Subcompartmentalized Nanoreactors as Artificial Organelle with Intracellular Activity

Article abstract of DOI:10.1002/smll.201502109

Cell mimicry is an approach which aims at substituting missing or lost activity. In this context, the goal of artificial organelles is to provide intracellularly active nanoreactors to affect the cellular performance. So far, only a handful of reports discuss concepts addressing this challenge based on single-component reactors. Here, the assembly of nanoreactors equipped with glucose oxidase (GOx)-loaded liposomal subunits coated with a poly(dopamine) polymer layer and RGD targeting units is reported. When comparing different surface modifications, the uptake of the nanoreactors by endothelial cells and macrophages with applied shear stress is confirmed without inherent cytotoxicity. Furthermore, the encapsulation and preserved activity of GOx within the nanoreactors is shown. The intracellular activity is demonstrated by exposing macrophages with internalized nanoreactors to glucose and assessment of the cell viability after 6 and 24 h. The macrophage viability is found to be reduced due to the intracellularly produced hydrogen peroxide by GOx. This report on the first intracellular active subcompartmentalized nanoreactors is a considerable step in therapeutic cell mimicry.

Full text of DOI:10.1002/smll.201502109

full papers
www.MaterialsViews.com
Biomimicry
Subcompartmentalized Nanoreactors as Artificial
Organelle with Intracellular Activity
Bo Thingholm, Philipp Schattling, Yan Zhang, and Brigitte Städler*
C ell mimicry is an approach which aims at substituting missing or lost activity. In this
context, the goal of artificial organelles is to provide intracellularly active nanoreactors
to affect the cellular performance. So far, only a handful of reports discuss concepts
addressing this challenge based on single-component reactors. Here, the assembly
of nanoreactors equipped with glucose oxidase (GOx)-loaded liposomal subunits
coated with a poly(dopamine) polymer layer and RGD targeting units is reported.
When comparing different surface modifications, the uptake of the nanoreactors by
endothelial cells and macrophages with applied shear stress is confirmed without
inherent cytotoxicity. Furthermore, the encapsulation and preserved activity of
GOx within the nanoreactors is shown. The intracellular activity is demonstrated by
exposing macrophages with internalized nanoreactors to glucose and assessment of
the cell viability after 6 and 24 h. The macrophage viability is found to be reduced
due to the intracellularly produced hydrogen peroxide by GOx. This report on the
first intracellular active subcompartmentalized nanoreactors is a considerable step in
therapeutic cell mimicry.
1
. Introduction
to synthesize, degrade, or convert a specific compound in
the required product. Since many diseases are related to the
Nature has perfected the way our cells operate and function malfunctioning or missing of a specific enzyme, TCM might
in our body. Multiple reactions happen in parallel, in a cas- have the potential to provide a platform to address a variety
cade manner, executed by highly specialized subunits under of medical conditions. Enzyme replacement therapy as such
[
2]
stringent control. The successful interplay between the nuclei is a well-known concept, which faces a lot of challenges
as the carrier of the genetic information, the ribosome as the predominantly due to the loss of enzymatic function in bio-
protein factory, and the mitochondria as the energy plant logical environments. TCM is a nature-inspired approach
among others is what defines a healthy living organism. How- that tries to mimic biological concepts to facilitate the inter-
ever, the treatment of malfunctioning organelles, either since action of synthetic assemblies (formulations) with biological
birth or due to environmental factors, is challenging.[1]
systems in a sustained manner. Options are to mimic either
Therapeutic cell mimicry (TCM) aims to substitute for the entire cells or their organelles. In the latter case, typi-
missing or lost cellular function by imitating a specific cel- cally small single compartment systems, encapsulating one
[
3]
lular activity, i.e., by performing encapsulated biocatalysis or more specific enzymes, are considered as simple mimics.
Among the most crucial aspects of artificial organelles is their
ability to be active inside of a biological cell. However, only
few published reports demonstrated intracellular activity
using polymersomes or hollow silica nanospheres. Ben-Haim
B. Thingholm, Dr. P. Schattling,
Dr. Y. Zhang, Dr. B. Städler
Interdisciplinary Nanoscience Center
Aarhus University
Gustav Wieds Vej 14, Aarhus 8000, Denmark
E-mail: bstadler@inano.au.dk
[
4]
[5]
et al. and later van Dongen et al. demonstrated the
uptake by cells and preserved activity of enzyme-loaded
polymersomes within the cells, employing model systems.
[6]
Tanner et al. assembled a reactor that could combat oxi-
DOI: 10.1002/smll.201502109
dative stress in THP-1 cells via an encapsulated enzymatic
1806 www.small-journal.com
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
small 2016, 12, No. 13, 1806–1814

www.MaterialsViews.com
cascade reaction. The group of Mou employed enzyme-
loaded silica nanospheres for the triggered conversion of
[
7]
prodrugs. These first reports highlight the potential of the
approach to equip natural cells with artificial organelles
toward a potential treatment of organelle dysfunction,
but the field is in its infancy with a magnitude of scientific
challenges to be overcome and fundamental aspects to be
elucidated.
From a different perspective, the assembly of subcompart-
mentalized microreactors is a promising approach toward the
[8]
creation of therapeutic artificial cells Such microreactors
mimic the hierarchical structure of a cell, which allow them
to perform multiple reactions simultaneously in a spatially
confined manner with high efficiency and accuracy. Sub-
compartmentalized systems have been fabricated using
predominantly a combination of liposomes, polymersomes,
[9]
and polymer capsules as building blocks. Among the most
[
10]
advanced approaches are polymersomes in polymersomes
and capsosomes, liposomes within polymer carrier capsules,
[11]
including the demonstration of encapsulated enzymatic cas-
cade reactions as the most complex functionality in both cases.
Janus microreactors, toward the mimicry of the polarity of
Scheme 1. Schematic illustration of i) the assembly of the
subcompartmentalized nanoreactors starting with PLL coating of
silica particles followed by the deposition of LGOx, PDA, and RGD, and
ii) the intracellularly active subcompartmentalized nanoreactors with
encapsulated glucose oxidase in the liposomal subunits (LGOx) including
the illustration of the enzymatic conversion of glucose into D-glucono-
[12]
cells, were recently reported.
Furthermore, we advanced
capsosomes toward their use as an oral treatment for phe-
nylketonuria by loading them with the enzyme phenylalanine
ammonialyase, which converts phenylalanine into benign
[13]
trans-cinnamic acid in the intestine. We demonstrated that 1,5-lactone and H O .
2
2
such capsosomes could be used as extracellular microreac-
tors with preserved activity in cell culture with applied peri-
staltic flow. Thus far, the concept of subcompartmentalization solution followed by their internalization by cell with pre-
has never been considered for intracellular active reactors, served structural integrity and function. The first aspect we
although natural organelles contain separated subunits to considered was the uptake efficiency of core–shell particles
facilitate parallel reactions while avoiding undesired interac- depending on the outermost surface coating to ensure high
tions. Furthermore, free liposomes suffer from instability in internalization. Herein, we chose 800 nm core–shell parti-
biological environment. In contrast, we have recently demon- cles due to their desirable high loading of enzyme-loaded
strated that capsosomes exhibit superior stability of the lipo- liposomes with preserved colloidal stability and possibility
somal subunits when exposed to the enzyme phospholipase.[14] for surface modification. Admittedly, the particles might be
Therefore, subcompartmentalized artificial organelles might fast filtered out by the organs when administered in vivo due
turn out to be superior over single compartment structures.
to their size and hardness, yet these particles are still able
Herein, we report the assembly of subcompartmentalized to illustrate the potential of subcompartmentalized artificial
nanoreactors with intracellular activity (Scheme 1). Specifi- organelles.
cally, we i) coated 800 nm silica particles with poly(dopamine)
(PDA), poly(ethylene glycol) (PEG), or (arginylglycylaspartic 2.1.1. Coating
acid) (RGD) and compared their uptake by endothelial cells
and macrophages without and with applied shear stress, First, the particles were coated with a layer of fluorescein
ii) assembled and characterized subcompartmentalized nano- isothiocyanate labeled poly(l-lysine) (PLL ) to visualize the
F
reactors by depositing enzyme-loaded liposomal subunits particles when internalized by the cells. Next, a thin layer of
onto the silica particles prior to the surface-modification with PDA was adsorbed. PDA, deposited by the self-polymeriza-
RGD, and iii) confirmed their intracellular enzymatic activity tion of dopamine (DA) at slightly basic pH, is an interesting
by the production of hydrogen peroxide (H O ), which led to material to be used for various biomedical applications as
2
2
[
15]
[16]
reduced cell viability.
recently extensively reviewed by us and others. Among
the many advantages is the opportunity for postmodification
with amines and thiols. We employed the graft copolymer
PLL-graft -PEG with unmodified or RGD end-functionalized
PEG chains, a copolymer with prior success in biomedicine,
2
. Results and Discussion
2
.1. Surface Coating-Dependent Interaction of Nanoparticles
[
17]
[18]
with Endothelial Cells and Macrophages
for instance, as surface coating,
for biolubrication,
or
[
19]
for DNA delivery. This polymer layer equipped the parti-
Successful artificial organelles need to fulfil a lot of different cles with cell-adhesive (RGD) or cell-repellent (PEG) prop-
requirements including their assembly and testing in buffer erties, since the amine groups on the PLL allowed for the
small 2016, 12, No. 13, 1806–1814
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.small-journal.com 1807

full papers
www.MaterialsViews.com
2
.1.2. Uptake by Endothelial Cells and Macrophages
Intracellularly active nanoreactors need to fulfil multiple
requirements including the internalization by the targeted
cell line. Therefore, we aimed at comparing the uptake of
PDA
PEG
RGD
P
, P , and P
by macrophages and endothelial cells
in static conditions and with applied shear stress. These two
cell types were chosen because they are interesting targets
for artificial organelles and are straight forward accessible
upon intravenous injection. Macrophages are the body’s
first line of defense against intruding pathogens. How-
ever, microorganisms have come up with ways to avoid or
escape the digestive organelles of immune cells leading to
severe medical conditions. For instance, adherent-invasive
Escherichia coli bacteria can resist macrophage degrada-
[
21]
tion in Crohn’s disease patients,
or Mycobacterium tuber-
culosis is a microorganism that has adapted to survive in
[
22]
macrophage phagosomes.
Since traditional treatment
strategies did not succeed, novel concepts such as TCM are
required, i.e., equipping macrophages with artificial digestive
reactors might present itself as an alternative approach to
battle resistant invaders.
Static Conditions: The first step for an artificial organelle
to support its host cell is to be internalized. With the aim to
PDA
PEG
RGD
compare the uptake of P
, P , and P
by endothelial
−1
cells and macrophages, 6250 and 2500 particles mL were
incubated with endothelial cells and macrophages, respec-
tively, at static conditions (τ = 0 dyn cm ). Different par-
Figure 1. a) ζ-Potential of the SiO particles measured after the different
2
1
coating stepsin HEPES buffer. b) Representative fluorescent microscopy
image of P . The scale bar is 2 µm. c) Flow cytometry histogram of
−2
RGD
0
ticle concentrations were used to keep the particles to cell
ratio constant at 10/1. The normalized cell mean fluorescence
RGD
2
P
. Experiments in (b) and (c) were performed in HEPES buffer.
(
nCMF; Figure 2) was assessed at different time points
interaction with the quinons of the PDA. The particles will be using flow cytometry. Endothelial cells exhibited similar
PDA
PEG
RGD
labeled P
, P , and P
from now on.
nCMF independent on the used particles. Unexpectedly,
RGD
The deposition of these polymer layers was confirmed the endothelial cells did not show any preference to P
by measuring the ζ-potential of the particles (Figure 1a). (Figure 2a). Furthermore, the nCMF was highest after
The initial negative ζ-potential of the silica particles in 2 h, followed by a constant decrease over the subsequent
−3
1
0 × 10 m 4-(2-hydroxyethyl)piperazine-1-ethanesulfonic 21 h. On the other hand, macrophages showed constant
1
acid (HEPES) buffer (pH 7.4, HEPES buffer) increased nCMF between 3 and 24 h exposure times to the particles
upon the deposition of PLL , followed by a slight drop (Figure 2b). Additionally, P
when the PDA layer was adsorbed. P
RGD
led to the expected sig-
F
PEG
RGD
PDA
PEG
and P
had a nificantly higher nCMF compared to P
and P
for all
ζ-potential of ≈0 and ≈ −10 mV, respectively. This observa- tested time points.
tion together with our prior report on the deposition of PLL-
Applied Shear Stress: Assessing the interaction of drug
[
20]
g-PEG onto PDA
suggested the successful deposition of carriers and cells, using predominantly endothelial cells, with
RGD
the final polymer layer. P
was visualized in buffer solution applied shear stress starts to be recognized as an important
2
−3
−3
[23]
(HEPES buffer: 10 × 10 m HEPES and 150 × 10 m NaCl factor as recently reviewed by Godoy-Gallardo et al.
(pH 7.4)) as a representative example using an epifluorescent Endothelial cells line our blood vessel and are therefore per-
microscope (Figure 1b). The homogenous fluorescence, due manently exposed to blood flow. Furthermore, circulating
to the presence of PLL , confirmed that the PDA layer was macrophages experience the same mechanical forces, and its
F
[
24]
[25]
thin enough to not entirely block the fluorescence emission. relevance has recently been reported by us
and others.
This aspect was important since we have prior observed that Since every potential artificial organelle administered
thick layers of PDA could entirely hinder the visualization intravenously will be exposed to the blood stream and the
[
11]
of fluorescently labeled underlying layers.
Furthermore, induced mechanical forces such as shear stress, the uptake of
PDA
PEG
RGD
the particles remained well dispersed without aggregation,
which was also supported when considering the flow cytom- were reassessed with applied shear stress (τ = 4 dyn cm )
etry histogram using particles suspended in HEPES buffer using a commercial microfluidic setup. The nCMF of the cells
P
, P , and P
by endothelial cells, and macrophages
−2
4
2
(Figure 1c). The nonaggregated state of the particles was upon exposure to the different particles was monitored after
important to ensure that differences in the biological evalua- 2 h for macrophages and 16 h for endothelial cells and com-
tion were only due to varied surface chemistries and not due pared to the nCMF obtained at τ (Figure 3). For endothelial
0
RGD
to changes in size, i.e., aggregates versus individual particles.
cells, the nCMF due to the exposure of the cells to P
was
1808 www.small-journal.com
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
small 2016, 12, No. 13, 1806–1814

www.MaterialsViews.com
Figure 3. Normalized CMF of endothelial cells and macrophages upon
PDA PEG
RGD
exposure to P , P , or P for 16 and 2 h, respectively, with applied
**
shear stress (τ ) compared to the static condition (τ ) (n = 4, p < 0.01,
4
0
***
p < 0.001).
that there were many noninternalized particles observed
when using endothelial cells, while this was not the case for
samples with macrophages. Employing the CLSM images
and flow cytometry results, we estimated the number of inter-
nalized particles for the macrophages to be ≈37 and ≈74 at τ0
and τ , respectively.
4
Following on, we assessed the cell viability of endothe-
lial cells and macrophages with applied shear stress τ when
4
RGD
exposed to P
in comparison to τ after 16 and 2 h, respec-
0
tively (Figure 5). The cell viability was normalized to cells
RGD
only at τ . First, exposure to P
did not negatively affect
0
the cells for τ and τ . Interestingly, a ≈60% higher number
0
4
Figure 2. The normalized CMFofa) endothelialcellsand b) macrophages of viable endothelial cells were detected at τ . The appear-
4
PDA
PEG
RGD
exposed to P , P , and P
depending on the incubation time is
ance of the cells depending on the applied shear stress and
*
shown (n = 3, p < 0.05).
RGD
the presence of P
was visualized using bright field micro-
scopy (Figure 6). More endothelial cells were observed at τ4
significantly higher compared to PPDA and PPEG at τ , while in agreement with the quantitative cell viability results. Since
4
the nonsignificant difference at τ was confirmed. For mac- this type of cells is permanently exposed to the blood flow
0
RGD
rophages, the nCMF due to the exposure of the cells to P
was significantly higher compared to P
and the induced mechanical forces, it is not surprising that
PDA
PEG
and P
at τ0 conditions which mimic the biological environment have
and τ . Furthermore, the nCMF of both cell types exposed to beneficial effect on the endothelial cells. Furthermore, the
4
RGD
P
was significantly higher at τ compared to τ , while no number of macrophages was similar but they were found to
4
0
PDA
PEG
significant difference was observed for P
and P
.
have a more rounded morphology at τ . We have prior found
4
With the aim to confirm the uptake of the particles by
the cells, endothelial cells and macrophages exposed to
RGD
P
with applied shear stress (τ ) for 16 and 2 h, respec-
4
tively, were fixed, stained, and visualized by confocal laser
scanning microscopy (CLSM). The endothelial cells had only
very few internalized particles (Figure 4, right). On the other
hand, macrophages had a high amount of internalized par-
ticles (Figure 4, left). The visualization confirmed the flow
cytometry results. Due to the size and mechanical properties
of the particles, the results were expected. Immune cells are
known to be able to internalize much larger particles than
other type of cells without the activation of a specific endo-
cytotic pathway such as the intercellular adhesion molecule
Figure 4. CLSM images of endothelial cells and macrophages exposed
[
26]
RGD
1
mediated uptake.
Furthermore, we would like to note to P with applied shear stress (τ ) for 16 and 2 h, respectively.
4
small 2016, 12, No. 13, 1806–1814
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.small-journal.com 1809

full papers
www.MaterialsViews.com
method to monitor the intracellular activity of the nanoreac-
tors by assessing the viability of the macrophages.
2
.2.1. Assembly
Inspired by our prior work,[27] we deposited glucose
oxidase-loaded 1,2-ditetradecanoyl-sn-glycero-3-phospho-
choline/1,2-dihexadecanoyl-sn-glycero-3-phosphocholine
liposomes (LGOx) as subunits on 800 nm PLL precoated
silica particles followed by the deposition of PDA overnight
RGD
and PLL-g-PEG/RGD (P
LGOx). As we have previously
shown, PLL allows for the immobilization of different types
[
11]
of liposomes without their rupturing,
and the subsequent
PDA deposition did not lead to their rupturing or displace-
[
27]
ment. CLSM was employed to confirm that the assembly of
the nanoreactors was successful. In this context, fluorescently
labeled GOx (GOx633) was incorporated into NBD-labeled
liposomes (LNBD). Figure 7ai,ii shows that the nanoreactors
were homogenously coated with both LNBD and GOx633. As
control, the PLL precoated silica particles were exposed to a
Figure 5. Normalized cell viability of endothelial cells and macrophages
RGD
upon exposure to P
in comparison to cell only (C) at τ0 and τ4 for
1
6 and 2 h, respectively, (n = 3, ***p < 0.001).
solution of GOxD followed by coating with PDA and PLL-
RGD
g-PEG/RGD (P
GOx) with the aim to demonstrate that
that shear stress affected the number of macrophages likely the activity comes from enzymes within the liposomal subu-
due to their semiadherent nature.2
4b
nits only. (GOxD refers to GOx treated the same way as for
encapsulation into liposomes but without using lipids.) The
−3
time-dependent conversion of 100 × 10 m glucose into H O
2
2
−1
2.2. Nanoreactor
by 100 reactors µL was monitored using the Amplex Red
assay and reading out of the fluorescent intensity (λ = 598 nm)
Following on, we aimed at assembling enzyme-loaded sub- using a multimode plate reader (Figure 7b). As expected,
compartmentalized nanoreactors and to test such a system for only the sample containing nanoreactors with GOx encap-
the first time toward its intracellular activity. Glucose oxidase sulated within the liposomal subunits exhibited increasing
RGD
was chosen as enzyme since it converts glucose into H O , a fluorescent intensities, i.e., only P
was able to con-
2
2
LGOx
RGD
cytotoxic compound. This aspect provides a straight-forward vert glucose into H O . On the other hand, P
showed
GOx
2
2
Figure 6. Representative bright field images of endothelial cells and macrophages upon exposure to PRGD at τ and τ for 16 and 2 h, respectively.
0
4
1
810 www.small-journal.com
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
small 2016, 12, No. 13, 1806–1814

www.MaterialsViews.com
inside of the macrophages using CLSM,
RGD
P
was assembled using LNBD
LGOx
and GOx633. We would like to note that
the PDA coating in this case was much
thinner (only 15 min PDA coating) in
order to be able to detect the fluorescence
by CLSM. This thinner PDA coating was
likely to affect the stability of the nano-
reactors inside of the cells and therefore,
we chose 6 h incubation time. Although
both, LNBD and GOx633 were observed,
only a few colocalized spots could be iden-
tified (Figure S3, Supporting Information).
We attributed this aspect to the fact that
lipids coated with a PDA capping layer
could be internalized by cells in a PDA
coating thickness depending manner as we
[
28]
have previously shown.
On the other
hand, increasing the thickness of the PDA
layer hindered the fluorescent imaging
[
11]
as we have previously reported.
As a
consequence, CLSM images of the fluo-
rescently labeled nanoreactors inside the
macrophages yielded poor quality images.
Following on, the subcompartmen-
RGD
talized reactor P
was incubated
LGOx
with macrophages for 2 h followed by
two washing steps with PBS and incuba-
−3
tion with 100 × 10 m glucose for 6 and
4 h. We would like to note that the PDA
2
coating time of these nanoreactors was
sufficiently long to ensure that the PDA
thickness was sufficiently high to avoid
the above-mentioned issue for the chosen
time points. As controls, the macrophages
Figure 7. a) CLSM images of nanoreactors coated with i) fluorescently labeled liposomes
(
LNBD) and ii) fluorescently labeled GOx (GOx ). b) Nanoreactor activity: The time-dependent
633
RGD
RGD
conversion of glucose into H O by P
and P
is shown.
GOx
2
2
LGOx
very low fluorescent intensity strongly suggesting the absence were exposed to PRGD
without the addition of glucose,
LGOx
or denaturation of GOx in the assembly. Furthermore, macrophages were only incubated with glucose but not with
RGD
employing the calibration curve for GOx (Figure S1, Sup-
P
LGOx, or the cells remained entirely untreated. The
porting Information) and assuming that 1 U corresponds to washing step was important to ensure that the reactors which
the production of 1 µmol H O per min, the amount of pro- were not internalized were removed. The success of this step
2
2
duced H O per nanoreactor was estimated on the basis of has been confirmed by CLSM in Figure 4. The viability of the
2
2
the recorded FI after 30 min (Figure 7) and a concentration cells was assessed and compared. The absorbance readings,
−
1
of 100 nanoreactors µL . Accordingly, the estimated H O
representing the not normalized cell viability, showed sig-
2
−1
2
−9
production rate per nanoreactor was ≈3 × 10 µmol min .
nificant lower viability when the macrophages were exposed
RGD
to both P
and glucose after 6 and 24 h compared to
LGOx
2
.2.2. Intracellular Activity
the untreated cells (Figure 8a). Importantly, there was no
significant difference in the absorbance reading between the
With the aim to demonstrate that the assembled nanoreactors untreated cells and the cells exposed to either PRGDLGOx or
exhibited intracellular activity, they were employed to kill the glucose only after 6 and 24 h. Furthermore, when normal-
host cells using the produced H O to illustrate the function izing the raw data to the untreated cells, the viability of cells
2
2
RGD
of the potential artificial organelles to counteract pathogenic exposed to P
and glucose after 6 and 24 h was signifi-
LGOx
invasion in macrophages. First, the dose response curve for cantly lower compared to the controls (Figure 8b). Addition-
macrophages exposed to glucose was measured (Figure S2, ally, the normalized viability of the macrophages incubated
−3
RGD
Supporting Information). 100 × 10 m glucose was chosen with P
and glucose for 6 h was significantly higher
LGOx
as a concentration which did not negatively affect the cells, compared to an incubation time of 24 h. These results
but enough H O should be produced to monitor the reduc- strongly suggest that H O2 was produced by the subcom-
2
2
2
tion in cell viability and by doing so, the intracellular activity partmentalized reactor intracellularly leading to cell death.
of the nanoreactors was confirmed. In an attempt to visu- Considering the assumption that one nanoreactor produces
−9
−1
alize both components of the subcompartmentalized reactor ≈3 × 10 µmol min H O and that all the 37 nanoreactors
2
2
small 2016, 12, No. 13, 1806–1814
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.small-journal.com 1811

full papers
www.MaterialsViews.com
reduced cell viability upon exposure to glucose compared to
the controls, demonstrating the intracellular function of the
nanoreactors. These results provide a considerable contribu-
tion to the field of TCM by assembling nanoreactors, which
could act as artificial organelles.
4. Experimental Section
Materials: Sodium chloride (NaCl), poly(L-lysine) hydrobromide
(PLL, 40–60 kDa), phosphate buffered saline (PBS), 4-(2-hydrox-
yethyl)piperazine-1-ethane-sulfonic acid (HEPES), chloroform
anhydrous (≥99%), dimethyl sulfoxide (DMSO), glucose oxidase
from Aspergillus niger (GOx, ≥15 units mg− solid, 160 kDa),
catalase from bovine liver (10 000 units mg− solid, 240 kDa),
peroxidase from horseradish (HRP, 250–330 units mg− solid),
1
1
1
paraformaldehyde
(PFA),
tris(hydroxymethyl)-aminomethane
(TRIS), dopamine hydrochloride, 4′,6-diamidino-2-phenylindole
dihydrochloride (DAPI), phalloidin-tetramethylrhodamine B iso-
thiocyanate (phalloidin-TRITC), cell counting kit-8 (CCK-8), o-dian-
isidine, polyethylene glycol-tert-octylphenyl ether (Triton X-100),
D-glucose (≥99.5%), and hydrogen peroxide (H O , 30% w/w) were
2
2
purchased from Sigma-Aldrich.
Zwitterioniclipids,1,2-ditetradecanoyl-sn-glycero-3-phosphocholine
DMPC, Tm = 23 °C), 1,2-dihexadecanoyl-sn-glycero-3-phosphocholine
(
(
DPPC, Tm = 41 °C), and 1-myristoyl-2-[12-[(7-nitro-2-1,3-benzoxadi-
azol-4-yl)amino]dodecanoyl]-sn-glycero-3-phosphocholine (NBD) were
purchased from Avanti Polar Lipids (Alabama, USA). Silica particles
(0.8 µm in diameter) were obtained from Microparticles GmbH (Berlin,
Germany). Amplex Red Glucose/Glucose Oxidase Assay Kit and Alexa
Flour 633 C5 maleimide (AF633) reactive dye were purchased from Life
Technologies (Carlsbad, USA).
Poly(L-lysine (20 kDa))-graft [3.5]-poly(ethylene glycol (2 kDa))
(PLL-g-PEG) and poly(L-lysine (20 kDa))-graft [3.5]-poly(ethylene
Figure 8. Intracellular active subcompartmentalized nanoreactor: glycol (2 kDa)/poly(ethylene glycol (2 kDa))/L-arginine-gly-
RGD
a) normalized cell viability of macrophages b) exposed to Glu, P
cine-L-aspartic acid (RGD)) (PLL-g-PEG/RGD) (6–15% of PEG
(3.4 kDa) is functionalized with RGD) were obtained from SoSuS AG
LGOx
RGD
or Glu, and P
p < 0.001,
compared to cell only after 6 and 24 h (n = 3,
LGOx
p < 0.0001).
*
**
****
(
Dübendorf, Switzerland).
Three buffers were used: (i) HEPES buffer consisting of
1
−3
2
per cell were functional and no H O2 was degraded by 10 × 10 M HEPES (pH 7.4), (ii) HEPES buffer consisting of
inherent cellular activity, 0.04 nmol H O2 was produced 10 × 10 M HEPES and 150 × 10 M NaCl (pH 7.4), and (iiii)
within one macrophage after 6 h leading to the observed TRIS buffer consisting of 10 × 10 M TRIS (pH 8.5). All buffers
reduction in viability. Furthermore, employing the dose were made using ultrapure water (Milli-Q gradient A 10 system,
response curve of macrophages exposed to H O (Figure S4, resistance 10 MΩ cm, TOC <4 ppb, Millipore Corporation, USA).
2
−3
−3
2
−3
2
2
Supporting Information), we estimated that the cell viability
Unilamellar liposomes were prepared by evaporation of 100 µL
was reduced to a similar amount as when the macrophages chloroform/lipid solution (25 mg mL− DMPC:DPPC 84:16 wt%)
1
−3
were exposed to media containing ≈1.4 × 10 m H O .
under vacuum for at least 1 h, followed by rehydration of the lipids
2
2
2
−1
in 200 µL HEPES containing 5 mg mL GOx (LGOx). LGOx was then
2
further hydrated with 800 µL HEPES . The solutions were extruded
through 100 nm filters (11×) to obtain monodisperse liposomes.
Nonencapsulated enzymes were removed by dialyzing the sample
3
. Conclusion
In summary, we assembled the first subcompartmentalized for 2 h under stirring against HEPES2 (MWCO 100 kDa mem-
nanoreactor with intracellular activity. The uptake efficiency branes). The liposome formation was confirmed by analyzing the
was found to be dependent on the surface coating and the solutions by dynamic light scattering (DLS, Zetasizer Nano S90)
presence of shear stress, with significant higher internaliza- using a material refractive index of 1.590 and a dispersant (water
tion for RGD-modified particles with applied shear stress for at 25 °C) refractive index of 1.330.
endothelial cells and macrophages. The loading of GOx into
the liposomal subunit of the nanoreactor was confirmed and (50 mg mL ) of 0.8 µm diameter silica particles was washed 2× in
Core–Shell Particles of Cell Uptake Experiments: 200 µL
−1
2
−1
the activity was demonstrated by the conversion of glucose. HEPES buffer, coated with PLL (1 mg mL , 15 min), and washed
FITC
RGD
Finally, macrophages with internalized P
exhibited 3× in TRIS buffer. All washing steps were performed in a bench top
LGOx
1812 www.small-journal.com
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
small 2016, 12, No. 13, 1806–1814

www.MaterialsViews.com
centrifuge (MiniSpin, Eppendorf) at 1500 rpm for 3 min. Next, the seeded in a 96-well plate. Each well was washed 2× in PBS and
particles were exposed to a dopamine (DA) solution (2 mg mL , 100 µL PPDA, PPEG, or PRGD suspended in the corresponding media
−1
1
5 min), followed by 3× washing in HEPES2 buffer yielding PPDA was added (particle/cells 10:1). The cells and particles were incu-
particles. If needed, these particles were further coated by incuba- bated at 37 °C in 5% CO for 2, 6, 12, and 24 h. Then, each well
2
−
1
2
tion in either a PLL-g-PEG (1 mg mL in HEPES buffer, 30 min) was washed 2× in PBS and the cells were harvested for analysis by
or a PLL-g-PEG/RGD (1 mg mL− in HEPES buffer, 30 min) solu- flow cytometry.
tion, followed by 3× washing in HEPES buffer to obtain P and
1
2
2
PEG
Uptake Experiments—Dynamic Conditions: The ibidi Pump
PRGD, respectively. The ζ-potential of the particles was analyzed by System (ibidi GmbH, Munich, Germany) was employed to apply
dispersing 1 µL of the final particles in 999 µL HEPES buffer. The controlled shear stress. Macrophages (100 000 cells per channel)
1
analysis was carried out at each deposition step using a Zetasizer and endothelial cells (40 000 cells per channel) were seeded in
Nano ZS (Malvern Instruments).
closed perfusion channels (µ-slide VI 0.4 six-well ibiTreat chan-
The fluorescence and concentration of all particles was nels) and allowed to attach at 37 °C in 5% CO for 16 and 4 h,
2
measured by flow cytometry prior to the cell experiments.
respectively. Afterwards, 7.5 mL of corresponding media con-
Nanoreactor Assembly: 200 µL (50 mg mL ) of 0.8 µm silica taining PPDA, PPEG, or PRGD (particle/cell 10:1) was added to the
−1
2
particles was washed 2× in HEPES buffer, coated with PLL (1 mg syringes of the pump system, which were connected to the chan-
mL− in HEPES buffer, 15 min), followed by 3× washing in HEPES
1
2
2
nels, and a shear stress (τ = 4 dyn cm ) was applied for 2 and
4
−2
buffer. Then, the particles were exposed to an LGOx, stock solution 16 h for macrophages and endothelial cells, respectively, at 37 °C
for 30 min followed by 3× washing in HEPES2 buffer. The parti- in 5% CO . As a control, 120 µL of medium with no shear stress
2
−2
cles were then washed 3× in TRIS buffer and incubated with in a (τ0 = 0 dyn cm ) and no particles was employed. The channels
−1
DA solution (2 mg mL in TRIS buffer) overnight, followed by 3× were washed 2× with 120 µL PBS, and the cells were harvested for
2
washing in HEPES buffer. Finally, the particles were coated with analysis by flow cytometry.
−
1
2
PLL-g-PEG/RGD (1 mg mL in HEPES buffer, 30 min) and washed
Cell Viability: The viability of the cells was assessed after
2
RGD
was exposure to PRGD as described above using dynamic conditions.
3
× washing in HEPES buffer. Fluorescently labeled P
assembled by using liposomes NBD lipids (2.5 mg mL ), AF633- Following on, the channels were washed 2× with the corre-
labeled GOx, and the PDA was allowed to deposit for 15 min. sponding media and 110 µL media and 10 µL of CCK-8 (Dojindo)
LGOx
−1
Nanoreactor Activity: The assembled nanoreactors were solution was added to each channel. The cells were incubated for
counted using flow cytometry. 50 µL of the particle samples 2 h at 37 °C in 5% CO . Then, 100 µL of the solution from each
2
−1
(
1
5
100 particles µL ) was incubated in a 96-well plate at 37 °C with channel was transferred to a 96-well plate and analyzed using a
00 × 10− M glucose at different time points. After incubation, multimode plate reader by measuring the absorbance at 450 nm.
3
0 µL of a working solution consisting of 100 × 10− M Amplex Red Three independent repeats were performed and the results were
6
(
Life Science) and 0.2 U mL−1 horse radish peroxidase in HEPES2 normalized to cells in the absence of PRGD at τ0.
was added to each well, and the fluorescence (λ = 598 nm) was
Intracellular Nanoreactor Activity: Macrophages were seeded
immediately measured using a multimode plate reader.
in 96 well plates (50 000 cells per well) and allowed to attach
Cell Work: RAW 264.7 mouse macrophage and human umbil- overnight at 37 °C in 5% CO . The cells were then washed 2× with
2
ical vein endothelium cells (HUVEC) cell lines were purchased from 100 µL PBS, incubated with PRGDLGOx (particles/cells 10:1) for 2 h
European Collection of Cell Cultures. The macrophages were cul- at 37 °C in 5% CO , and washed 2× with PBS before adding media
2
tured in 75 cm culture flasks (1 100 000 cells per flask in 20 mL with 100 × 10−3 M glucose or without added glucose for 6 or 24 h.
2
medium) at 37 °C and 5% CO using Dulbecco’s modified Eagle’s As controls, cells without exposure to PRGD
were considered.
2
LGOx
medium (DMEM, with 4500 mg mL− glucose, sodium pyruvate, The cells were washed 2× in PBS and 110 µL cell medium con-
1
and sodium bicarbonate) and added 10% fetal bovine serum taining 10 µL of CCK-8 solution was added to each well and incu-
−
1
−1
(
FBS), penicillin (50 µg mL ), and streptomycin (50 µg mL ). The bated for 2 h at 37 °C in 5% CO . Finally, 100 µL of the solution
2
2
endothelial cells were cultured in 75 cm culture flasks (250 000 from each well was transferred to a 96-well plate and analyzed
cells per flask in 20 mL medium) at 37 °C and 5% CO using using a multimode plate reader (absorbance at 450 nm). Three
2
−1
Medium 200 (Invitrogen) supplemented with 50 µg mL penicillin, independent repeats were conducted. The statistical significance
0 µg mL− streptomycin, and 2% large vessel endothelial supple- used to compare the distribution was determined using a two-way
1
5
ment (Invitrogen).
ANOVA with a confidence level of 95% (α = 0.05), followed by a
*
At least 2000 cells were analyzed and at least three inde- Tukey’s multiple comparison posthoc test ( p < 0.05).
pendent repeats were performed for all the reported flow cytom-
Glucose Dose Response: Macrophages were seeded in a
etry analysis (C6 Flow Cytometer, Accuri Cytometer, Inc.) using an 96-well plate (50 000 cells per well) and allowed to attach over-
excitation wavelength of 488 nm. The autofluorescence of the cells night at 37 °C in 5% CO . The cells were then washed 2× with
2
was subtracted by analyzing control samples where no particles 100 µL PBS and incubated with different concentrations of glucose
were added. All data were normalized to the fluorescence of the for 6 or 24 h. The cells were washed 2× in PBS and 110 µL cell
particles which were measured using a multimode plate reader. medium containing 10 µL of CCK-8 solution was added to each
Unless otherwise noted, the statistical significance used to com- well and incubated for 2 h at 37 °C in 5% CO . Finally, 100 µL of
2
pare the distribution was determined using a one-way ANOVA the solution from each well was transferred to a 96-well plate and
with a confidence level of 95% (α = 0.05), followed by a Tukey’s analyzed using a multimode plate reader (absorbance at 450 nm).
*
multiple comparison posthoc test ( p < 0.05).
Three independent repeats were performed.
Uptake Experiments—Static Conditions: Macrophages (50 000
Imaging: Macrophages (100 000 cells per channel) were
cells per well) and endothelial cells (20 000 cells per well) were seeded in closed perfusion channels (µ-slide VI 0.4 six-well
small 2016, 12, No. 13, 1806–1814
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.small-journal.com 1813

full papers
www.MaterialsViews.com
ibiTreat channels) and allowed to attach at 37 °C in 5% CO over-
Vol. 3 (Ed: T. M. S. Chang), World Science Publisher/Imperial
College Press, Singapore, 2013; b) M. Marguet, C. Bonduelle,
S. Lecommandoux, Chem. Soc. Rev. 2013, 42, 512.
2
night. The cells were then washed 2× with PBS and incubated with
fluorescently labelled PRGD
for 6 h in a particle to cell ratio of
LGOx
[
9] R. Chandrawati, F. Caruso, Langmuir 2012, 28, 13798.
10] R. J. R. W. Peters, M. Marguet, S. Marais, M. W. Fraaije,
1
0/1. After incubation, the cells were washed twice with 100 µL
[
PBS; 100 µL PFA (4%) was added to each channel, washed twice
J. C. M. van Hest, S. Lecommandoux, Angew. Chem., Int. Ed. 2014,
with 100 µL PBS and washed once with 100 µL of 0.1% Triton-
5
3, 146.
X in PBS (T-PBS). 100 µL DAPI (1 µg mL− in PBS) was added to [11] L. Hosta-Rigau, M. J. York-Duran, Y. Zhang, K. N. Goldie, B. Städler,
1
each channel and incubated for 30 min at room temperature in
aluminum foil. The channel was then washed thrice in T-PBS and
finally 100 µL PBS was added for storage. Confocal laser scanning
fluorescence microscopy was conducted using a Zeiss Axiovert
microscope coupled to a LSM 700 confocal scanning module (Carl
Zeiss, Germany) and a 100× oil-immersion objective.
ACS Appl. Mater. Interfaces 2014, 6, 12771.
[
[
12] P. Schattling, C. Dreier, B. Stadler, Soft Matter 2015, 11, 5327.
13] L. Hosta-Rigau, M. J. York-Duran, T. S. Kang, B. Stadler, Adv. Funct.
Mater. 2015, 25, 3860.
14] R. Chandrawati, S.-F. Chong, A. N. Zelikin, L. Hosta-Rigau,
B. Stadler, F. Caruso, Soft Matter 2011, 7, 9638.
[
[
15] M. E. Lynge, P. Schattling, B. Städler, Nanomedicine 2015, 10,
2
725.
[
[
16] Y. Liu, K. Ai, L. Lu, Chem. Rev. 2014, 114, 5057.
17] a) B. Stadler, D. Falconnet, I. Pfeiffer, F. Hook, J. Voros, Lang-
muir 2004, 20, 11348; b) J. W. Lussi, D. Falconnet, J. A. Hubbell,
M. Textor, G. Csucs, Biomaterials 2006, 27, 2534; c) R. Ogaki,
M . F o s s , Langmuir 2014, 30, 7014; d) S. G. Ricoult, G. Thompson-
Steckel, J. P. Correia, T. E. Kennedy, D. Juncker, Biomaterials 2014,
35, 727.
Supporting Information
Supporting Information is available from the Wiley Online Library
or from the author.
[18] A. Dedinaite, Soft Matter 2012, 8, 273.
[
[
[
19] T. Luehmann, M. Rimann, A. G. Bitterman, H. Hall, Bioconj. Chem.
008, 19, 1907.
20] B. M. Teo, R. van der Westen, L. Hosta-Rigau, B. Städler, BBA-Gen.
Subjects 2013, 1830, 4838.
21] M.-A. Bringer, A.-L. Glasser, C.-H. Tung, S. Méresse, A. Darfeuille-
Michaud, Cell. Microbiol. 2006, 8, 471.
2
Acknowledgements
This project was supported by a grant from the Danish Research
Council for Independent Research-Technology and Production [22] D. Schnappinger, S. Ehrt, M. I. Voskuil, Y. Liu, J. A. Mangan,
I. M. Monahan, G. Dolganov, B. Efron, P. D. Butcher, C. Nathan,
G. K. Schoolnik, J. Exp. Med. 2003, 198, 693.
(DFF-FTP).
[23] M. Godoy-Gallardo, P. K. Ek, M. M. T. Jansman, B. M. Wohl,
L. Hosta-Rigau, Biomicrofluidics 2015, 9, 052605.
[
24] a) K. Panneerselvam, M. E. Lynge, C. F. Riber, S. Mena-Hernando,
A. A. A. Smith, K. N. Goldie, A. N. Zelikin, B. Städler, Biomicroflu-
idics 2015, 9, 052610; b) K. Panneerselvam, S. Mena-Hernando,
B. M. Teo, K. N. Goldie, B. Stadler, RSC Adv. 2014, 4, 44769.
25] a) L. Castro-Núñez, I. Dienava-Verdoold, E. Herczenik, K. Mertens,
A. B. Meijer, J. Thromb. Haemost. 2012, 10, 1929; b) N. Wohner,
P. Legendre, C. Casari, O. D. Christophe, P. J. Lenting, C. V. Denis,
J. Thromb. Haemost. 2015, 13, 815.
[
[
[
1] A. Dillin, D. E. Gottschling, T. Nyström, Curr. Opin. Cell Biol. 2014,
6, 107.
2] A. Jurecka, A. Tylki-Szymanska, Expert Opin. Orphan Drugs 2015,
, 293.
3] a) R. Peters, I. Louzao, J. C. M. van Hest, Chem. Sci. 2012, 3, 335;
b) P. Tanner, P. Baumann, R. Enea, O. Onaca, C. Palivan, W. Meier,
Acc. Chem. Res. 2011, 44, 1039.
2
3
[
[
[
4] N. Ben-Haim, P. Broz, S. Marsch, W. Meier, P. Hunziker, Nano Lett.
26] S. Muro, C. Garnacho, J. A. Champion, J. Leferovich, C. Gajewski,
E. H. Schuchman, S. Mitragotri, V. R. Muzykantov, Mol. Ther.
2
008, 8, 1368.
[
5] S. F. M. van Dongen, W. P. R. Verdurmen, R. J. R. W. Peters,
2
008, 16, 1450.
R. J. M. Nolte, R. Brock, J. C. M. van Hest, Angew. Chem., Int. Ed.
[
[
27] L. Hosta-Rigau, P. Schattling, B. M. Teo, M. E. Lynge, B. Stadler,
J. Mater. Chem. B 2014, 2, 6686.
28] M. E. Lynge, R. Ogaki, A. O. Laursen, J. Lovmand, D. S. Sutherland,
B. Städler, ACS Appl. Mater. Interfaces 2011, 3, 2142.
2
010, 49, 7213.
[
[
[
6] P. Tanner, O. Onaca, V. Balasubramanian, W. Meier, C. G. Palivan,
Chemistry – A Eur. J. 2011, 17, 4552.
7] F.-P. Chang, Y. Hung, J.-H. Chang, C.-H. Lin, C.-Y. Mou, ACS Appl.
Mater. Interfaces 2014, 6, 6883.
Received: July 16, 2015
Revised: December 19, 2015
Published online: February 8, 2016
8] a) L. Hosta-Rigau, B. Stadler, in Regenerative Medicine, Artifi-
cial Cells and Nanomedicine: Selected Topics in Nanomedicine,
1814 www.small-journal.com
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
small 2016, 12, No. 13, 1806–1814