovic et al., 2008).

Transport of exosomes through the endocytic pathway requires the RAB family of small G proteins, and trafficking of MVEs to the plasma membrane for release of exosomes has therefore been associated with a number of these RAB proteins. RAB7 proteins appear to be required both for transport of MVEs to lysosomes for degradation (Rocha et al., 2009; Vanlandingham and Ceresa, 2009) and for transport of MVEs to the cell surface (Baietti et al., 2012). RAB31 is active in both loading of proteins into ILVs and preventing the transport of MVEs to lysosomes, thereby promoting the eventual release of ILVs as exosomes (Ferreira et al., 2022; Wei et al., 2021). RAB27A and RAB27B regulate exocytosis of MVEs to release exosomes (Bobrie et al., 2012; Ostrowski et al., 2010) and have been confirmed to control the release of miRNA in exosomes from endothelial and cancer cells (Jaé et al., 2015; Ostenfeld et al., 2014). RAB27B specifically controls movement of MVEs towards the plasma membrane (Ostrowski et al., 2010), whereas both RAB27A and RAB27B are involved in the docking of MVEs at the plasma membrane that is required for fusion and subsequent release of exosomes (Ostrowski et al., 2010; Sinha et al., 2016). However, not all cells constitutively express RAB27A and RAB27B. RAB35 and RAB11 proteins, which are canonical regulators of recycling endosomes, have also been demonstrated to regulate docking of MVEs to the plasma membrane (Hsu et al., 2010; Savina et al., 2005). Additionally, RAB11 proteins appear to be important in exosome release associated with recycling endosomes (Marie et al., 2023). Lastly, RAB39 proteins have been reported to mediate the transport of MVEs specifically to the basolateral plasma membrane of polarized cells for exosome secretion (Matsui et al., 2022).

Ectosomes

Ectosomes are defined as EVs generated directly from the plasma membrane of healthy cells. Ectosome shedding can generate EVs that range in size from 30 nm to 10,000 nm (Fordjour et al., 2022; Jeppesen et al., 2019, 2023; Mathieu et al., 2021; Nabhan et al., 2012). Many different types of ectosomes have been described (see poster). For example, microvesicles, which range in size from 150 nm to 1000 nm and exhibit lower flotation densities than exosomes, have variously been characterized by expression of the proteins annexin A1 and annexin A2 (Jeppesen et al., 2019; Matsui et al., 2021, 2022; Shiri et al., 2022; Yan et al., 2022), α-actinin-4 (Jeppesen et al., 2019; Kowal et al., 2016) and ARF6 (Clancy et al., 2022; Muralidharan-Chari et al., 2009). Small ectosomes (30–150 nm) express CD147 (also known as basigin), as well as the tetraspanin proteins CD63, CD81 and CD9 (Fordjour et al., 2022; Mathieu et al., 2021). Arrestin domain-containing protein 1 (ARRDC1)-mediated microvesicles (ARMMs) are small (<150 nm) specialized ectosomes released from most cell types, are characterized by the expression of the proteins ARRDC1 and TSG101, and require activity of the ESCRT accessory protein VPS4 for their release (Jeppesen et al., 2019; Nabhan et al., 2012; Wang and Lu, 2017; Wang et al., 2018). ARMMs can deliver NOTCH2, which upon activation by γ-secretase cleavage can activate NOTCH signaling in recipient cells (Wang and Lu, 2017). Large oncosomes are, as the name implies, EVs that are 1–10 µm in size. These EVs are released from tumor cells marked by overexpressed or constitutively active oncoproteins, and characteristically express annexin A1 and ARF6 (Di Vizio et al., 2012; Jeppesen et al., 2019; Minciacchi et al., 2017).

Ectosomes can also be generated from specialized areas of plasma membrane extension. For example, migrasomes are 500–3000 nm vesicles that are released from the retraction fibers of migrating cells – a process mediated by large macrodomain assemblies enriched for TSPAN4 and cholesterol (Huang et al., 2019; Ma et al., 2015; Zhao et al., 2019). Additionally, both small and large ectosomes are shed from microvilli, cilia and filopodia in diverse cell types, including melanocytes, podocytes, enterocytes, neuroepithelial cells, tracheobronchial epithelial cells and cancer cells (Nishimura et al., 2021; Rilla, 2021).

Autophagy-related EVs are a currently poorly understood class of EVs containing autophagy-related proteins such as lipidated LC3 (also known as MAP1LC3B); they range in size from small EVs (<200 nm) to larger vesicles with diameters ranging from 350 nm to 500 nm (Chen et al., 2015; Hessvik et al., 2016; Jeppesen et al., 2019; Leidal et al., 2020; Sirois et al., 2012). The LC3-conjugation machinery is responsible for loading of specific RNA-binding proteins within autophagy-related EVs (Leidal et al., 2020). Lysosomal inhibition results in secretion of autophagy proteins and autophagy cargo receptors associated with the outer surface of EVs and/or associated with NVEPs, likely in order to buffer the cell against accumulation of autophagic machinery and maintain proteostasis (Solvik et al., 2022).

In line with the original hypothesis of EV function (Johnstone et al., 1987), some EVs are indeed utilized for expelling unwanted material. For example, exophers are large 3.5–4 µm EVs that are released from Caenorhabditis elegans neurons and murine cardiomyocytes in response to stress in order to remove damaged mitochondria and protein aggregates (Melentijevic et al., 2017; Nicolás-Avila et al., 2020). Migrasome release has additionally been proposed as a cellular disposal mechanism for damaged mitochondria (Jiao et al., 2021). However, damaged mitochondria can also be removed from cells independently of EVs through an autophagy-dependent process known as mitophagy (Rodriguez-Enriquez et al., 2009).

Apoptotic EVs

Apoptosis is a widely used form of programmed cell death. Cells undergo apoptosis in three morphological steps: blebbing of the plasma membrane, formation of membrane protrusions such as apoptopodia (Poon et al., 2014) and, finally, fragmentation. In the final fragmentation step, apoptotic bodies (∼1–5 µm) are released to the extracellular environment (Atkin-Smith and Poon, 2017; Poon et al., 2014; Tixeira and Poon, 2019). During apoptosis, smaller apoptotic vesicles (∼100–1000 nm) are also generated (Dieude et al., 2015; Jeppesen et al., 2019; Park et al., 2018; Zhang et al., 2022). Apoptotic bodies and vesicles differ in molecular composition (Dieude et al., 2015; Phan et al., 2021; Tucher et al., 2018); for example, DNA-binding histones and ribosomal proteins are enriched in apoptotic bodies, whereas lysosomal proteins, proteasome core complex proteins and extracellular matrix proteins are enriched in apoptotic vesicles (Dieude et al., 2015). Apoptotic bodies and vesicles are distinct from exosomes and ectosomes in their biogenesis (see poster) (Dieude et al., 2015; Jeppesen et al., 2019; Schiller et al., 2008; Zhang et al., 2022). As apoptotic bodies and vesicles are formed from fragmenting cells, they would be expected to contain cargo that reflects the contents of those cells, including DNA, nuclear proteins and endoplasmic reticulum proteins, which have all been used as markers of apoptotic EVs (Dieude et al., 2015; Schiller et al., 2008; Sisirak et al., 2016). However, the presence of DNA alone does not provide ironclad evidence that an EV is of apoptotic origin, as non-apoptotic large EVs can also contain DNA (Vagner et al., 2018). Specific proteins involved in apoptosis, including Rho-associated coiled-coil-containing protein kinase 1 (ROCK1) and pannexin-1, have been proposed as more informative apoptotic marker proteins (Poon et al., 2019). Another frequently used marker is phosphatidylserine, a plasma membrane phospholipid that is flipped to the outer leaflet during apoptosis and functions as an ‘eat me’ signal for engulfment of apoptotic cells by phagocytes (Bratton et al., 1997). As annexin V binds to phosphatidylserine exposed on the outer leaflet, it is commonly used to detect apoptotic EVs (Dieude et al., 2015; Hristov et al., 2004; Sisirak et al., 2016), and annexin V-positive apoptotic sEVs have been reported to contain very low or undetectable levels of the exosomal marker proteins CD63, CD81, CD9, ALIX and TSG101 (Jeppesen et al., 2019; Lai et al., 2016). However, exposure of phosphatidylserine is not an exclusive feature of apoptotic EVs, as this also occurs on activated platelets, where it serves as a scaffold for the assembly of complexes involved in blood coagulation (Fujii et al., 2015). The extent of release of apoptotic EVs and their attendant cargo can be altered under specific conditions. For example, treatment with EGF receptor (EGFR) kinase inhibitors can result in release of small apoptotic EVs containing EGFR and DNA (Montermini et al., 2015).

At least two functional roles of apoptotic EVs have been proposed. First, disassembly of apoptotic cells into smaller apoptotic EVs might facilitate more efficient engulfment and clearance of apoptotic debris by phagocytes (Atkin-Smith et al., 2019; Atkin-Smith and Poon, 2017). Second, apoptotic EVs might engage in intercellular communication through their cargo of DNA, RNA and proteins; for example, Wnt8a-containing apoptotic EVs derived from epidermal basal stem cells have been reported to drive proliferation in neighboring stem cells to maintain epithelial tissue homeostasis (Brock et al., 2019).

Non-vesicular extracellular nanoparticles

NVEPs are a diverse group of particles but typically have distinguishing features that set them apart from EVs. Most NVEPs lack a lipid bilayer membrane: some have an outer shell comprised of a lipid monolayer and others are completely amembranous (see poster). Recently, it has become clear that NVEPs are major vehicles for the release of proteins, RNA and DNA to the extracellular space (Zhang et al., 2018, 2021b). A surge of interest in NVEPs has been triggered, at least in part, by the recent discovery of exomeres and supermeres. Exomeres are 28–50 nm amembranous nanoparticles that were first isolated using asymmetric flow field-flow fractionation (AF4), a technique in which injected samples are focused towards a semi-permeable membrane and a tangential crossflow separates particles by size (Zhang and Lyden, 2019). Exomeres contain both RNA and DNA and are enriched for metabolic enzymes (Zhang and Lyden, 2019). We have subsequently reported a simpler method for isolation of exomeres and identified functional cargo (Zhang et al., 2019) such as angiotensin-converting enzyme 2 (ACE2), which binds the SARS-CoV-2 spike protein (Zhang et al., 2019, 2021a). Supermeres (supernatant of exomeres), discovered by our group (Jeppesen et al., 2022; Zhang et al., 2021b), are smaller (22–32 nm) than exomeres and, unlike sEVs and exomeres, are able to efficiently cross the blood–brain barrier. Supermeres are more highly enriched for metabolic enzymes than exomeres, and they are replete with disease biomarkers and therapeutic targets (Zhang et al., 2021b), including DDR1 and TGFBi, which are part of an immune-exclusion signature in microsatellite-stable colorectal cancer (Heiser et al., 2023). Supermeres are a greater source of extracellular RNA than EVs or exomeres and also contain a large number of RNA-binding proteins (Zhang et al., 2021b). Unlike EVs, both exomeres and supermeres lack a lipid bilayer membrane and appear to consist predominantly of proteins and nucleic acids. They are released by many cell types in mice and humans and are present in circulation (Hoshino et al., 2020; Zhang et al., 2018, 2019, 2021a,b). The mechanisms of their biogenesis and cellular release are unknown but under active investigation.

Other extracellular carriers of RNA include subclasses of lipoproteins that circulate in the blood. Lipoproteins are small micelle-like, phospholipid-rich particles that are synthesized in the liver and small intestine (Vickers and Remaley, 2014). High-density lipoprotein (HDL; 5–14 nm) returns excess cholesterol to the liver, thus controlling cholesterol homeostasis. Low-density lipoprotein (LDL; 18–25 nm) delivers cholesterol to cells that upregulate the LDL receptor in response to an increased cellular need for cholesterol. Both HDL and LDL transport and transfer functional small RNAs, including miRNAs and bacterial small RNAs (sRNAs), to recipient cells (Allen et al., 2022; Vickers et al., 2011). Very-low-density lipoprotein (VLDL; 30–80 nm) transports endogenous lipids to adipose and muscle tissues, after which they return to circulation as intermediate-density lipoprotein (IDL; 25–35 nm) (Mehta and Shapiro, 2022). Chylomicrons and their remnants are 75–1200 nm lipoproteins produced by intestinal enterocytes for transport of exogenous (derived from food) lipids, including cholesterol, from the intestine to the circulation (Ko et al., 2020). Vaults are large (41 nm by 72.5 nm) cytoplasmic ribonucleoprotein NVEPs that consist primarily of major vault protein (MVP) but also contain small non-coding RNAs (see poster) (Rome and Kickhoefer, 2013). Vaults might be released from ruptured cells, but it is possible that they might also be actively secreted through amphisomes (Carter et al., 2020; Jeppesen et al., 2019).

NVEPs are also important in immunity. Infected cells release viral particles as enveloped viruses with lipid bilayer membranes (similar to EVs), or as non-enveloped viruses (similar to NVEPs) (Kerviel et al., 2021; Nolte-'t Hoen et al., 2016; Raab-Traub and Dittmer, 2017). As a mechanism to enhance replication, infectivity and immune evasion, viruses can also be released inside EVs (Chen et al., 2015; Kerviel et al., 2021; Santiana et al., 2018). Furthermore, cytotoxic T lymphocytes can release 120 nm supramolecular attack particles (SMAPs), which consist of a thrombospondin-1 (TSP1) shell and a core of cytotoxic proteins, including perforin-1 and granzyme B (Bálint et al., 2020). SMAPs are found in multicore granules and are released from the cytotoxic T lymphocytes to kill target cells (Chang et al., 2022).

Extracellular DNA, in the form of nucleosomes (DNA wound around histones) or strings of nucleosomes (chromatin), can be considered an NVEP; DNA may also be released within, or on the outside, of EVs. Cells can release extracellular DNA either passively or by active secretion. Necrotic, apoptotic or ruptured cells passively spill DNA into the extracellular space (Choi et al., 2005). During apoptosis, apoptotic EVs may contain extracellular DNA, and larger types of EVs have also been associated with the presence of extracellular DNA (Buzas et al., 2014; Clancy et al., 2022; Jeppesen et al., 2019; Sisirak et al., 2016; Vagner et al., 2018). Active secretion of cytosolic DNA from cells also occurs through an amphisome-dependent mechanism where nucleosome particles are released independently of EVs through endosomal and autophagic pathways (Jeppesen et al., 2019).

Lipid droplets (LDs) are found in almost all eukaryotic cells and are composed of an inner hydrophobic neutral lipid core and an outer phospholipid monolayer coated with proteins. LDs are actively secreted from mammary gland cells as 1–10 µm milk fat globules, surrounded by an additional phospholipid bilayer, by budding from the plasma membrane, and are abundant in milk (Lu et al., 2022). Adipocytes can release small LDs contained inside exosomes, termed adipocyte-derived exosomes (AdExos), within adipose tissue to modulate local immune cell function (Flaherty et al., 2019). The phospholipid monolayer of LDs characteristically contains members of the perilipin (PLIN) family of lipid enzymes, but it is not clear to what degree ‘naked’ LDs without additional surrounding membranes are actively secreted from cells. However, exchange of LDs between cells has been observed (Collot et al., 2018), and LDs are likely released passively from ruptured cells.

Methods for isolation of EVs and NVEPs

Many different methods are commonly employed to isolate EVs and NVEPs based on their size, density or affinity to a specific marker of the particle of interest (Table 1). The most commonly applied method has been differential centrifugation, which is still widely utilized, although increasingly in combination with more advanced techniques for increased EV sample purity (Dhondt et al., 2020; Huang et al., 2020; Zhang et al., 2023). It is unlikely that one single EV isolation method will prevail, as each technique has advantages and limitations (Table 1). Furthermore, the optimal choice of method will depend not only on the type of EV to be purified but also on the downstream analyses and functional studies that are desired (Hendrix et al., 2023).

Table 1.

Common EV and NVEP isolation methods

Interest in isolation methods has increased as it has become apparent that samples of presumed EVs can contain NVEPs, including lipoproteins (Jeppesen et al., 2019; Karimi et al., 2018; Li et al., 2018; Sódar et al., 2016), nucleosomes (Jeppesen et al., 2019; Palma et al., 2012), vaults (Jeppesen et al., 2019) and viral particles (Kerviel et al., 2021; Nolte-'t Hoen et al., 2016; Raab-Traub and Dittmer, 2017). Lipoproteins in particular, including HDL and LDL, are highly abundant in many plasma EV preparations as they will be co-purified by many of the commonly used EV isolation methods (Das et al., 2019; Jeppesen et al., 2019; Karimi et al., 2018; Sódar et al., 2016). This co-isolation of NVEPs with EVs has the potential to confound studies on RNA and DNA purported to be present in EVs. The common presence of vault RNA in EV samples (Lässer et al., 2017; Nolte-'t Hoen et al., 2012) highlights the need to separate vaults from EVs (Jeppesen et al., 2019). Furthermore, because of their overlapping size and molecular characteristics, viral particles can co-purify with EVs (Kerviel et al., 2021; McNamara and Dittmer, 2020; Raab-Traub and Dittmer, 2017).

The adoption of more advanced purification methods in recent years is a step forward, but these newer methods must still be critically assessed. For example, techniques based on particle size, such as size-exclusion chromatography (SEC) and AF4, might not be able to separate EVs from similarly-sized NVEPs, and density gradient fractionation might not be able to separate different subpopulations of EVs that overlap in flotation density. Consequently, great care should be taken when interpreting the results of previous EV studies and close attention given to the exact purification methods used before investigators draw conclusions that ascribe functions to a specific EV population.

Conclusions and perspectives

As discussed above, all cells can release a multitude of EVs and NVEPs that differ in size, molecular composition and mode of biogenesis. Some EVs and NVEPs may be limited to specific cell types and serve distinct, well-defined roles, but even specialized cells release a bewildering array of vesicles and particles. As EVs and NVEPs are present together in the extracellular space and can overlap in both size and density, they are often co-purified from cell cultures, biofluids and tissues. However, the distinct molecular composition of EVs and NVEPs provides them with distinct properties that have been used to separate them. Adding to this complexity, it is also possible for NVEPs to be fully contained with an EV, as is the case for nucleosomes inside apoptotic EVs (Schiller et al., 2008; Sisirak et al., 2016). Other examples are the presence of LDs inside exosomes (Flaherty et al., 2019) and viral particles inside EVs (Chen et al., 2015; Santiana et al., 2018). Thus, it seems likely that NVEPs can also be released to the extracellular space by direct secretion of EVs from cells followed by passive release as a result of EV rupture or disintegration.

Care must be taken in interpreting studies that try to attribute specific biomolecules or functions to a specific type of EV or NVEP. Intervention studies, either basic or translational, that attempt to inhibit or alter secretion of EVs carrying a specific biomolecule of interest will have to contend with the fact that this target might be expressed only on a small subset of EVs. Efforts are ongoing to discover new EVs and NVEPs, to develop better isolation technologies and techniques, and to further our ability to distinguish between different types of EVs and NVEPs to truly elucidate their multifaceted roles in intercellular communication.