Extracellular vesicles (EVs) are nanometer-to-micrometer sized vesicles released from diverse cells (Raposo and Stoorvogel, 2013); these are found in most biofluids including blood, urine, saliva, and bile juice (Pardini et al., 2019; Cheong et al., 2022). Depending on origin, size, and biogenesis, EVs can be subdivided as apoptotic bodies, microvesicles, and exosomes (Doyle and Wang, 2019). Exosomes are 30-100 nm in diameter, originate from multivesicular bodies (MVBs) via endosomal pathways, and possess surface markers, such as tetraspanins (CD9, CD63, and CD81), Alix, and TSG101. Recent findings have suggested that exosomes play an important role as natural conveyors of genetic information between cells and across various tissues through horizontal transfer of macromolecules (Momen-Heravi et al., 2015a). They contain various bioactive molecules, such as DNA, RNA, proteins, peptides, lipids, and carbohydrates (Merino et al., 2014), and transfer these to neighboring cells.

As damaged cell-derived exosomes play a crucial role as mediators of intercellular communication to target cells, several biological processes are influenced by exosome-mediated transfer (Valadi et al., 2007; Belting and Wittrup, 2008). In the progression of toxicant-mediated hepatic injuries, including those induced by alcohol, free fatty acids release a large number of exosomes, which are internalized into neighboring hepatic nonparenchymal cells, thereby promoting inflammation, fibrogenesis, and angiogenesis in the liver. Here, we review the recent findings in exosome-related research. Understanding the biogenesis and biological roles of exosomes will provide insights into the progression of alcoholic liver disease (ALD) and nonalcoholic fatty liver disease (NAFLD).

Nomenclature and classification

The biological role of EVs in mediating novel mechanisms of intercellular communication has been demonstrated over the past two decades. EVs are easily found in various body fluids, including saliva, blood, milk, urine, cerebrospinal fluid, bronchoalveolar lavage fluid, and semen. They comprise a heterogenous population and are known to transport various genetic materials (Malkin and Bratman, 2020).

Based on their biogenesis, EVs are classified into apoptotic bodies, microvesicles, and exosomes (Table 1) (Todorova et al., 2017; Doyle and Wang, 2019). During apoptosis, dying cells release apoptotic bodies (0.5-2 μm), which are generated by plasma membrane budding. The size, structure, and composition of apoptotic bodies are variable, and they contain micronuclei, remaining chromatin, cytosol, damaged proteins, DNA fragments, and entire cellular organelles (Holdenrieder et al., 2001; Jahr et al., 2001). In addition, apoptotic bodies can be efficiently removed by the surrounding phagocytes (Battistelli and Falcieri, 2020). Microvesicles have a diameter of 100-1,000 nm, but their density remains unknown. They are produced by plasma membrane budding (Tricarico et al., 2017), and their size is larger than that of exosomes (50-100 nm in diameter). Exosomes are generated by exocytosis of multivesicular bodies (MVBs; also known as multivesicular endosomes [MVE]) and contain proteins including class II Major Histocompability Complex and tetraspanins (Avcı and Balcı-Peynircioğlu, 2016; Lawson et al., 2016). Microvesicles and exosomes share similar particle sizes and surface proteins (Kastelowitz and Yin, 2014). In general, the size and molecular markers of microvesicles and exosomes overlap, which makes it difficult to distinguish the two populations. In recent research articles, the most frequently used term is “exosome,” but it does not necessarily refer to exosomes of MVB origin. Currently, the lack of clarity in using this terminology across numerous research articles has caused confusion among researchers (Witwer and Thery, 2019). To reduce this confusion, the 2018 revised minimal information for studies of EVs guidelines have stated that “extracellular vesicle” is a general term for particles that are naturally released from cells, surrounded by a lipid bilayer, and do not replicate themselves (Thery et al., 2018). Therefore, the use of a the comprehensive term “extracellular vesicles” is suggested, and it is further recommended to characterize EVs based on distinct and measurable characteristics, such as cell of origin, molecular markers, size, density, and function. To the best of our knowledge, no suitable exosome isolation technique has been established yet owing to the similarities between microvesicles and exosomes. EVs are characterized using an electron microscope, nanoparticle tracking analysis, and immunoblots for EV markers (e.g., CD63, CD81, and CD9). EV characterization is usually performed by combining various exosome isolation methods to achieve high purity and yield (Stam et al., 2021).

Table 1 Apoptotic bodies, microvesicles, and apoptotic bodies

Type	Size (nm)	Origin	Pathways	Composition
Apoptotic bodies	100-5,000	Plasma membrane	Apoptosis-related	Cell organelles, proteins, nuclear fractions, RNA, noncoding RNA, and DNA
Microvesicles	100-1,000	Plasma membrane	Ca2+-dependent Stimuli- and cell-dependent	Protein, lipids, cell organelles, RNA, noncoding RNA, and DNA
Exosomes	50-100	Multivesicular body	ESCRT-dependent Tetraspanin-dependent	Protein, lipids, coding RNA, noncoding RNA, and DNA

Biogenesis of exosomes

In 1996, it was discovered that exosomes play important immunological roles; until then, they were believed to be plasma membrane debris (Raposo et al., 1996). Although exosomes are currently known to play a significant role in pathophysiology through intercellular communication, their biogenesis remains largely unexplored. A recent study clearly demonstrated that the secretion of harmful cytoplasmic DNA via exosomes plays an important role in maintaining cellular homeostasis (Takahashi et al., 2017). Thus, the production of exosomes is important not only for intercellular communication but also as a cellular homeostatic response to external stimuli.

Exosomes are produced via the endolysosomal pathway, which involves lysosomes, early endosomes, and endocytic vesicles (Keller et al., 2006). Proteins on the cell surface are endocytosed through the inward budding process, and early endosomes are formed on the cell plasma membrane (van der Goot and Gruenberg, 2006). Subsequently, MVBs are produced by the invagination of endosomal membranes and subsequent fusion of endocytic vesicles, which are associated with the endosomal sorting complex responsible for transport (ESCRT) as well as other components, such as ceramides and tetraspanins (Devhare and Ray, 2018). Among these MVBs, cholesterol-rich MVBs are released into the exosomes, whereas low-cholesterol MVBs alongside lysosomes can dissolve proteins (Avcı and Balcı-Peynircioğlu, 2016).

Regardless of their cellular origin, all exosomes have a lipid bilayer containing a lipid raft composed of cholesterol, sphingolipid, ceramide, and phosphoglyceride. Each exosomal particle is believed to contain approximately 4,400 proteins (Vlassov et al., 2012). Surface proteins such as tetraspannins (CD9, CD63, CD81, CD82, and CD151) and integrins, proteins involved in membrane fusion (annexins I, II, IV, V, and VI; Rab5; Rab7; Rabp1B; and RabGDI), heat shock proteins (Hsc70, and Hsp90), ESCRT complex proteins (Alix and TSG101), and cytoskeletal proteins (actin, cofilin, tubulin, and moesin) are commonly identified in exosomes (Hemler, 2005; Keller et al., 2007). In addition to these surface/internal proteins, exosomes are also known to contain a large amount of genetic material, such as mRNA, miRNA, and small noncoding RNA, to facilitate intercellular communication (Fig. 1) (Behbahani et al., 2016; Ding et al., 2020). However, despite several efforts, the complexity of biogenesis of exosomes has not yet been fully understood.

Figure 1. Biogenesis of exosomes and the classification of extracellular vesicles. This figure was created with BioRender (http://biorender.com).

Recently, the biological roles of exosomes in the pathogenesis of ALD and NAFLD have gained considerable attention. Hepatic parenchymal and nonparenchymal cells release exosomes in both physiological and pathological conditions. Therefore, all hepatocytes become sources and/or targets of endogenous exosomes. Interestingly, in pathological conditions, significant changes are observed in the concentration and composition of exosomes. Recent evidence has clearly demonstrated that following exposure to ethanol or free fatty acids, exosomes derived from hepatocytes play a significant role in the progression of liver disease (Povero et al., 2015; Cho et al., 2017; Seo et al., 2019; Bala et al., 2021). In pathological conditions, a large number of exosomes are released from damaged hepatocytes (Povero et al., 2015; Seo et al., 2016; Cho et al., 2018). Subsequently, secreted exosomes are internalized by various neighboring cells, such as hepatocytes, Kupffer cells, endothelial cells, and hepatic stellate cells, which play a critical role in inflammatory responses, angiogenesis, and cancer metastasis (Othman et al., 2019). In addition, damage-associated molecular patterns (DAMPs) or miRNAs derived from injured hepatocytes are delivered via exosomes, thereby activating hepatic stellate cells and increasing the expression of profibrogenic genes (Seo et al., 2016; Eguchi et al., 2017; Chang et al., 2021; Luo et al., 2021). These findings indicate that exosome-associated intercellular interaction has an impact on the fate of target cells and liver microenvironment. Here, we briefly demonstrate the involvement of exosomes in the progression of ALD and NAFLD (Fig. 2).

Figure 2. Biological roles of exosomes in the alcoholic liver disease and non-alcoholic fatty liver disease. This figure was created with BioRender (http://biorender.com).

Exosomes in alcoholic liver disease

Several studies have clearly indicated the significant role of hepatocyte-derived exosomes in the progression of ALD after exposure to ethanol (Cho et al., 2017; Seo et al., 2019; Bala et al., 2021). Previous studies have demonstrated that exosome production is enhanced by alcohol in acute/chronic alcohol drinking murine models as well as patients with alcoholic hepatitis (Momen-Heravi et al., 2015a, 2015b). The expression of various biomolecules, such as miRNA, mitochondrial DNA, CD40 ligand, and HSP90, are upregulated, and they interact with nonparenchymal cells, including Kupffer cells, endothelial cells, circulating monocytes, and hepatic stellate cells, and consequently promote inflammation and fibrosis (Momen-Heravi et al., 2015a; Verma et al., 2016; Eguchi et al., 2017; Saha et al., 2018). The levels of miR-122 were enriched in exosomes after alcohol binge drinking, and miR-122 was horizontally transferred to monocytes via exosomes. Thus, exosome-associated miRNA delivery to monocytes stimulated the production of proinflammatory cytokines (Momen-Heravi et al., 2015a). In addition, alcohol-associated hepatic stress produces large numbers of exosomes containing DAMPs (including mitochondrial double-stranded RNA) from damaged hepatocytes. These mitochondrial double-stranded RNA-enriched exosomes are internalized by Kupffer cells, and exosome-mediated Kupffer cell activation stimulates the production of proinflammatory cytokines (Lee et al., 2020). In the NIAAA model, the production of mitochondrial DNA–containing exosomes was enhanced by the activation of hepatic ER stress and inflammasome, which led to neutrophilic inflammation via TLR9 activation (Osna et al., 2022). A recent study revealed that prolonged, excessive intake of ethanol promotes the release of inflammatory mitochondrial DNA–enriched exosomes from hepatocytes. These exosomes were further internalized into hepatocellular carcinoma cells, thereby contributing to the tumor microenvironment and activating oncogenes (Seo et al., 2019). Furthermore, HepG2 cells treated with ethanol and overexpressing CYP2E1, which is involved in ethanol metabolism, were observed to release a considerable number of hepatocyte-derived exosomes containing CD40 ligand in a caspase-3–dependent manner. These exosomes switch macrophages to the M1 type inflammatory phenotype. In the NIAAA model, CD40-knockout mice exhibited little effect from ethanol-induced liver injury due to the reduction of macrophage infiltration (Osna et al., 2022).

In addition to hepatocyte-derived EVs, studies have been conducted on nonparenchymal cell-derived EVs in mice with alcohol intoxication. Saha et al. (2016) revealed that alcohol-treated monocytes produce exosome-enriched miR-27a, which play a role in the activation, polarization, and increased phagocytic activity of naïve monocytes. A previous study showed an increase in the levels of CYP2E1 and other p450 isoforms in the circulating exosomes of alcohol-exposed mice and alcoholics. In both alcoholics and alcohol-exposed rodents, a marked elevation was observed in the number of exosomes and levels of exosomal CYP2E1, CYP2A, CYP1A1/2, and CYP4B. These exosomes can eventually promote hepatocyte death through the apoptosis signaling pathway (Cho et al., 2017).

Exosomes in nonalcoholic fatty liver disease

NAFLD is characterized by excessive fat deposition in hepatocytes and refers to a clinicopathological disorder that is closely associated with metabolic syndrome (Zhang et al., 2021). NAFLD, the most prevalent chronic liver disease worldwide, is primarily caused by abnormal lipid metabolism. The functions of exosomes have increasingly gained prominence as a mechanism for regulating NAFLD. Due to massive free fatty acid influx from peripheral tissues or enhanced de novo lipogenesis, hepatic lipotoxicity has been associated with the severity of NAFLD. Hepatic lipotoxicity influences ER and causes oxidative stress, apoptosis, and inflammation (Rada et al., 2020). Palmitic acid-treated hepatocytes showed significantly increased production of CD36 and exosomes. Microarray analysis revealed distinct miRNA expression in exosomes derived from vehicle- and palmitic acid-treated hepatocytes (Lee et al., 2017). Exposure of hepatocytes to palmitic acid considerably increased the number of exosomes and CD36 expression. Under lipotoxic conditions, hepatocyte-derived exosomes harbor vanin-1 on their surface and encapsulates miR-128-3p. These exosomes activate hepatic stellate cells via suppression of PPAR-γ, resulting in the induction of profibrogenic genes (Povero et al., 2013; Lee et al., 2017; Lee and Kim, 2022). Furthermore, hepatocytes treated with saturated fatty acids, such as palmitic acid, promote two processes required for angiogenesis: endothelial cell migration and vascular tube formation. Mice treated with RNA interference vanin-1 were protected from steatohepatitis-induced pathological angiogenesis of the liver and proangiogenic effects of exosomes (Povero et al., 2013). Moreover, hepatocyte-derived exosomes under lipotoxic environment activate NLRP3 inflammasome in both hepatocytes and macrophages, leading to caspase 1 activation and IL-1β (Cannito et al., 2017). Nonesterified fatty acids stimulate hepatocytes to secrete exosomes containing TNF-related apoptosis-inducing ligand, CXC motif ligand 10, sphingosine-1-phosphate, mitochondrial DNA, miRNAs, and ceramides (Hirsova et al., 2016; Ibrahim et al., 2016; Malhi, 2019). More than 70% of hepatic miRNAs is found in the blood, and miR-122 contributes to hepatic lipid homeostasis by regulating fatty acid and cholesterol levels (Girard et al., 2008; Lewis and Jopling, 2010; Pirola et al., 2015). High levels of miR-122 in exosomes are closely associated with the progression of NAFLD (Pirola et al., 2015). Further research is warranted to determine whether miR-122-enriched exosomes are a useful biomarker for predicting the occurrence of chronic liver disorder (Conde-Vancells et al., 2008). Hepatocyte-derived exosomes containing miR-192-5p also increase in patients with NAFLD. These exosomes activate proinflammatory macrophages and hepatic inflammation via the negative regulation of the Rictor/Akt/Fox01 signaling pathway (Garcia-Martinez et al., 2020; Liu et al., 2020). Neutrophil infiltration, a hallmark of nonalcoholic steatohepatitis (NASH), is believed to further promote hepatocyte damage in NASH (He et al., 2021). NAFLD progression is significantly regulated through neutrophil-to-hepatocyte communication, which occurs when circulating miR-223-enriched exosomes from neutrophils are selectively taken up by hepatocytes. The selective transfer of miR-223 into hepatocytes, which is dependent on low-density lipoprotein receptor in hepatocytes and apolipoprotein E (ApoE) in neutrophils, plays an important role in preventing NAFLD progression (He et al., 2021).

Exosomes, which are secreted from almost all types of cells, contain various molecules for intercellular communication. They can therefore be applied to numerous research purposes, such as disease diagnosis and treatment, drug delivery, cell-free vaccines, and regenerative medicine (Barros et al., 2018; Tran et al., 2019; Zhao et al., 2019). Despite the development of numerous techniques to efficiently separate exosomes from biological fluids, no standardized exosome isolation method has been established yet. The exosomes isolated from various methods exhibit diverse features, such as yield, purity, and size distribution (Kang et al., 2017). However, exosome-related research faces two major limitations: 1) low yield, which requires an easy exosome extraction method and improved yield and 2) low purity, which requires distinction of exosomes from other EVs (especially microvesicles) (Yang et al., 2020). It is difficult to isolate pure exosomes using existing techniques, and several researchers have considered combinations of two or more isolation methods to increase their purity and yield (Lobb et al., 2015; Patel et al., 2019). However, this combination may increase the cost and/or complexity of the isolation method. Therefore, the exosome isolation technique should be carefully selected according to the sample type and purpose of the study.

Ultracentrifugation

Ultracentrifugation is a classical method that operates under strong centrifugal force to isolate exosomes (Konoshenko et al., 2018). Owing to its high processing capacity, suitability for large volume preparation, and cost-effectiveness, ultracentrifugation is regarded as the “gold standard” for exosome isolation and an optimal process for separating small particles under centrifugal forces (100,000-150,000 ×g for exosome separation) (Yang et al., 2020). However, the quantity and functional analysis of exosome samples obtained using this technique are severely hampered by protein aggregation and lipoprotein contamination (Li et al., 2017). Ultracentrifugation for exosome isolation can be largely classified as differential ultracentrifugation and density-gradient ultracentrifugation.

Differential ultracentrifugation (simple ultracentrifugation) is the most common method for exosome isolation (Konoshenko et al., 2018; Chen et al., 2021). Under centrifugal force, extracellular components are sequentially separated according to their density, size, and shape. First, large-sized particles are removed under low centrifugal force (~300 g); thereafter, a number of sequential centrifugation steps under different centrifugal forces are performed to remove cell debris, apoptotic bodies, and protein aggregates (Szatanek et al., 2015). After the final centrifugation, the supernatant is removed, and the exosome pellet and contaminant proteins are obtained at the bottom of the tubes (all centrifugation procedures are performed at 4°C). Due to its suitability for high-volume preparation, low cost, and ease of operation, differential ultracentrifugation has been widely used over the past decades (Doyle and Wang, 2019). However, the quantity of EVs obtained is dependent on the applied centrifugal force and duration of the procedure.

The density-gradient ultracentrifugation method has been commonly used in hematological studies for separating subpopulations of blood cells (Kim et al., 2019). Because proteins accumulate in different density layers than exosomes, density-gradient ultracentrifugation is effective in isolating exosomes with high purity (Lobb et al., 2015). In a typical density-gradient ultracentrifugation procedure, a layer of biocompatible medium (e.g., iodoxinol or sucrose) that covers the density range of the particles is used. In general, OptiPrep^TM (60 w/v% iodixanol in distilled water) is used to prepare dilutions of 50%, 30%, and 10% in 0.25 M sucrose buffer (1 mM EDTA and 1 mM Tris–HCl, pH 7.4), and a discontinuous gradient is formed by layering of each solution in centrifugation tubes (Onodi et al., 2018). Large- and medium-sized particles are removed from biofluids, which are then placed onto 10%-30%-50% iodixanol gradient layers, with density gradually decreasing from the bottom to the top of the tube (Onodi et al., 2018). After adding the sample to the top of the biocompatible medium, it is centrifuged at 100,000 ×g for 16 h (Tran et al., 2019). Exosomes, apoptotic bodies, and protein aggregates gradually aggregate in the isopycnic layer, i.e., at the same density. Protein aggregates are concentrated at the bottom of the tube, and exosomes remain in the media layer between the concentrations of 1.10 and 1.18 g/mL (Onodi et al., 2018). In recent years, density-gradient ultracentrifugation has become the most popular isolation method as it can provide exosome samples with high purity. However, the commonly used isopycnic ultracentrifugation is completely dependent on the difference between densities; therefore, it is not possible to separate EVs with similar densities and different sizes. In other words, the differentiation of exosomes and microvesicles remains unattainable (Yang et al., 2020).

Polymer precipitation

In polymer precipitation, a polymer with high hydrophilicity is used for isolating exosomes. A highly hydrophilic polymer is used to competitively bind to water molecules around the exosomal membrane, which reduces the solubility of exosomes (Yu et al., 2022). Subsequent centrifugation under low force causes precipitation of exosomes to the bottom of the tube. Among various hydrophilic polymers, polyethylene glycol, a nontoxic polymer that can remodel the water solubility of surrounding materials, is the most commonly used polymer for exosome isolation (Garcia-Romero et al., 2019; Yu et al., 2022). Rider et al. (2016) suggested the use of a fast-enriching exosome purification technique called ExtraPEG to obtain adequate exosomes and exosomal contents for further investigation. The polymer precipitation method is easy to use, time-efficient, and requires no special device. The principal advantage of polymer precipitation is its high yield (Yang et al., 2020). However, it has the disadvantage of contaminating other water-soluble substances. Impurities of precipitates such as protein aggregates, other types of EVs, and polymeric contaminants affect the results of exosome quantification. Currently, several commercial kits based on polymer precipitation have been developed for exosome isolation and enrichment (Yu et al., 2022). According to previous studies, ExoQuick Plus (System Biosciences, Palo Alto, CA, USA) and ExoEasy (Qiagen, Venlo, The Netherlands) can produce a reasonably high yield and purity of exosomes compared with other kits (Ding et al., 2018; Macias et al., 2019).

Size exclusion chromatography

Exosome separation using size-based techniques is a method for separating particles according to their size by allowing a liquid to flow through a column containing porous particles as the stationary phase. Using porous filters with specific pore sizes for exosome isolation is beneficial in terms of ease of use and efficient purification (Taylor and Shah, 2015). Smaller particles enter the pores in the stationary phase and leave the column at a later stage because they travel longer in the stationary phase. Conversely, large particles do not enter the stationary phase and leave the column rapidly, indicating that nanosized particles are separated by size (Yang et al., 2020). After the passage of large particles, the smaller exosome-enriched vesicles are eluted. Once exosomes are eluted, nonexosomal proteins are concentrated in the last several fractions. Therefore, size exclusion chromatography (SEC) could separate small and large vesicles as well as remove contamination of nonexosomal soluble proteins, such as albumin and globulins (Kuo and Jia, 2017). The chromatography column is packed with gel polymers, such as cross-linked dextrans, agarose, polyacrylamide, or allydextran (Sidhom et al., 2020), which are commercially available. Advantages of SEC include the preservation of exosome biological activity, high purity of the isolated exosome, time- and labor-efficiency, and high reproducibility and separation of large or small sample volumes depending on column capacity (Kuo and Jia, 2017; Hu et al., 2022). However, the instrument is expensive and SEC cannot differentiate exosomes and microvesicles of the same size. In cases where the identification of exosome subtype is crucial, the combination of SEC with immunocapture methods is highly recommended (Sidhom et al., 2020).

Ultrafiltration

Ultrafiltration is used to separate samples according to vesicle size using filters with different size exclusion limits or molecular weight cutoffs. It is based on the principle of centrifugation and can purify and concentrate exosomes using cellulose membranes; exosomes are retained above the filter, whereas smaller impurities and other proteins pass through the cellulose membrane (Lobb et al., 2015; Diaz et al., 2018; Konoshenko et al., 2018). Filters with pore sizes of 0.8 and 0.45 μm are used to remove larger particles first, producing a filtrate that is relatively rich in exosomes. Subsequently, smaller vesicles are removed from the filtrate by passing it through membranes with pore sizes that are smaller than the desired exosomes (Sidhom et al., 2020). Two ultrafiltration devices are currently used to separate exosomes: tandem-configured microfilters and sequential ultrafiltration system; these can effectively separate particles in the range of 20-200 nm (Xie et al., 2022). A tandem-configured microfilter comprises two tandem-configured microfilters with defined size exclusion limits of approximately 20-200 nm. Smaller particles such as proteins pass through the 20-nm microfilter, whereas larger vesicles (apoptotic bodies or microvesicles) remain at the bottom of the two membranes and are trapped in the 200-nm membrane (Yang et al., 2020). Alternatively, sequential ultrafiltration can be used for exosome isolation (Popović and de Marco, 2018). To remove large particles from extracellular fluids, a 1000-nm filter is used before further processing (i.e., removal of cell debris or apoptotic bodies). Free proteins and other small particles are removed from the filtrate using a second filter with a molecular weight cutoff of 500 kD; a filter with a cutoff of 200 nm can then be used to separate exosomes with diameters of 50-200 nm from the filtrate (Yang et al., 2020). Ultrafiltration-based exosome isolation drastically reduces processing time without requiring any specialized equipment (Heinemann and Vykoukal, 2017; Yu et al., 2018). Its disadvantages include the possibility of vesicle clogging and trapping, which can shorten the life of the membrane and lower the separation yield. Using tangential flow filtration may overcome these issues because the liquid in tangential flow filtration flows parallel to the membrane, thereby reducing the pressure on the membrane and chances of membrane clogging (Schwartz and Seeley, 2014). Another disadvantage is low purity because particles that are similar in size to exosomes are not separated well; further, the biological activity of exosomes may be impaired because of the pressure on the membrane. However, the low purity can be improved using two or more separation methods, and we can overcome the loss of exosome function through careful pressure control.

Immunoaffinity capture

In immunoaffinity-based exosome isolation, exosomes are isolated and purified by exploiting specific interactions between antigens and antibodies (Chen et al., 2021). For immunoaffinity capture, immunomagnetic beads—antibody-coated beads that specifically bind to the corresponding exosomes and separate them from impurities through magnetic force are commonly used (Clayton et al., 2001; Chen et al., 2021). The lipid bilayer of exosomes (exosomal surface) has various proteins and receptors, including both common and specific proteins. Exosomal surface proteins (e.g., tetraspanins and annexins) bind to the corresponding antibodies, allowing specific isolation of the exosomes (Ruivo et al., 2017). Other exosomal markers utilized for exosome isolation include lysosomal associated membrane protein 2B, transmembrane proteins, heat shock proteins, platelet-derived growth factor receptors, fusion proteins (flotillins, annexins, and GTPases), lipid droplet-associated proteins, and phospholipases (Brzozowski et al., 2018; Smolarz et al., 2019; Zara et al., 2019; Zhang et al., 2019). Owing to its high specificity and purity, the immunoaffinity method is suitable for isolating a specific subpopulation of exosomes. Commercially available immunoaffinity capture-based exosome isolation kits (namely, exosome isolation and analysis kit [Abcam, Cambridge, UK], exosome-human CD63 isolation reagent [Thermofisher, CA, USA], and exosome isolation kit CD81/CD63 [Miltenyi Biotec, Bergisch-Gladbach, Germany]) target transmembrane proteins, such as Rab5, CD81, CD82, annexin, and Alix, for exosome isolation. As immunoaffinity-based exosome isolation needs a large amount of antibody-conjugated beads, it is one of the most expensive techniques for isolating exosomes from a large volume of sample, which can restrict its application (Sidhom et al., 2020). Therefore, the immunoaffinity-based exosome isolation method is not an optimal alternative for researchers who do not need high purity or specific subpopulation of exosomes (Chen et al., 2021).

Purification/concentration of exosomes

We compared the most commonly used exosome isolation techniques, the isolation methods were summarized in Table 2 and Fig. 3. In exosome-related research, there are several challenges that hinder the exploration of exosome dynamics, including the lack of established markers and heterogeneity of isolation protocols (Wei et al., 2021). Obtaining exosomes with high purity from body fluids or culture media is essential for using them for diagnostic or therapeutic purposes (Shtam et al., 2018; Cho et al., 2020). The evaluation of exosome purity is important for searching biomarkers or determining functional properties of exosomes, which are free from co-isolated contaminants. With this regard, Webber and Clayton (2013) proposed a simple approach for assessing the purity of exosomes by comparing the ratio of nanovesicle counts to protein concentration using nanoparticle tracking analysis (NanoSight platform) and colorimetric protein assay (bicinchoninic acid). A ratio of ≥3×10¹⁰ particles/μg of protein was considered to represent high purity. The ratio between 2×10⁹ and 2×10¹⁰ particles/μg indicated an increase in protein contaminants (low purity), and any ratio below 1.5×10⁹ particles/μg indicated an impure sample (Webber and Clayton, 2013).

Table 2 Comparison of exosome isolation techniques

Isolation method	Isolation technique	Advantage	Disadvantage
Ultracentrifugation	Uses strong centrifugal force to isolate exosomes	- Suitable for large volume preparation - Cost-effectiveness - No special techniques required	- Protein aggregation - Lipoprotein contamination
Polymer precipitation	Use a polymer with high hydrophilicity	- Easy to use - Time-saving - No requirement of device	- Contamination of other water-soluble substances - Kits are expensive
Size exclusion chromatography	Allows the flow a liquid through a column containing porous particles	- Time- and labor-efficient - High reproducibility - Large or small sample volumes are available	- Isolation instrument or column is expensive - Similar sized particles are not differentiated
Ultrafiltration	Separate samples by vesicle size using filters with different size exclusion limits	- Drastically reduces processing time - No requirement of equipment	- Vesicle clogging and trapping, which affects the membrane lifespan and lowers the separation yield - Low purity - Similar sized particles are not separated well - Biological activity of exosomes may be influenced by pressure
Immunoaffinity capture	Antibody-coated beads specifically bind to exosomes	- Suitable to isolate a specific subpopulation of exosomes	- The most expensive technique for exosome isolation

Figure 3. Exosome isolation techniques using biofluid. This figure was created with BioRender (http://biorender.com).

In addition to measuring the purity of exosomes, ensuring the exosome concentration for further analyses is critical. The use of “super absorbent polymer beads” with the ability to absorb water, also known as absorbent polymers, was suggested as a rapid method for exosome isolation. Super absorbent polymer beads have been used as an alternative to filtration methods for concentrating microorganisms from water samples (Xie et al., 2016). Superabsorbent polymer beads utilize nanoscale channels to absorb small molecules and water. The superabsorbent polymer beads dramatically increase the concentration and purity of exosomes by reducing the solution volume without affecting exosome characteristics (Yang et al., 2021).

Intercellular communications can occur in various forms, such as direct cell–cell interactions, ligand–receptor interactions, cytokines/chemokines, hormones, and EVs (microvesicles, exosomes, and apoptotic bodies) (Osna et al., 2022). Among them, exosomes are mediators of cell–cell interactions in an autocrine or paracrine manner and are involved in cellular homeostasis, immune responses, angiogenesis, and cancer cell metastasis (Becker et al., 2016). Exosomes can be internalized by their neighboring cells, following which act as cargo for delivering genetic information (Valadi et al., 2007; Belting and Wittrup, 2008). Exosome-mediated exchange of genetic information contributes to the progression of various liver diseases. Generally, damaged hepatocytes secrete large numbers of exosomes with specific biomarkers reflecting the pathophysiological conditions of their donor cells. These transferred exosomes alter the function of the target cells and activate other pathways to promote the progression of ALD and NAFLD. The biogenesis and underlying mechanisms of the exosomes have not been fully elucidated, thus it is worth to identify the biological characteristics of exosomes as a diagnostic biomarker as well as an efficient gene delivery mediator (Luo et al., 2022). Therefore, future research should focus on the analysis of exosomes to suggest novel insights for understanding the underlying mechanisms of disease progression and potential therapeutic targets. In the meantime, exosome-associated liquid biopsy will serve noninvasive tools to diagnose the pathophysiological condition of liver diseases.

To facilitate in-depth research on the role of exosomes in various liver diseases, there is a need for more efficient methods that can achieve exosome isolation with high yield and purity. Despite proposals for several exosome isolation techniques and devices, no standardized procedure has been established yet due to the heterogeneity of biological samples. Combinations of two or more exosome isolation techniques have been suggested to increase the purity of the exosome; the combination of isolation methods, however, increases the complexity and/or cost of the procedure and reduces the yield and/or reliability of subsequent studies. Therefore, the choice of exosome isolation technique should be based on the type of sample and purpose of the study. As a result, future research must be directed to the development of standardized procedures for exosome isolation and characterization (Muñoz-Hernández et al., 2022). The identification of disease-specific exosome biomarkers and high-purity exosome isolation methods will benefit research focused on understanding the roles of exosomes in various liver diseases.

This work was supported by the National Research Foundation of Korea (NRF) grant (2018R1A5A2025286, 2021R1C1C1009445 and 2022R1C1C1008912), Korea Mouse Phenotyping Project (2014M3A9D5A01073556), Korea Basic Science Institute grant (National research Facilities and Equipment Center; 2021R1A6C101A442) and Supporting Program of The Korean Association for the Study of the Liver and The Korean Liver Foundation.