ArticlePDF Available

Nanostructure enabled extracellular vesicles separation and detection

Authors:

Abstract

Extracellular vesicles (EVs) have recently attracted significant research attention owing to their important biological functions, including cell-to-cell communication. EVs are a type of membrane vesicles that are secreted into the extracellular space by most types of cells. Several biological biomolecules found in EVs, such as proteins, microRNA, and DNA, are closely related to the pathogenesis of human malignancies, making EVs valuable biomarkers for disease diagnosis, treatment, and prognosis. Therefore, EV separation and detection are prerequisites for providing important information for clinical research. Conventional separation methods suffer from low levels of purity, as well as the need for cumbersome and prolonged operations. Moreover, detection methods require trained operators and present challenges such as high operational expenses and low sensitivity and specificity. In the past decade, platforms for EV separation and detection based on nanostructures have emerged. This article reviews recent advances in nanostructure-based EV separation and detection techniques. First, nanostructures based on membranes, nanowires, nanoscale deterministic lateral displacement, and surface modification are presented. Second, high-throughput separation of EVs based on nanostructures combined with acoustic and electric fields is described. Third, techniques combining nanostructures with immunofluorescence, surface plasmon resonance, surface-enhanced Raman scattering, electrochemical detection, or piezoelectric sensors for high-precision EV analysis are summarized. Finally, the potential of nanostructures to detect individual EVs is explored, with the aim of providing insights into the further development of nanostructure-based EV separation and detection techniques.
View
Online
Export
Citation
CrossMark
REVIEW ARTICLE | SE PT EM BE R 28 2 02 3
Nanostructure enabled extracellular vesicles separation and
detection
Xinyuan He; Wei Wei; Xuexin Duan
Nanotechnol. Precis. Eng. 6, 045002 (2023)
https://doi.org/10.1063/10.0020885
05 October 2023 01:58:34
Nanotechnology and
Precision Engineering REVIEW pubs.aip.org/aip/npe
Nanostructure enabled extracellular vesicles
separation and detection
Cite as: Nano. Prec. Eng. 6, 045002 (2023); doi: 10.1063/10.0020885
Submitted: 11 April 2023 Accepted: 19 May 2023
Published Online: 28 September 2023
Xinyuan He,1Wei Wei,1and Xuexin Duan1,2,a)
AFFILIATIONS
1State Key Laboratory of Precision Measuring Technology and Instruments, Tianjin University, Tianjin 300072, China
2Department of Chemistry, School of Science, Tianjin University, Tianjin 300072, China
a)Author to whom correspondence should be addressed: xduan@tju.edu.cn
ABSTRACT
Extracellular vesicles (EVs) have recently attracted significant research attention owing to their important biological functions, including
cell-to-cell communication. EVs are a type of membrane vesicles that are secreted into the extracellular space by most types of cells. Several
biological biomolecules found in EVs, such as proteins, microRNA, and DNA, are closely related to the pathogenesis of human malignancies,
making EVs valuable biomarkers for disease diagnosis, treatment, and prognosis. Therefore, EV separation and detection are prerequisites for
providing important information for clinical research. Conventional separation methods suffer from low levels of purity, as well as the need
for cumbersome and prolonged operations. Moreover, detection methods require trained operators and present challenges such as high opera-
tional expenses and low sensitivity and specificity. In the past decade, platforms for EV separation and detection based on nanostructures have
emerged. This article reviews recent advances in nanostructure-based EV separation and detection techniques. First, nanostructures based on
membranes, nanowires, nanoscale deterministic lateral displacement, and surface modification are presented. Second, high-throughput sep-
aration of EVs based on nanostructures combined with acoustic and electric fields is described. Third, techniques combining nanostructures
with immunofluorescence, surface plasmon resonance, surface-enhanced Raman scattering, electrochemical detection, or piezoelectric sen-
sors for high-precision EV analysis are summarized. Finally, the potential of nanostructures to detect individual EVs is explored, with the aim
of providing insights into the further development of nanostructure-based EV separation and detection techniques.
©2023 Author(s). All article content, except where otherwise noted, is licensed under a Creative Commons Attribution (CC BY) license
(http://creativecommons.org/licenses/by/4.0/). https://doi.org/10.1063/10.0020885
KEYWORDS
Nanostructure, Extracellular vesicle, Separation, Detection, Individual
I. INTRODUCTION
A. Introduction to extracellular vesicles
Cells from diverse domains of life secrete membrane-enclosed
vesicles known as extracellular vesicles (EVs).1EVs carry spe-
cific protocellular components, including DNA, RNA, lipids, pro-
teins, and metabolites.2,3 These components mediate intercellu-
lar communication, especially in oncology.4EVs participate in
cell–microenvironment interactions and influence cell prolifera-
tion, migration, immune regulation, pregnancy, and cardiovascular
diseases, among other conditions.5EVs are ubiquitous in various
human body fluids such as blood, saliva, semen, sputum, urine,
cerebrospinal fluid, amniotic fluid, and breast milk.6EV formation
involves a complex process in which cells endocytose cell surface
proteins and extracellular soluble proteins to form early endosomes
(ESEs). ESEs mature into late endosomes (LSEs), which eventually
become multivesicular bodies (MVBs). MVBs fuse with the plasma
membrane and release their intraluminal vesicles (ILVs) as EVs by
exocytosis.7EVs are classified into exosomes, microvesicles (MVs),
and apoptotic bodies based on their origin and size.8Exosomes are
small lipid bilayer vesicles with an average thickness of about 5 nm,
derived from MVBs with diameters of 30–150 nm and densities of
1.08–1.22 g/ml.9MVs have diameters of 50–1000 nm and densi-
ties of 1.12–1.16 g/ml, and they bud directly from the outer plasma
Nano. Prec. Eng. 6, 045002 (2023); doi: 10.1063/10.0020885 6, 045002-1
© Author(s) 2023
05 October 2023 01:58:34
Nanotechnology and
Precision Engineering REVIEW pubs.aip.org/aip/npe
membrane. Apoptotic bodies have diameters of 500–5000 nm and
are secreted by apoptotic cells during the late stages of programmed
cell death.10
In recent years, EVs have attracted much research interest. EVs
carry specific biomarkers, such as tetraspanins (CD9, CD63, and
CD81), integrins, immunomodulatory proteins, and nucleic acids
[DNA, mRNA, and microRNA (miRNA)].11 These biomarkers act
as a biological fingerprint of the source cell, reflecting its biologi-
cal information.12 EVs are intimately involved in tumor initiation,
proliferation, and metastasis.7,13,14 Studies have shown that tumor
cells secrete more EVs than normal tissue cells.15,16 The proteins and
nucleic acids in EVs differ not only between cancer patients and
healthy individuals, but also at different stages of cancer or other
disease.17,18 Therefore, EVs are potential biomarkers for early diag-
nosis and prognosis of cancer and other diseases and have great
clinical diagnostic value.19,20 Dash et al.21 explored the potential of
plasma EV membrane proteins from tumor cells as biomarkers for
early colorectal cancer (CRC) detection. They found that the expres-
sion levels of ADAM10, CD59, and TSPAN9 were 2.19–5.26-fold
higher in plasma EVs of CRC patients. Li et al.22 examined cir-
cular RNA (circRNA) in EVs secreted by high-grade astrocytoma.
They constructed a circRNA panel of serum EVs and discovered
that circRNA could serve as a liquid biopsy target for monitoring
and treating high-grade astrocytoma (HGA). Moreover, EVs can
be used not only for diagnosis but also for therapy. In addition,
EVs are an excellent natural delivery vehicle, which can be used to
load biomolecules for targeted therapy. Liang et al.23 designed EVs
derived from HEK293T cells as vectors targeting CRC cells. Anti-
cancer miR-21 inhibitors were loaded onto these EVs, mediated by
electroporation. It was observed that tumors were reduced in mice
with colon cancer by intravenous injection of these EVs in vivo.
B. Existing separation and detection techniques
Traditional techniques for separating EVs include ultracen-
trifugation, precipitation, immunoaffinity, ultrafiltration, and size
exclusion chromatography.24 Ultracentrifugation (UC) is consid-
ered the gold standard for separating EVs.19 This method extracts
EVs by centrifuging at low speeds (300g–2000g) and gradually
increasing to high speeds (100000g–200 000g) to separate cells, cell
debris, vesicles, and proteins.25 However, UC also has clear draw-
backs. First, the requirement for a large sample volume (>10 ml)
makes it unsuitable for processing trace biological samples.26 Sec-
ond, the long centrifugation time and high rotational speeds damage
the integrity of EVs, potentially affecting the feasibility of their
downstream analysis. This method also causes co-precipitation of
proteins, leading to poor recovery.19 Third, UC requires expen-
sive equipment and trained operators, which limits its clinical
application.19,26
Size exclusion chromatography (SEC) is a method that sepa-
rates EVs based on their size.20,27 SEC utilizes a stationary phase
column consisting of spherical polymer porous beads with varying
pore sizes to separate EVs.10,28 This technique enables smaller EVs to
enter the pores and be eluted with phosphate-buffered saline (PBS),
while larger EVs cannot diffuse into the pores and move between
the porous beads. Compared with UC, SEC does not require expen-
sive equipment, it is easy to operate, and it separates EVs more
completely.29 However, contaminants such as lipoprotein may still
remain in the EVs after separation.30 Furthermore, the SEC sta-
tionary phase may interact nonspecifically with the sample, which
can alter the selectivity of separation. SEC is also limited by low
concentration and low size resolution.27,31
Another approach to separate EVs using solubility or dis-
persibility is a precipitation-based technique. Polymer or saline
solution, such as polyethylene glycol (PEG), is usually used as a
precipitating agent.32 The polymer binds water molecules and forces
poorly soluble EVs out of solution.33 Precipitants and samples are
incubated overnight under certain conditions, and then the precipi-
tate containing EVs is separated by low-speed centrifugation (1500g)
or filtration.34,35 Finally, the residue is washed with PBS solution for
further downstream analysis. Reports in the literature mention that
the pH value of the EV separation environment is one of the key
parameters affecting the separation quality and yield. Although this
method does not require specialized equipment, is easy to use, and
can handle large sample volumes, the precipitating agent interferes
with downstream analysis and the precipitation process damages
the membrane structure of EVs. Moreover, the co-precipitation
of non-EV particles (e.g., proteins) can reduce separation
efficiency.36
One of the more popular methods for separating EVs using
physical properties is ultrafiltration (UF).37 The working principle
of UF is similar to that of traditional filtration, which is based on
the size or molecular weight of particles for classification. During
the filtration process, particles larger than the pore size are trapped
on the membrane, while particles smaller than the pore size pass
through the membrane.38,39 The operational steps of UF are sim-
ple and easy, and it is time-saving and low-cost, but it also has
prominent disadvantages. During the passage of EVs through the
membrane pores, they will be subjected to high pressures, leading to
extrusion deformation, which can result in morphological changes
or damage, thus interfering with downstream analysis.40 Also, the
membrane will become clogged after filtration, both by large par-
ticles on the membrane surface and by small particles inside the
membrane pores, resulting in a decrease in flux and a decrease in
recovery.41
The immunoaffinity method relies on interaction between
antigen and antibody or between receptor and ligand to separate
EVs.10,27 This technique is supported by abundant antigen mark-
ers on the membrane surface of EVs, such as CD9, CD63, CD81,
and others.25,42 A common approach is to immobilize specific
antibodies on magnetic beads, chromatographic matrices, multi-
well plates, and microfluidics to capture EVs.43,44 Immunoaffinity-
based chromatography methods resemble traditional chromato-
graphic separations by immobilizing antibodies in the stationary
phase and EVs in the mobile phase. Different components in
the sample can be separated by exploiting their different elution
rates.45 Some literature reports indicate that the immunoaffin-
ity method has much higher capture efficiency and yield of EVs
than UC.40,46
In addition, traditional detection technologies are divided into
many categories, which are only listed here and not described
in detail. Methods for characterizing the size and concentration
of EVs include nanoparticle tracking analysis (NTA),47 dynamic
light scattering (DLS),48 and flow cytometry (FCM).49 Meth-
ods for characterizing EV morphology include scanning electron
microscopy (SEM), transmission electron microscopy (TEM), and
Nano. Prec. Eng. 6, 045002 (2023); doi: 10.1063/10.0020885 6, 045002-2
© Author(s) 2023
05 October 2023 01:58:34
Nanotechnology and
Precision Engineering REVIEW pubs.aip.org/aip/npe
atomic force microscopy (AFM).37 Enzyme-linked immunosorbent
assay (ELISA)50 and immunoblotting are used to characterize the
types of proteins contained in EVs.51 Methods for detecting RNA
in EVs include quantitative reverse transcription polymerase chain
reaction (qRT-PCR). In addition, surface plasmon resonance (SPR)
imaging,52 tunable resistive pulse sensing (TRPS),53 laser tweezers
Raman spectroscopy (LTRS),54 and other methods are used for EV
detection.
In the past decade, researchers have developed a variety of com-
mercial instruments or kits to separate and detect EVs from cell cul-
tures or body fluids.55 However, owing to the complex composition
of biological samples and where EV components intersect in den-
sity and size range, most available methods cannot accurately isolate
or analyze specific constituent parts of EVs—particularly proteins
and nucleic acids.56,57 Consequently, most of the EV-related research
literature currently available is typically based on mixed subsets
of EVs secreted by different cells, including additional EV com-
ponents and biologically active acellular components. Therefore,
the true nature of EVs with their specific functions and structural
dimensions has been challenging to elucidate clearly, thus hindering
the investigation of their biological and clinical applications.24 The
limitations of the methods mentioned above have brought signifi-
cant challenges to the development of EV separation and detection
techniques.
Advances in nanotechnology and micro- and nanoprocess-
ing techniques have enabled the production of nanostructures that
are precisely sized to fit EVs. Incorporating these structures into
microfluidic chips increases the EV exposure probability by expand-
ing the surface area available. Nanostructures not only provide
higher separation efficiency, purity, and throughput for EV sepa-
ration, but also enable enhanced accuracy, specificity, and sensi-
tivity through nanoscale detection.58–60 Through nanotechnology,
researchers have designed a variety of functional nanostructures
SCHEME 1. Nanostructure-based EV separation and detection methods.
with the advantages of low sample consumption, low cost, and
portable operation. The combined advantages of nanotechnology
and microfluidic technology provide precious resources for obtain-
ing suitable EV samples for further research and clinical application.
Although microfluidic systems for separating and detecting EVs are
still in their infancy, the fusion of nanotechnology and microfluidics
has achieved outstanding results. In this review, we first divide the
nanostructures used to separate EVs into three categories, namely,
nanostructures without external field, nanostructures with an exter-
nal field, and surface-modified nanostructures, and we review the
development of nanostructure-based separation methods in typical
examples of biomedical and clinical analysis. Second, we intro-
duce the nanostructure-based EV detection methods and review
the innovative applications of nanostructures in the detection field.
Third, we review the capture and detection of individual EVs
based on nanostructures, fully demonstrating the advantages of
nanostructures in the analysis of individual EVs (Scheme 1). In
each section, we will introduce the principles and mechanisms
in detail and provide an in-depth analysis of their advantages
and challenges. Finally, we discuss the future development trends
of nanostructures and their potential applications in biomedical
fields.
II. NANOSTRUCTURE-BASED EV SEPARATION
METHODS
The separation of EVs based on nanostructures is an emerg-
ing technique that is compatible with microfluidic chips to create
nanoscale lab-on-a-chip devices. Nanostructures provide a high
area-to-volume ratio, increasing the contact probability with EVs,
and requiring smaller sample volumes with fewer reagents. This
approach also eliminates the need for the bulky equipment required
by conventional methods, since it combines multiple separation
methods and components on a single chip. Techniques for sep-
arating EVs based on nanostructures are classified according to
whether they use physical properties or biological characteristics.
Physical property-based mechanisms use nanostructures with or
without external fields, taking advantage of EV density, size, and
electrical properties. Such microfluidic platforms rely on filtra-
tion, nanowires, acoustics, deterministic lateral displacement, and
electricity. Meanwhile, biological property-based separation relies
on specific interactions of EV surface biomarkers with capture
agents. These approaches employ immunoaffinity-based antibodies
or aptamers, as well as polymeric reagents.
A. Nanostructure without external fields
Generally speaking, the platforms based on microfluidics com-
bined with nanostructures without external fields use pressure or
centrifugal force to achieve the separation of EVs. These platforms
rely on differences in physical properties to separate different kinds
of EVs.24,58 According to the current literature, exosomes typically
range in diameter from 30 to 150 nm and have densities ranging
from 1.08 to 1.22 g/ml.9Nanostructures without external fields sep-
arate EVs with the same physical properties, sacrificing purity while
improving size resolution, and thus the requiring further down-
stream analysis.61 In this subsection, we will introduce platforms
to achieve separation based on the above EV physical properties,
Nano. Prec. Eng. 6, 045002 (2023); doi: 10.1063/10.0020885 6, 045002-3
© Author(s) 2023
05 October 2023 01:58:34
Nanotechnology and
Precision Engineering REVIEW pubs.aip.org/aip/npe
divided into (1) the dead-end filtration method, (2) the tangen-
tial flow filtration method, (3) the nanowire filtration method, and
(4) the nanoscale deterministic lateral displacement (nano-DLD)
method (Table I).
1. Dead-end filtration method
Dead-end filtration is a classic filtering mode often used in
a variety of application scenarios. Particles and solutions smaller
than the membrane pore diameter pass through the membrane,
while particles larger than the pore diameter are trapped on the
membrane.62 There are two ways to create the required pressure
difference: the first is to add positive pressure upstream of the mem-
brane (i.e., at the inlet), and the second is to add negative pressure
downstream of the membrane (i.e., at the outlet). Combining two
or more membranes can be used for sequential filtration to remove
large particles and proteins, thereby retaining EVs in a specific size
range.63
On a microfluidic-based filtration platform, Liang et al.14 devel-
oped an integrated dual-filter microfluidic device to separate EVs
from the urine of bladder cancer patients [Fig. 1(a)]. The device
contained polymethyl methacrylate (PMMA) layers with two fixed
polycarbonate (PC) membranes of sizes 200 and 30 nm, respec-
tively. Subsequent to positive pressure filtration, microbeads and
particles exceeding 200 nm were trapped on the larger membrane,
while nucleic acids and proteins below 30 nm passed through the
smaller one, enabling EV retention. Anti-CD63 antibody labeled and
formed an immune complex with EVs, which was followed by inter-
action with streptavidin–horseradish peroxidase (SA-HRP), leading
to on-chip separation, enrichment, and quantification of EVs in
urine. The nonspecific adsorption of EVs on the membrane sur-
face in such as device may reduce the recovery rate. As a result,
the membrane surface becomes clogged, decreasing filtration effi-
ciency. Also using a dual-filtration system, Dong et al.64 devised an
integrated microfluidic chip named ExoIDChip for efficient sepa-
ration and high-sensitivity detection of EVs [Fig. 1(b)]. The chip
had a slidable mechanical structure connecting two fluid channels.
After completion of filtration in the left channel, the upper chan-
nel was moved to the right, and the excess AptCD63 aptamer passed
through. Enrichment of unbound aptamers was performed by pho-
tonic crystal nanostructures, which enhanced fluorescence when
bound to nitrocellulose membranes. In the final detection stage,
infused SA-HRP quantified EVs by a competitive immunoassay,
enabling sensitive EV detection. The photonic crystal nanostructure
had a fluorescence enhancement effect. ExoID-Chip was used to dif-
ferentiate serum samples from breast cancer patients and healthy
individuals.
On the basis of the dead-end filtration method, Liu et al.63
developed a new type of filter chip called ExoTIC [Fig. 1(c)]. This
used PC membranes with different pore sizes as filters to separate
EVs from different clinical samples of blood, lungs bronchoalveolar
lavage fluid, cell culture medium, saliva, and urine. On this plat-
form, the membranes were modularized to allow easy connection
to syringes and syringe pumps. After filtration for some time, the
device was rotated through 180to prevent clogging from causing
a drop in flux. ExoTIC gave a four times higher yield than UC, and
a three to four times higher yield than a PEG precipitation kit, with
a separation efficiency exceeding 90%. Furthermore, it was possible
to separate EV subpopulations in different size ranges by cascading
together filters of different sizes (e.g., 30, 50, 80, 100, and 200 nm).
Seder et al.65 fabricated an automated non-microfluidic plat-
form [Fig. 1(d)]. The working principle of this platform was similar
to that of ExoTIC, cascading membranes with different pore sizes
(30, 50, 80, 100, and 200 nm) and using positive pressure to filter
TABLE I. Nanostructure-based EV separation method without external fields.
EV separation method Separated size (nm) Purity (%) Recovery yield (%) Throughput (μl/min) Reference
Dead-end filtration:
Dual-filter microfluidic device 155 NA 74.2 40 14
ExoIDChip 20–200 NA NA 10 64
ExoTIC 30–100 75 >90 83 63
Movable-layer device 30–200 NA 89 13 65
Tangential flow filtration:
Exodisc 20–600 55 >95 36 68
Exo-Hexa 20–450 83 >95 133 67
Serpentine channel device 100 >97 80 50 70
Ultrathin silicon nitride nanomembrane 80 NA 17.2 5 71
Nanowire separation:
Ciliated microcolumn arrays 83–120 NA 45–60 10 38
Nanowire-induced electrostatic collection 30–200 >99 NA 50 74
3D carbon nanotube arrays 80–300 NA 47–55 200 75
Nanoscale deterministic lateral
displacement (nano-DLD):
Nano-DLD sorting using pillar array <100 NA 99 0.0001–0.0002 76
Nano-DLD sorting 30–200 NA 50 15 77
i-nanoDLD 20–1000 >95 NA 283 85
Nano. Prec. Eng. 6, 045002 (2023); doi: 10.1063/10.0020885 6, 045002-4
© Author(s) 2023
05 October 2023 01:58:34
Nanotechnology and
Precision Engineering REVIEW pubs.aip.org/aip/npe
FIG. 1. Platforms for EV separation using dead-end filtration. (a) Microfluidic dual-filtration platform for EV separation.14 Reproduced from Liang et al., Sci. Rep. 7, 46224
(2017). (b) Schematic of the ExoIDchip.64 Reproduced from Dong et al., Lab Chip 19, 2897–2904 (2019). (c) Schematic and workflow of ExoTIC.63 Reproduced from
Liu et al., ACS Nano 11, 0712–10723 (2017). (d) EV separation device with a vertically moving piston and a rotating chip to separate EVs according to their sizes.65
Reproduced from Seder et al., Lab Chip 22, 3699–3707 (2022).
EVs. The main structure of this automatic device was composed of a
vertically movable piston and a rotatable disc. Seven open chambers
were distributed on the disk, with each chamber being connected
by a microchannel and a one-way valve, which also contained PC
membranes with different pore sizes. Experiments with this device
showed that when the piston was pressurized at a speed of 5 μm/s,
EVs did not deform significantly. Compared with UC, the purity of
the isolate was ten times higher, and the recovery rate was as high as
89%. The results of EV characterization showed that the greater the
number of EVs secreted by the cells, the higher were the expressions
of CD63 protein and TSG101 protein. It was found that the device
could separate EVs with different sizes of 30, 50, 80, 100, and 200 nm
at high resolution and could separate EV subpopulations according
to size.
Although dead-end filtration has a simple structure, easy oper-
ation, and low cost, as the filtration time becomes longer, large
particles will accumulate on the membrane, forming a cake structure
and causing membrane pollution. The cake will cause the filtra-
tion resistance to increase continuously, the permeability of the
membrane will decrease, and this will eventually lead to failure of
separation. In addition, higher pressures may damage the integrity
of EVs.
2. Tangential flow filtration method
Tangential flow filtration (TFF) is a filtration method where
the flow direction of the sample is perpendicular to the filtration
direction, so that the shear force of the tangential fluid reduces the
accumulation of large particles on the membrane surface.66 TFF
equipment has been developed, such as LabSpinner’s lab-on-a-disc,
which uses centrifugal force to achieve TFF.67 Woo et al.68 developed
Exodisc for label-free, rapid, and sensitive EV separation and detec-
tion. This platform enabled fully automated EV separation of 1 ml
samples within 30 min from cell culture supernatant (CCS) or cancer
patient urine, with the sample being under a centrifugal force (about
500g) much less than that in UC. The sample passed successively
through two membranes, with the combination of a 20 nm anodic
aluminum oxide (AAO) membrane and a 600 nm PC membrane
proving most suitable. Microvalves automatically controlled differ-
ent chambers, and the 20 nm AAO membrane was able to filter out
protein. Although EV recovery from CCS was found to be greater
than 95%, the purity of the EVs was affected by co-segregation of
larger EVs (size range 200–600 nm) [Fig. 2(a)]. Subsequently, the
same group made adjustments to Exodisc and developed Exodisc-
B.69 With this, they were able to enrich EVs from whole blood
samples with a volume of 30–600 μl within 40 min. The device had a
Nano. Prec. Eng. 6, 045002 (2023); doi: 10.1063/10.0020885 6, 045002-5
© Author(s) 2023
05 October 2023 01:58:34
Nanotechnology and
Precision Engineering REVIEW pubs.aip.org/aip/npe
FIG. 2. Platforms for EV separation using tangential flow filtration. (a) Schematic of Exodisc.68 Reproduced from Woo et al., ACS Nano. 11, 1360–1370 (2017). (b) Schematic
of Exo-Hexa.67 Reproduced from Woo et al., Lab Chip 19, 87–97 (2019). (c) Tangential flow filter with serpentine microchannels.70 Reproduced from Han et al., Sensors Act.
B Chem. 333, 129563 (2021). (d) Ultrathin silicon nitride membrane filtration system.71 Reproduced from Dehghani et al., Adv. Mater. Technol. 4, 1900539 (2019).
capture efficiency of 75% at a lower centrifugal force (66g). Further-
more, Exodisc extracted mRNA from EVs with a 100-fold increase
compared with UC.68 Finally, on-chip ELISA was used to detect EVs
from the urine of bladder cancer patients and showed high levels of
CD9 and CD81 expression. This platform is clinically significant for
detecting EVs in urine for cancer diagnosis. In 2019, Woo et al.67
developed Exo-Hexa, which had an improved chamber structure
compared with Exodisc, increasing the purity of the isolate from 55%
to 83%. Throughput, was also improved with this device, with one
disc being able to process six samples simultaneously [Fig. 2(b)].
Apart from centrifugal force, the driving force of tangential
flow can also be provided by a syringe pump or a vacuum pump.
Han et al.70 successfully separated and purified EVs from human
blood using tangential flow via a symmetrical two-layer PMMA ser-
pentine flow channel and a 100 nm PC membrane, which separated
over 97% of protein and other impurities. The collected EVs were
analyzed using matrix-assisted laser desorption/ionization time-of-
flight (MALDI-TOF) mass spectrometry (MS), proving effective
clearance of plasma proteins, with only the protein spectrum of EVs
present [Fig. 2(c)]. Dehghani et al.71 developed a tangential flow ana-
lyte capture (TFAC) EV separation and purification method using
an ultrathin silicon nitride nanomembrane of 10 μm thickness and
80 nm pore size to increase capture efficiency and avoid cake for-
mation, reducing pressure drop. Compared with PC membranes,
ultrathin silicon nitride membranes demonstrate low clogging and
transmembrane pressure drop, sustaining continuous filtration for
a longer period while also enhancing the separation and release
efficiency of EVs. The device also supported in situ EV analysis,
outperforming dead-end filtration [Fig. 2(d)].
3. Nanowire separation method
Nanowires represent another size-based technology that lever-
ages the physical properties of EVs for separation by integrating
mechanical structures into microfluidic platforms. A prime example
of nanowire implementation is an EV filtration system compris-
ing a structural array of ciliated microcolumns, developed by Wang
et al.38 This configuration possessed hierarchical filtration capabil-
ity [Fig. 3(a)]. Traditional micromachining techniques were used
to fabricate a microcolumn array on a silicon substrate, permitting
the use of an electrodeposited silver nanoparticle catalyst in electro-
less etching on the sidewall of the microcolumn, finally causing cilia
formation. Since the spacing of the nanowires was adjustable, rang-
ing from 30 to 200 nm, this enabled the capture of EVs within the
corresponding size range (40–100 nm). Furthermore, EVs could be
retrieved without damaging the nanowires, by soaking them in PBS
solution.38 However, owing to its low throughput, this platform had
limited prospects for practical application. Rahong et al.72 fabricated
a 3D nanowire structure embedded in a PDMS microchannel via
a bottom-up approach, allowing them to grow tin oxide nanowires
with an average diameter of 18.9 nm on a quartz substrate. This
method was able to quickly separate biomolecules, such as proteins,
DNA, and RNA. The concentration of EVs in urine is extremely
Nano. Prec. Eng. 6, 045002 (2023); doi: 10.1063/10.0020885 6, 045002-6
© Author(s) 2023
05 October 2023 01:58:34
Nanotechnology and
Precision Engineering REVIEW pubs.aip.org/aip/npe
FIG. 3. Platforms for separating EVs based on size. (a) EV filtration system composed of ciliated microcolumn arrays that is capable of hierarchical filtration.38 Reproduced
from Wang et al., Lab Chip 13, 2879–2882 (2013). (b) Microfluidics of ZnO nanowires for separation of low-concentration EVs from urine.74 Reproduced from Yasui et al., Sci.
Adv. 3, e1701133 (2017). (c) Nanoscale micropillar array chips with gap sizes ranging from 25 to 235 nm.76 Reproduced from Wunsch et al., Nat. Nanotechnol. 11, 936–940
(2016). (d) Chip composed of 1024 parallel nano-DLD arrays.77 Reproduced from Smith et al., Lab Chip 18, 3913–3925 (2018).
low, posing specific challenges to EV extraction.73 To address this,
Yasui et al.74 used a device composed of zinc oxide nanowires and
herringbone substrates immobilized in a microchannel to efficiently
separate low-concentration EVs from 1 ml urine [Fig. 3(b)]. They
verified that urine EVs carried a negative charge, while nanowires
displayed a positive charge at pH 6–8, thus increasing EV cap-
ture efficiency. Consequently, within 40 min of in situ extraction
of miRNAs, in an improvement compared with UC and commer-
cially available kits, this platform was able to harvest 1000 types of
miRNAs.
Besides utilizing nanowires, researchers have also employed
carbon nanotubes for the separation of EVs. Yeh et al.75 reported a
microfluidic platform consisting of 3D carbon nanotube arrays that
separated differently sized EVs in a label-free and high-throughput
manner. These arrays were made up of 3D nitrogen-doped car-
bon nanotubes (CNxCNTs) that formed two regions with different
pitches, capturing larger (300 nm diameter) and smaller (80 nm
diameter) EVs. However, both types of EVs had low capture rates.
In summary, while separating EVs using nanowires or carbon nan-
otubes offers label-free and highly efficient separation benefits, the
manufacturing process can be complex, and issues with clogging also
pose a challenge for this technique.
4. Nanoscale deterministic lateral displacement
method (nano-DLD)
Deterministic lateral displacement (DLD) is a fluid dynamics-
based method through which particles of varying sizes are separated
according to differences in their trajectories between periodically
spaced micropillars.78 This method uses rectangular, circular, or tri-
angular arrays of micropillars distributed throughout the channel
of a microfluidic chip.78,79 Nano-DLD involves setting the distance
between the microcolumns at the nanoscale to achieve higher sepa-
ration accuracy and resolution. The critical cutoff diameter (critical
dimension) of DLD is determined by the distance between micropil-
lars and the displacement angle, which refers to the ratio of the
lateral offset of the micropillar array to the center-to-center distance
between two pillars. While particles smaller than the critical size
exhibit a zigzag pattern as they pass through the array of micropil-
lars, larger particles undergo lateral displacement due to the collision
mode.79 Nano-DLD facilitates the separation of particles that are
larger and smaller than the critical size along the direction of fluid
motion. The DLD technique has been employed to separate cir-
culating tumor cells (CTCs),80 blood cells,81 bacteria,82 and EVs.83
Santana et al.83 designed a DLD platform to study the separation
of EVs from the culture supernatant of cancer cells. Their experi-
Nano. Prec. Eng. 6, 045002 (2023); doi: 10.1063/10.0020885 6, 045002-7
© Author(s) 2023
05 October 2023 01:58:34
Nanotechnology and
Precision Engineering REVIEW pubs.aip.org/aip/npe
ments revealed that particles larger than the critical size (250 nm)
displayed lateral displacement, while those smaller than 250 nm
exhibited minimal lateral displacement. Although the efficiency of
particle recovery from this microfluidic platform was only about
39%, its purity was 98.5%. However, owing to diffusion, this non-
nanostructured device faced difficulties in accurately separating EVs
with a diameter less than 200 nm. To effectively separate EVs, nano-
structures featuring smaller gaps and finer structures or external
field-assisted separation measures are necessary. Zeming et al.84
exploited an electrostatic force in conjunction with adjustments to
the ionic strength of the solution. This allowed them to achieve the
separation of particles measuring 51 and 190 nm by modifying the
electrostatic force between the particles and microcolumns. Wun-
sch et al.76 developed a nano-DLD technique that exploited a passive
microfluidic action instead of relying on an external field. They cre-
ated nanoscale micropillar arrays on silicon substrates with gap sizes
ranging from 25 to 235 nm [Fig. 3(c)]. Their experiments revealed
that at low Péclet numbers Pe, where deterministic displacement and
diffusion compete, the resolution of the nano-DLD arrays was ade-
quate enough to separate particles sized between 20 and 110 nm.
Although the platform had the potential to separate biocolloids on a
chip, the flow rate remained limited to 0.1–0.2 nl/min. To address
the challenge of sample throughput, Smith et al.77 transformed a
single-array nano-DLD device into a parallel array comprising 1024
separate nano-DLD arrays integrated on a single chip [Fig. 3(d)].
Following calibration with fluorescent particles, intact EVs were
successfully separated from human serum and urine. The platform
demonstrated a recovery rate of 50%, surpassing conventional tech-
niques such as UC and SEC. Wunsch et al.85 developed i-nanoDLD,
an integrated nano-DLD device. They fabricated 31160 parallel
arrays at high density, resulting in a throughput of 17 ml/h, a first
for this technique. They then utilized the device to enrich and purify
EVs smaller than 200 nm from urine, significantly reducing pro-
tein content and other impurities. While the recovery rate of the
nano-DLD technique is high, its clinical applicability is limited by
the complexity of the manufacturing process, and crucial concerns
remain regarding equipment clogging and co-precipitation of EVs
and proteins.
B. Nanostructure with external fields
Integrating acoustic and electric fields into nanostructures pro-
vides higher-throughput separation techniques that do not harm
EVs (Table II). The acoustic field is comparable to some med-
ical ultrasound techniques and can efficiently separate biological
particles of interest.86 Electric fields typically exploit the electrical
properties of EVs. The zeta potential of EVs differs slightly depend-
ing on their origin, with EVs from plasma and those from breast
cancer cells (MCF-7) displaying zeta potentials of approximately
11 and 13.4 mV, respectively.87,88 These property differences
facilitate the separation of EVs.
1. Acoustic field
The combination of an acoustic field with nanotechnology
provides a label-free, reagent-free, and contact-free EV separa-
tion method. Acoustic field methods perform separation based
on differences in mechanical properties such as size, density, or
compressibility.89,90 This approach enhances cell and bioparticle
manipulation accuracy while maintaining biocompatibility.86 The
sample solution is subjected to acoustic streaming effects when
acoustic waves are applied, allowing the acoustic radiation force and
acoustic drag force to act on particles.91–93 Nanostructures such as
microtips, microcolumns, and filter membranes have been paired
with acoustic waves to achieve EV separation functionality. Chen
et al.94 developed an ultrafast EV separation device called EXO-
DUS by combining piezoelectric transducers (PZTs) with porous
alumina (AAO) membranes. This device utilized a negative-pressure
oscillation system and a double-coupled harmonic oscillation
TABLE II. Nanostructure-based EV separation method with external fields.
EV separation method Separated size (nm) Purity (%) Recovery yield (%) Throughput (μl/min) Reference
Acoustic field:
EXODUS 30–200 99 90 1000 94
Electrical field:
Pressure-driven filtration 150 NA >1.5 1 95
Electrophoretic separation
on nanoporous membrane 10–400 84 65 20 96
ExoSMP 30–120 90 94.2 10 103
Ion-depletion zone sorting 30–200 NA 98 1 104
Electric field-driven filtration <150 NA 60–80 150–200 97
Surface modified nanostructures:
Fe3O4magnetic nanoparticles coated with PEG 30–200 60 NA NA 108
ExoTENPO 150 NA NA 166 109
Wedge-shaped nanopore 30–400 NA 85% 83 110
Peptide-modified ZnO nanowires 80–160 NA 70 50 111
NanoVilli 30–300 NA 63–82 3112
Magnetic polypyrrole nanowire 40–150 NA NA 4113
Y-shaped micropillars 185 NA 83 50 114
Nano. Prec. Eng. 6, 045002 (2023); doi: 10.1063/10.0020885 6, 045002-8
© Author(s) 2023
05 October 2023 01:58:34
Nanotechnology and
Precision Engineering REVIEW pubs.aip.org/aip/npe
system to generate high- and low-frequency vibrations. EXODUS
not only effectively circumvented the clogging problem caused
by the accumulation of proteins or large vesicles on the mem-
brane surface, but also provided enhanced purification and separa-
tion efficiency. Moreover, compared with traditional EV separation
methods, this platform displayed significant improvements in pro-
cessing throughput (1 ml/min), purity (99%), and recovery (90%).
EVs were extracted from the urine of 113 patients displaying uri-
nary system diseases via RNA sequencing analysis using EXODUS,
demonstrating its high efficiency, effectiveness, and repeatability
[Fig. 4(a)].
2. Electric field
In a homogeneous electric field, only charged particles migrate
toward oppositely charged electrodes, leaving neutral and uncharged
particles unaffected by the electric field forces.98 When a particle
is in a nonuniform electric field, it becomes polarized, inducing
a directional movement under the dielectrophoretic (DEP) force,
irrespective of the particle’s charge state.99,100 Generally, the magni-
tude of the DEP force is affected by the volume, dielectric constant,
electrophoretic mobility, and electric field intensity of a particle.87,101
In EV separation methods based on the DEP effect, EVs are
also subjected to thermal effects caused by the electrodes. To avoid
direct contact occurring between electrodes and EVs, Ibsen et al.102
used 400 alternating current electrokinetic (ACE) electrodes on a
microfluidic platform. They overlayed the electrodes with a porous
hydrogel layer, enabling concentration of larger cells and EVs under
a low field strength and the effective recovery of separated EVs.
Nanostructures such as membranes can also effectively minimize
the contact between electrodes and EVs. Davies et al.95 devel-
oped a filtration system based on a pressure-driven tangential flow
filtration method along with a DC electrophoresis drive for effi-
cient separation of EVs from whole blood samples collected from
melanoma mice [Fig. 4(b)]. They synthesized a porous polymer
monomer (PPM) into membrane structures and integrated these
within PMMA microfluidic channels via in situ photolithography.
On application of a DC voltage, the separation of EVs based on their
distinct differences in electrophoretic mobility compared with pro-
teins increased both the purity and efficiency of the separated EVs
and enabled greater extraction of RNA. Cho et al.96 proposed an
electromigration-based EV separation technique [Fig. 4(c)]. They
employed an electrical field together with a dialysis membrane, fea-
turing a pore size of 30 nm, to extract EVs from diluted mouse
blood samples. pH and solution osmolarity buffers were employed
to prevent the potential problems caused by Joule heating and elec-
trochemical reaction products. The platform extracted 65% of the
EVs and excluded about 84% of the related proteins. Chen et al.103
developed a microfluidic platform (ExoSMP) based on a dual filter
device driven by an electrical field to separate EV subpopulations
via size-selectivity. Mogi et al.104 proposed a separation mechanism
based on electrophoresis and employing a microfluidic chip design,
utilizing a synthetic Nafion membrane sandwiched between two
channels. Cations were drawn into this membrane, ultimately gen-
erating an ion-depletion region that imposed an electric field force
FIG. 4. Nanostructures combined with external fields for EV separation. (a) Schematic of the EXODUS device.94 Reproduced from Chen et al., Nat. Methods 18, 212–218
(2021). (b) Platform for the separation of EVs by synthesizing porous membranes combined with electric field filtration in microfluidic channels.95 Reproduced from Davies
et al., Lab Chip 12, 5202–5210 (2012). (c) Electromigration-based EV separation chip.96 Reproduced from Cho et al., Sensors Actuators B Chem. 233, 289–297 (2016).
(d) Microfluidic platform combining ion-exchange membrane and gel electrophoresis.97 Reproduced from Marczak et al., Electrophoresis 39, 2029–2038 (2018).
Nano. Prec. Eng. 6, 045002 (2023); doi: 10.1063/10.0020885 6, 045002-9
© Author(s) 2023
05 October 2023 01:58:34
Nanotechnology and
Precision Engineering REVIEW pubs.aip.org/aip/npe
on EVs, resulting in their concentration within the main channel of
the sample batch. Marczak et al.97 developed a microfluidic platform
incorporating ion-exchange membranes and gel electrophoresis to
separate EVs [Fig. 4(d)]. In a combination of pressure- and electric
field-based approaches, agarose gel was used to filter out large-scale
material such as cell debris. A lateral local electric field of 100 V/cm
was applied to the sample flow within the microchannel, separating
EVs from other components. Ultimately, the enriched EVs perme-
ated the ion-exchange membranes and were separated within the
agarose gels.
3. Surface-modified nanostructures
Immunologically based separation techniques are extensively
utilized for EV separation because of their high specificity, selec-
tive character, and biocompatible compatibility.105,106 In contrast to
traditional immunological methods, this strategy can be applied on
a microfluidic platform. Specific antibodies targeting antigen fea-
tures on the EV surface can be immobilized on the surfaces of either
magnetic beads or microfluidic channels. Generally, nanostructures
amalgamated with magnetic nanomaterials demonstrate advantages
with regard to diagnostic applications, since they can function as
effective carriers or probes that capture intended EVs directly.107
Utilizing a chemical co-precipitation approach, Chang et al.108
synthesized PEG-coated Fe3O4magnetic nanoparticles with a size
of 20 nm that reduced the protein content (e.g., serum albumin and
immunoglobulins) of fetal bovine serum by 39.89%, without dam-
age to EV integrity [Fig. 5(a)]. Ko et al.109 designed the ExoTENPO
chip containing six series-connected membranes. A PC membrane
including a layer of magnetic material was adopted for extracting
nucleic acids, proteins, and cellular debris. By utilizing biotinylated
antibodies conjugated with streptavidin-coated magnetic nanoparti-
cles (MNPs), this innovative method successfully reduced antibody
amounts at a low cost. The formation of a magnetic trap by creating
a wedge-shaped nanopore edge rather than cylindrical nanopores
increased capture efficiency by generating a high nanoscale magnetic
field gradient. Zhang et al.110 deposited gold and NiFe succes-
sively on track-etched membranes, forming a structure capable of
detecting monopoles and dipoles and consequently elevating trap-
ping forces by 10 times, resulting in a 99% capture efficiency on
magnetic beads functionalized with anti-rabbit IgG antibody. The
device was able to separate EVs from human plasma mixed with
high-density lipoproteins (HDL) [Fig. 5(b)].
The abovementioned nanowire configurations can generally
be combined with immunoaffinity techniques to separate EVs.
Suwatthanarak et al.111 developed a microfluidic chip with bifunc-
FIG. 5. Platforms for surface-modified nanostructures to separate EVs. (a) EV purification method using PEG-coated Fe3O4MNPs.108 Reproduced from Chang et al., PLoS
One 13, e0199438 (2018). (b) Separation of EVs and HDL by wedge-shaped magnetic nanoporous membranes.110 Reproduced from Zhang et al., Commun. Biol. 5, 1358
(2022). (c) Microfluidic chip with bifunctional peptide-modified ZnO nanowires.111 Reproduced from Suwatthanarak et al., Lab Chip 21, 597–607 (2021). (d) Device with
Y-shaped micropillars and EV trapping through a gelatin layer and PS spheres.114 Reproduced from Zhou et al., Small 16, 2004492 (2020).
Nano. Prec. Eng. 6, 045002 (2023); doi: 10.1063/10.0020885 6, 045002-10
© Author(s) 2023
05 October 2023 01:58:34
Nanotechnology and
Precision Engineering REVIEW pubs.aip.org/aip/npe
tional peptide-modified ZnO nanowires, featuring EWI-2 amino
acid sequence-functionalized arrays constructed via the bond-
ing sequence P238. By using CD9 and CD81 markers, the chip
was able to separate breast cancer cell-secreted EVs [Fig. 5(c)].
Inspired by the microvilli on the surfaces of intestinal epithelial
cells, Dong et al.112 fabricated an array of nanowires on a silicon
substrate called NanoVilli. Anti-EpCAM (epithelial cell adhesion
molecule) modification intensified the binding force, leading to
reproducible separation of tumor cell-derived EVs from the serum
of non-small-cell lungs cancer patients. Cho and co-workers113
demonstrated a magnetic polypyrrole nanowire combining differ-
ent types of EV antibodies with high-density MNPs. Elongated and
biotinylated polypyrrole nanowires facilitated CD9-, CD63-, and
CD81-specific antibody binding and consequently EV separation.
Supplementing nanowires with magnetic materials or surface modi-
fication can specifically separate EVs effectively with reproducibility
and convenience. However, distinguishing EV subtypes remains
challenging.
Immobilizing EV-specific antibodies on the surface of
microfluidic devices is another promising approach. Zhou
et al.114 designed a microfluidic platform comprising Y-shaped
micropillars. Both a biotinylated gelatin layer and streptavidin-
modified polystyrene (PS) microspheres with a diameter of 50 nm
were integrated into the microchannel. Utilizing the Y-shaped
structure boosted the contact rate between EVs and CD63 antibod-
ies, while the modified 50 nm PS spheres enhanced the binding sites
for CD63 antibodies. Taking advantage of the thermoresponsive
properties of the gelatin membrane, Zhou et al.114 elevated the
chip temperature to 37C to release the captured EVs [Fig. 5(d)].
Although introducing antibodies improves the capture rate of EVs,
the high cost of antibodies limits their practical application, and the
complicated modification process is another drawback.
III. EV DETECTION METHODS BASED
ON NANOSTRUCTURES
Nanotechnology has enabled the development of novel
approaches to EV detection. Over the past few decades, numerous
methods have emerged that combine nanotechnology and charac-
terization of EVs based on their physical or biological properties.
Given the complexity of components carrying EVs, detection meth-
ods tend to involve multiple techniques working in tandem to
TABLE III. Exosome detection methods based on nanostructures.
EV separation method Limit of detection Throughput (μl/min) Sample Analyte type Reference
Immunofluorescence detection:
ExoAPP 160 EVs/μl NA Blood Protein 121
Peptide nucleic acid probes NA NA Cell culture medium RNA 122
MoS2multiwalled carbon nanotubes 1480 EVs/μl NA Blood Protein 123
Nano-IMEX 50 EVs/μl 0.5 Plasma Protein 124
3D herringbone nanopatterns 10 EVs/μl 0.5 Plasma Protein 125
exoNA sensor chip 58.3 fM0.8–1.3 Blood RNA 126
Integrative microfluidic device 1 EV/μl166 Urine Protein 127
Surface plasmon resonance
(SPR) detection:
Nanowall arrays 4.87 ×107EVs/cm2300 Cell culture medium Protein 129
nPLEX 3000 EVs 10 Ascites Protein 130
Cu-TCPP 2D MOF 16.7 EVs/ml 5 Serum Protein 132
APEX 200 EVs 3 Blood Protein 133
ExoSCOPE 1000 EVs 0.08 Blood Protein 134
Surface-enhanced Raman scattering
(SERS) detection:
Butterfly-shaped structure 30 EVs/μm2NA Cell culture medium Liposome 137
Multiplex SERS detection 1 EV/μl2 Serum Protein 138
3D plasmonic nanostructures <10 aMNA Urine RNA 139
Bi-functionalized SERS immunoassay 0.5 EVs/ml NA Serum Protein 140
MIO structure NA NA Plasma Protein 141
Electrochemical detection:
Aptasensor 1012 EVs/μl 10–400 Cell culture medium Protein 144
Electrochemical aptasensor 954 EVs/ml NA Plasma Protein 145
Aptamer strategy 70 EVs/μl NA Fetal bovine serum DNA 146
Integrating nano-interdigitated electrodes 5 EVs/μl NA Blood Protein 147
DeMEA chip 17 EVs/μl 0.5 Plasma Protein 148
Piezoelectric biosensor detection:
SAW sensor 1.1 EVs/μl 40 Blood Protein 150
QCM-D 1.4 ×105EVs/μl 10 Serum Protein 152
Nano. Prec. Eng. 6, 045002 (2023); doi: 10.1063/10.0020885 6, 045002-11
© Author(s) 2023
05 October 2023 01:58:34
Nanotechnology and
Precision Engineering REVIEW pubs.aip.org/aip/npe
characterize EVs. Recently, EV-based point-of-care testing (POCT)
techniques have been introduced, offering advanced cancer diag-
nosis, treatment monitoring, and assessment of prognosis. Lab-on-
chip systems have been devised that incorporate nanostructures for
EV separation, enrichment, purification, and detection. These sys-
tems include those based on detection using immunofluorescence,
surface plasmon resonance (SPR), surface-enhanced Raman scatter-
ing (SERS), and electrochemical methods (Table III). Additionally,
single-EV detection techniques have great potential for addressing
the question of EV heterogeneity.
A. Nanostructure-enabled immunofluorescence
detection
Fluorescence detection is a widely used EV detection tech-
nique in biochemical analysis and clinical diagnosis, boasting high
accuracy, sensitivity, and rapid response.115,116 It typically works in
tandem with highly specific immunoaffinity in detecting EV surface
proteins or constructing surface protein maps.117 EVs are frequently
labeled with nonspecific lipophilic membrane fluorescent dyes such
as PKH26. However, these dyes are not ideal markers for detect-
ing and analyzing surface proteins.118 Both antibodies and nucleic
acid aptamers can label EV surface proteins. Given the low cost and
excellent chemical stability of nucleic acid aptamers, they serve as
potential substitutes for traditional antibodies in high-throughput
protein analyses.119 Combining fluorescent immunoassays with
nanostructures facilitates the detection of EV subpopulations.
Graphene oxide (GO) is a 2D material with the capacity for
fluorescence quenching. The ππstacking interaction facilitates
solid affinity binding of single-stranded DNA (ss-DNA) to the
GO surface, and fluorescence quenching takes place by fluores-
cence resonance energy between conjugated dyes and GO.120 Jin
et al.121 employed GO as a nanointerface to develop ExoAPP, an
EV analysis system predicated on aptamer nanoprobes. Following
GO membrane quenching of specifically labeled FAM aptamers, tar-
get EVs bound these aptamers preferentially. The efficacy of both
EpCAM and prostate-specific membrane antigen (PSMA) in facil-
itating epithelial–mesenchymal transition detection was demon-
strated through identifying EVs coming from prostate cancer cells.
EpCAM and CD63 were also proven to be effective in differentiat-
ing tumor-derived EVs from benign tissues [Fig. 6(a)]. Oh et al.122
quenched FAM-labeled peptide nucleic acid probes to detect miR-
193 expression in EVs, enabling precise visualization of intercellular
EVs. Carbon nanotubes serve as another nanostructure. Tayebi
et al.123 employed MoS2multiwalled carbon nanotubes (MWC-
NTs) as a fluorescence quenching material to quench anti-human
CD63PE and produced a fast, sensitive, and quantifiable EV detec-
tion platform. Following the combination of CD63PE antibody with
EVs, the complex subsequently reacted with MWCNTs.
Zeng’s group124 devised an EV analysis platform that featured
modified nanointerfaces on microchannel surfaces. For the first
time, they integrated microfluidics by embedding polydopamine
(PDA) films onto GO, thereby significantly increasing the depo-
sition rate of PDA. A Y-shaped micropillar array along with the
GO/PDA nanointerface structure served to elevate the contact
area in the flow channel, thereby elevating EV capture efficiency
along with suppressing nonspecific adsorption. Utilizing this novel
nanostructure, captured EVs were labeled using biotinylated anti-
bodies, which subsequently underwent reaction with streptavidin-
conjugated β-galactosidase (SβG). The catalytic action of SβG on
2-β-D-galactopyranoside produces fluorescent signal amplification.
This assay is ultrasensitive and elegantly detected EVs at volumes
as low as 2 μl of plasma from women suffering from ovarian can-
cer. Building upon their previous work, the same research group
created an integrated self-assembled 3D nanopatterned microfluidic
chip with highly sensitive capabilities for detecting EV surface pro-
teins.125 In this new work, the authors replaced GO/PDA with a silica
colloid. Nanocolloids exhibit the capability of autonomous assembly
into 3D porous structures in microfluidic channels, while facilitat-
ing low hydrodynamic resistance to enable extensive EV capture and
detection. The authors proceeded to employ SβG again for enzy-
matic signal amplification to detect EVs. Finally, EV detection in
patient plasma determined that folate receptor αon EVs is a poten-
tial biomarker for early diagnosis and monitoring of ovarian cancer
patients [Fig. 6(b)].
Hydrogels serve as a biocompatible materials that can be
employed in the creation of nanostructures for detecting EVs.126
Lim et al.126 developed a microfluidic device, the exoNA sensor chip,
to detect EV mRNA. They engineered hydrogels equipped with 3D
nanostructures that enabled mRNA detection in EVs from HER2+
breast cancer mice blood within 2 h. The vacuum-driven exoNA
sensor chip enabled precise control over flow rate. Their 3D nanos-
tructured hydrogel contained two probes, ERBB2 and GAPDH,
which triggered enzyme-free amplification of the fluorescent signal
via in situ hybridization chain reaction at room temperature. Encap-
sulation of the probe in liposomes resulted in no fluorescent signal
being generated under static conditions and ensured stability and
response time during detection. Upon contact between liposomes
and EVs, membrane degradation ultimately amplified the generated
fluorescent signal. Comparison of mRNA expression levels between
diseased mice and control groups was carried out using their plat-
form, with outcomes that corresponded with those of qRT-PCR
analysis [Fig. 6(c)].
In another study,127 Liu’s group proposed an integrated EV
in situ detection device. They utilized ion sputtering to deposit gold
nanoparticles on an AAO membrane to produce 50 nm nanoporous
gold nanoclusters (AuNCs). Captured modified CD63 antibody
was then used to hybridize EVs on the AuNCs with a secondary
antibody-conjugated gold nanorod (AuR) probe. In a fluorescent
field, AuNCs scatter green light, whereas AuRs scatter red. How-
ever, when the gap between the two scattered wavelengths becomes
less than 200 nm, resonance coupling occurs between them, with
the creation of plasmons. This shift of the spectrum of scattered
light to yellow amplifies the scattering intensity. The device detected
EVs directly in urine samples (500 μl) obtained from lungs can-
cer patients and controls [Fig. 6(d)]. Dong et al.64 developed a
nanostructure based on a photonic crystal for EV detection. EVs
were captured by an AAO nanofiltration membrane, which was
then exposed to CD63-targeting aptamers (AptCD63). Some of the
aptamers bound to the EVs, while the remainder passed through
the membrane and were captured by a nitrocellulose membrane
endowed with a photonic crystal nanostructure that boosted the
fluorescent signal.
Nano. Prec. Eng. 6, 045002 (2023); doi: 10.1063/10.0020885 6, 045002-12
© Author(s) 2023
05 October 2023 01:58:34
Nanotechnology and
Precision Engineering REVIEW pubs.aip.org/aip/npe
FIG. 6. Platforms for immunofluorescence detection of EVs. (a) Aptamer nanoprobe-based EV analysis system (ExoAPP).121 Reproduced from Jin et al., Anal Chem 90,
14402–14411 (2018). (b) GO/PDA as a nanointerface EV analysis platform.125 Reproduced from Zhang et al., Nat. Biomed. Eng. 3, 438–451 (2019). (c) Schematic of exoNA
sensor chip.126 Reproduced from Lim et al., Biosens. Bioelectron. 197, 113753 (2022). (d) Integrated urine EV in situ detection device.127 Reproduced from Yang et al.,
Biosens. Bioelectron. 163, 112290 (2020).
B. Nanostructure-enabled surface plasmon
resonance (SPR) detection
Surface plasmon resonance (SPR) has emerged as an opti-
cal technique to detect molecular interactions at sensing interfaces
by tracking surface refractive index fluctuations.52 This technique
facilitates label-free and real-time monitoring of optical contrast
transformations caused by the adsorption of EVs on the plasma
layer. SPR represents a novel potential approach for detecting EV
subsets with increased sensitivity and specificity.128
SPR devices are ideal for EV detection owing to their sensitiv-
ity to contact responses that occur within 200 nm of the surface.
Zhu et al.129 developed nanowall array-based microfluidic chips
bound to specific antibodies to extract and purify EVs from cell
culture supernatants. Antibodies to CD9, glycoprotein CD41b, and
the tyrosine kinase receptor MET were adhered to gold-coated glass
chips and used to detect EVs through SPR imaging [Fig. 6(a)].
Weissleder’s team130 incorporated periodic gold nanopore arrays
(nPLEX) into microfluidic chips using a light interference lithogra-
phy technique to provide high-throughput and label-free EV detec-
tion through SPR. When functionalized with CD63 antibody, EVs
could be detected from the ascites of ovarian cancer patients on spe-
cific binding to the nPLEX sensor. Measurement of the wavelength
shift in the spectrum or of the intensity change at fixed wave-
lengths enabled observation of changes in the local refractive index
around pore complexes, giving a limit of detection of 3000 EVs.
This technique was faster and more sensitive and required a smaller
sample size when compared with the conventional gold standard
ELISA. Drawing on the results of their previous work, Weissleder’s
team131 also developed an automated detection and analysis system
for EVs derived from circulating tumor cells that could diagnose
pancreatic cancer via measurement of five protein markers. This
approach proved efficient, with 84% accuracy, 81% specificity, a
price of only $60, and a mere ten-minute measurement and analy-
sis time. Chen’s team132 utilized a hydrothermal synthesis approach
to create a 2D metal–organic framework (MOF), copper tetrakis(4-
carboxyphenylporphyrin) (Cu-TCPP), which they deposited on the
surface of a gold chip to provide improved optoelectronic properties.
This was then used to manufacture an SPR biosensor with substan-
tially enhanced detection accuracy, sensitivity, and quality factor. A
multifunctional peptide, comprising four structural domains, was
employed as a probe in an experiment to detect programmed death
ligand 1 (PD-L1) EVs [Fig. 6(b)].
Shao’s team133 created an amplified plasma exosome (APEX)
platform based on enzymatic deposition of insoluble optical pre-
cipitates and subsequent signal amplification through SPR for
detection of EVs. This platform included a silicon nitride sur-
face with a patterned array of gold nanoholes. EVs were cap-
tured on an array surface modified by CD63 antibody, displaying
400% spectral improvement in signal by binding amyloid (Aβ).
The detection of EV-bound Aβprovided an improved reflection
of PET imaging of plaques within the brain, enabling diagnosis
of Alzheimer’s disease in under an hour. Following this success,
the same group increased the structural complexity of the plat-
form by modifying the nanohole array into a nanoring array.134
Using molecular reactions within a plasmonic nanoring resonator
and bio-orthogonal probe amplification, in a technique they referred
to as ExoSCOPE, they were able to measure EVs in the blood
of a patient undergoing targeted cancer therapy, accurately clas-
sify the patient’s condition, and effectively determine treatment
outcomes within 24 h, representing an improvement compared
Nano. Prec. Eng. 6, 045002 (2023); doi: 10.1063/10.0020885 6, 045002-13
© Author(s) 2023
05 October 2023 01:58:34
Nanotechnology and
Precision Engineering REVIEW pubs.aip.org/aip/npe
with conventional pharmacokinetic/pharmacodynamic analysis
methods [Fig. 7(c)].
C. Nanostructure-enabled surface-enhanced Raman
scattering (SERS) detection
Raman spectroscopy is a molecular spectroscopy technique that
relies on the inelastic scattering of photons to identify biomolecules
such as nucleic acids, lipids, and proteins.135 Nevertheless, the
Raman signal is commonly faint owing to the minute inherent
Raman scattering sizes of the molecules involved and hindrance
by fluorescent signals in the target sample. By employing surface-
enhanced Raman scattering (SERS) techniques to analyze coarse
metal surfaces or structures, the Raman signal of the molecule
being measured can be intensified (103–1015-fold), thereby improv-
ing detection sensitivity.136 Consequently, an integrated microfluidic
chip based on SERS techniques has the advantages of high specificity,
multiplexing, and sensitivity.
For the detection of nanoscale EVs, Jalali et al.137 devised
a microfluidic platform with butterfly-shaped structure to differ-
entiate and characterize non-immune and label-free EVs through
SERS. In this approach, PS spheres were employed in bottom-
up self-assembly for the manufacture of plasmonic nano-bow-tie
configurations. This structure boosted optical and electrical prop-
erties while bolstering Raman signals. EV monolayer dispersion was
accomplished using the microfluidic system. A miniature portable
Raman device was combined with a microfluidic chip to distinguish
EVs from two different glioma cells and noncancerous glial cell EVs
[Fig. 8(a)]. Wang et al.138 reported an EPAC-II-based microfluidic
chip incorporating multiplex SERS detection. This platform iden-
tified 28 serum samples (5 μl per sample) in 65 min, with an
asymmetrical ring-shaped electrode formation enhancing the effi-
cacy and sensitivity of antibody capture of EVs. Enrichment of
EVs on the surface of a modified electrode through antibodies was
followed by multiplex SERS detection and analysis. EV variations
in sera from early-stage melanoma patients and healthy individ-
uals were analyzed, leading to detection of high expression levels
of melanoma-associated chondroitin sulfate proteoglycan (MCSP),
melanoma cell adhesion molecule (MCAM), CD61, and CD63 in
EVs in the former in comparison with the latter. Thus, on the basis of
these four selected biomarkers, early-stage melanoma patients were
successfully distinguished from healthy individuals [Fig. 8(b)].
miRNAs are commonly low in abundance in body fluids and
exhibit a high degree of sequence homology. Kim et al.139 developed
a SERS system utilizing 3D plasmonic nanostructures for detect-
ing miRNAs in urine. Gold nanopillars and self-assembled DNA
probe-conjugated gold nanoparticles (SAP-AuNPs) were employed
FIG. 7. Detection of EVs by surface plasmon resonance techniques. (a) Detection platform for EVs in cell culture supernatant combined with antibody microarray and
SPR.129 Reproduced from Zhu et al., Anal. Chem. 86, 8857–8864 (2014). (b) Schematic of SPR sensor based on 2D MOF structure.132 Reproduced from Wang et al.,
Biosens. Bioelectron. 201, 113954 (2022). (c) ExoSCOPE detection platform with nanoring structure.134 Reproduced from Pan et al., Nat. Nanotechnol. 16, 734–742 (2021).
Nano. Prec. Eng. 6, 045002 (2023); doi: 10.1063/10.0020885 6, 045002-14
© Author(s) 2023
05 October 2023 01:58:34
Nanotechnology and
Precision Engineering REVIEW pubs.aip.org/aip/npe
FIG. 8. Detection of EVs by surface-enhanced Raman scattering (SERS) techniques. (a) Microfluidic platform with butterfly-shaped structure.137 Reproduced from Jalali
et al., Lab Chip 21, 855–866 (2021). (b) EPAC-II platform combining microfluidic chip and multiplex SERS detection.138 Reproduced from Wang et al., Adv. Funct. Mater.
32, 2010296 (2022). (c) SERS system for detecting miRNA in urine.139 Reproduced from Kim et al., Biosens. Bioelectron. 205, 2010296 (2022). (d) SERS detection of EVs
captured in a microfluidic channel.140 Reproduced from Li et al., Chem. Sci. 9, 5372–5382 (2018).
to construct a 3D nano hierarchical structure that significantly
boosted the SERS signal of miRNAs present in urine. This hier-
archical plasmonic nanostructure served to extend the plasmonic
hotspot between SAP-AuNPs and nanopillars, thereby increasing
sensor sensitivity and enhancing clinical applicability. When the tar-
get miRNA was present, it was confined within a 3D plasmonic
hotspot formed between a nanopillar head and a AuNP bearing a
semi-complementary DNA probe. The sensor exhibited an exem-
plary detection performance, with a detection limit as low as 10 aM
and a linear detection range from 10 aMto 100 nMfor miRNAs
(miR-10a and miR-21) in EVs of prostate cancer, surpassing that of
qRT-PCR [Fig. 8(c)].
A traditional technique for EV detection involves modify-
ing specific antibodies in microchannels. Li et al.140 developed a
pioneering SERS nanotag for sensitive EV detection utilizing a
polydopamine-coated antibody–reporter–Ag(layer)–Au(core) mul-
tilayer structure. By injecting EVs derived from pancreatic cancer
into the flow channel along with antibody to the macrophage
migration inhibitory factor (MIF), a high-intensity SERS signal was
achieved. Remarkably, only 2 μl of clinical serum samples were
required to differentiate pancreatic cancer patients from healthy
individuals [Fig. 8(d)]. Inspired by honeycombs, Dong et al.141
developed a 3D gold-coated titania microporous inverse opal (MIO)
structure exhibiting remarkable EV-enhanced SERS signal perfor-
mance. The MIO could directly separate and detect EVs from cancer
patients’ plasma. This method has the potential for swift and label-
free diagnosis of cancer by tracking the protein phosphorylation
process in EVs.
D. Nanostructure-enabled electrochemical detection
Electrochemical detection analyzes changes in the electrical
signals of target EVs and has the advantages that it is highly sen-
sitive and low in cost, requiring only simple equipment and small
samples. Typically, electrochemical detection amplifies the electri-
cal signal of the targeted molecule through aptamer or antibody
binding.142,143 Zhou et al.144 developed an aptamer-based CD63-
specific EV electrochemical biosensor to quantitatively measure EVs
in cell supernatants. They integrated CD63 aptamers and modified
gold electrodes into a microfluidics chip and applied a methylene
blue label to the aptamers via DNA probing strands. After the CD63
aptamer was hybridized into a DNA duplex, displacement of the
antisense strand led to a weakened electrochemical signal with-
out the need for antibody labeling or fluorescent image acquisition
[Fig. 8(a)]. Li and co-workers145 proposed a label-free EV sensor
that utilized the characteristics of antibody and aptamer binding to
amplify the electrochemical signal. The sensor used CD63 antibody-
modified gold electrodes and specific aptamers for binding gastric
cancer EVs. The addition of a heme/G-quadruplex system with
rolling circle amplification (RCA) allowed for sensitive and selec-
Nano. Prec. Eng. 6, 045002 (2023); doi: 10.1063/10.0020885 6, 045002-15
© Author(s) 2023
05 October 2023 01:58:34
Nanotechnology and
Precision Engineering REVIEW pubs.aip.org/aip/npe
tive detection of EVs in gastric cancer. Exposure of gastric cancer
EVs to RCA stimulated abundant production of G-quadruplexes.
After incubation of the reaction product with heme, the result-
ing heme/G-quadruplex structure catalyzed the reduction of H2O2,
with the consequent generation of electrochemical signals. Dong
et al.146 utilized a similar strategy in which aptamer-modified mag-
netic beads were used to bind EVs from tumor cells. Messenger
DNA (mDNA) released by these EVs then hybridized with probe
DNA immobilized on gold electrodes, generating an electrochemical
signal.
The use of nanostructures for electrochemical detection of EVs
is a novel technique. Mathew et al.147 developed a microfluidic
electrochemical detection system that integrated nano-interdigitated
electrodes, utilizing both a redox cycle and enzymatic amplifica-
tion to augment electrochemical signals. The sensor comprised two
sets of interdigitated nanoelectrode arrays with nanometer-scale
spacing, which improved the signal amplification capability of the
redox cycle. The method used an anti-EpCAM antibody on the elec-
trode surface to capture EVs, together with a reporter anti-EpCAM
antibody conjugated to alkaline phosphatase (ALP) via biotin–SAV
interactions. Subsequently, captured EVs were selected sequentially
twice. Reaction of the ALP with a p-aminophenyl phosphate sub-
strate provided a first electrochemical signal amplification, which
was followed by a second amplification due to redox cycling of the
product of this reaction. This approach was successful in analyzing
different concentrations of prostate cancer cells [Fig. 9(b)]. Jung’s
group148 introduced a detachable aptamer-based electrochemical
sensor, known as the DeMEA chip. They prepared a composite of
chitosan, graphene nanosheets, and MoS2nanosheets. This com-
posite material enhanced electrode conductivity and improved the
EV capture rate owing to the increased surface area. The sensor was
modified with an EpCAM aptamer and utilized a minute sample size
of only 10 μl for detection purposes.
E. Nanostructure-enabled piezoelectric
biosensor detection
Advances in acoustic technology have provided opportuni-
ties for the development of novel detection strategies based on
piezoelectric biosensors and the fact that the frequencies of acous-
tic waves change in response to fluctuations in the propagation
medium. Acoustic techniques that generate acoustic pressure and
streaming effects in fluids have the advantage of good biocom-
patibility. Piezoelectric biosensors enable capture of extracellular
vesicles (EVs), as well as functioning as tools for processing EVs
(e.g., lysing EVs).149 Wang et al.150 developed a surface acoustic
wave (SAW) sensor based on the use of gold nanoparticles as signal
enhancers for the highly sensitive detection of EVs in blood samples
obtained from cancer patients. After pre-treating the sensor, they
modified its surface with an anti-CD63 antibody to immobilize EVs.
Streptavidin-modified AuNPs and biotin-conjugated anti-EpCAM
antibodies were then utilized to bind and recognize EVs. The SAW
phase shift caused by the introduction of the AuNPs amplified the
FIG. 9. Platforms for detecting EVs based on nanostructures. (a) Electrochemical detection platform with modified aptamers on the electrode surface.144 Reproduced from
Zhou et al., Methods 97, 88–93 (2016). (b) Electrochemical method for the detection of tumor-derived EVs using nano-interdigitated electrodes.147 Reproduced from Mathew
et al., Nano. Lett. 20, 820–828 (2020). (c) SAW sensor with gold nanoparticles as signal amplification materials.150 Reproduced from Wang et al., ACS Sensors 5, 362–369
(2020). (d) Acoustic immunodetection of EVs with QCM-D.152 Reproduced from Suthar et al., Anal. Chem. 92, 4082–4093 (2020).
Nano. Prec. Eng. 6, 045002 (2023); doi: 10.1063/10.0020885 6, 045002-16
© Author(s) 2023
05 October 2023 01:58:34
Nanotechnology and
Precision Engineering REVIEW pubs.aip.org/aip/npe
signal, giving this sensor an improved detection sensitivity (through
the mass effect). Testing demonstrated that this sensor was able to
detect EVs in blood samples within 30 min [Fig. 9(c)].
Conventional methods for EV nucleic acid examination suffer
from laborious operation, excessive time consumption, and limited
sample retrieval rates.151 Suthar et al.152 developed a bulk acoustic
wave (BAW) resonator-based sensor for EVs that employed a quartz
crystal microbalance with dissipation detection (QCM-D) for label-
free capture and analysis of CD63-positive EVs. An affinity-based
method was used to functionalize the sensor surface with anti-CD63
antibodies. Selective separation of EVs was achieved by tracking
changes in sensor resonance frequency (which depended on sur-
face antigen species, mass, and viscoelasticity) when analytes were
adsorbed onto the sensor surface. Following binding of EVs express-
ing CD63 to the sensor surface, the resonance frequency underwent
specific changes that correlated with the physical properties and
concentration of the EVs. By utilizing dissipation monitoring and
frequency, this technique enabled direct evaluation of EVs at their
native concentrations in complex biological samples [Fig. 9(d)].
F. Nanostructure-enabled EV analysis
at the level of individual particles
The majority of the available techniques mentioned above focus
on analyzing bulk EVs rather than individual particles. However,
EVs show pronounced heterogeneity in their sources, compositions,
and functions. Therefore, analyzing individual EVs is essential to
unveil their diversity and complexity. In-depth analysis of single
EVs can provide crucial insights into their biological roles, inter-
cellular communication, and potential as diagnostic and therapeutic
tools.153 Moreover, single-EV analysis is a more effective strategy
than bulk analysis in determining specific molecular and pheno-
typic features of cancer.154 Consequently, innovative methods that
enable precise and comprehensive single-EV analysis are in great
demand.
Research on single EVs also focuses on their proteins, nucleic
acids, and physical and chemical properties, similar to studies on
EV populations. However, examining single EVs presents techni-
cal challenges that differ from those encountered when dealing with
EV populations. The minute size of EVs (30–150 nm) and their
low abundance in body fluids necessitate specialized techniques for
single-EV detection and trapping. Additionally, the limited cargo
of a few molecules per marker poses a challenge for signal ampli-
fication and detection.155 Microfluidic techniques based on nano-
structures have demonstrated significant advantages in separating
and detecting single EVs. On the one hand, nanostructures can
provide tiny channels and pores with similar sizes to EVs, thus
enabling the separation and capture of a single EV. On the other
hand, nanostructures provide superior optical detection capabilities
for digital-based techniques.156
FIG. 10. Platforms for EV analysis at the level of individual particles based on nanostructures. (a) NPN membrane-based platform for investigation of EV properties in real
time.158 Reproduced from Riazanski et al., Commun. Biol. 5, 13 (2022). (b) Acoustic nanocavity trapping concept and result.160 Reproduced from Tayebi et al., Small 16,
2000462 (2020). (c) Digital single-EV counting detection (DECODE) chip to differentiate malignant and benign lungs nodules.162 Reproduced from Li et al., Adv. Sci. 10,
2000462 (2023).
Nano. Prec. Eng. 6, 045002 (2023); doi: 10.1063/10.0020885 6, 045002-17
© Author(s) 2023
05 October 2023 01:58:34
Nanotechnology and
Precision Engineering REVIEW pubs.aip.org/aip/npe
Capturing and separating single EVs is a prerequisite for their
analysis.157 Techniques based on nanostructures and microfluidics
can be divided into two categories: those based on physical prop-
erties and those based on immune affinity. Physical characteristic-
based techniques generally employ nanoscale structures directly.
Nelson’s group158 developed a novel approach to allow visualiza-
tion and ion-sensitive dye measurement of single EVs [Fig. 10(a)].
A filtration-based platform was used to capture and stabilize EVs
based on their size. After selective capture of secreted EVs within
the pores of ultrathin nanoporous silicon nitride (NPN) membranes
based on their size, real-time fluorescence imaging was used to mea-
sure the kinetics of pH changes in single vesicles brought about by
the activity of the Na/H antiporter. This study has provided a new
platform for studying possible mechanisms for cargo loading and
maintenance at the single-EV level. In addition to single-size nano-
structures, a combination of multiple-size nanostructures enabled
the preparation of subpopulations of EVs in three different size frac-
tions and an analysis of the presence of HER2 and PSMA in breast
and prostate cancer cell-derived EVs, respectively, at the single-
EV level.159 Along with passive fluid interaction, active techniques
have also been utilized to capture single EVs.160 Nanostructures
triggered by acoustic waves generate robust nanoscale force gra-
dients for the trapping of single EVs [Fig. 10(b)]. A structured
elastic layer placed between a microfluidic channel and a traveling
SAW device provided submicrometer acoustic traps capable of cap-
turing individual submicrometer particles. The acoustically driven
deformation of the nanocavities produced time-averaged acoustic
fields that directed suspended particles toward the nanocavities and
trapped them there. SAWs permit massively multiplexed particle
manipulation with deterministic patterning at the single-particle
level.Immune-affinity traps capture EVs by binding a selected mem-
brane protein to its corresponding antibody immobilized on a nano-
structure.161 The small size of the nanostructure and an appropriate
optical enhancement effect then enable a lower detection limit to be
reached.130 Trau and co-workers162 introduced the DECODE chip,
which captured EVs on a nanostructured pillar chip, confined indi-
vidual EVs, and used SERS to detect three lungs cancer-associated
biomarkers as well as a generic single-EV biomarker [Fig. 10(c)].
The DECODE chip was used to generate digitally acquired single-
EV molecular profiles in a cohort of 33 subjects including both those
with malignant and benign lungs nodules, as well as healthy indi-
viduals. DECODE was able to provide specific molecular profiles
for single EVs that enabled differentiation between malignant and
benign nodules with an area under the curve (AUC) of 0.85.
Single-EV analysis utilizing nanostructures presents several
advantages over traditional bulk analysis methods. By analyzing
EVs individually, researchers can obtain more detailed information
about their characteristics and better understand the relationship
between different subtypes of vesicular molecules and diseases. The
high-precision counting and improved detection limits provided by
nanostructure-based methods can also help detect disease biomark-
ers with greater accuracy, potentially leading to earlier and more
accurate diagnoses. Additionally, these methods may be useful for
monitoring disease progression and treatment response. While cur-
rent nanostructure-based methods may be cumbersome and have
low throughput, ongoing research is focused on developing more
efficient and scalable techniques. As these methods continue to
improve, they hold promise as valuable diagnostic tools for cancer
and other diseases.
IV. CONCLUSION AND OUTLOOK
As essential biomarkers in liquid biopsies, EVs have shown sig-
nificant potential in clinical diagnosis. Much work has been done
on the separation and detection of these complex nanoscale parti-
cles and the detailed biological information that they encapsulate.
This review commenced by outlining the constituents and biogene-
sis of EVs, and the traditional methods used for their separation and
detection. This was followed by a summary of the latest advances in
nanostructure-based procedures for EV separation and detection, as
well as their application in bioanalysis and disease diagnosis. Thanks
to advances in nanofabrication technology, novel techniques based
on nanostructures are able to handle complex bodily fluid samples
and provide extensive insights into the function and structure of
EVs. Nevertheless, these nanostructure-based techniques are still in
their early stages of development.
1. Owing to the inherently small sample volume of microfluidics
in nanostructures, the resulting separation throughput is very
low (of the order of just microliters per minute), which falls far
below the liter per minute requirement of clinical applications.
Nonetheless, even such low sample sizes and throughputs have
minimal impact on detection results. Furthermore, EVs in
samples (especially bodily fluids) are biologically heteroge-
neous, with a mixture of multiple subtypes of EVs. Thus, the
combined implementation of several separation techniques
will be needed to achieve high-resolution separation. How-
ever, separating EVs of pathological cells from those of normal
cells poses a challenge. Moreover, employing passive or active
methods that combine acoustic and electric fields during EV
separation inevitably results in EV damage. Amplified elec-
tric and acoustic fields can produce thermal effects, which
irreversibly affect the biological activity of EVs. Overcom-
ing these limitations will enable a deeper appreciation to be
achieved of the roles of different EV subsets in early-stage can-
cer and in disease prognosis, allowing for personalized therapy
development.
2. Regarding detection, conventional methods often necessitate
an all-inclusive EV characterization through NTA, western
blotting, SEM, and qRT-PCR. Contemporary studies require
greater EV purity. The integration of nanostructures and
optical detection significantly reduces the detection limits,
enabling more precise and sensitive EV detection. Different
research groups utilize different antibodies or aptamers for
cancer diagnoses, and hence have varying views on analy-
sis and characterization techniques. Thus, there is a pressing
demand to establish standardized EV detection processes.
3. Integrating nanostructures with microfluidic technology in
the biomedical domain assists in system miniaturization. The
use of lab-on-chip equipment, including sample pretreat-
ment, separation, and detection, efficiently reduces exces-
sive processing periods, exorbitant costs, and laborious clin-
ical sample management. Physicians have acknowledged the
advances that can be achieved through the use of EV-based
point-of-care (POC) devices. Also, notably, machine-learning
Nano. Prec. Eng. 6, 045002 (2023); doi: 10.1063/10.0020885 6, 045002-18
© Author(s) 2023
05 October 2023 01:58:34
Nanotechnology and
Precision Engineering REVIEW pubs.aip.org/aip/npe
algorithms and other artificial intelligence techniques can
provide novel approaches for high-dimensional EV data pro-
cessing and analysis. Such algorithms should enable enhanced
cancer detection, diagnosis, and classification, boosting the
potential for clinical application of multifunctional POC
devices.
ACKNOWLEDGMENTS
The authors gratefully acknowledge financial support
from the National Key R& D Program of China (Grant No.
2018YFE0118700), the National Natural Science Foundation of
China (NSFC Grant No. 62174119), the 111 Project (No. B07014),
and the Foundation for Talent Scientists of Nanchang Institute for
Micro-technology of Tianjin University.
AUTHOR DECLARATIONS
Conflict of Interest
The authors have no conflicts to disclose.
DATA AVAILABILITY
Data sharing is not applicable to this article as no new data were
created or analyzed in this study.
REFERENCES
1Lobb RJ, Becker M, Wen Wen S, et al. Optimized exosome isolation protocol for
cell culture supernatant and human plasma. J Extracell Vesicles 2015;4(1):27031.
https://doi.org/10.3402/jev.v4.27031.
2Gibbings DJ, Ciaudo C, Erhardt M, Voinnet O. Multivesicular bodies associate
with components of miRNA effector complexes and modulate miRNA activity.
Nat Cell Biol 2009;11(9):1143–1149. https://doi.org/10.1038/ncb1929.
3Yuan K, Shamskhou EA, Orcholski ME, et al. Loss of endothelium-derived
Wnt5a is associated with reduced pericyte recruitment and small vessel loss in pul-
monary arterial hypertension. Circulation 2019;139(14):1710–1724. https://doi.
org/10.1161/circulationaha.118.037642.
4György B, Módos K, Pállinger É, et al. Detection and isolation of cell-
derived microparticles are compromised by protein complexes resulting from
shared biophysical parameters. Blood 2011;117(4):e39–e48. https://doi.org/10.
1182/blood-2010-09-307595.
5Raposo G, Stoorvogel W. Extracellular vesicles: Exosomes, microvesicles, and
friends. J Cell Biol 2013;200(4):373–383. https://doi.org/10.1083/jcb.201211138.
6Properzi F, Logozzi M, Fais S. Exosomes: The future of biomarkers in medicine.
Biomarkers Med 2013;7(5):769–778. https://doi.org/10.2217/bmm.13.63.
7Chen G, Huang AC, Zhang W, et al. Exosomal PD-L1 contributes
to immunosuppression and is associated with anti-PD-1 response. Nature
2018;560(7718):382–386. https://doi.org/10.1038/s41586-018-0392-8.
8Gardiner C, D. D. Vizio, Sahoo S, et al. Techniques used for the isolation and
characterization of extracellular vesicles: Results of a worldwide survey. J Extracell
Vesicles 2016;5(1):32945. https://doi.org/10.3402/jev.v5.32945.
9Samuel M, Chisanga D, Liem M, et al. Bovine milk-derived exosomes from
colostrum are enriched with proteins implicated in immune response and growth.
Sci Rep 2017;7(1):5933. https://doi.org/10.1038/s41598-017-06288-8.
10Shirejini SZ, Inci F. The Yin and Yang of exosome isolation methods:
Conventional practice, microfluidics, and commercial kits. Biotechnol Adv
2022;54:107814. https://doi.org/10.1016/j.biotechadv.2021.107814.
11Jenjaroenpun P, Kremenska Y, Nair VM, Kremenskoy M, Joseph B, Kurochkin
IV. Characterization of RNA in exosomes secreted by human breast cancer
cell lines using next-generation sequencing. PeerJ 2013;1:e201. https://doi.org/10.
7717/peerj.201.
12Théry C, Zitvogel L, Amigorena S. Exosomes: Composition, biogenesis and
function. Nat Rev Immunol 2002;2(8):569–579. https://doi.org/10.1038/nri855.
13Nabet BY, Qiu Y, Shabason JE, et al. Exosome RNA unshielding couples
stromal activation to pattern recognition receptor signaling in cancer. Cell
2017;170(2):352–366.e13. https://doi.org/10.1016/j.cell.2017.06.031.
14Liang LG, Kong MQ, Zhou S, et al. An integrated double-filtration microfluidic
device for isolation, enrichment and quantification of urinary extracellular vesicles
for detection of bladder cancer. Sci Rep 2017;7(1):46224. https://doi.org/10.1038/
srep46224.
15Szczepanski MJ, Szajnik M, Welsh A, Whiteside TL, Boyiadzis M. Blast-
derived microvesicles in sera from patients with acute myeloid leukemia suppress
natural killer cell function via membrane-associated transforming growth factor-
β1. Haematologica 2011;96(9):1302–1309. https://doi.org/10.3324/haematol.2010.
039743.
16Whiteside TL. Tumor-derived exosomes and their role in cancer progression. in
Advances in Clinical Chemistry, edited by Makowski GS. Elsevier. 2016. Vol. 74.
pp. 103–141.
17Zhan Y, Du L, Wang L, et al. Expression signatures of exosomal long non-coding
RNAs in urine serve as novel non-invasive biomarkers for diagnosis and recur-
rence prediction of bladder cancer. Mol Cancer 2018;17(1):142. https://doi.org/
10.1186/s12943-018-0893-y.
18Li C, Lv Y, Shao C, et al. Tumor-derived exosomal lncRNA GAS5 as a
biomarker for early-stage non-small-cell lung cancer diagnosis. J Cell Physiol
2019;234(11):20721–20727. https://doi.org/10.1002/jcp.28678.
19Wang J, Ma P, Kim DH, Liu BF, Demirci U. Towards microfluidic-based exo-
some isolation and detection for tumor therapy. Nano Today 2021;37:101066.
https://doi.org/10.1016/j.nantod.2020.101066.
20Chiriacò MS, Bianco M, Nigro A, et al. Lab-on-Chip for exosomes and
microvesicles detection and characterization. Sensors 2018;18(10):3175. https://
doi.org/10.3390/s18103175.
21Dash S, Wu CC, Wu CC, et al. Extracellular vesicle membrane protein profiling
and targeted mass spectrometry unveil CD59 and tetraspanin 9 as novel plasma
biomarkers for detection of colorectal cancer. Cancers 2022;15(1):177. https://doi.
org/10.3390/cancers15010177.
22Li P, Xu Z, Liu T, et al. Circular RNA sequencing reveals serum exosome circular
RNA panel for high-grade astrocytoma diagnosis. Clin Chem 2022;68(2):332–343.
https://doi.org/10.1093/clinchem/hvab254.
23Liang G, Zhu Y, Ali DJ, et al. Engineered exosomes for targeted co-delivery
of miR-21 inhibitor and chemotherapeutics to reverse drug resistance in colon
cancer. J Nanobiotechnol 2020;18(1):10. https://doi.org/10.1186/
s12951-019-0563-2.
24Hassanpour Tamrin S, Sanati Nezhad A, Sen A. Label-free isolation of exosomes
using microfluidic technologies. ACS Nano 2021;15(11):17047–17079. https://doi.
org/10.1021/acsnano.1c03469.
25Li P, Kaslan M, Lee SH, Yao J, Gao Z. Progress in exosome isolation techniques.
Theranostics 2017;7(3):789–804. https://doi.org/10.7150/thno.18133.
26Patel GK, Khan MA, Zubair H, et al. Comparative analysis of exosome
isolation methods using culture supernatant for optimum yield, purity
and downstream applications. Sci Rep 2019;9(1):5335. https://doi.org/10.1038/
s41598-019-41800-2.
27Gandham S, Su X, Wood J, et al. Technologies and standardization in research
on extracellular vesicles. Trends Biotechnol 2020;38(10):1066–1098. https://doi.
org/10.1016/j.tibtech.2020.05.012.
28Fekete S, Beck A, Veuthey JL, Guillarme D. Theory and practice of size exclu-
sion chromatography for the analysis of protein aggregates. J Pharm Biomed Anal
2014;101:161–173. https://doi.org/10.1016/j.jpba.2014.04.011.
29Blans K, Hansen MS, Sørensen LV, et al. Pellet-free isolation of human and
bovine milk extracellular vesicles by size-exclusion chromatography. J Extracell
Vesicles 2017;6(1):1294340. https://doi.org/10.1080/20013078.2017.1294340.
30Sódar BW, Kittel Á, Pálóczi K, et al. Low-density lipoprotein mimics blood
plasma-derived exosomes and microvesicles during isolation and detection. Sci
Rep 2016;6(1):24316. https://doi.org/10.1038/srep24316.
31Tengattini S. Chromatographic approaches for purification and analytical char-
acterization of extracellular vesicles: Recent advancements. Chromatographia
2019;82(1):415–424. https://doi.org/10.1007/s10337-018-3637-7.
Nano. Prec. Eng. 6, 045002 (2023); doi: 10.1063/10.0020885 6, 045002-19
© Author(s) 2023
05 October 2023 01:58:34
Nanotechnology and
Precision Engineering REVIEW pubs.aip.org/aip/npe
32Yamamoto KR, Alberts BM, Benzinger R, Lawhorne L, Treiber G. Rapid bacte-
riophage sedimentation in the presence of polyethylene glycol and its application
to large-scale virus purification. Virology 1970;40(3):734–744. https://doi.org/10.
1016/0042-6822(70)90218-7.
33Kim D, Nishida H, An SY, Shetty AK, Bartosh TJ, Prockop DJ. Chromato-
graphically isolated CD63+CD81+extracellular vesicles from mesenchymal
stromal cells rescue cognitive impairments after TBI. Proc Natl Acad Sci USA
2016;113(1):170–175. https://doi.org/10.1073/pnas.1522297113.
34Zeringer E, Barta T, Li M, Vlassov AV. Strategies for isolation of exosomes. Cold
Spring Harb Protoc;2015(4):319. https://doi.org/10.1101/pdb.top074476.
35Batrakova E V, Kim MS. Using exosomes, naturally-equipped nanocarriers, for
drug delivery. J Controlled Release 2015;219:396–405. https://doi.org/10.1016/j.
jconrel.2015.07.030.
36Zarovni N, Corrado A, Guazzi P, et al. Integrated isolation and quantitative
analysis of exosome shuttled proteins and nucleic acids using immunocap-
ture approaches. Methods 2015;87:46–58. https://doi.org/10.1016/j.ymeth.2015.
05.028.
37Gurunathan S, Kang MH, Jeyaraj M, Qasim M, Kim JH. Review of the isolation,
characterization, biological function, and multifarious therapeutic approaches of
exosomes. Cells 2019;8(4):307. https://doi.org/10.3390/cells8040307.
38Wang Z, Wu H, Fine D, et al. Ciliated micropillars for the microfluidic-based
isolation of nanoscale lipid vesicles. Lab Chip 2013;13(15):2879–2882. https://doi.
org/10.1039/c3lc41343h.
39Alvarez ML, Khosroheidari M, Kanchi Ravi R, DiStefano JK. Comparison
of protein, microRNA, and mRNA yields using different methods of urinary
exosome isolation for the discovery of kidney disease biomarkers. Kidney Int
2012;82(9):1024–1032. https://doi.org/10.1038/ki.2012.256.
40Greening DW, Xu R, Ji H, Tauro BJ, Simpson RJ. A protocol for exosome isola-
tion and characterization: Evaluation of ultracentrifugation, density-gradient sep-
aration, and immunoaffinity capture methods. in Proteomic Profiling: Methods
and Protocols, edited by Posch A. Springer. New York. 2015. pp. 179–209.
41Vergauwen G, Dhondt B, Van Deun J, et al. Confounding factors of ultrafiltra-
tion and protein analysis in extracellular vesicle research. Sci Rep 2017;7(1):2704.
https://doi.org/10.1038/s41598-017-02599-y.
42Koliha N, Wiencek Y, Heider U, et al. A novel multiplex bead-based
platform highlights the diversity of extracellular vesicles. J Extracell Vesicles
2016;5(1):29975. https://doi.org/10.3402/jev.v5.29975.
43Li F, You M, Li S, et al. Paper-based point-of-care immunoassays: Recent
advances and emerging trends. Biotechnol Adv 2020;39:107442. https://doi.org/
10.1016/j.biotechadv.2019.107442.
44Li B, Zang G, Zhong W, et al. Activation of CD137 signaling promotes
neointimal formation by attenuating TET2 and transferrring from endothelial
cell-derived exosomes to vascular smooth muscle cells. Biomed Pharmacother
2020;121:109593. https://doi.org/10.1016/j.biopha.2019.109593.
45Cutler P. Immunoaffinity chromatography. in Protein Purification Protocols,
edited by Cutler P. Humana Press. 2004. pp. 167–177.
46Tauro BJ, Greening DW, Mathias RA, et al. Comparison of ultracentrifu-
gation, density gradient separation, and immunoaffinity capture methods for
isolating human colon cancer cell line LIM1863-derived exosomes. Methods
2012;56(2):293–304. https://doi.org/10.1016/j.ymeth.2012.01.002.
47Soo CY, Song Y, Zheng Y, et al. Nanoparticle tracking analysis moni-
tors microvesicle and exosome secretion from immune cells. Immunology
2012;136(2):192–197. https://doi.org/10.1111/j.1365-2567.2012.03569.x.
48Hoo CM, Starostin N, West P, Mecartney ML. A comparison of atomic force
microscopy (AFM) and dynamic light scattering (DLS) methods to characterize
nanoparticle size distributions. J Nanopart Res 2008;10(S1):89–96. https://doi.org/
10.1007/s11051-008-9435-7.
49Erdbrügger U, Lannigan J. Analytical challenges of extracellular vesicle detec-
tion: A comparison of different techniques. Cytometry Part A 2016;89(2):123–134.
https://doi.org/10.1002/cyto.a.22795.
50Khan S, Bennit HF, Turay D, et al. Early diagnostic value of survivin and its
alternative splice variants in breast cancer. BMC Cancer 2014;14(1):176. https://
doi.org/10.1186/1471-2407-14-176.
51Fan Z, Yu J, Lin J, Liu Y, Liao Y. Exosome-specific tumor diagnosis via
biomedical analysis of exosome-containing microRNA biomarkers. Analyst
2019;144(19):5856–5865. https://doi.org/10.1039/c9an00777f.
52Sina AAI, Vaidyanathan R, Wuethrich A, Carrascosa LG, Trau M. Label-free
detection of exosomes using a surface plasmon resonance biosensor. Anal Bioanal
Chem 2019;411(7):1311–1318. https://doi.org/10.1007/s00216-019-01608-5.
53Vogel R, Willmott G, Kozak D, et al. Quantitative sizing of nano/microparticles
with a tunable elastomeric pore sensor. Anal Chem 2011;83(9):3499–3506.
https://doi.org/10.1021/ac200195n.
54Dey S, Trau M, Koo KM. Surface-enhanced Raman spectroscopy for cancer
immunotherapy applications: Opportunities, challenges, and current progress in
nanomaterial strategies. Nanomaterials 2020;10(6):1145. https://doi.org/10.3390/
nano10061145.
55Tzaridis T, Bachurski D, Liu S, et al. Extracellular vesicle separation tech-
niques impact results from human blood samples: Considerations for diag-
nostic applications. Int J Mol Sci 2021;22(17):9211. https://doi.org/10.3390/
ijms22179211.
56Mahgoub EO, Razmara E, Bitaraf A, et al. Advances of exosome isolation tech-
niques in lung cancer. Mol Biol Rep 2020;47(9):7229–7251. https://doi.org/10.
1007/s11033-020-05715-w.
57Yang D, Zhang W, Zhang H, et al. Progress, opportunity, and perspective on
exosome isolation - efforts for efficient exosome-based theranostics. Theranostics
2020;10(8):3684–3707. https://doi.org/10.7150/thno.41580.
58Le MCN, Fan HZ. Exosome isolation using nanostructures and microflu-
idic devices. Biomed Mater 2021;16(2):22005. https://doi.org/10.1088/1748-605x/
abde70.
59Talebjedi B, Tasnim N, Hoorfar M, Mastromonaco GF, De Almeida Monteiro
Melo Ferraz M. Exploiting microfluidics for extracellular vesicle isolation and
characterization: Potential use for standardized embryo quality assessment. Front
Vet Sci 2021;7:620809. https://doi.org/10.3389/fvets.2020.620809.
60Iliescu FS, Vrtaˇ
cnik D, Neuzil P, Iliescu C. Microfluidic technology for clini-
cal applications of exosomes. Micromachines 2019;10(6):392. https://doi.org/10.
3390/mi10060392.
61Voldman J. Electrical forces for microscale cell manipulation. Annu Rev Biomed
Eng 2006;8(1):425–454. https://doi.org/10.1146/annurev.bioeng.8.061505.095739.
62Kirschner AY, Cheng YH, Paul DR, Field RW, Freeman BD. Fouling mech-
anisms in constant flux crossflow ultrafiltration. J Membr Sci 2019;574:65–75.
https://doi.org/10.1016/j.memsci.2018.12.001.
63Liu F, Vermesh O, Mani V, et al. The exosome total isolation chip. ACS Nano
2017;11(11):10712–10723. https://doi.org/10.1021/acsnano.7b04878.
64Dong X, Chi J, Zheng L, et al. Efficient isolation and sensitive quantification of
extracellular vesicles based on an integrated ExoID-Chip using photonic crystals.
Lab Chip 2019;19(17):2897–2904. https://doi.org/10.1039/c9lc00445a.
65Seder I, Moon H, Kang SJ, Shin S, Rhee WJ, Kim SJ. Size-selective
filtration of extracellular vesicles with a movable-layer device. Lab Chip
2022;22(19):3699–3707. https://doi.org/10.1039/d2lc00441ks.
66Díaz-Reinoso B. Concentration and purification of seaweed extracts using
membrane technologies. in Sustainable Seaweed Technologies, edited by Torres
MD, Kraan S, Dominguez H. Elsevier. 2020. pp. 371–390.
67Woo HK, Park J, Ku JY, et al. Urine-based liquid biopsy: Non-invasive and sen-
sitive AR-V7 detection in urinary EVs from patients with prostate cancer. Lab
Chip 2019;19(1):87–97. https://doi.org/10.1039/c8lc01185k.
68Woo HK, Sunkara V, Park J, et al. Exodisc for rapid, size-selective, and efficient
isolation and analysis of nanoscale extracellular vesicles from biological samples.
ACS Nano 2017;11(2):1360–1370. https://doi.org/10.1021/acsnano.6b06131.
69Sunkara V, Kim CJ, Park J, et al. Fully automated, label-free isolation of extracel-
lular vesicles from whole blood for cancer diagnosis and monitoring. Theranostics
2019;9(7):1851–1863. https://doi.org/10.7150/thno.32438.
70Han Z, Peng C, Yi J, et al. Highly efficient exosome purification from human
plasma by tangential flow filtration based microfluidic chip. Sens Actuators B
2021;333:129563. https://doi.org/10.1016/j.snb.2021.129563.
71Dehghani M, Lucas K, Flax J, McGrath J, Gaborski T. Tangential flow microflu-
idics for the capture and release of nanoparticles and extracellular vesicles on
conventional and ultrathin membranes. Adv Mater Technol 2019;4(11):1900539.
https://doi.org/10.1002/admt.201900539.
72Rahong S, Yasui T, Yanagida T, et al. Three-dimensional nanowire struc-
tures for ultra-fast separation of DNA, protein and RNA molecules. Sci Rep
2015;5(1):10584. https://doi.org/10.1038/srep10584.
Nano. Prec. Eng. 6, 045002 (2023); doi: 10.1063/10.0020885 6, 045002-20
© Author(s) 2023
05 October 2023 01:58:34
Nanotechnology and
Precision Engineering REVIEW pubs.aip.org/aip/npe
73Barutta F, Tricarico M, Corbelli A, et al. Urinary exosomal MicroRNAs in incip-
ient diabetic nephropathy. PLoSOne 2013;8(11):e73798. https://doi.org/10.1371/
journal.pone.0073798.
74Yasui T, Yanagida T, Ito S, et al. Unveiling massive numbers of cancer-related
urinary-microRNA candidates via nanowires. Sci Adv 2017;3(12):e1701133.
https://doi.org/10.1126/sciadv.1701133.
75Yeh YT, Zhou Y, Zou D, et al. Rapid size-based isolation of extracellular vesi-
cles by three-dimensional carbon nanotube arrays. ACS Appl Mater Interfaces
2020;12(11):13134–13139. https://doi.org/10.1021/acsami.9b20990.
76Wunsch BH, Smith JT, Gifford SM, et al. Nanoscale lateral displacement arrays
for the separation of exosomes and colloids down to 20 nm. Nat Nanotechnol
2016;11(11):936–940. https://doi.org/10.1038/nnano.2016.134.
77Smith JT, Wunsch BH, Dogra N, et al. Integrated nanoscale deterministic lateral
displacement arrays for separation of extracellular vesicles from clinically-relevant
volumes of biological samples. Lab Chip 2018;18(24):3913–3925. https://doi.org/
10.1039/c8lc01017j.
78Huang LR, Cox EC, Austin RH, Sturm JC. Continuous particle separa-
tion through deterministic lateral displacement. Science 2004;304(5673):987–990.
https://doi.org/10.1126/science.1094567.
79Inglis DW, Davis JA, Austin RH, Sturm JC. Critical particle size for frac-
tionation by deterministic lateral displacement. Lab Chip 2006;6(5):655–658.
https://doi.org/10.1039/b515371a.
80Liu Z, Huang F, Du J, et al. Rapid isolation of cancer cells using microfluidic
deterministic lateral displacement structure. Biomicrofluidics 2013;7(1):11801.
https://doi.org/10.1063/1.4774308.
81Davis JA, Inglis DW, Morton KJ, et al. Deterministic hydrodynamics: Taking
blood apart. Proc Natl Acad Sci USA 2006;103(40):14779–14784. https://doi.org/
10.1073/pnas.0605967103.
82Beech JP, Ho BD, Garriss G, Oliveira V, Henriques-Normark B, Tegen-
feldt JO. Separation of pathogenic bacteria by chain length. Anal Chim Acta
2018;1000:223–231. https://doi.org/10.1016/j.aca.2017.11.050.
83Santana SM, Antonyak MA, Cerione RA, Kirby BJ. Microfluidic isolation
of cancer-cell-derived microvesicles from hetergeneous extracellular shed vesi-
cle populations. Biomed Microdev 2014;16(6):869–877. https://doi.org/10.1007/
s10544-014-9891-z.
84Zeming KK, Thakor N V, Zhang Y, Chen CH. Real-time modulated nanopar-
ticle separation with an ultra-large dynamic range. Lab Chip 2016;16(1):75–85.
https://doi.org/10.1039/c5lc01051a.
85Wunsch BH, Hsieh KY, Kim SC, et al. Advancements in throughput, lifetime,
purification, and workflow for integrated nanoscale deterministic lateral dis-
placement. Adv Mater Technol 2021;6(4):2001083. https://doi.org/10.1002/admt.
202001083.
86Wang Z, Li F, Rufo J, et al. Acoustofluidic salivary exosome isolation. J Mol
Diagn 2020;22(1):50–59. https://doi.org/10.1016/j.jmoldx.2019.08.004.
87Shi L, Rana A, Esfandiari L. A low voltage nanopipette dielectrophoretic
device for rapid entrapment of nanoparticles and exosomes extracted from
plasma of healthy donors. Sci Rep 2018;8(1):6751. https://doi.org/10.1038/
s41598-018-25026-2.
88Ayala-Mar S, Perez-Gonzalez VH, Mata-Gómez MA, Gallo-Villanueva RC,
González-Valdez J. Electrokinetically driven exosome separation and concentra-
tion using dielectrophoretic-enhanced PDMS-based microfluidics. Anal Chem
2019;91(23):14975–14982. https://doi.org/10.1021/acs.analchem.9b03448.
89Lenshof A, Magnusson C, Laurell T. Acoustofluidics 8: Applica-
tions of acoustophoresis in continuous flow microsystems. Lab Chip
2012;12(7):1210–1223. https://doi.org/10.1039/c2lc21256k.
90Zhang P, Bachman H, Ozcelik A, Huang TJ. Acoustic microflu-
idics. Annu Rev Anal Chem 2020;13(1):17–43. https://doi.org/10.1146/
annurev-anchem-090919-102205.
91Wu H, Tang Z, You R, et al. Manipulations of micro/nanoparticles using giga-
hertz acoustic streaming tweezers. Nanotechnol Precis Eng 2022;5(2):023001.
https://doi.org/10.1063/10.0009954.
92Wei W, Wang Y, Wang Z, Duan X. Microscale acoustic streaming for biomed-
ical and bioanalytical applications. TrAC Trends Anal Chem 2023;160:116958.
https://doi.org/10.1016/j.trac.2023.116958.
93Wu M, Mao Z, Chen K, et al. Acoustic separation of nanoparticles in contin-
uous flow. Adv Funct Mater 2017;27(14):1606039. https://doi.org/10.1002/adfm.
201606039.
94Chen Y, Zhu Q, Cheng L, et al. Exosome detection via the ultrafast-isolation
system: EXODUS. Nat Methods 2021;18(2):212–218. https://doi.org/10.1038/