Probing Intracellular Nanoscale Barriers
Fluorescent probes
Probing Cell-Type-Specific Intracellular Nanoscale
Barriers Using Size-Tuned Quantum Dots
Yvonne Williams,* Alyona Sukhanova, Małgorzata Nowostawska, Anthony
M. Davies, Siobhan Mitchell, Vladimir Oleinikov, Yurii Gun’ko, Igor Nabiev,
Dermot Kelleher, and Yuri Volkov
The compartmentalization of size-tuned luminescent semiconductor nanocrystal quantum dots (QDs) in four distinctive cell lines, which would be
representative of the most likely environmental exposure routes to nanoparticles in humans, is studied. The cells are fixed and permeabilized prior to the
addition of the QDs, thus eliminating any cell-membrane-associated effects
due to active QD uptake mechanisms or to specificity of signaling routes in
different cell types, but leaving intact the putative physical subcellular barriers.
All quantitative assays are performed using a high content analysis (HCA)
platform, thereby obtaining robust data on large cell populations. While
smaller QDs 2.1 nm in diameter enter the nuclei and localize to the nucleoli in
all cell types, the rate and dynamics of their passage vary depending on the cell
origin. As the QD size is increased to 4.4 nm, penetration into the cell is
reduced but each cell line displays its own cutoff size thresholds reflecting celltype-determined cytoplasmic and nuclear pore penetration specificity. These
results give rise to important considerations regarding the differential compartmentalization and susceptibility of organs, tissues, and cells to nanoparticles, and may be of prime importance for biomedical imaging and drugdelivery research employing nanoparticle-based probes and systems.
1. Introduction
The rules governing particle properties at the nanoscale
relate more to the laws of quantum mechanics than classical
[] Y. Williams, M. Nowostawska, Dr. A. M. Davies, Dr. S. Mitchell,
Prof. D. Kelleher, Prof. Y. Volkov
Department of Clinical Medicine
Trinity College Dublin
Dublin 8 (Ireland)
E-mail: [email protected]
Y. Williams
Children’s Research Centre
Our Lady’s Children’s Hospital Crumlin
Dublin 12 (Ireland)
Dr. A. Sukhanova, Prof. I. Nabiev
EA3798, University of Reims Champagne-Ardenne
Reims (France)
DOI: 10.1002/smll.200900744
small 2009, 5, No. 22, 2581–2588
Keywords:
cells
fluorescent probes
nanoparticle uptake
quantum dots
mechanics.[1] Biologists have been keen to harness nanoparticles not only for these properties but also because there are
now numerous engineered particles that are similar in size to
those occurring naturally in the environment.[2] The potential
M. Nowostawska, Dr. Y. Gun’ko
School of Chemistry
Trinity College Dublin
Dublin 2 (Ireland)
V. Oleinikov
Shemyakin-Ovchinnikov Institute of Bioorganic Chemistry
Russian Academy of Sciences
Moscow (Russia)
Dr. A. Sukhanova, Prof. I. Nabiev
CIC nanoGUNE Consolider Research Center
Donostia–San Sebastian (Spain)
ß 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2581
full papers
Y. Williams et al.
in vivo benefits of nanoparticles have been explored, including
delivery of site-specific drugs[3] and specific cell killing,[4] which
have long been the holy grail of cancer therapy. Nanoparticles
may also have a role in fundamental research as robust reagents
in bioimaging.[5–7] However, as recent reports have indicated, in
vivo applications have been beset with problems of stability and
toxicology.[8–11]
Furthermore, engineered nanoparticles are proving to
be useful tools in probing size-related barriers.[12,13] Along
with shape and charge, size plays an important part in how a
molecule is processed in any biological system. Nanoparticles
of size greater than 6 nm will accumulate in the liver, lungs,
and reticuloendothelial system, while smaller ones accumulate directly into the kidneys and bladder.[3] Recently, Soo
Choi et al. have shown that nanoparticles with a hydrodynamic diameter greater than 15 nm cannot be cleared by
the kidneys, while anything under 5 nm can be cleared quite
rapidly.[8] It has also been shown that nanoparticles of size
20–50 nm will travel straight to the lymph nodes, whereas
larger particles (>500 nm) need to be taken up by dendritic
cells and then brought to the lymph nodes for processing.[14]
Any foreign particle has the potential to stimulate the
immune response, another barrier. Depending on particle
size, a cell-mediated or antibody-mediated response can be
induced.[15] The mechanism of nanoparticle uptake by living
cells is largely unknown;[9] however, it is believed that
receptor-mediated endocytosis is the most likely scenario.[15]
It has been shown that cells of macrophage lineage engulf
nanoparticles within a few minutes of exposure, which
indicates phagocytosis of aggregated particles,[15] while
epithelial cells are more likely to take the particles in by
endocytosis in a size-dependent manner.[16] Particles taken
up by phagocytosis and contained in lysosomes can initiate
size-dependent signaling that induces programmed cell
death.[10] As early as 1986, the size-specific fenestrae in
liposomes were exploited for the passive delivery of cancertreating drugs.[17] Fluorescent nanoparticles, without conjugated biomolecules, can be taken up by human cells and
have unique cellular distribution patterns, which are largely
dependent on the particle size[13] and charge.[18] For a
particle to enter the nucleus there may be a transient increase
in nuclear pore size due to an accompanying signal-mediated
transport.[19] More recently, it has been shown that passive
diffusion of ions, metabolites, and other small molecules can
occur across the nucleocytoplasmic membrane reaching a
Michaelis–Menton-type equilibrium,[20] while only molecules
of >40 kD require active transport mechanisms.[21]
Utilization of fluorescent probes in tracking molecules
within cells has made a valuable contribution in cell biology
studies. However, conventional probes are prone to bleaching, interfering with cell biology processing, and have
different absorption bands, thus making multiplexing difficult.[7,22] Semiconductor nanocrystal quantum dots (QDs)
are proving to be very useful in vitro probes, for example, in
the study of endocytosis where accumulating evidence shows
that they are often able to mimic the behavior of viruses
within the cells.[23] Changing conditions during synthesis,
such as temperature, duration, and the addition of functional
groups, can influence the resulting size and shape of QDs.
Useful properties include photostability, broad absorption
bands, and signals that can be distinguished from those with a
shorter half-life, for example autofluorescence.[24,25] Multicolor QDs are useful fluorescent beacons as they can emit
light when excited by low-energy light that can be absorbed
harmlessly by living cells,[26] and because their narrow
emission wavelengths are very suitable for multiplexing
assays.[27]
We have previously shown variations in QD uptake by live
cells from lineages representative of those in vivo that would be
initially encountered by particles.[13] We have also demonstrated that the smaller QDs (2–3 nm) target histones in the
nucleoli of macrophages in a process involving endocytosis,
active cytoplasmic transport, and nucleocytoplasmic exchange
via the nuclear pore complex.[13,28]
Herein, by using QDs with a subtle but wide range of
nanosize variation, we demonstrate the existence of intracellular barriers specific to cell type, which confine discrete
particle penetration from plasma membrane to cytosol and
further towards the perinuclear space, into the nucleus, and
eventually the nucleoli. A unique opportunity to investigate
these events at the level of whole-cell populations is presented
by high content analysis (HCA) technologies. Cell-based HCA
utilizing fluorescent QDs enables unbiased quantitative
information to be obtained at the high-resolution level.[29]
Moreover, the process is very rapid with minimum exposure
time and simultaneous acquisition of multicolored emissions
from QDs, thus limiting possible photodamage to the cells.[30]
We have used CdTe and CdSe/ZnS QDs ranging in
size from 2.1 to 4.4 nm and in emission from 492 to 592 nm
(see Table 1), synthesized as described in our previous
Table 1. Summary of QDs, wavelength, diameter, fluorescence, and location within cells.
QD
Emission wavelength [nm]
Diameter [nm]
TEM
CdTe
CdTe
CdSe/ZnS
CdTe
CdSe/ZnS
CdSe/ZnS
CdSe/ZnS
492
536
542
580
562
582
592
2.1
3.1
3.3
3.4
3.7
3.9
4.4
[a]
DLS
2.6 0.1
3.6 0.1
3.8 0.1
4.0 0.1
4.3 0.2
4.6 0.2
5.4 0.1
Cell lines
THP-1
HEp-2
AGS
nucleoli
nucleus
nucleus
cytoplasm
cytoplasm
cytoplasm
plasma membrane
nucleoli
nucleus
cytoplasm
cytoplasm
plasma membrane
plasma membrane
negative
nucleus
nucleus
NT[b]
plasma membrane
NT
NT
NT
[a] DLS results are presented as the mean of triplicate measurements standard deviation. [b] NT: not tested.
2582 www.small-journal.com
ß 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
small 2009, 5, No. 22, 2581–2588
Probing Intracellular Nanoscale Barriers
Figure 1. Time-dependent penetration of differently sized CdTe QDs into intracellular compartments of AGS (blue), HEp-2 (green), and THP-1 (red) cell
lines. a) QDs 2.1 nm in diameter in the cell nucleoli measured in relative fluorescent units (RFU). b) The percentage of fluorescence intensity increases as
a function of time for penetration of 2.1-nm QDs in the nuclei. c) Penetration of 3.1-nm QDs in the nuclei. d) Penetration of 3.4-nm QDs in the cytoplasm.
reports,[6,31,32] to examine particle accumulation in four cell
lineages, namely THP-1 (macrophage) cells, HEp-2 (epithelial)
cells, AGS (gastric adenocarcinoma) cells, and A549 (lung
epithelial) cell line. These cells were chosen as representative of
in vivo sites that encounter incoming foreign particles. The cells
were seeded into 96-well plates and cultured under physiological conditions. By fixing the cells with the nondenaturing
fixative paraformaldehyde (PFA), which operates by crosslinking proteins,[33] and by permeabilizing the cell plasma
membrane with Triton X-100[34] prior to the addition of QDs,
we ensured that the barriers to particle localization were mainly
a function of size. While fixed cells undoubtedly represent a
different system compared to living cells, they could be referred
to as cells irreversibly ‘‘frozen’’ at a certain stage of their life
cycle, at the same time preserving their key morphological and
structural features. Moreover, we have previously shown that
the choice of fixative is crucial as it can influence the location of
QDs within the cell.[25] By using HCA with supporting software
to analyze the resulting images, we were
able to acquire extensive and robust data
reflecting the responses of each individual
cell within the population under study.
Accordingly, we obtained results that
give rise to important considerations
regarding the differential compartmentalization and susceptibility of organs, tissues,
and cells to nanoparticles, which may be of
prime importance for biomedical imaging
and drug-delivery research employing fluorescent semiconductor QDs.
2. Results and Discussion
2.1. Time-Course Assay
Figure 2. Intracellular distributions of CdTe QDs of different diameters. Distribution of QDs in
AGS (left column), HEp-2 (middle), and THP-1 cells (right). Row A, QDs of diameter 2.1 nm
(green), row B, 3.1 nm (yellow), and row C, 3.4 nm (red). Indicative QD exposure times are shown
under the images. Scale bar: 10 mm.
small 2009, 5, No. 22, 2581–2588
We initially looked at QD uptake in
the three cell lines over a 1 h time period
at intervals of 5, 10, 15, 20, 30, and 60 min.
For this experiment we used thiolcapped CdTe QDs of three different
diameters: 2.1 nm (emitting fluorescence
in the green region, lem ¼ 491.6 nm),
3.1 nm (yellow, lem ¼ 535.6 nm), and
ß 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.small-journal.com
2583
full papers
Y. Williams et al.
3.4 nm (red, lem ¼ 580 nm).[32] The data show that the
localization in the different compartments occurred at
different rates in each of the cells (Figure 1).
The smallest CdTe QDs (2.1 nm in diameter) entered all
three cell types quite rapidly and were localized to the nuclear
membrane of the HEp-2 and THP-1 cells within 10 min, while
taking 15 to 20 min in the AGS cells. After 60 min they had not
only penetrated the nuclei but had also appeared in the nucleoli.
The fluorescence intensity increased by 100% in AGS cells,
150% in HEp-2 cells, and 200% in THP-1 cells by 1 h after time
zero. Interestingly, HEp-2 nuclear fluorescence reached
maximum levels within 30 min (Figure 1B), thus demonstrating
the maximum QD penetration rate between the three studied
cell lines. The QDs accumulated in the nucleoli of both the
THP-1 cells (900 relative fluorescence units (RFU)) and the
HEp-2 cells (700 RFU), occurring again at an earlier time point
in the HEp-2 cells, that is, 20 min compared to the 60 min in the
THP-1 cells. The final fluorescence signal from QDs in the
nucleoli of the AGS cells was far less intense (200 RFU), clearly
indicating that the process of QD intranuclear penetration was
much slower in these cells (Figure 1A).
The slightly larger yellow QDs (3.1 nm in diameter)
penetrated all cells as rapidly as the smaller green QDs within
the first 30 min. However, they remained mainly in the
cytoplasm for all cell types, with some entering the nuclei
but not the nucleoli. By 60 min an increase in nuclear
fluorescence intensity of only 30% in AGS cells and 80% in
HEp-2 and THP-1 cells was noted when compared to initial
readings (Figure 1C).
Although the largest, red-fluorescence-emitting CdTe QDs
(3.4 nm in diameter) had concentrated at the perinuclear space
between 15 and 20 min in all the cell types, there was no further
increase in fluorescence intensity. Interestingly, the signal from
HEp-2 cells reached maximum intensity by 15 min, but then
there was a decrease in fluorescence, which suggests the
reversibility of QD accumulation (Figure 1D).
It would appear that the permeabilized plasma membrane
did not have any impact on the infiltration of nanoparticles,
while the nuclear membrane was very size specific. The
fluorescence from nanoparticles displayed a meshlike pattern
in the cytosol of THP-1 cells, which indicated likely retention at
the level of the endoplasmic reticulum barrier.[35] On the other
hand, a fainter diffuse homogeneous pattern was detected for
QDs in the HEp-2 cells similar to that seen when inserting QDs
by microinjection.[36] The concentration of QDs at the
perinuclear membrane within minutes has also been noted in
this cell line by other authors.[1,30] This would indicate that there
are components within the macrophage that bind strongly to the
nanoparticles slowing down their entry into the nucleus. These
components are likely not to be present in epithelial-like cells,
and this would be consistent with the scenario when the smaller
QDs enter the nucleus more rapidly and are subsequently
diffused out of the cytoplasm in epithelial cells. The homogeneous pattern is also seen in the AGS cells, which is not
surprising as these have a similar epithelioid lineage as the
HEp-2 cells. However, the nanoparticle progression into and
through the AGS is much slower than that of the other two cell
lines, which is probably due to the specific physiology of gastric
cells (Figure 2).[37]
2584 www.small-journal.com
Figure 3. Images of THP-1 and HEp-2 cells fixed with methanol and
incubated with 2.1-nm QDs. Left: grayscale fluorescence images of QD
distribution. Right: color overlay of QD fluorescence and nuclear staining.
Scale bar: 10 mm.
A similar cytoplasmic pattern could be seen when the cells
were prefixed with methanol prior to the addition of the green
QDs. It should be noted that fixation by methanol causes
proteins to precipitate and lipids to solubilize, thus destroying
the integrity of the nucleo–cytoplasmic pore membrane.[38]
Therefore, in contrast to the PFA and Triton X-100-treated
cells, no nuclear localization of QDs was registered under these
conditions (Figure 3).
We further analyzed the CdTe QD intracellular distribution
in live cells exposed to a mild detergent treatment with 0.0094%
Triton X-100 in THP-1 cells and lung epithelial cell line A549
(Figure 4). Such treatment preserves cell viability but increases
membrane penetration capacity for nanoparticles. As seen
from Figure 4, detergent treatment did not significantly alter the
ultimate QD localization patterns in these distinctive cell types,
Figure 4. Intracellular distribution of QDs within live detergent-treated cell
lines. Images of THP-1 macrophage cells (A, B) and A549 epithelial cells
(C,D)arepresented.Confocalfluorescenceimages(A,C)arecomparedwith
an overlay of the same microscopic field with transmitted light (B, D).
ß 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
small 2009, 5, No. 22, 2581–2588
Probing Intracellular Nanoscale Barriers
wavelength of their fluorescence emission
that ranged from the yellow (lem ¼ 542 nm)
to red (lem ¼ 592 nm) colors of the visible
spectrum. The size of one CdSe/ZnS QD
(3.3 nm, lem ¼ 542 nm) falls between the two
CdTe QDs (3.1 nm, lem ¼ 536 nm and
3.4 nm, lem ¼ 580 nm), and the other
CdSe/ZnS QDs with lem ¼ 562, 582, and
592 nm have the largest diameters of 3.7, 3.9,
and 4.4 nm, respectively (Table 1).
These subtle changes in QD size
showed that the QD distribution differed
substantially between the two cell types.
Differential localization was already occurFigure 5. Distribution of CdSe/ZnS QDs in THP-1 (A–D) and HEp-2 cells (E–H). Images of cells
ring between THP-1 and HEp-2 cells with
treatedwithQDsfluorescingatA,E)542 nm(3.3 nmindiameter),B,F)562 nm(3.7 nmindiameter), the QDs emitting at 542 nm. These QDs
C, G) 582 nm (3.9 nm in diameter), and D, H) 592 nm (4.4 nm in diameter). Scale bar: 10 mm.
were located in both the cytoplasm and the
nucleus of the THP-1 cells but only the
cytoplasm of the HEp-2 cells (Figure 5A,
E). As the size of the QDs increased, the
penetration of QDs into the cell diminished. The largest CdSe/ZnS QDs
(lem ¼ 592 nm) could only be seen on the
cell membrane of the THP-1 cells and were
entirely absent in the HEp-2 cells
(Figure 5D, H).
Using HCA, objective quantitative
values correlated well with the subjective
findings described above. The data from
both the CdTe and CdSe/ZnS QDs are
combined in Figure 6, which shows the
nuclear–cytoplasmic fluorescence intensity
differential that indicates quantitatively
the size-dependent penetration of the
QDs through the cytoplasm to the nucleus.
From this study we can say that QDs up to
3.1 nm in diameter can enter the nuclei of
Figure 6. Distribution of CdTe and CdSe/ZnS QDs in THP-1 and HEp-2 cells. The graph
demonstrates the nuclear–cytoplasmic fluorescence difference. Positive results indicate that all cell types. However, it would appear
that as QD size is increased, penetration
the QDs are located mainly in the nuclei, whereas negative results indicate cytoplasmic
distribution.
into the cell is reduced and each cell line
has its own cutoff size reflecting cell-typewhich showed a nuclear, nucleolar, and perinuclear distribution determined nuclear pore size specificity. Epithelial cells have
of green-emitting nanoparticles. This clearly indicates that the an important role as barriers and therefore would be
enhancement of membrane permeability per se does not impervious to the larger particles that could permeate into
change the fundamental QD intracellular transport mechan- macrophage or scavenger-type cells.
The nuclear pore has long been known to allow passive
isms, and their specific localization sites are predominantly
diffusion of particles of diameter less than 9 nm[39] while
dictated by the nature of a particular cell type.
particles of a greater size up to approximately 39 nm require
active transport.[40] However, the effective nuclear pore sizepenetration barrier needs to be revisited in relation to
2.2. Size-Tuned Assay
nanomaterials compared to biomolecule derivatives. Particle
We then proceeded to look more closely at a gradual aggregation may prevent the larger particles passing through
increase in size of QDs using the same CdTe QDs and, the pores, but there should still be the occasional single QD
additionally, CdSe/ZnS core/shell QDs that were capped with getting through although this has never been observed. The
many proteins that line the inner surface of the nuclear pore
DL-cysteine to render them soluble and stable in a physiological
environment.[31] As AGS cells were refractory to QD uptake complex form a selective transport barrier.[41] Therefore, while
within the one-hour timeframe, we opted to use just the THP-1 passive diffusion is the potential mechanism of QD uptake, the
and HEp-2 lineages for the remainder of the study. The sizes of fact that they are strongly charged means the true passive
the CdSe/ZnS QDs varied from 3.3 to 4.4 nm, as indicated by the mechanism can no longer be applied.
small 2009, 5, No. 22, 2581–2588
ß 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.small-journal.com
2585
full papers
Y. Williams et al.
3. Conclusions
By using HCA as a quick, quantitative, and objective
method of locating the position of QDs in cells, we have been
able to confirm that the size of QDs influences not only their
emission wavelength but also where they locate within cells. If
QDs are small enough they can get into the nucleus and
eventually will locate at the nucleoli in all cell types examined.
This affinity for the nucleoli is likely to be determined by the
interaction of the negatively charged QDs with histones as
previously described.[13,28] Although macrophages are more
capable of allowing a wider range of QD sizes across their
membranes, QDs travel through epithelial cells at a more rapid
rate. This fact suggests that the different cell lines differ in QD
transport through the cytoplasm, which is not related to
phagocytosis or endocytosis. The difference in the membrane
permeability among the cell lines may be a result of how the
detergent Triton X-100 interacts with the constitutive lipids of
the membranes.[42] Also, the nuclear pore complexes are
known to vary between different cell types,[21] and this too may
influence which QDs enter the nuclei and which do not. Our
results also support the suggestion that the uptake of QDs may
reach an equilibrium in epithelial cells[20] and that there may be
a lower affinity for QDs within the epithelial cell cytoplasm in
comparison to phagocytes. Therefore, since the size as well as
shape and charge of nanoparticles have a considerable
influence on bioactivity and toxicity,[10] it is important to
ensure that proper use can be made of the opportunity to
manipulate nanoparticles to explore the barriers within the
body and within the cell. Therefore, while there is an
accumulation of evidence showing that QDs may not be
suitable for in vivo applications and there is a reluctance to
continue with any medical research applying semiconductor
nanocrystals, these fascinating particles still have an important
role as robust and versatile probes for examining cellular events
at the nanoscale.
4. Experimental Section
Cell lines: The four cell lines used were: HEp-2 epithelial cell
line (ECACC, Salisbury, England) grown in minimum essential
medium (Eagle) with Earles salts (Sigma–Aldrich, Dublin, Ireland);
AGS endothelial cell line derived from a gastric adenocarcinoma
and lung epithelial cell line A549 (ECACC, Salisbury, England)
grown in F12 HAM medium; and THP-1 monocytic cell line (ECACC,
Salisbury, England) grown in RPMI 1640 medium. In all cases
the medium was supplemented with 10% fetal bovine serum,
1
1
L-glutamine (200 m M L ), penicillin (10 000 U mL ), and
1
streptomycin (10 mg mL ). The cells were seeded out into 96well microtiter plates to form a confluent monolayer. The THP-1
cells were co-cultured with phorbol 12-myristate 13-acetate
(100 ng mL1; Sigma–Aldrich, Dublin, Ireland) to enable monocyte
to macrophage differentiation.[43] All cells were incubated under
controlled atmospheric conditions of 37 8C, 5% CO2 until
confluency was reached (72 h for the THP-1 cell line and 48 h for
both HEp-2 and AGS cell lines). After incubation the cells were
2586 www.small-journal.com
washed in phosphate-buffered saline (PBS), fixed with 2% PFA,
washed again, and permeabilized with 1.5% Triton X-100 (Sigma–
Aldrich, Dublin, Ireland). The plates were washed again, left in
PBS, then sealed and stored at 4 8C until required. All assays were
carried out in triplicate. In the part of the study involving live cells,
following a 1-h treatment with a 0.0094% solution of Triton X-100,
cells were exposed to QDs for up to 3 h, fixed as above, and
immediately used for confocal microscopy analysis.
QDs: The CdTe QDs had a core of cadmium and telluride and
were capped with a stabilizer, thioglycolic acid.[32] The CdSe/ZnS
QDs had a core of cadmium and selenium with a zinc sulfide shell
and were treated with DL-cysteine to render them soluble and
stable in aqueous solution.[31] Such surface treatment with lowmolecular-weight mercapto compounds, instead of the generally
accepted encapsulation of nanoparticles within an additional
organic polymer shell, yielded water-soluble CdSe/ZnS QDs of the
smallest possible diameters.[8] CdTe QDs are easier and cheaper
to synthesize than CdSe/ZnS core/shell QDs, especially for use in
biological systems;[23] however, they are generally less stable
than CdSe/ZnS QDs.[7]
The ‘‘physical’’ size (or diameter) of the QDs was calculated
according to Peng et al. [44] and confirmed by transmission
electron microscopy (TEM) as described. [31] Hydrodynamic
diameters of QDs in aqueous solutions were determined by the
dynamic light scattering (DLS) approach. Light-scattering analysis
was performed with a Zetasizer Nano-ZS device from Malvern
Instruments using the protocols provided by the supplier. All stock
solutions (2 mM) of QD samples were prepared in distilled MilliQ
water and filtered through a 0.02-mm filter before analysis. Typical
count rates were around 200 kHz. Each autocorrelation function
(ACF) was acquired for 10 s and averaged for 10 min per
measurement. A software filter was employed to discard all ACF
fits with sum of square errors >15. Hydrodynamic diameters were
obtained from a mass-weighted size distribution analysis. Results
presented in Table 1 are the mean of triplicate measurements.
Assay protocol: The diluted QDs were added to cells resulting
in a final concentration of 0.1 mg mL1 and incubated at room
temperature for up to 1 h. Unincorporated QDs were removed by
washing in PBS, and then the cell nuclei were stained with
Hoescht 33342 (5 mg mL1; Molecular Probes, Karlsbad, CA) for
3 min, then washed with PBS and analyzed using a Cellomics
KineticScan instrument (Thermo Fisher, Pittsburgh, PA). Initial
examination was carried out with a Nikon Eclipse TE 300
epifluorescence microscope.
Image analysis: The images from the microtiter plates were
acquired using the Cellomics KineticScan instrument and analyzed
on the Cellomics Toolbox Scan with the Compartmental Analysis
Bioapplications. The samples were illuminated by a mercury/
xenon white light source. A quadruple-band fluorescence excitation filter (Omega XF93) was used in acquiring fluorescence, and
images were captured using a CCD Quantix camera 32.1 with a
40 objective. Each particular acquisition channel was allocated
with filters set to detect specific emission wavelengths (Table 2).
The number of QDs appearing in different parts of the cell was
represented by an increase in fluorescence intensity at the
corresponding emission wavelength. A user-defined gate was
applied to the nucleus (as indicated by Hoechst staining) in
channel 1 (lex (360 50), lem (515 20) nm) that identified
ß 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
small 2009, 5, No. 22, 2581–2588
Probing Intracellular Nanoscale Barriers
Table 2. Compartmental analysis algorithm as applied for QD cellular distribution.
Channel
Excitation/emission [l]
Spectra
Function
1 focusing channel[a]
360(50)/515(20)
blue
nuclear stain; defines
areas to gate for nucleus
and cytoplasm in
channels 2 and 3
2
475(40)/515(20)
green
information on
fluorescence distribution
and intensity in channel 2
3
560(15)/600(25)
red
information on
fluorescence distribution
and intensity in channel 3
Image
[a] In channel 1 (focusing channel), Hoescht dye fluorescence shows up the nuclei and defines the target object (nucleus). This gated area is then
used to locate the nucleus (blue outline) and the cytoplasm (yellow outline) in channels 2 and 3. Inclusions in the nucleus (purple outline) and
cytoplasm (turquoise outline) can also be located and quantified.
the areas relating to the nucleus and cytoplasm in channel 2
(lex (475 40), lem (515 20) nm) and channel 3 (lex (560 15),
lem (600 25) nm). Information could then be obtained on
parameters such as number, size, shape, and fluorescence
intensities on the cells of interest. Organelles, which were shown
as discrete fluorescent inclusions within the cytoplasm or nucleus,
could also be identified. The cell monolayer was located by
focusing on the nuclei within channel 1 (Hoescht), and the cells
were accepted or rejected depending on their size, shape,
intensity, and clumping.
All assays were carried out in triplicate and all results were
verified by epifluorescence microscopy, although analysis of the
red QDs by microscopy was difficult as the fluorescence faded
almost immediately. However, this was overcome by using
automated acquisition as fading was avoided due to the shorter
exposure time.
It should be noted that a shortcoming of the HCA technologies is
that in cells where staining was confined to the cytoplasm, falsely
raised levels of RFU were found in the nuclei. This was due to
a) dead cells that concentrated to the size of the nucleus and
b) cytoplasmic fluorescence overlaying the area of the nuclei. This
emphasizes the importance of visually examining the images as
well as quantitative data.
small 2009, 5, No. 22, 2581–2588
Acknowledgements
This work was supported by the Health Research Board of
Ireland (HRB), Science Foundation of Ireland, as part of the
BioNanoInteract Strategic Research Cluster, and by the FP-6
European Consortium NanoInteract. I.N. and A.S. were supported by the French National Research Agency (Agence
Nationale de Recherche–ANR) programs under the grants
ANR-07-PNANO-051-01 and ANR-07-RIB-012-03. Partial support
from the NATO SfP-983207, RFBR/CNRS (07-04-92164/
PICS3868), and RFBR grants is also acknowledged. I.N. was a
recipient of the Walton Award from the Science Foundation of
Ireland. A.S. was a recipient of the senior Marie Curie
fellowship from EC.
[1] W. J. Parak, D. Gerion, T. Pellegrino, D. Zanchet, C. Micheel,
S. C. Williams, R. Boudreau, M. A. Le Gros, C. A. Larabell,
A. P. Alivisatos, Nanotechnology 2003, 14, R15–R27.
[2] S. E. McNeil, J. Leukoc. Biol. 2005, 78, 585–594.
[3] M. Ferrari, Nat. Nanotechnol. 2008, 3, 131–132.
ß 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.small-journal.com
2587
full papers
Y. Williams et al.
[4] R. Weissleder, K. Kelly, E. Y. Sun, T. Shtatland, L. Josephson, Nat.
Biotechnol. 2005, 23, 1418–1423.
[5] P. Debbage, W. Jaschke, Histochem. Cell Biol. 2008, 130, 845–
875.
[6] A. Sukhanova, L. Venteo, J. Devy, M. Artemyev, V. Oleinikov,
M. Pluot, I. Nabiev, Lab. Invest. 2002, 82, 1259–1261.
[7] A. Sukhanova, I. Nabiev, Expert Opin. Med. Diagn. 2008, 2, 429–
447.
[8] H. Soo Choi, W. Liu, P. Misra, E. Tanaka, J. P. Zimmer, B. Itty Ipe,
M. G. Bawendi, J. V. Frangioni, Nat. Biotechnol. 2007, 25, 1165–
1170.
[9] M. A. Dobrovolskaia, S. E. McNeil, Nat. Nanotechnol. 2007, 2, 469–
478.
[10] W. Jiang, B. Y. Kim, J. T. Rutka, C. W. C. Warren, Nat. Nanotechol.
2008, 3, 145–150.
[11] L. Balogh, S. S. Nigavekar, B. M. Nair, W. Lesniak, C. Zhang,
L. Y. Sung, M. S. Kariapper, A. El-Jawahri, M. Llanes, B. Bolton,
F. Mamou, W. Tan, A. Hutson, L. Minc, M. K. Khan, Nanomedicine
2007, 3, 281–296.
[12] Y. H. Sun, Y. S. Liu, P. T. Vernier, C. H. Liang, S. Y. Chong, L. Marcu,
M. A. Gundersen, Nanotechnology 2006, 17, 4469–4476.
[13] I. Nabiev, S. Mitchell, A. Davies, Y. Williams, D. Kelleher, R. Moore,
Y. K. Gun’ko, S. Byrne, Y. P. Rakovich, J. F. Donegan, A. Sukhanova,
J. Conroy, D. Cottell, N. Gaponik, A. Rogach, Y. Volkov, Nano Lett.
2007, 7, 3452–3461.
[14] V. Manolova, A. Flace, M. Bauer, K. Schwarz, P. Saudan,
M. F. Bachmann, Eur. J. Immunol. 2008, 38, 1404–1413.
[15] S. D. Xiang, A. Scholzen, G. Minigo, C. David, V. Apostolopoulos,
P. L. Mottram, M. Plebanski, Methods 2006, 40, 1–9.
[16] F. Osaki, T. Kanamori, S. Sando, T. Sera, Y. Aoyama, J. Am. Chem.
Soc. 2004, 126, 6520–6521.
[17] Y. Matsumura, H. Maeda, Cancer Res. 1986, 46, 6387–6392.
[18] J. Lovric, H. S. Bazzi, Y. Cuie, G. R. Fortin, F. M. Winnik,
D. Maysinger, J. Mol. Med. 2005, 83, 377–385.
[19] C. M. Feldherr, D. Akin, R. J. Cohen, J. Cell Sci. 2001, 114, 4621–
4627.
[20] R. B. Kopito, M. Elbaum, Proc. Natl. Acad. Sci. USA 2007, 104,
12743–12748.
[21] L. J. Terry, E. B. Shows, S. R. Wente, Science 2007, 318, 1412–
1416.
[22] I. L. Medintz, H. T. Uyeda, E. R. Goldman, H. Mattoussi, Nat. Mater.
2005, 4, 435–446.
[23] Z. Deng, Y. Zhang, J. Yue, F. Tang, Q. Wei, J. Phys. Chem. B 2007,
111, 12024–12031.
[24] X. Michalet, F. F. Pinaud, L. A. Bentolila, J. M. Tsay, S. Doose, J. J. Li,
G. Sundaresan, A. M. Wu, S. S. Gambhir, S. Weiss, Science 2005,
307, 538–544.
2588 www.small-journal.com
[25] Y. Williams, S. Byrne, M. Bashir, A. Davies, A. Whelan, Y. Gun’ko,
D. Kelleher, Y. Volkov, J. Microsc. 2008, 232, 91–98.
[26] A. P. Alivisatos, W. Gu, C. Larabell, Annu. Rev. Biomed. Eng. 2005,
7, 55–76.
[27] A. Sukhanova, I. Nabiev, Crit. Rev. Oncol. Hematol. 2008, 68, 39–
59.
[28] J. Conroy, S. J. Byrne, Y. K. Gun’ko, Y. P. Rakovich, J. F. Donegan,
A. Davies, D. Kelleher, Y. Volkov, Small 2008, 4, 2006–2015.
[29] M. J. Ignatius, J. T. Hung, Methods Mol. Biol. 2007, 356, 233–244.
[30] J. K. Jaiswal, E. R. Goldman, H. Mattoussi, S. M. Simon, Nat.
Methods 2004, 1, 73–78.
[31] A. Sukhanova, J. Devy, L. Venteo, H. Kaplan, M. Artemyev,
V. Oleinikov, D. Klinov, M. Pluot, J. H. Cohen, I. Nabiev, Anal.
Biochem. 2004, 324, 60–67.
[32] N. Gaponik, D. V. Talapin, A. L. Rogach, K. Hoppe,
E. V. Shevchenko, A. Kornowski, A. Eychmuller, H. Weller, J. Phys.
Chem. B 2002, 106, 7177–7185.
[33] J. A. Kiernan, Microsc. Today 2000, 00-1, 8–12.
[34] S. Schuck, M. Honsho, K. Ekroos, A. Shevchenko, K. Simons, Proc.
Natl. Acad. Sci. USA 2003, 100, 5795–5800.
[35] E. Gagnon, S. Duclos, C. Rondeau, E. Chevet, P. H. Cameron,
O. Steele-Mortimer, J. Paiement, J. J. M. Bergeron,
M. Desjardins, Cell 2002, 110, 119–131.
[36] B. Dubertret, P. Skourides, D. J. Norris, V. Noireaux, A. H. Brivanlou,
A. Libchaber, Science 2002, 298, 1759–1762.
[37] S. C. Barranco, C. M. Townsend, Jr, C. Casartelli, B. G. Macik,
N. L. Burger, W. R. Boerwinkle, W. K. Gourley, Cancer Res. 1983, 43,
1703–1709.
[38] D. DiDonato, D. L. Brasaemle, J. Histochem. Cytochem. 2003, 51,
773–780.
[39] T. D. Allen, J. M. Cronshaw, S. Bagley, E. Kiseleva, M. W. Goldberg,
J. Cell Sci. 2000, 113, 1651–1659.
[40] N. Pante, M. Kann, Mol. Biol. Cell 2002, 13, 425–434.
[41] F. Alber, S. Dokudovskaya, L. M. Veenhoff, W. Zhang, J. Kipper,
D. Devos, A. Suprapto, O. Karni-Schmidt, R. Williams, B. T. Chait,
M. P. Rout, A. Sali, Nature 2007, 450, 683–694.
[42] E. London, D. A. Brown, Biochim. Biophys. Acta 2000, 1508, 182–
195.
[43] S. Tsuchiya, Y. Kobayashi, Y. Goto, H. Okumura, S. Nakae,
T. Konno, K. Tada, Cancer Res. 1982, 42, 1530–1536.
[44] W. W. Yu, L. Qu, W. Guo, X. Peng, Chem. Mater. 2003, 15, 2854–
2860.
ß 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Received: May 5, 2009
Published online: August 14, 2009
small 2009, 5, No. 22, 2581–2588