Deciphering the mechanisms of cellular uptake of engineered nanoparticles by accurate evaluation of internalization using imaging flow cytometry
© Vranic et al.; licensee BioMed Central Ltd. 2013
Received: 5 December 2012
Accepted: 3 February 2013
Published: 6 February 2013
The uptake of nanoparticles (NPs) by cells remains to be better characterized in order to understand the mechanisms of potential NP toxicity as well as for a reliable risk assessment. Real NP uptake is still difficult to evaluate because of the adsorption of NPs on the cellular surface.
Here we used two approaches to distinguish adsorbed fluorescently labeled NPs from the internalized ones. The extracellular fluorescence was either quenched by Trypan Blue or the uptake was analyzed using imaging flow cytometry. We used this novel technique to define the inside of the cell to accurately study the uptake of fluorescently labeled (SiO2) and even non fluorescent but light diffracting NPs (TiO2). Time course, dose-dependence as well as the influence of surface charges on the uptake were shown in the pulmonary epithelial cell line NCI-H292. By setting up an integrative approach combining these flow cytometric analyses with confocal microscopy we deciphered the endocytic pathway involved in SiO2 NP uptake. Functional studies using energy depletion, pharmacological inhibitors, siRNA-clathrin heavy chain induced gene silencing and colocalization of NPs with proteins specific for different endocytic vesicles allowed us to determine macropinocytosis as the internalization pathway for SiO2 NPs in NCI-H292 cells.
The integrative approach we propose here using the innovative imaging flow cytometry combined with confocal microscopy could be used to identify the physico-chemical characteristics of NPs involved in their uptake in view to redesign safe NPs.
KeywordsLung epithelial cells ImageStreamX Endocytosis Clathrin Macropinocytosis TiO2 SiO2
The number of current and future applications of nanotechnology is growing exponentially, making possible the production of many innovative high-quality items as well as a lot of promising biomedical implementations. Great efforts are made in order to understand and estimate the risk for human health coming from the unintentional occupational or environmental exposure to nanoparticles (NPs). Biodistribution studies in animals have shown that NPs accumulate not only at the site of exposure but can reach the circulation by crossing the epithelial barriers . In order to understand the mechanisms of transcytosis and potential NP toxicity to perform a reliable risk assessment it is of crucial importance to study NP uptake by epithelial cells. The capacity of NPs to pass the biological barriers has also let to foresee the application of nanotechnologies in the biomedical field for imaging and targeted drug delivery. The quantification of NP uptake is thus of crucial importance to evaluate the fate of tuned NPs in order to increase drug delivery by nanovectors, but also to develop NPs “safe by design”. Consequently, knowledge about the endocytic pathways involved in NP uptake is important to design safe NPs. Such issues are not simple since physico-chemical characteristics of NPs influence the interaction mechanisms leading to NP uptake [2–5].
Unfortunately, quantitative measurement of NP uptake still poses a technical challenge. It is still difficult to evaluate accurately real NP uptake because of the adsorption of NPs on the cell surface by interaction with the plasma membrane. This is one of the main limitations of spectroscopic methods that moreover are restricted to metallic particles. NPs can be observed inside the cells by transmission electron microscopy, but quantification of their uptake remains time and cost consuming . NP uptake is generally quantified by flow cytometric analysis of fluorescently labeled NPs,  or of NPs that have the capacity to scatter the light . In the literature, surprisingly, little attention has been paid to differentiate intracellular signal from the fluorescence or scattering due to NPs adsorbed on the cell surface, thus just providing a global signal.
In this article, we used fluorescent SiO2 NPs since they are easily labeled with different fluorochromes but have poor ability to scatter the light as well as non fluorescent TiO2 NPs as an example of NPs that are difficult to label with fluorochromes but that can well scatter the light. They are used in many industrial applications leading to the potential occupational exposures. Both NPs are also foreseen for nanomedical applications as drug delivery systems or in biomedical imaging. Our study was performed on a pulmonary epithelial cell line (NCI-H292) as the respiratory system is one of the most exposed to NPs.
We characterized the mechanisms of NP uptake by combining different approaches. Firstly, we focused on methods to distinguish adsorbed from internalized NPs. External fluorescence from adsorbed fluorescently labeled NPs was excluded by quenching using the vital dye Trypan Blue (TB) that is incapable of penetrating intact cell membranes. However, there is a whole panel of fluorochromes used to label NPs that cannot be quenched and this method did not allow analyzing fixed cells. Imaging flow cytometry, a newly developed technology, may allow not only determining the cellular localization of fluorescent, but also light diffracting, non fluorescent NPs . This innovative technique permitted us to define the inside of the cell by eroding the cell borders to determine the ratio of internalization and to study NP uptake on a large number of cells.
These two flow cytometric techniques allowed us to study the time course, dose-dependency as well as the influence of surface charges on the uptake. They were also used to examine the endocytic pathways involved in NP uptake firstly by a rather broad approach using energy depletion and pharmacological inhibitors and secondly more specifically by using siRNA induced gene silencing. These functional studies were strengthened by confocal microscope observations used also for colocalization studies of NPs with proteins specific for different endocytic vesicles.
We provide evidence that ImageStreamX, that associates flow cytometry and high resolution image analysis, allows an accurate evaluation of fluorescent or light diffracting NP uptake. By setting up an integrative approach to decipher the mechanisms of NP endocytosis, we showed that macropinocytosis is the predominant pathway of SiO2 NP internalization by NCI-H292 pulmonary epithelial cells and that TiO2 uptake is charge-dependent.
Results and discussion
Interactions of SiO2-NPs with NCI-H292 cells
Cells were treated with NPs at non cytotoxic concentrations (2.5 and 5 μg/cm2 for 50 nm-FITC-SiO2 NPs and 5 and 25 μg/cm2 for 100 nm-Por-SiO2 NPs) that were determined by WST-1 test (Additional file 1: Figures S1A and S1B in Supporting information).
Comparison with 50 nm-FITC-SiO2 NPs let us to conclude that the adsorption of NPs on the cell surface is NP dependent and should be carefully verified before interpretation of the results obtained by FCM. Numerous studies have shown interactions of fluorescent NPs with different cell lines by FCM and/or confocal microscopy [10–12] and adsorption of some SiO2 NPs on the cell surface has also been reported, [13–15] but have rarely been taken into account for the quantification of their uptake. Confocal microscopy permits to localize NPs, while FCM gives statistical quantification of the interactions by evaluation of MFI of the treated cells. This quantification allows a relative comparison between treatment conditions. By the analysis of NP-cell interactions with these two techniques in parallel we showed that this global analysis of cell fluorescence by FCM is not suitable to quantify the uptake in case of adsorbed NPs.
Elimination of the fluorescent signal from adsorbed NPs by quenching
Elimination of the fluorescent signal from adsorbed NPs by Imaging flow cytometry
Using these two parameters (IS and MFI inside the eroded cell mask) analyzed by ImageStreamX Ideas software we have obtained important information about NP uptake. Comparing IS values after treatment at 37° and 4°C allowed us to define an eroded cell mask to distinguish adsorbed from internalized part. Using MFI inside the eroded cell mask we could establish the time and dose-dependency of SiO2 NP internalization. Combining these two parameters allowed us to demonstrate the saturation of endocytic mechanisms of NPs.
Mechanisms of NP uptake: energy dependence
Concerning 100 nm-Por-SiO2 NPs, energy depletion with NaN3 induced the same decrease of the IS as the treatment at 4°C, suggesting complete inhibition of the uptake by NaN3 and thus no passive diffusion (Figure 7E). This is confirmed by the corresponding images obtained by the imaging flow cytometer (Figure 7E). The absence of passive diffusion of 100 nm-Por-SiO2 NPs can be explained by the bigger size of the NPs as it has already been shown that passive uptake of SiO2 is size dependent . Furthermore these NPs form aggregates that may not permit passive processes to occur.
Effect of pharmacological inhibitors
Figure 8B shows the inhibition of uptake of 100 nm-Por-SiO2 NPs studied by imaging flow cytometry. Negative values of the internalization score were obtained for CP, E and A suggesting low internalization after treatment with these inhibitors. After treatment with CP and E, the IS was similar to the IS observed at 4°C indicating complete inhibition of NPs internalization with these inhibitors. The second inhibitor for clathrin dependent endocytosis (MDC) did not prevent NP uptake neither did the caveolae dependent endocytosis inhibitors (N and F).
These results using pharmacological inhibitors indicate that both types of SiO2 NPs seem to be taken up by macropinocytosis and clathrin dependent endocytosis in NCI-H292 cells. However, for both SiO2 NPs studied we observed contradictory results with the two clathrin dependent endocytosis inhibitors as MDC had no effect on their uptake. Utilization of pharmacological inhibitors is a common approach for studying endocytic pathways, hence, their efficacy has been largely questioned [25, 27]. They can also lack specificity for defined pathways  and are shown to be cell type dependent . Balance should be found between the concentration of inhibitor high enough to inhibit endocytosis but not to induce cytotoxicity. For 100 nm-Por-SiO2 NPs the efficacy of Amiloride was lower than that of EIPA. This may be explained by different potencies of these two inhibitors as Amiloride blocks macropinocytosis when used at millimolar concentrations,  whereas EIPA is effective in the range of 50–100 μM[29, 30].
Effect of siRNA induced gene silencing
In order to confirm that SiO2 NPs uptake predominantly involves formation of macropinosomes and not clathrin or caveolae coated vesicles, we studied the colocalization of NPs with proteins that are specific for these vesicles. Cells were treated with SiO2 NPs during 4 h and observed for colocalization with clathrin heavy chain, caveolin-1 and sorting nexin-5 which is involved in macropinocytosis [37, 38].
These different approaches combining confocal microscopy and imaging flow cytometry using inhibitors or siRNA silencing reveal that both SiO2 NPs are taken up by macropinocytosis. SiO2 NP uptake has also been shown to be independent from clathrin and caveolae dependent endocytosis in A549 cells [7, 39]. However, Chung et al. analyzed mechanisms of silica NPs uptake in mesenchymal cells after inhibition of clathrin dependent pathway with Phenylarsine Oxide (PAO) and of macropinocytosis by Cytochalasin D and found that the uptake can occur not only by actin dependent but also by clathrin dependent endocytosis . However PAO is not specific for the clathrin pathways and has also been shown to inhibit macropinocytosis in adipocytes  and phagocytosis in mast cells . Meng et al. have demonstrated that macropinocytosis is implied in the internalization of silica NPs with high aspect ratio by small GTPase-dependent macropinocytosis mechanism . These studies support our conclusion that studies using inhibitors have to be analyzed with caution.
On the other hand these differences between studies on SiO2 uptake mechanisms may also be due to size variations of aggregates formed. In the study of Gyenge et al. TEM revealed that single silica NPs were most likely internalized by clathrin coated pits, while larger aggregates of NPs were internalized by membrane ruffling characteristic for macropinocytosis . These findings are in agreement with the size of our NP aggregates evaluated by DLS analysis showing 1μm and 300 nm aggregates for 100 nm-Por-SiO2 NPs and 50 nm- FITC-SiO2 NPs respectively. Clathrin coated vesicles have indeed diameters of about 120 nm while macropinocytotic vesicles can accommodate molecules of more than 1 μm size. Although recent publications emphasize the size of NPs as a critical factor for entering the respective uptake pathways,  it cannot be applied as a general rule as in some studies it has been shown that larger NPs (500 nm) can be taken up exclusively by caveolae-dependent pathway while these vesicles are not bigger than 80 nm .
Applying of the imaging flow cytometry to study internalization of non fluorescent NPs
Until now internalization of non fluorescent NPs has been based on global estimations of side scatter values obtained by flow cytometry, [8, 46] an approach that cannot discriminate between adsorbed and internalized NPs. The same stands for studying internalization by different quantitative spectroscopic or chemical detection methods . On the other hand estimations obtained by electron microscopy remain complex to perform, time consuming and rather qualitative . The approach using ImageStreamX platform we have developed here for the first time can be helpful in fulfilling critical need for the uptake assessment of TiO2 or other non fluorescent but light scattering NPs and can be implemented in a battery of screening tests, especially as the cytotoxic effects of NPs could be closely linked to their internalization. The efficient inhibition of NP uptake assessed with ImageStreamX may allow determining whether adverse effects are dependent on NP internalization.
An accurate evaluation of NP uptake is necessary for a better understanding of NP toxicity. We have demonstrated that identification of the mechanism of NP uptake has to be studied by different approaches as all of them have their limitations. One of the most complete approaches to study NP uptake is a combination of confocal microscopy for determining the localization of NPs and FCM for a statistically reliable quantification of uptake. However, to estimate the internalized fraction of applied NPs it should be taken into consideration that certain nanomaterials tend to firmly adsorb onto cellular membranes. This highlights the importance to discriminate between internalized and extracellular particles when assessing particle uptake. Using ImageStreamX, which combines FCM with high resolution microscopy, we propose a new method to distinguish between adsorbed and internalized NPs that can be applied not only to fluorescent NPs, but also to NPs that scatter light.
Applying this method we have shown that the uptake of SiO2 NPs by NCI-H292 cells is a time and dose-dependent process that can be saturated. Moreover in combination with qualitative observations by confocal microscopy, we have demonstrated that this internalization is predominantly an energy dependent process but that 50 nm-FITC-SiO2 can also be taken up by passive diffusion. The use of pharmacological inhibitors of endocytosis suggests that SiO2 NPs are taken up by macropinocytosis and the clathrin dependent pathway. However, using more specific siRNA-clathrin heavy chain silencing, definitive evidence was found that clathrin mediated endocytosis is not involved in the uptake of SiO2 NPs by NCI-H292 cells. This also underlines the importance of a cautious interpretation of results obtained with pharmacological inhibitors. Experiments of NP colocalization with proteins specific for vesicles of each of three principal endocytic pathways further confirmed the involvement of macropinocytosis in SiO2 NP uptake by NCI-H292. Adapting imaging FCM to non fluorescent TiO2 NPs allowed us to demonstrate the influence of surface charges on their uptake which is reduced by the presence of negative charges.
Our study is the first to propose an integrative approach to characterize NP uptake by combining functional studies with confocal microscopy and the imaging flow cytometer ImageStreamX. This powerful detection system allows distinguishing between adsorbed and internalized fluorescent as well as non-fluorescent NPs. ImageStreamX provides thus a reliable statistical quantification of the intracellular signal and the ratio of internalization based on the analysis of a large population of living or even fixed cells. This integrative approach could be used to identify the physico-chemical characteristics of NPs involved in their uptake and the role of the internalization in their adverse effects in view to redesign safe NPs.
Fluorescent SiO2 NPs of two sizes (50 and 100 nm) were synthesized. Fluorochromes were incorporated inside the NP core during synthesis. 50 nm-FITC-SiO2-NPs were labeled with fluorescein isothiocyanate (FITC) that is characterized by an excitation wavelength at 488 nm and maximum emission wavelength at 520 nm. The NPs were synthesized as described in Supporting Information (section 12). 100 nm-Por-SiO2-NPs were synthesised as already described . They were coupled with 5, 10, 15, 20 -Tetrakis-(1-methyl-4pyridino) porphyrine tetra (toluene-4-sulfonate) or shortly porphyrine characterized by an excitation wavelength at 422 nm and maximum emission wavelengths at 666 nm and 716 nm. Non fluorescent TiO2 NPs of 10 nm size with neutral, positive and negative surface charges were synthesized as described previously .
Nanomaterial characterization (TEM, DLS, fluorescence excitation/emission spectrums)
Morphology and aggregation of SiO2 NPs were verified by Transmission Electron Microscopy (TEM, JEOL 1200 EXII (OXFORD LINK ISIS 300)) as well as by Dynamic Light Scattering (DLS) analysis in RPMI 1640 cell culture medium (Life Technologies). For DLS characterization, NPs were prepared at the highest concentrations used in experiments. DLS values as well as the values of zeta potential were measured by Zetasizer (nano ZS, Malvern Instruments, USA). NP excitation/emission spectrums were verified by confocal laser scanning microscope (Zeiss 710 confocal microscope). All the results are presented in the Supporting Information (sections 13, 14 and 15). Non fluorescent TiO2 NPs of 10 nm size were characterized as described previously .
Human lung adenocarcinoma (NCI-H292) cells were purchased from the American Type Culture Collection (Sigma-Aldrich, Saint Quentin Fallavier, France) and grown in RPMI 1640 (Roswell Park Memorial Institute) medium without phenol red (Life Technologies), containing 10% fetal calf serum (FCS, Life Technologies) and 1% glutaMAXTM (Life Technologies), subsequently referred to as complete cell culture medium. The NCI-H292 cell line was derived from a lymph node metastasis of a pulmonary mucoepidermoid carcinoma. All experiments were performed with cells from passages 13 to 20. Cells were grown in T75-flasks (Costar, Sigma) as a monolayer. Exponentially growing cultures were maintained in a humidified atmosphere of 5% CO2 and 95% air at 37°C and were passaged twice weekly using 0.05% Trypsin-EDTA (Life Technologies) whose action was stopped with 10% FCS.
Depending on experiment, cells were seeded into cell culture plates (Costar, Sigma) and treated when they had reached 70-80% confluence. Before NP exposure, cells were rinsed with PBS to eliminate trace amounts of FCS. Treatments were performed in the absence of FCS as it has been found previously that the serum can form a protein corona modulating NP uptake and as the bronchial cells are not directly exposed to serum proteins conversely to internal organs. 50 nm-FITC-SiO2-NP stock (23.2 mg/mL in water) was vortexed shortly before making final dilution for the treatment. 100 nm-Por-SiO2-NPs stock suspension (5.25 mg/mL in water) was sonicated in an ultrasonic bath (Branson Cleaner, Ultrasonic, B200) at 20 W for 10 min then vortexed before making dilutions for treatment. TiO2 NPs stock suspension (2.56 mg/mL in water) was sonicated at 60 W (Ultrasonic processor, Bioblock Scientific) for 8 min. Cells were exposed for different times with either FITC labeled SiO2 NPs at 2.5 and 5 μg/cm2, porphyrine-labeled SiO2 NPs at 5, 15 and 25 μg/cm2 or non fluorescent TiO2 NPs at 20 and 40 μg/cm2.
For endocytic inhibition experiments, cells were pre-incubated for 30 min with different endocytic inhibitors at the following concentrations: chlorpromazine (CP) 25 μM, monodansylcadaverine (MDC) 75 μM, EIPA 75 μM, amiloride (A) 1.5 mM, nystatine (N) 75 μM and filipine (F) 4.5 μM (all from Sigma). Energy dependence experiments were performed by pre-incubating the cells at 4°C or with sodium azide (NaN3, 100 mM, Sigma) for 30 min prior to exposure to NPs. After these pre-incubations, NPs were added to cell cultures and incubated for 3.5 h, either in the presence of drugs or at 4°C.
Flow cytometry (FCM) analysis
Cells were seeded in 12-well plates at 10,000 cells/cm2 in complete cell culture medium and incubated for 48 h before treatment. After treatment with 1.2 mL/well of 50 nm-FITC-SiO2-NPs at indicated concentrations, medium was removed, cultures were thoroughly washed with PBS and cells were harvested by trypsination whose action was stopped with 10% FCS. Shortly before FCM analysis cells were incubated with 0.11% Trypan Blue for 1 min in order to quench the FITC-fluorescent signal coming from NPs adsorbed to the cell surface. Cell-associated fluorescence was detected using a CyAn ADP LX (Dako Cytomation, Beckman Coulter, Villepinte, France) flow cytometer. Laser excitation and emission bandpass wavelengths were 488 nm and 575 ± 25nm respectively. Minimum of 10,000 cells was analyzed after exclusion of the cellular debris from the analysis by gating on the 575 nm Log versus FS area graph. The results are reported as the median of the distribution of cell fluorescence intensity obtained by analyzing 10,000 cells in the gate.
Imaging flow cytometry analysis
Cells were seeded in 6-well plates at 10,000 cells/cm2 in complete cell culture medium and incubated for 48 h before treatment with 2.9 mL/well of 100 nm-Por-SiO2-NPs at indicated concentrations. At the end of the exposure to NPs, the media was removed, cells were thoroughly washed with PBS and cells were harvested. Cell suspension was centrifuged for 5 min at 200 g at 4°C and the cell pellet was resuspended in 500 μl of 4% para-formaldehyde (PFA, Santa Cruz Biotechnology Inc, Heidelberg, Germany). After 20 min of incubation in PFA, cells were rinsed three times with PBS and finally suspended in at least 50 μl of PBS. At least 2,500 cells were analyzed using Amnis ImageStreamx platform (Amnis, Proteigene, Saint Marcel, France) and InspireTM system software (Amnis). Camera magnification was 40x, 488 nm excitation laser was at 100 mW and 785 nm excitation laser was at 2.33 mW, except for non fluorescent NPs where it was set at 0.01 mW. The images were acquired with a normal depth of field, providing a cross-sectional image of the cell with a 4 μm depth of focus. A mask representing the whole cell was defined by the bright-field image, and an internal mask was defined by eroding the whole cell mask by 6 pixels (equivalent to 3 μm, as the size of 1 pixel is 0.5 μm) in order to eliminate the fluorescent signal coming from the NPs attached to the cell surface thus measuring only the internalized part. The results are analyzed by IDEAS software (Amnis),. Values of the internalization score, mean fluorescence intensity and mean side scatter intensity were calculated for at least 500 cells per sample.
Cells were seeded in 8 well Lab-TekTM chambered coverglasses (Nunc, Thermo Scientific, Dominique Dutscher, Brumath, France) at 40,000 cells/well in complete cell culture media. After treatment with 0.22 mL/well of NPs at indicated concentrations, cells were fixed in 4% PFA for 20 min at 25°C, rinsed 3 times with PBS and incubated for 10 min with NH4Cl (50 mM, Sigma) before permeabilization in 0.05% PBS-Tween20 (Sigma). To stain the actin filaments, cells were incubated for 30 min with FITC or TRITC-phalloïdine (0.9 nM in PBS,), then rinsed 4 times in PBS. Cells were mounted in Polyvinyl alcohol mounting medium with DABCO® (Sigma).
For immunolabelling experiments, cells were fixed with methanol at −20°C for at least 20 min, and rinsed three times with PBS. After permeabilization in 0.05% PBS-Tween20 and saturation in 0.01% PBS-Tween20–3% Bovine Serum Albumin (BSA, Sigma) cells were incubated for 60 min with goat polyclonal clathrin heavy chain antibody, C-20 (sc-6579, 1:50, Santa Cruz), goat polyclonal SNX5 D-18 (sc-10625, 1:50, Santa Cruz) or rabbit polyclonal caveolin1 N-20 (sc-894, 1:50, Santa Cruz) antibodies in 0.01% PBS-Tween20-3% BSA. Secondary anti goat antibodies Alexa fluor 488-IgG or 647-IgG (Life Technologies), anti rabbit antibodies Alexa fluor 488-IgG or 546-IgG (Life Technologies) were diluted in 0.01% PBS-Tween20-3% BSA at 1:400 and incubated for 45 min. Cell nuclei were stained with DAPI (4',6-Diamidino-2-Phenylindole, Dihydrochloride, Sigma, 0.25 μg/mL in PBS) for 1 min. Cells were examined under a Zeiss 710 confocal microscope using 63x objective (NA of 1.4) and a 1 to 1.5x zoom. The refractive index of immersion oil was 1.512. Considering optical laws the theoretical resolution was calculated and instrument settings adapted to obtain the best possible resolution in our images. Image treatment was done with Image J software (Image J 1.42 NIH, USA). The three-dimensional (3D) structure of the cells treated with NPs was reconstructed from corresponding confocal images using IMARIS software 7.5 (Bitplane). Pearson’s correlation coefficient was calculated using JACoP (Just Another Colocalization Plugin) for images with lateral resolution of 0.09 μm and axial resolution of 0.5μm.
siRNA knockdown experiments
The expression of clathrin heavy chain (CHC) protein in NCI-H292 cells was knocked down by transfection with specific siRNA targeted against this protein (SI00299880, Qiagen, Courtaboeuf, France). Cells were seeded in 6-well plates at 15,000 cells/cm2 in complete culture medium. 24 h after seeding cells were treated with 100 nM siRNA-control or siRNA-clathrin heavy chain using 18 μl/well of Hiperfect transfection reagent (Qiagen) as indicated in the transfection protocol provided by the supplier. Cells were retransfected every 24 h until 72 h when transfection efficiency was monitored by confocal microscopy and Western blot. After treatment with siRNA, cells were incubated with NPs during 3.5 h.
Every experiment was repeated at least twice with duplicates or triplicates of each condition. Data are represented as means ± SD or SEM and were analyzed on commercially available software SigmaStat (version 3.0, Systat software Inc, San Jose, California, USA) analysis of variance (one-way ANOVA) followed by Bonferroni post hoc test for multiple comparisons with p < 0.05 (two tailed) considered as significant.
This study was funded by the E.C. FP7 ENPRA (n°228789) grant. Support for this study was also provided by Nanotrans (N°EST-2010/2/079) and TiSiTrans (N°PNR-EST-2010/2/79) grants. We acknowledge the confocal microscope platform in the Institute Jacques Monod, Paris, France. The authors wish to thank Dr Tad George for fruitful scientific discussions in the field of imaging flow cytometry. We acknowledge Dr Florent Busi for helping preparation of the manuscript, Kevin Moreau and Emile Petit for technical assistance and Pierre Hugounenq for DLS analysis. We wish to thank Jonathon Hauser for proof reading the manuscript.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License ( http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.