Recent Advances in the Study of Extracellular Vesicles in Colorectal Cancer

Abstract

There has been significant progress in the study of extracellular vesicles (EVs) since the 2017 American Gastroenterological Association-sponsored Freston conference Extracellular Vesicles: Biology, Translation and Clinical Application in GI Disorders. The burgeoning interest in this field stems from the increasing recognition that EVs represent an understudied form of cell-to-cell communication and contain cargo replete with biomarkers and therapeutic targets. This short review will highlight recent advances in the field with an emphasis on colorectal cancer (CRC). Following a short introduction to secreted particles, we will describe how our lab became interested in EVs, which led to refined methods of isolation and identification of two secreted nanoparticles. We will then summarize the cargo found in small (s)EVs released from CRC cells and other cells in the tumor microenvironment (TME), as well as those found in the circulation of CRC patients. Finally, we will consider the continuing challenges and future opportunities in this rapidly evolving field.

EVs as a class of secreted particles

EVs are lipid bilayer-enclosed vesicles derived from the cell membrane either through direct budding or through cell surface fusion of an endosomally-derived multivesicular body (MVB)¹. Figure 1 depicts several specialized classes of EVs, as well as other categories of secreted mediators like lipoproteins and two recently described nanoparticles, exomeres and supermeres^2–4. While many of these particles have been studied for decades, exomeres and supermeres are newly discovered amembranous nanoparticles of unknown biogenesis that are smaller than extracellular vesicles with distinct cargo and functional properties^3–4. There are many different ways in which EVs are formed: apoptotic bodies bleb from the plasma membrane during cell death processes, microvesicles bud outward from the plasma membrane, and exosomes arise from inward invagination and pinching off within a subclass of late endosomes to form multiple vesicles within an MVB^{2, 5}. In the process of this inward budding, the topology of transmembrane proteins in these intraluminal vesicles is reoriented with the ectodomain facing outwards⁶. MVBs can fuse with the plasma membrane rather than lysosomes, releasing their cargo in a signaling-competent manner with the extracellular domain facing the extracellular environment and cytoplasmic components enclosed within the exosome⁷. Classical exosomes have a distinct set of commonly associated proteins reflecting their endosomal-linked biogenesis, including tetraspanins such as CD63 and CD81, endosomal sorting complexes required for transport (ESCRT)-related proteins like TSG101, and scaffold proteins like syntenin-1 that are less typical of other types of EVs or nanoparticles^{1, 4, 8–10}. It is important to note that exosomes represent only a subset of EVs. Given the heterogeneity of EVs, the International Society of Extracellular Vesicles currently recommends the terms small (s) and large (l)EVs, which have a diameter of < 200 nm and > 200 nm, respectively¹. For this review, we will focus on sEVs and have chosen to use the term exosomes for sEVs that contain tetraspanins and/or other well-accepted exosomal markers. In general, EVs, regardless of their cargo or biogenesis, are released from cells and can interact locally with other cells and travel to distant sites to impart cellular changes.

Figure 1. Schematic of secreted particles with typical size ranges.

Extracellular vesicles have been classified by their origin and size into apoptotic bodies, microvesicles, and exosomes. Additional secreted particles include lipoproteins and two recently described amembranous nanoparticles, exomeres and supermeres.

Our entrée into the EV field and refinement of isolation methods

Our initial interest in EVs arose from our long-standing studies of EGFR ligand trafficking in polarized epithelial cells^11–16. We discovered that these type I transmembrane ligands are not only trafficked to the plasma membrane but are also packaged in exosomes, building on previous work that identified EGFR as an EV cargo¹⁷. Amphiregulin (AREG), one of the seven mammalian EGFR ligands, shows a propensity for being endocytosed and packaged in intraluminal vesicles within MVBs with its ectodomain facing outwards. Signaling competent, AREG-containing vesicles are released during MVB-plasma membrane fusion. We found that AREG-containing exosomes were much more potent in enhancing invasiveness of recipient breast cancer cells than equivalent amounts of recombinant AREG¹⁸. The effects of these exosomes were mediated, at least in part, by binding to EGFR, leading us to introduce exosomal targeted receptor activation (ExTRAcrine) as a new mode of EGFR ligand signaling.

After this initial foray into the EV field, a major focus became the impact of mutant KRAS on the sEV cargo produced by CRC cells. In a series of reports, we systematically examined the sEV protein and RNA content that was differentially produced by KRAS^G13D DLD-1 cells and their isogenic derivatives engineered to express only a single wild-type KRAS allele (DKs-8) or a single mutant KRAS allele (DKO-1)^19–23. In contrast to mutant KRAS cells, DKs-8 cells no longer exhibit a transformed phenotype. We were able to detect mutant KRAS in sEVs isolated from DLD-1 and DKO-1 cells and confirmed that transfer of mutant KRAS via sEVs could lead to transformation of DKs-8 and rat intestinal epithelial cells¹⁹.

We subsequently reported that the glucose transporter GLUT-1 (SLC2A1) was enriched in sEVs isolated from DKO-1 and DLD-1 cells, but not DKs-8 cells, and was functional in these mutant KRAS-derived sEVs as they were able to take up ¹⁸F-fluorodeoxyglucose (FDG)²⁰. Addition of these sEVs to wild-type KRAS cells led to increased aerobic glycolysis and growth in the recipient cells. More recent reports highlight exosomal circular RNAs, which will be elaborated on as potential biomarkers in this review, as mediating enhanced glycolysis in CRC^{24, 25}. Our results identified a cell-nonautonomous effect of mutant KRAS, suggesting that mutant KRAS-derived sEVs may “seed the soil” by altering the composition and metabolis state of cells within the tumor microenvironment. This work builds upon the pioneering contributions of Richard Simpson in the proteomic analysis of CRC EVs²⁶. Another important contribution by Dr. Simpson and his colleague Suresh Mathivanan was the creation of early searchable EV databases, EVpedia and ExoCarta, as well as Vesiclepedia, which is a web-based repository of proteins, RNA, lipids and metabolites identified in EVs that acts as a critical resource for the EV community^28–30.

As the work matured, we began to develop more refined methods of EV isolation. A wide range of isolation options exist, including ultracentrifugation (UC) with or without a density gradient, size-exclusion chromatography (SEC), immunoaffinity capture methods, microfluidic isolation, high-performance liquid chromatography (HPLC), and polymer precipitation^31–33. In 2019, our lab described rigorous methods for exosome isolation from non-vesicular material by generating a high-speed ultracentrifugation exosomal pellet (EV-P), loading it at the bottom of an optimized iodixanol gradient, performing another high-speed ultracentrifugation step, and collecting individual fractions for large scale proteomic and RNA profiling⁸. This work complemented and extended proteomic studies performed by Clotilde Théry’s group¹⁰. Our ongoing refinement of methods of isolation and parsing of particles led to our characterization of exomeres and supermeres^{3, 4}. David Lyden’s group was the first to identify exomeres using a costly, low-yield asymmetric flow field fractionation (AF4) method; however, we found exomeres could be isolated by simply performing high-speed ultercentrifugation of the supernatant from the EV-P³⁴. We recently identified yet another amembranous nanoparticle, the supermere, by performing a higher speed ultracentifugation of the supernatant from the exomere pellet⁴. We conducted a comprehensive protein and RNA profiling of supermeres, exomeres and sEVs from a CRC cell line, DiFi, along with correlating these findings to other cancer cell lines and human plasma⁴. While the field is consistently challenged by the heterogeneity intrinsic to EVs and other particles, we have made important strides towards characterization of these secreted mediators in CRC that will be useful for further functional studies as well as biomarker discovery.

EVs as a source of CRC biomarkers

With advances in isolation and characterization, a shift in thinking has occurred from DNA being in exosomes to its presence on the outside of EVs or in non-vesicular fractions⁸. We do not detect DNA within exosomes. Rather we have found that amphisomes, representing a fusion of autophagasomes and MVBs, are a source of extracellular dsDNA⁸. Major sources of circulating DNA biomarkers are thought to be dying cancer cells releasing naked DNA or circulating tumor cells themselves. However, a recent report found mutant KRAS and BRAF DNA on the outside of exosomes isolated from the plasma of CRC patients whose tumors harbor these mutations³⁵. The amount of either wild-type or mutant KRAS associated with exosomes was reported to be a prognostic marker for CRC³⁶.

KRAS protein also can affect micro (mi)RNA sorting into exosomes with miR-100 being preferentially packaged and released in exosomes from mutant KRAS CRC cells ²¹. Other reported miRNA biomarkers span from miR-17–5p and 125b-5p being associated with liver metastasis to a combination of let7b-3p and miR-145–3p predicting early-stage CRC^{37, 38}. A recent report used a bioinformatics approach of TCGA data to determine a CRC-specific, exosomal miRNA-mRNA network, and found that five miRNA hubs, encompassing miRs 141, 126, 139, 29c, and 423, were predictive for CRC³⁹. Along with miRNAs, circular RNAs (circRNAs) are attracting the attention of the EV research community. CircRNAs are single-stranded RNAs that form a closed loop and have gene regulatory roles. In CRC in particular, ciRs-7 has been shown to be a biomarker and therapeutic target, acting as a sponge for a tumor-suppressing miRNAs, leading to EGFR activation and association with poor patient survival⁴⁰. Circ-PNN, circ_0005963, and circ_0000338 have been shown to be EV-related circRNA biomarkers with the first being isolated from plasma exosomes and the latter two having a role in chemoresistance^{24, 41, 42}. Important considerations for RNAs as functional biomarkers include how many molecules are present in an EV, what quantity is needed to be functionally active, are the necessary RNA-binding proteins present, and how can isolation and quantification methods be improved to reduce the time, cost, and processing for clinical translation.

In addition to nucleic acids serving as EV biomarkers in CRC, a number of groups have examined the utility of EV-associated proteins for this purpose. Hakho Lee and co-workers have developed a technique that uses an integrated magneto-electrochemical device to capture and analyze EVs from 20 microliters of plasma within an hour⁴³. When this approach was applied to a bank of CRC plasma samples, it revealed that multi-marker combinations of EGFR, EpCAM, GPA33 and CD24 led to a 98% accuracy in diagnosing CRC⁴³. In our own studies, we have identified DPEP1, a dipeptidase normally expressed in the kidney, as a potential CRC biomarker⁴⁴. It is the most abundant protein in sEVs from a CRC line, DiFi, and is highly enriched in a subset of exosomes sorted on the basis of CD81 and EGFR positivity by fluorescence-activated vesicle sorting (FAVS)⁴. In addition to EGFR, this subset also contained EpCAM and GPA33, as well as CEACAM5 and CD73, highlighting an overlap with Hakho Lee’s group that fosters confidence that these proteins may be relevant CRC biomarkers⁴. Using a clinically well-annotated tissue microarray of over 150 CRC samples, we found that over 70% of CRCs express DPEP1 and that diffuse staining portended a worse overall and progression-free survival⁴. By examining DPEP1/CEACAM5 double-positive vesicles in plasma samples from three CRC patients and three healthy individuals, we have preliminary evidence that EV-bound DPEP1 could be a clinically relevant biomarker⁴. Figure 2 is a cartoon of a CRC exosome with classical tetraspainins and recently reported potential protein and RNA CRC biomarkers and Table 1 annotates their reported functions⁴⁵. One striking feature is that a number of these proteins are GPI-linked, which make them attractive biomarkers based on their efficient sorting into exosomes, cell surface localization, and overexpression in CRC⁴⁶. Isolating EVs from a biofluid is advantageous to simply profiling the overall patient sample since proteins such as albumin or other contaminants are so abundant that they can frustrate one’s ability to identify clinically relevant biomarkers. EVs also have other markers that can help determine the cell type of origin and can increase overall stability of the biomarker. Comparing clinical samples to validated reference materials and an adherence to accepted purification protocols will be necessary for clinical translation of basic research findings. Overall, profiling EVs is a tractable method for minimally invasive liquid biopsies, and further parsing of EV populations offers promise as a strategy to identify potential diagnostic, prognostic and predictive biomarkers.

Figure 2. Representative exosome with classical tetraspanins and cargo identified as potential CRC biomarkers from the recent literature.

Tetraspanins listed are in order of their specificity to classical exosomes¹³. Lipid modifications depicted can be farnesylation and geranylgeranylation for KRAS and myristoylation and palmitoylation for Src.

Table 1.

Colorectal Biomarkers Found in EVs and Their Disease Relevance

Biomarker category	Biomarker	Source	EV Isolation Method	Function in CRC*	References
Protein	GPA33	DiFi cells, human CRC plasma	Ultracentrifugation and FAVS	barrier function, biomarker	4,50
	EpCAM	DiFi cells, human CRC plasma	Ultracentrifugation and FAVS	proliferation, migration, EMT, cancer stemness	4,51
	EGFR	DiFi cells, human CRC plasma	Ultracentrifugation and FAVS	proliferation, invasion, metastasis	4, 52
	DPEP1	DiFi cells, human CRC plasma	Ultracentrifugation and FAVS	proliferation, metastasis	4, 53, 54
	CD73	DiFi cells	Ultracentrifugation and FAVS	immunosuppression, metastasis	4, 55
	CEACAM5	DiFi cells, human CRC plasma	Ultracentrifugation and FAVS	blocks differentiation, inhibits apoptosis	4, 56
	CD24	human CRC plasma	Ultracentrifugation and HiMEX	proliferation, invasion	43, 57
	Mutant KRAS	DKO-1 cells, DLD-1 cells	Ultracentrifugation	proliferation, cetuximab resistance	19, 20, 58
	Src	HT29 cells, HCT116 cells	Ultracentrifugation	proliferation, tumor formation	59
miRNA	miR-21	LS174, human CRC plasma	Ultracentrifugation	proliferation, invasion, 5-FU resistance	38, 60
	miR-29c	human CRC plasma	Ultracentrifugation	inhibits EMT, migration, invasion	39, 61
	miR-92a	human CRC plasma	Ultracentrifugation	proliferation, invasion migration	39, 62
	miR-126	human CRC plasma	Ultracentrifugation	inhibits proliferation, invasion, and migration	39, 63
	miR-135b	human CRC plasma	Ultracentrifugation and exosome precipitation	proliferation, inhibits apoptosis, chemoresistance	64, 65
	miR-139	human CRC plasma	Ultracentrifugation	inhibits EMT, enhances drug sensitivity, inhibits proliferation and invasion	38, 66
	miR-141	human CRC plasma	Ultracentrifugation	reduces migration and metastasis	39, 67
	miR-423	human CRC plasma	Ultracentifugation	cell growth	39, 68
mRNA	Rab13	DKO-1 cells	Density-gradient ultracentrifugation	regulates sEV secretion	22
circRNA	circ-PNN	human CRC plasma	Exoquick exosome precipitation	no reported function	41
	circ_0000338	SW480 cells, HCT116 cells, human CRC plasma	Ultracentrifugation	5-FU resistance	42
DNA	mutant KRAS (outside surface)	human CRC plasma	Ultracentrifugation and exosome precipitation	production of mutant KRAS protein	35
	mutant BRAF (outside surface)	human CRC plasma	Ultracentrifugation and exosome precipitation	production of mutant BRAF protein	35

^*Functions in CRC are based on cellular function and not consequences of EV transfer

Cancer-promoting properties of CRC EVs

EVs are not only an indicator of the state of a cell, but also functional entities that can impart cellular changes on recipient cells. Cargo transfer, cell-EV surface interactions to activate signaling cascades, and EV-mediated decoy mechanisms are all processes that can promote cancer. CRC cells can release EVs into the tumor microenvironment to modulate themselves or other stromal components for cancer promotion and progression⁴⁷. Typically, cancer cells secrete more EVs than normal cells, and in CRC a variety of factors contribute to altered secretion patterns⁴⁸. Activated Wnt signaling through loss-of-function mutations in APC, as well as collagen deposition, increases CRC sEV release⁴⁹. RAB27B, a regulator of the late endocytic pathway, has been implicated as a downstream target of Wnt signaling that promotes EV secretion as well as SNAP23, a component of the SNARE complex^{69, 70}. Wnt signaling not only stimulates EV release but also can be activated by transferred cargo. Mutant β-catenin was transferred by CRC EVs to recipient cells, leading to nuclear localization of β-catenin and activated Wnt signaling²⁷. This transfer led to increased tumor burden, highlighting potential resistance mechanisms and sEV-mediated tumor heterogeneity²⁷.

Another important signaling hub in CRC is initiated by EGFR activation⁷¹. Regulators of EGFR ligand sorting into EVs may be clinically relevant as this could activate EGFR signaling throughout a tumor. A recent study discovered that loss of tetraspanin 6 (Tspan6) increased packaging of TGF-α, an EGFR ligand, into exosomes, resulting in increased EGFR signaling⁷². The mechanism involves syntenin-1, a common exosome marker, linking Tspan6 to TGF-α⁷². High levels of Tspan6 in CRC patients led to a better response to cetuximab, an EGFR neutralizing monoclonal antibody, and portended a better prognosis for mutant KRAS patients⁷². If confirmed, these findings have important treatment implications since individuals with mutant KRAS CRC are excluded from receiving anti-EGFR monoclonal antibodies.

Yet another characteristic of most CRCs is an immunosuppressive immune environment. Roughly 15% of CRC patients have a deficient mismatch repair system and exhibit microsatellite instability (MSI), resulting in an influx of immune cells that recognize neoantigens, although the majority of CRCs are microsatellite stable (MSS) with an overall immunosuppressive phenotype⁷³. EVs, particularly exosomes, are critical contributors to these TME differences as serum exosomal miR-146a levels are correlated with decreased CD8+ T-cell numbers and increased neutrophil counts⁶⁹. CD8+ T-cell exclusion is associated with a worse prognosis, and neutrophil infiltration is emerging as a potential marker of CRC progression⁷⁴. A handful of studies have equated neutrophil infiltration, and in some cases subsequent lymphocyte exclusion, with the onset or progression of CRC ^{75, 76}. In terms of mechanism, neutrophils can secrete MMP9, which, in turn, activates TGF-β to suppress T-cell proliferation⁷⁷. While the connection between neutrophils and tumor progression is beginning to take shape, much less is known about how neutrophil EVs contribute. One model involves exosomes from CRC stem cells migrating to the bone marrow where RNA transfer led to prolonged neutrophil survival and conversion to a pro-tumorigenic phenotype⁷⁸. Yet another study discovered that exosomes from mutant KRAS CRC cells promote IL-8 and Neutrophil Extracellular Trap (NET) production in neutrophils that subsequently increases CRC proliferation, migration, and invasion⁷⁹.

CRC EVs not only modulate neutrophil behavior, but also fibroblast and macrophage function. CRC sEVs have been shown to educate fibroblasts towards a cancer-associated fibroblast (CAF) phenotype through enhanced Rho-Fak signaling⁸⁰. Another group reported that CRC EVs transferred integrin beta-like 1(ITGBL1) to liver and lung fibroblasts, leading to their activation and NFκB-mediated release of pro-inflammatory cytokines for metastatic niche formation⁸¹. These investigators detected ITGBL1 in the EV-P so further work is needed to confirm that ITGBL1 is in EVs rather than non-vesicular material; however, they did confirm the previous observation by David Lyden’s group that plasma CRC EVs enriched for ITGαv/ITGβ5 and ITGα6/ITGβ4 or ITGα6/ITGβ1 correlated with liver and lung metastasis, respectively⁸². CRC EVs can also induce pro-inflammatory cytokine production and PD-L1 induction in macrophages⁸³. In summation, CRC EVs can travel throughout circulation to interact with a variety of cell types to promote tumor progression.

Stromal cell-derived EVs promotion of CRC

Along with cancer cells releasing EVs for tumor progression, cells within the TME can also release EVs. In this review, we have chosen to focus on well-studied components that contribute to CRC such as CAFs and macrophages, as well as emerging players such as EVs derived from neutrophils and bacteria. Although CAFs are a heterogenous population of cells with no one distinct marker, a number of studies have established their ability to promote CRC⁸⁴. This also applies to CAF-derived exosomes as one study attributes metastatic capacity, EMT, and oxaliplatin and 5-fluoroacil resistance to a single exosomal miRNA, miR-92a-3p⁸⁵. Evidence for sEVs producing these cancer-promoting changes is supported by studies showing that sEVs derived from CAFs with lncRNA H19 enhance CRC stemness and chemoresistance, and CAF sEVs containing circEIF3K and miR-224–5p increase invasion and proliferation of CRC cells^86–88. AREG, which we have found in CRC exosomes, is present in CAF sEVs and can increase CRC proliferation, giving an alternative source of an EGFR ligand from a stromal component^{18, 89}.

Another cell type known to have roles in CRC is the macrophage⁹⁰. Recent studies regarding how macrophage sEVs impact CRC characteristics are focused on miRNA cargo and the M2 pro-tumorigenic subset^{91, 92}. miR-21–5p and miR-155–5p from M2 macrophage-derived exosomes regulate the levels of BRG1, a chromatin remodeling component, to increase metastatic capacity of recipient CRC cells, suggesting that stromal-derived exosomes influence CRC cell plasticity⁹¹. Another miRNA, miR-183–5p, was reported to be upregulated and released in exosomes from M2 macrophages⁹². This miRNA was reported to increase migration, invasion, colony formation, and reduce apoptosis in vitro as well as increase tumor volume and number of lung metastasis in vivo⁹².

A recent study reported that yet another stromal component, activated neutrophils, release exosomes enriched in neutrophil elastase (NE), a serine protease with broad substrate specificity⁹³. Exosomal NE was resistant to inhibition and could degrade extracellular matrix in the lung, triggering emphysema⁹³. It is intriguing to consider how activated neutrophils might act in a similar way in CRC to degrade matrix, leading to tumor cell invasion and metastasis. Another study described a feedforward loop whereby mRNA for the transcription factor salmonella pathogenicity island 1 (SPI1/PU.1) is released by neutrophil sEVs and taken up by CRC cells where it works in concert with SPI1-related protein (SPIB) to increase expression of hexose kinase 2 (HK2) and phosphoglycerate kinase 1 (PGK1)⁹⁴. This results in aerobic glycolysis with increased lactate production by CRC cells that drives neutrophils to become tumor-promoting⁹⁴.

Bacteria such as E. coli, Fusobacterium nucleatum, and Bacteroides fragilis have been associated with CRC incidence, presumed to be due to inflammation and oxidative stress⁹⁵. Mechanisms underlying bacterial tumor promotion are an active area of investigation and might include a role for bacterial EVs. EVs released by gut bacteria differ in composition and diversity between CRC and normal patients, as well as in early and late-stage disease with EVs from Firmicutes and Proteobacteria phyla being significantly changed^{96, 97}. Not only does CRC impact the microbiome, but bacterial vesicles can impact CRC progression. Fusobacterium nucleatum releases outer membrane vesicles that stimulate proinflammatory cytokine production in colonic epithelial cells, which could contribute to a pro-tumorigenic microenvironment⁹⁸. Bacterial EVs have been reported since the 1960s, but true characterization and functional studies have lagged due to difficulties in purification and a lack in consensus for common identifying markers^{99, 100}. Expanding the study of EVs to include the tumor microenvironment and microbiome will enhance our understanding of how EVs promote CRC.

Challenges and Future Opportunities

The EV field continues to be confronted by issues of isolation, characterization, and nomenclature. One outstanding issue within the field is the concept of physiological relevant concentrations. While there is no consensus about reporting absolute particle numbers or protein amount when adding EVs to cells, a larger issue looms. Hundreds of micrograms of EV protein are often used to observe effects in vivo, raising the concern that non-phyiological amounts of material are being introduced^{101, 102}. Studies are needed to determine the local concentrations of EVs released from cells and the concentrations that exist in circulation to better inform both in vitro and in vivo experiments. A variety of approaches are being taken to address these issues. From bioluminescence resonance energy transfer (BRET) imaging to pH-sensitive reporters coupled with dynamic correlative light-electron microscopy and TIRF to defining optical sigantures of EVs using label-free methods, technological innovations are being combined to address concentration concerns. It is anticipated that progress will be made as the field continues to grapple with the definition of a physiologically-relevant amount of EVs^103–105.

Use of the EV pellet, which contains non-vesicular material, has been used to identify a variety of extracellular biomarkers in cancer, but this broad stroke approach obscures relevant biomarkers that are seen only in a subpopulation of EVs or nanoparticles¹⁰⁶. By only analyzing top proteomic hits from a crude starting material, rare proteins that are enriched in specific subclasses of extracellular particles may be lost in the milieu. We have demonstrated by FAVS that parsing of sEVs into further subsets is of value as it not only informs the cell biology of cargo that are sorted together but also allows for identification of new biomarkers previously unrecognized as important for the initiation and progression of CRC^{4, 107}. These biomarkers though require rigorous validation for clinical translation. Our present labor-intensive and time-consuming methods do not lend themselves to easy scaling. A biomarker confirmation process specific to EVs should include verifying the biomarker in the tumor in comparison to normal samples and other inflammatory non-neoplastic states, as well as its presence in circulating EVs, and its reduction in plasma levels post-resection or after an effective therapeutic intervention¹⁰⁸. This validation process also hints at the clinical application of biomarkers for assessing CRC disease management, where a blood draw followed by isolation of the biomarker-containing EV population could be used to track the course of disease or response to therapy. Strategies that reduce the amount of starting material needed or specifically enrich for a certain biomarker may lead to improved disease monitoring.

Further technological developments and refinements, as demonstrated by the isolation of exomeres first by AF4, but eventually by our group by a simple sequential centrifugation spin of the supernatant from an EV-P, will lead to parsing of the heterogeneity present in EV and nanoparticle classes for accurate subclass categorization^{3, 34}. Another improvement is the use of microfluidics devices that require as little as 20 microliters of plasma sample in order to quantify miRNAs in sEVs¹⁰⁹. Since many commercial flow cytometers are not designed to detect sEVs and related nanoparticles, the focus of round 2 of the Extracellular RNA Communication Consortium (ERCC), of which we are a part, is to design EV-centric devices, such as a flow cytometer designed for the purpose of analyzing and sorting EVs. Our own studies have shown that further refinements often lead to reassessing carrier contents as we have recently shown that supermeres, and not exomeres, as originally reported by Lyden and our group, are a more abundant source of metabolic cargo and glycolytic enzymes^{3, 4, 34}. Furthermore, at least in some contexts, superemeres contain a higher percentage of secreted RNAs including miRNAs than do EVs or exomeres, highlighting the possibility that effects ascribed to secreted RNAs will be greater when measuring their function within supermeres⁴. As this field evolves and matures, further refinements in nomenclature and parsing of complex EV fractions will lead to a clearer analysis when attributing function to a particular EV class.