Page 81 - Read Online

Page 175 Gupta et al. Extracell Vesicles Circ Nucleic Acids 2023;4:170-90 https://dx.doi.org/10.20517/evcna.2023.12

cells with 5 nm SPIONs [88,89] . The advantage of radiolabelling and MRI is the exceptional sensitivity in vivo
over other light-based reporters, but due to high infrastructure costs, these technologies are hard to
implement in basic science research.

Fluorescent and bioluminescent proteins
In addition to a variety of exogenous labelling strategies, EVs can be genetically engineered with fluorescent
or bioluminescent proteins to label all the EVs or a specific population thereof [80,82,86,90-92] . For labelling a
specific population, producer cells are genetically engineered to express a reporter protein fused to an EV
sorting domain to permit the loading of the reporter protein during EV biogenesis. For example, a fusion
protein of CD63 and EGFP can drive the sorting of EGFP into EVs and label 30%-40% of the EV
[91]
population, each carrying 30-60 EGFP molecules on average . This approach is not limited to CD63 since
other EV sorting domains can also be exploited for labelling EVs (e.g., CD9, CD81, syntenin, and Gag) [91,92] .
Although it seems straightforward, the EV engineering efficiency with certain EV sorting domains such as
[91]
ALIX, SIMPLE and syndecan is relatively low . This could be due to the potential loss of the protein’s
function due to the fusion of a reporter protein. Overall, genetic engineering approaches provide an efficient
way of labelling a specific EV population either with fluorescent proteins (GFP, RFP, etc.) or bioluminescent
proteins (Gaussia-, Firefly-, and Nano-luciferase). However, these approaches fail to label all EVs and
require genetic engineering of the producer cells, which could be challenging for some cell sources. In
addition, overexpression of certain EV sorting domains may alter the EV biogenesis pathway or EV
proteome, which can influence their biodistribution. An ideal EV reporter or labelling strategy does not
exist at this moment. With the multitude of EV labelling strategies available, each has some degree of
advantage and disadvantage; therefore, it is essential to select a method based on indication and feasibility.

IN VIVO UPTAKE OF EVS
The general trend of EVs in vivo biodistribution
Since the first study showing CNS delivery of synthetic siRNA by EVs, interest in EVs as drug delivery
[93]
technology has gained increasing attention . For translational research, it is of uttermost importance to
understand the distribution of EVs in vivo. Numerous studies in the past decade have showcased EVs in
vivo biodistribution [76,80,82,94] . Since the early use of lipophilic dyes, EV imaging modalities have evolved, and
with the development of endogenous labelling strategies, biodistribution can be evaluated with much higher
accuracy. The majority of studies have shown that upon systemic administration either by intravenous,
intraperitoneal, or subcutaneous injections, EVs tend to primarily accumulate in the liver, lungs, and spleen
and to a lesser extent in other organs and tissues such as the brain, muscle, heart, and kidneys [80,95,96] . This
distribution occurs within minutes of systemic administration of EVs, which is also reflective of the short
plasma half-life of EVs. Notably, across literature and based on a meta-study, the peak time of EV
accumulation varies drastically from 5 mins to 12 h across studies, but surprisingly the plasma half-life of
[96]
EVs is always reported short across studies . This variability could be due to the differences in EV labelling
methods employed in different studies, as different dyes, particularly lipophilic dyes, can shuttle from an EV
membrane to a cellular membrane and may not display EV tissue distribution [7,97,98] . Secondly, and
importantly, the tissue accumulation should be reflective of plasma levels of the EVs, as plasma is primarily
the main reservoir of injected exogenous EVs. Therefore, biodistribution models with fast plasma clearance
and rapid biodistribution profiles are more logical. In addition, the short half-life of EVs and hepatic
tropism is very similar to the majority of synthetic nanoparticles and viral vectors, which indicates that in
terms of in vivo biodistribution, they all experience similar biological barriers . Importantly, this short
[99]
plasma half-life is observed by various studies utilizing different cell sources of EVs and may imply that cell
sources may not play a significant role in regulating EV plasma half-life [76,80,82,85,94,95] . The different factors
potentially regulating the short plasma half-life of EVs will be discussed later in this review. In contrast, EVs
show different half-lives in tissues after being taken up from plasma. These varying levels could be different