Abstract: Cell-free RNAs and extracellular vesicles (EVs) are valuable
biomarkers
in liquid biopsies, but they are prone to preanalytical variabilities
such as nonstandardized centrifugation or <i>ex vivo</i> blood degradation. Herein, we report a high-throughput and label-free
inertial microfluidic device (ExoArc) for isolation of platelet-free
plasma from blood for RNA and EV analysis. Unlike conventional inertial
microfluidic devices widely used for cell sorting, a submicrometer
size cutoff (500 nm) was achieved which completely removed all leukocytes,
RBCs, platelets, and cellular debris based on differential lateral
migration induced by Dean vortices. The single-step operation also
reduced platelet-associated miRNAs (∼2-fold) compared to centrifugation.
We clinically validated ExoArc for plasma miRNA profiling (39 samples)
and identified a 7-miRNA panel that detects non-small cell lung cancer
with ∼90% sensitivity. ExoArc was also coupled with size exclusion
chromatography (SEC) to isolate EVs within 50 min with ∼10-fold
higher yield than ultracentrifugation. As a proof-of-concept for EV-based
transcriptomics analysis, we performed miRNA analysis in healthy and
type 2 diabetes mellitus (T2DM) subjects (<i>n</i> = 3 per
group) by coupling ExoArc and ExoArc+SEC with quantitative polymerase
chain reaction (RT-qPCR) assay. Among 293 miRNAs detected, plasmas
and EVs showed distinct differentially expressed miRNAs in T2DM subjects.
We further demonstrated automated in-line EV sorting from low volume
culture media for continuous EV monitoring. Overall, the developed
ExoArc offers a convenient centrifugation-free workflow to automate
plasma and EV isolation for point-of-care diagnostics and quality
control in EV manufacturing.
Abstract: Monodisperse microbubbles with diameters less than 10
μm
are desirable in several ultrasound imaging and therapeutic delivery
applications. However, conventional approaches to synthesize microbubbles,
which are usually agitation-based, produce polydisperse bubbles that
are less desirable because of their heterogeneous response when exposed
to an ultrasound field. Microfluidics technology has the unique advantage
of generating size-controlled monodisperse microbubbles, and it is
now well established that the diameter of microfluidically made microbubbles
can be tuned by varying the liquid flow rate, gas pressure, and dimensions
of the microfluidic channel. It is also observed that once the microbubbles
form, the bubbles shrink and eventually stabilize to a quasi-equilibrium
diameter, and that the rate of stabilization is related to the lipid
solution. However, how the lipid solution concentration affects the
degree of bubble shrinkage, and the stable size of microbubbles, has
not been thoroughly examined. Here, we investigate whether and how
the lipid concentration affects the degree of microbubble shrinkage.
Namely, we utilize a flow-focusing microfluidic geometry to generate
monodisperse bubbles, and observe the effect of gas composition (2.5,
1.42, and 0.17 wt % octafluoropropane in nitrogen) and lipid concentration
(1–16 mg/mL) on the degree of microbubble shrinkage. For the
lipid system and gas utilized in these experiments, we observe a monotonic
increase in the degree of microbubble shrinkage with decreasing lipid
concentration, and no dependency on the gas composition. We hypothesize
that the degree of shrinkage is related to lipid concentration by
the self-assembly of lipids on the gas–liquid interface during
bubble generation and subsequent lipid packing on the interface during
shrinkage, which is arrested when a maximum packing density is achieved.
We anticipate that this approach for creating and tuning the size
of monodisperse microbubbles will find utility in biomedical applications,
such as contrast-enhanced ultrasound imaging and ultrasound-triggered
gene delivery.