1 23

Environmental Science and PollutionResearch ISSN 0944-1344 Environ Sci Pollut ResDOI 10.1007/s11356-012-1436-5

Internalization of SiO2 nanoparticles byalveolar macrophages and lung epithelialcells and its modulation by the lungsurfactant substitute Curosurf®

Sandra Vranic, Ignacio Garcia-Verdugo,Cécile Darnis, Jean-Michel Sallenave,Nicole Boggetto, Francelyne Marano,Sonja Boland, et al.

1 23

Your article is protected by copyright and

all rights are held exclusively by Springer-

Verlag Berlin Heidelberg. This e-offprint is

for personal use only and shall not be self-

archived in electronic repositories. If you

wish to self-archive your work, please use the

accepted author’s version for posting to your

own website or your institution’s repository.

You may further deposit the accepted author’s

version on a funder’s repository at a funder’s

request, provided it is not made publicly

available until 12 months after publication.

ECOTOXICOLOGYAND ENVIRONMENTAL TOXICOLOGY : NEW CONCEPTS, NEW TOOLS

Internalization of SiO2 nanoparticles by alveolar macrophagesand lung epithelial cells and its modulation by the lungsurfactant substitute Curosurf®

Sandra Vranic & Ignacio Garcia-Verdugo & Cécile Darnis &

Jean-Michel Sallenave & Nicole Boggetto & Francelyne Marano &

Sonja Boland & Armelle Baeza-Squiban

Received: 14 September 2012 /Accepted: 13 December 2012# Springer-Verlag Berlin Heidelberg 2012

Abstract Because of an increasing exposure to environ-mental and occupational nanoparticles (NPs), the potentialrisk of these materials for human health should be betterassessed. Since one of the main routes of entry of NPs is viathe lungs, it is of paramount importance to further charac-terize their impact on the respiratory system. Here, we havestudied the uptake of fluorescently labeled SiO2 NPs (50 and100 nm) by epithelial cells (NCI-H292) and alveolar macro-phages (MHS) in the presence or absence of pulmonarysurfactant. The quantification of NP uptake was performed

by measuring cell-associated fluorescence using flowcytometry and spectrometric techniques in order to identifythe most suitable methodology. Internalization was shown tobe time and dose dependent, and differences in terms ofuptake were noted between epithelial cells and macro-phages. In the light of our observations, we conclude thatflow cytometry is a more reliable technique for the study ofNP internalization, and importantly, that the hydrophobicfraction of lung surfactant is critical for downregulating NPuptake in both cell types.

Keywords NCI-H292 . Mouse alveolar macrophages .

Uptake . Flow cytometry . Microplate reader . BrilliantBlack . Quenching . Silica nanoparticles

Introduction

Nanotechnologies are emerging technologies which allowmanufacturing of molecular structures smaller than 100 nmin at least one dimension, which are referred as nanoparticles(NPs). These nanometric structures have characteristics thatare different from those of same material at microscopic sizesuch as higher surface reactivity and the acquisition of in-creased catalytic potential (Oberdörster et al. 2005). Thesecharacteristics endow NPs with new specific properties allow-ing them to be broadly applied in the fields of medicine,biotechnology, informatics, and telecommunications.

As the exploitation of nanomaterials increases, potentialrisks coming from environmental and occupational expo-sure should be assessed. Indeed, concern is justified by thehigh reactivity of NPs compared with ultrafine particlespresent in atmospheric pollution which are similar in size(Oberdörster et al. 2005).

Responsible editor: Philippe Garrigues

S. Vranic (*) : C. Darnis : F. Marano : S. Boland :A. Baeza-SquibanLaboratory of Molecular and Cellular Responses to Xenobiotics,Unit of Functional and Adaptive Biology (BFA) EAC CNRS 4413,Sorbonne Paris Cité, Univ Paris Diderot, 5 rue Thomas Mann,75013 Paris, Francee-mail: sandra.vranic@univ-paris-diderot.fr

I. Garcia-Verdugo : J.-M. SallenaveUnité de Défense Innée et Inflammation, Institut Pasteur,25 rue du Dr Roux,75015 Paris, France

I. Garcia-Verdugo : J.-M. SallenaveINSERM U874, Institut Pasteur, 25 rue du Dr Roux,75015 Paris, France

I. Garcia-Verdugo : J.-M. SallenaveSorbonne Paris Cité, Cellule Pasteur, Univ Paris Diderot,rue du Dr Roux,75015 Paris, France

N. BoggettoInstitut Jacques Monod, CNRS, UMR 7592,ImagoSeine BioImaging Core Facility, Sorbonne Paris Cité,Univ Paris Diderot, 75205 Paris, France

Environ Sci Pollut ResDOI 10.1007/s11356-012-1436-5

Author's personal copy

One of the main routes of entry of NPs is the respiratorysystem. Because of their size, NPs deposit in differentregions of the lungs: NPs of 5 nm in diameter all along therespiratory tract, those over 5 nm mainly in the alveoli,while those below 5 nm deposit in the upper airways(Oberdörster et al. 2005). After their entry in the respiratorysystem, NPs encounter two main cell types: lung epithelialcells and macrophages. Naïve and intact lung epithelial cellsbecause of the presence of a network of tight junctions act asa physical barrier and prevent the paracellular entry ofmicroorganisms, dusts, and particles. However, the integrityof this barrier can be compromised in the case of epitheliumdamage. In addition, phagocytic cells are critical actors forparticle clearance as they phagocytose microorganisms anddusts that reach distal parts of the lungs. Among these,airway macrophages, present in the upper mucosae, and alve-olar macrophages (AM), present in the alveoli, are particularlyimportant, with a special emphasis on AM, which are in closecontact with pulmonary surfactant (see below). In addition,both epithelial cells and macrophages respond to particleexposure by secreting inflammatory cytokines (Mohamed etal. 2011; Nel et al. 2006; Wu et al. 2010; Xia et al. 2008).Furthermore, surfactant, an alveolar-derived lining fluid incontact with AMs and epithelial cells, is instrumental inavoiding alveolar collapse. The latter is mainly performedby the lipid component (90 %) of surfactant whereas associ-ated proteins SP-A and SP-D (the remainder 10 %) are mainlyinvolved in immune defence (Chroneos et al. 2010).

One of the main properties of NPs in living systems istheir capacity to go through the epithelial barrier and enterthe systemic circulation. Indeed, biodistribution studies afterNP inhalation have shown minimal NP leaking into thesystemic circulation and have demonstrated mainly accumu-lation in the lungs (Schleh et al. 2012; Semmler-Behnke etal. 2007, 2008). To better understand the potential translo-cation or biopersistence of NPs in the lungs, it is importantto understand the mechanisms of NP uptake. This uptake isgenerally an energy-dependent process called endocytosisthat is subdivided into phagocytosis and pinocytosis.Phagocytosis consists of internalization of large particlesand is performed by professional phagocytes such as macro-phages and dendritic cells. Pinocytosis is defined as theinternalization of fluids by formation of vesicles of differentsizes but can also involve particles. Internalization of NPscan also take place by passive diffusion, as it has beenshown in energy-depleted cells treated with metabolicinhibitors as well as in cells exposed to NPs at lowtemperatures. In addition, red blood cells, although con-sidered as non-phagocyting cells (they lack phagocyticreceptors on their surface and have no actin-myosinsystem) have also been shown to interact with NPs(Geiser et al. 2005; Kruth 2002; Mu et al. 2012;Rothen-Rutishauser et al. 2006).

Importantly, NP uptake is probably dependent on their sur-face properties. In biological fluids, proteins associate with NPsbecause of their high surface energy, forming the so-calledcorona that has different properties and could lead to a differentresponse, compared to an uncoated particle (Casals et al. 2010;Lesniak et al. 2012; Lundqvist et al. 2011). The composition ofthe corona depends on the biological system that is exposed toNPs. In the lungs, NPs will likely interact with surfactant beforereaching target cells. It has already been shown that NPsmodify the capacity of surfactant to regulate surface tension atthe alveolar air–liquid interface (Schleh et al. 2009). However,the ability of lung surfactant to modulate NP uptake and fate isstill an important topic, and several questions remain open.

In this manuscript, we have studied the internalization ofSiO2 NPs of two different sizes (50 and 100 nm) in two lungcell types: human lung epithelial cells (NCI-H292) and mousealveolar macrophage cells (MHS) in presence or absence ofCurosurf®—a commercially available natural porcine lungsurfactant that contains the hydrophobic fraction of the com-plete surfactant. We have used fluorescently labeled SiO2 NPsto investigate the interactions and the localization of NPs incells by confocal microscopy. The quantification of NP uptakewas performed using different methodologies to identify themost suitable tool. Fluorescence intensity was measured byflow cytometry and spectrophotometry microplate reading toallow the measurement of individual fluorescence of a largenumber of cells or that of the global fluorescence of the wholecell population, respectively.

Materials and methods

Cell culture

Human lung mucoepidermoid carcinoma cells (NCI-H292)and murine macrophage cell line (MHS) were purchased fromthe American Type Culture Collection (Sigma-Aldrich, SaintQuentin Fallavier, France). Cells were grown in RPMI 1640medium without phenol red (Life Technologies, Saint Aubin,France), containing 10 % fetal calf serum (FCS, LifeTechnologies), 1 % glutaMAXTM (Life Technologies), 1 %penicillin and streptomycin (Life Technologies), and 0.5 %fongizone (Life Technologies) referred as a complete cellculture medium. Complete cell culture medium for MHS cellline was furthermore supplemented with 10 mM HEPES(Sigma) and 1 mM pyruvate (Sigma).

Cells were grown in T75 flasks (Costar, Life Technologies)as a monolayer. Exponentially growing cultures were main-tained in a humidified atmosphere of 5 % CO2 and 95 % air at37 °C and were passaged twice weekly using 0.05 % trypsin-EDTA (Life Technologies) whose action was stopped with10 % FCS. All experiments were performed with cells frompassages 13 to 20.

Environ Sci Pollut Res

Author's personal copy

Surfactant

We used poractant (Curosurf®, Chiesi Farmaceutici, Parma,Italy) as seminatural pulmonary surfactant of porcine origin.This surfactant is composed of phospholipids (80 mg/mL),surfactant protein B (SP-B), and SP-C (1 % by weight) asindicated by the manufacturers.

NP synthesis and characterization

SiO2 NPs of two different sizes were used: 50 and 100 nmNPs. These NPs were coupled to different fluorochromes inorder to facilitate the measurement of NP–cells interactions.50 nm SiO2 NPs were synthesized following a slightlymodified method described by Van Blaaderen and Vrij(1992) and were thoroughly characterized as already de-scribed (Vranic et al., submitted). Briefly, 50 nm fluoresceinisothiocyanate (FITC)-SiO2 NPs made small agglomeratesin water (258 nm). In the cell culture medium with orwithout lung surfactant, hydrodynamic diameter of theseNPs was 242.7 and 210 nm, respectively. Potential zeta inwater was 12.8 mV. In RPMI cell culture medium, it shiftedfrom −9.8 mV in the absence of surfactant to 2.4 mV in itspresence. FITC was sequestrated in the 50 nm SiO2 NPs,allowing detection by excitation at 488 nm, with maximumof emission at 522 nm. Stock solution of NPs at 5.1 mg/mLwas vortexed before making final dilutions for treatments.

One hundred nanometers SiO2 NPs were synthesized asalready described (Sergent et al. 2012). They were coupled to5,10,15,20-tetrakis-(1-methyl-4-pyridino) porphyrin tetra (tol-uene-4-sulfonate) or shortly porphyrin (Por). Hydrodynamicdiameter and potential Zeta of these NPs in water and RPMI inthe absence and presence of lung surfactant were verified.Briefly, 100 nm Por-SiO2 NPs made larger agglomerates inwater (942 nm) as well as in the cell culture medium in theabsence of surfactant (1,216 nm) and in its presence(1,250 nm). Potential zeta in water was −15.2 mV. In RPMIcell culture medium, it shifted from −20.8 mV in the absenceof surfactant to −1.2 mV in its presence. Stock solution of100 nm Por-SiO2 NPs at 5.25 mg/mL was sonicated in anultrasonic bath (Branson Cleaner, Ultrasonic, B200) at 20 Wfor 10 min then vortexed before making dilutions fortreatments.

Cell culture treatment

Depending on the experiment, NCI-H292 and MHS cellswere seeded into cell culture plates (Costar, LifeTechnologies) and treated when reaching 70–80 % conflu-ence. Before treatment with NPs, cells were rinsed withphosphate-buffered saline (PBS) and incubated with RPMImedium for 1 h in order to eliminate trace amounts of serumthat could modulate the uptake. Cells were treated with 5

and 25 μg/cm2 of 100 nm Por-SiO2 NPs and either 2.5 or5 μg/cm2 of 50 nm FITC-SiO2 NPs diluted in RPMI medi-um in the absence of serum at different time points at either37 or 4 °C. For treatments with lung surfactant, Curosurf®was mixed with different NP concentrations to obtain0.25 % dilution of surfactant and sonicated in the ultrasonicbath (Branson Cleaner, Ultrasonic, B200) at 20 W for10 min before adding the mixtures to the cells.

Flow cytometry

Cells were seeded in 12-well plates (Costar) at 10,000 cells/cm2 in complete cell culture medium and incubated for 48 hbefore treatment with 1.2 mL of NPs at different concen-trations and time points, with or without surfactant. Aftertreatment, cells were thoroughly rinsed with PBS and har-vested using 0.05 % trypsin-EDTA (Life Technologies),whose action was stopped with 10 % FCS. Cell-associatedfluorescence was detected using a CyAn ADP LX (DakoCytomation, Beckman Coulter, Villepinte, France) flowcytometer. Laser excitation and emission bandpass wave-lengths were 488 and 575±25 nm, respectively, for 50 nmFITC-SiO2 NPs and 488 and 670±30 nm for 100 nm Por-SiO2 NPs. Minimum of 10,000 cells was analyzed afterexclusion of the cellular debris from the analysis by gatingon the 575 nm log versus FS area graph for 50 nm FITC-SiO2 NPs and on the 670 nm log versus FS area graph for100 nm Por-SiO2 NPs. The results are reported as themedian of the distribution of cell fluorescence intensityobtained by analyzing 10,000 cells in the gate.

Microplate reader

Cells were seeded in 96-well plates at 8,000 cells/well forNCI-H292 and 12,800 cells/well for MHS, in correspondingcomplete cell culture medium and incubated for 48 h beforetreatment with 100 μL of NPs at different concentrationsand time points, with or without surfactant. After treatment,cells were thoroughly rinsed with PBS and fixed using 4 %paraformaldehyde (PFA), at 4 °C, overnight. Fluorescenceintensity was analyzed using Microplate Reader TECANinfinite M1000 (Lyon, France). Fluorescence intensity wasmeasured from the bottom of the plate before and afteradding 100 μL of Brilliant Black at 1 mg/mL (Sigma)in order to quench the fluorescence of NPs attached tothe cellular membrane or remaining on the bottom ofthe well after rinsing. Values are expressed as relativefluorescence units.

Confocal microscopy

Cells were seeded in eight-well Lab-TekTM chambered coverglasses (Nunc, Thermo Scientific, Dominique Dutscher,

Environ Sci Pollut Res

Author's personal copy

Brumath, France) at 40,000 cells/well in complete cell culturemedia. After treatment with 0.22 mL/well of NPs at indicatedconcentrations, cells were fixed in 4 % PFA for 20 minat 25 °C, rinsed three times with PBS and incubated for 10minwith NH4Cl (50 mM, Sigma) before permeabilization in0.05 % PBS-Tween 20 (Sigma). To stain the actin filaments,cells were incubated for 30min at 25 °C with FITC or TRITC-phalloidin (0.9 nM in PBS, Life Technologies), thenrinsed four times in PBS. Cell nuclei were stained with4′,6-diamidino-2-phenylindole dihydrochloride (DAPI, Sigma,0.25 μg/mL in PBS) for 1 min. Cells were mounted in polyvi-nyl alcohol mounting medium with DABCO® (Sigma) andwere examined under a Zeiss 710 confocal microscopeusing ×63 objective and a ×1.5 zoom. Image treatment wasdone with Image J software (Image J 1.42 NIH, USA).

Statistical analysis

Data are represented as mean ± SD (n=3) and were analyzedon commercially available software SigmaStat (version 3.0,Systat software Inc, San Jose, CA, USA) analysis of vari-ance (one-way ANOVA) followed by Dunnett’s test formultiple comparisons with p<0.05 (two tailed) consideredas significant.

Results

Demonstration of NP internalization by confocal microscopy

In order to qualitatively appreciate the internalization andfollow the uptake kinetics of two fluorescently labeled SiO2

NPs, lung epithelial (NCI-H292) and mouse alveolar mac-rophage (MHS) cells were treated with 50 nm FITC-SiO2

NPs and 100 nm Por-SiO2 NPs at a concentration of 5 and25 μg/cm2, respectively, for 15 min, 1 h, or 4 h. Theseconcentrations were shown not to interfere with the meta-bolic activity of the cells at the tested time points (data notshown). For a more precise localization of NPs (either insidethe cells or attached to the plasma membrane), actin fila-ments were stained using fluorescent phalloidin. Clear in-ternalization of 50 nm FITC-SiO2 and 100 nm Por-SiO2 NPswas detected in NCI-H292, as evidenced from thecorresponding Z-slices (Fig. 1), as well as in MHS (datanot shown). Interestingly, smaller SiO2 NPs (50 nm FITC-SiO2 NPs) were shown to be predominantly internalized asagglomerates after 15 min of exposure, while the larger NPs(100 nm Por-SiO2 NPs) remained on the cellular surface.However, the latter were also shown to be internalized at alater time point (4 h).

Determination of NP internalization by two approaches

To measure labeled NPs cell internalization, we used bothflow cytometry (FCM) and spectrophotometric microplateanalysis. First, we analyzed by FCM the mean fluorescenceintensity (MFI) of the two cell types treated with 50 nmFITC-SiO2 NPs and 100 nm Por-SiO2 NPs for 1 and 4 h.MFI increased with time and concentration of NPs for bothcell types (Fig. 2a, b). Treatment with NPs at the same cellconfluence and acquisition of the same number of cells inthe analysis allowed us to compare interactions of NPs withthe cells for both cell lines. For 50 nm FITC-SiO2 NPs, MFIvalues were not profoundly different between two cell types:

Fig. 1 Internalization of SiO2

NPs by lung epithelial cells.Cells were treated with 25 μg/cm2 of 100 nm Por-SiO2 NPs(a–c) or 5 μg/cm2 of 50 nmFITC-SiO2 NPs (d–f) for 15min,1 h, and 4 h. The upper imagescorrespond to the projection ofall images acquired in the stack.Corresponding lower imagesrepresent x, z-slices of the sec-tion indicated by the arrowhead.Staining of the cells is as follows:100 nm Por-SiO2 NPs: red—porphyrin-labeled SiO2 NPs,green—FITC-phalloidin-stainedactin filaments, blue—DAPI-stained nuclei; 50 nm FITC-SiO2

NPs: red—TRITC-phalloidin-la-beled actin filaments, green—FITC-labeled 50 nm SiO2 NPs,blue—DAPI-stained nuclei.Scale bars are 10 μm

Environ Sci Pollut Res

Author's personal copy

they were higher for macrophages than epithelial cells atlower doses (2.5 μg/cm2), at both time points but at higherdose (5 μg/cm2) similar MFI values were obtained for bothcell lines (Fig. 2a). MFI values for larger 100 nm Por-SiO2

NPs were clearly higher in macrophages than in epithelialcells at both tested concentrations and time points, except atlower dose after 4 h of treatment, where MFI value was notdifferent between the two cell types (Fig. 2b). As observedon images of confocal microscopy (Fig. 1), NPs can remainattached to the cellular membrane after rinsing and thisfluorescence of adsorbed NPs was also be measured by flowcytometry. This analysis thus evaluates the global fluores-cence signal of the cells treated with NPs, includingadsorbed and internalized NPs.

To quantify internalized NPs using a microplate reader, itis necessary to use a fluorescence quencher that will elimi-nate the signal coming from NPs adsorbed on the cellularsurface or on the bottom of the well (as observed on theFig. 1d) that have not been eliminated by rinsing. Weobserved increased internalization of NPs with time andconcentration for both epithelial cells and macrophages.Comparing the two cell types, in the same conditions oftreatment with 50 nm FITC-SiO2 NPs, internalization ofNPs by MHS cells was significantly higher than the onemeasured after treatment of the NCI-H292 cell line after 4 hof exposure (Fig. 2c). After treatment with 100 nm Por-SiO2

NPs, there was no significant difference in NP internaliza-tion between the two cell types except for a slightly higherinternalization by NCI-H292 cells after 1 h of treatment at25 μg/cm2 (Fig. 2d).

After 4 h of exposure to 50 and 100 nm SiO2 NPs at 5and 25 μg/cm2 respectively, results obtained by the two

techniques were not in agreement. Smaller SiO2 NPsseemed to be better internalized by macrophages if studiedby a microplate reader, while if studied by FCM, the inter-actions of epithelial cells with NPs were slightly higher thanwith macrophages. On the other hand, larger SiO2 NPsseemed to better interact with macrophages if studied byFCM, while the uptake was the same for both cell types ifstudied by a microplate reader. As this could be due to theadsorbed part of NPs on the cellular surface, we have furtheranalyzed this adsorbed part by exposing the cells to NPs at4 °C. Indeed, at 4 °C, active as well as passive processes ofinternalization are stopped. As cells are depleted from ener-gy, the activity of enzymes is decreased at low temperaturewhich impairs active, energy-dependent uptake of NPs. Onthe other hand, the rigidity of plasma membrane is in-creased, and this disables passive processes. Fluorescentsignal that is measured at 4 °C is thus originated from NPsthat were adsorbed on cellular surface and this could beconfirmed by confocal microscopy imaging (data notshown).

Contribution of the fluorescent signal coming from NPsadsorbed on the cellular surface

Whatever the cell type, the concentration of NPs, and thetime point studied, we observed by FCM that the MFI of thecells treated with 50 and 100 nm SiO2 NPs at 4 °C wasgenerally lower (Fig. 3a, b) than the signal measured aftertreatment at 37 °C.

Adsorption of 100 nm Por-SiO2 NPs after 1 h oftreatment was more important on the surface of MHS at25 μg/cm2 (57 % of the signal obtained after the treatment

Fig. 2 Interactions of NPs withlung epithelial (NCI-H292) andmacrophage (MHS) cells. a, bCells were treated with 50 or100 nm SiO2 NPs at indicatedconcentrations and time pointsand then analyzed by FCM. c, dCells were treated with 50 or100 nm SiO2 NPs at indicatedconcentrations and time pointsand then analyzed by fluores-cence microplate reader. Short-ly before analysis, BrilliantBlack was added to the sam-ples. Representative results ofat least three independentexperiments. *p<0.05, statisti-cally different from MHS celltype, (two tailed)

Environ Sci Pollut Res

Author's personal copy

at 37 °C) than NCI-H292 (33 %) cells (Fig. 3b). Thistrend was confirmed after 4 h of exposure to these NPs asMHS still adsorbed more NPs (52 %) at 25 μg/cm2

compared to NCI-H292 cell line (43 % at 25 μg/cm2).When the values of signals obtained at 4 °C were sub-tracted from the values obtained at 37 °C, it seemed thatboth cell types internalized to the same extent after 4 h oftreatment with 25 μg/cm2 of these NPs (ΔMFI, 313 forMHS; ΔMFI, 268 for NCI-H292).

A similar trend (although to a lesser extent) was observedfor 50 nm FITC-SiO2 NPs as 40 % of NPs was adsorbed onthe surface of MHS cells compared to 19 % on NCI-H292cells after 1 h of treatment and 47 % of NPs wereadsorbed on MHS compared to 12 % on NCI-H292after 4 h of treatment with 5 μg/cm2 of 50 nm FITC-SiO2 NPs (Fig. 3a). When the values of signalsobtained at 4 °C were subtracted from the valuesobtained at 37 °C, it appeared that the NCI-H292 cellline internalized 50 nm SiO2 NPs better (ΔMFI, 560),than MHS cell line (ΔMFI, 308) after 4 h of treatmentat 5 μg/cm2 with these NPs.

When analyzed by the microplate reader technique,the contribution of adsorbed NPs was eliminated by theuse of a fluorescence quencher, Brilliant Black.However, surprisingly, fluorescence was still detectedat 4 °C in this condition (Fig. 3c, d). This signalrepresented at least 50 % of the fluorescence comparedto what was detected at 37 °C.

Effect of lung surfactant on NPs internalization

In order to study the effect of lung surfactant on NP inter-nalization, cells were treated with NPs mixed with 0.25 % of

Curosurf®—a commercially available natural lung surfac-tant that contains the hydrophobic fraction (lipids, SP-B,and SP-C). Whatever the cell type, the presence of lungsurfactant strongly diminished the cell-associated fluores-cence. This was firstly observed by confocal microscopy(Fig. 4a, b). After 4 h of treatment, it was evident that in thepresence of lung surfactant, few NPs were observed incontact with the cells for 100 nm Por-SiO2 NPs and evenno NPs were observed when the cells were treated with50 nm FITC-SiO2 NPs in the presence of lung surfactant.

This was confirmed by flow cytometry analysis, whichshowed that the MFI of the cells treated with NPs andCurosurf® was low compared to cells treated with NPsalone and even close to control cells for 50 nm FITC-SiO2

NPs (Fig. 4c, f). Reducing Curosurf® concentrations to0.06 % allowed for a very modest increase of 50 nmFITC-SiO2 NPs uptake (Fig. 5a).

Using the microplate reader, we observed a similar trend.In the presence of lung surfactant, the fluorescence of cellstreated with 50 nm FITC-SiO2 NPs was similar to thecontrol values in both cell lines (Fig. 4e). After addition ofthe lung surfactant, the fluorescence intensity of NCI-H292cells treated with 100 nm Por-SiO2 NPs was 80 % lower atboth concentrations and time points tested, while the de-crease was of 65 % in MHS cells (Fig. 4f).

Microscopic observations of the cells treated with NPs inpresence of Curosurf® revealed that NPs floated above thecell culture and formed fewer agglomerates which seemedto delay their sedimentation on the culture and diminishedtheir contact with cellular membranes. This could thereforeexplain the reduced internalization of NPs observed here. Tofurther confirm this hypothesis, cells were treated withCurosurf® 24 h before depositing NPs, in order to allow

Fig. 3 Interactions of NPs withlung epithelial (NCI-H292) andmacrophage (MHS) cells at 37and 4 °C. a, b Cells were trea-ted with 50 or 100 nm SiO2 NPsat indicated concentrations andtime points at 37 or 4 °C, thananalyzed by FCM. c, d Cellswere treated with 50 or 100 nmSiO2 NPs at indicated concen-trations and time points at 37 or4 °C, than analyzed by micro-plate reader. Shortly beforeanalysis, Brilliant Black wasadded to the samples. Data arerepresented as MFI ± SD (n=3).Representative results of at leastthree independent experiments.*p<0.05, statistically differentfrom results obtained aftertreatment at 4 °C, (two tailed)

Environ Sci Pollut Res

Author's personal copy

the sedimentation of the lung surfactant before applyingNPs. In these conditions, NPs uptake was more important(37 %) than when NPs were directly dispersed withsurfactant, but still significantly lower than the one

obtained in the absence of surfactant (Fig. 5b). Thisresult indicates that a lung surfactant layer formed during24 h of surfactant sedimentation could indeed diminish theuptake of NPs by the cells.

Fig. 4 Effect of lung surfactant on the uptake of 50 and 100 nm SiO2

NPs by NCI-H292 and MHS cell line. a, b Cells were treated with5 μg/cm2 of 50 nm FITC-SiO2 NPs without Curosurf® (a, left image)or with Curosurf® (a, right image) and 25 μg/cm2 of 100 nm Por-SiO2

NPs without Curosurf® (b, left image) or with Curosurf® (b, rightimage) for 4 h. Images correspond to the projection of all imagesacquired in the stack. Staining of the cells is as follows: 100 nm Por-SiO2 NPs: red—porphyrin-labeled SiO2 NPs, green—FITC-phalloi-din-stained actin filaments, blue—DAPI-stained nuclei; 50 nm FITC-SiO2 NPs: red—TRITC-phalloidin-labeled actin filaments, green—

FITC-labeled 50 nm SiO2 NPs, blue—DAPI-stained nuclei. c, d Cellswere treated with 50 or 100 nm SiO2 NPs premixed with or without0.25 % Curosurf® at indicated concentrations and time points, thananalyzed by FCM. e, f Cells were treated with 50 or 100 nmSiO2 NPs with or without 0.25 % Curosurf® at indicated concen-trations and time points, than analyzed by microplate reader.Shortly before analysis, Brilliant Black was added to the samples.Representative results of at least three independent experiments.*p<0.05, statistically different from results obtained after treat-ment with Curosurf®, (two tailed)

Fig. 5 Effect of lung surfactant on the uptake of 50 nm FITC-SiO2

NPs by NCI-H292. a Cells were treated with 5 μg/cm2 of 50 nm FITC-SiO2 NPs premixed with 0.25, 0.12, 0.06 % or without Curosurf® for4 h. Cellular associated fluorescence was measured by FCM. Repre-sentative results of at least three independent experiments. *p<0.05,statistically different from results obtained after treatment without

Curosurf®, (two tailed); b 0.25 % Curosurf® was deposited on thecells with NPs or 24 h before adding 50 nm FITC-SiO2 NPs at 5 μg/cm2 for 4 h. Cellular associated fluorescence was measured by FCM.Representative results of at least three independent experiments. *p<0.05, statistically different from results obtained after treatment withCurosurf® deposited 24 h before treatment with NPs, (two tailed)

Environ Sci Pollut Res

Author's personal copy

Discussion

Our study intended to characterize NP internalization bytarget cells in the lungs as inhalation is one of the mainroutes of NP entry in the organism. We also determined therole of lung surfactant in this internalization. For this pur-pose, we used different methodologies to study the uptake offluorescently labeled NPs. We chose SiO2 NPs as a modelNP since they can be easily labeled with fluorochromes andbecause of their wide use. Indeed, SiO2 NPs are produced inhigh quantities and have important industrial applicationssuch as the improvement of the elastic properties of tyresand their increased resistance to wear and tear. They alsohave important biomedical applications as drug transporters,for gene transfer and for imaging as they are easily fluores-cently labeled (Roy et al. 2005). Anti-agglomerating prop-erties of SiO2 NPs have led to their applications in foodindustry. In this context, we chose to study the fate offluorescently labeled SiO2 of two sizes (50 and 100 nm)and their interactions with lung epithelial cells (NCI-H292)and macrophages (MHS) as well as the potential modulatoryeffect of pulmonary surfactant. Using confocal microscopy,we observed that SiO2 NPs were internalized as agglomer-ates and that this uptake was time dependent. The internal-ization was strongly reduced in the presence of Curosurf®even though its presence did not change the agglomerationproperties of NPs (Figs. 1 and 4). We also observed thepresence of NPs on the cell surface where they wereadsorbed, even after thorough washing.

Since confocal microscopy does not allow for uptakequantification of NPs, we used for this purpose flow cytom-etry and fluorescence measurements in a microplate reader.Indeed, flow cytometry analysis allows for the measurementof the individual fluorescence of a large number of cells,which need to be detached and analyzed in suspension.Microplate reader measures the global fluorescence of thewhole cell population that has been treated but the analysisis performed directly on the plates where cells are growthand treated. It can be used as an alternative to flow cytom-etry as it avoids trypsinization to detach cells from theirsupport and the generation of potential artifacts.Microplate reader analysis also enables for high throughputanalysis and is useful for experiments involving a largenumber of conditions or for the evaluation of the effect ofa large number of NPs.

For both cell types and both NPs used, it was clear thatthe uptake was time and dose dependent. Our first measure-ments revealed a discrepancy in the results which dependedon the technique used. Indeed, smaller NPs seemed tointeract slightly better with the NCI-H292 cell line thanMHS after 4 h of exposure if studied by flow cytometry,while the trend was reversed when studied with a microplatereader. Conversely, larger SiO2 NPs were shown to better

interact with the MHS cell line when studied by flowcytometry, but the fluorescent signal was similar when mea-sured by a microplate reader. This could be due to the factthat single cells are analyzed by FCM, taking into accountNPs that are internalized as well as those adsorbed on thecell surface. We have indeed observed by confocal micros-copy that NPs still attach to the cells even after thoroughwashing. This makes the accurate quantification of internal-ized NPs difficult as only a global signal is detected. Using amicroplate reader on the other hand, adsorbed NPs are notconsidered as they are eliminated from the analysis byaddition of Brilliant Black.

To further appreciate the contribution of NPs adsorbed onthe cellular surface, uptake was analyzed when the cellswere exposed to NPs at 4 °C, where active as well as passiveendocytotic processes are blocked. Flow cytometry datarevealed that both sizes of NPs were more adsorbed onMHS than on NCI-H292 cell line (Fig. 3). Taking intoaccount the difference of MFI values measured at 37 and4 °C, the uptake can be measured and we showed that100 nm Por-SiO2 NPs were similarly internalized by epi-thelial cells and macrophages at high doses after 4 h oftreatment, while 50 nm FITC-SiO2 NPs were more internal-ized by epithelial cells under the same conditions. It hasindeed already been shown that macrophages are moreefficient in phagocyting larger particles than smaller ones(Geiser 2010; Geiser et al. 2008) which could explain whydifferences between epithelial cells and macrophages arereduced when comparing larger particles.

With the microplate reader analysis, the results obtainedat 4 °C suggest that an important part of NPs is internalizedindicating that passive diffusion of NPs could take place,confirming previous results. This is not in agreement withthe results obtained by flow cytometry and confocal micros-copy, since the images of the cells treated with NPs at 4 °Cshowed that NPs stayed exclusively on the cell surface andwere not seen inside the cells (data not shown). An alterna-tive explication could be that quenching of the fluorescencecoming from NPs is not efficient, especially when they areagglomerated. Efficient use of Brilliant Black in quenchingextracellular fluorescence has already been reported(Wemhöner et al. 2006); however as shown here, its use inquenching the extracellular fluorescence of fluorescentlylabeled NPs is questionable. A possible explanation couldbe aggregation of NPs hindering the access of the quencheror the position of the fluorochrome on the NPs with respectto the quenching agent. Moreover, in the case of a stronginteraction of NPs with the cellular membrane, it is possiblethat this also prevents access to the quencher.

It is known that protein or lipid corona formed aroundNPs in the biological fluids is important for further inter-actions of NPs with the cells (Lynch et al. 2007, 2009). Thiscorona differs according to the biological fluids in contact

Environ Sci Pollut Res

Author's personal copy

with NPs. In the lungs, NPs will interact with pulmonarysurfactant components as surfactant is produced by alveolarepithelial cells and is present on their surface (Kapralov etal. 2012). We showed that the hydrophobic fraction ofsurfactant (Curosurf®) strongly diminished the uptake ofNPs, whatever the cell line or NPs analyzed (Fig. 4). Twohypotheses could explain this observation: either hydropho-bic fraction of lung modifies the properties of NP surfacesby diminishing the interaction of NPs with the cell mem-brane that is essential for internalization, or the phospholi-pids complexed to hydrophobic proteins changes thedispersion of NPs in the solution, disabling their sedimen-tation and interaction with the cell membrane. Indeed, mi-croscopic observation revealed that in the presence ofCurosurf®, NPs floated above the cell culture and formedfewer agglomerates which likely retarded their sedimenta-tion in the culture and diminished contact between NPs andthe cell membrane. The observed lack of internalizationcould thus be explained by a reduced contact with the cellsin the presence of this surfactant. By contrast, when theCurosurf® was applied 24 h before the exposure to NPs,the latter sediment quicker and the cellular uptake wasindeed increased. In addition, the effects of surfactant inNPs cellular uptake may also be influenced by the NPsurface. Indeed, recently, Ruge et al. have shown that naturalsurfactant can reduce or enhance the association of magne-tite NPs to macrophages (MHS cell line) depending onwhether NPs are coated with phosphatidylcholine or starch,respectively (Ruge et al. 2012). In our study, we used acommercially available natural surfactant, as lung surfactantmodel (Curosurf®). Use of this product is advantageous, asit has formally been used in clinics to prevent alveolarcollapse in premature newborns; therefore, the preparationsare relatively stable and controlled by manufacturers.However, Curosurf® lacks hydrophilic surfactant-associatedproteins, like SP-A. Since surfactant lipids and SP-A couldexert opposite effects on NP uptake (Ruge et al. 2012),surfactant composition in terms of lipid/protein ratiomight be another important factor that controls theglobal effect of surfactant on NP internalization.

Conclusions

In the present article, we reported internalization of fluores-cently labeled 50 and 100 nm SiO2 NPs in lung epithelialcells as well as in mouse alveolar macrophages. This inter-nalization was firstly observed by confocal microscopy andfurther quantified by two approaches: flow cytometry andmicroplate fluorescence reading. Internalization was shownto be time and dose dependent by both approaches. In thelight of these observations, we can conclude that flowcytometry gives more reliable results. The presence of

surfactant Curosurf® is shown to diminish NP uptake byboth cell lines tested. We have shown that this surfactantinfluences to a certain extent sedimentation of NPs andprevents their contact with cell membranes and theirefficient internalization.

Acknowledgments This study was funded by the E.C. FP7 ENPRA(no. 228789) grant. Support for this study was also provided by Nano-trans (no. EST-2010/2/079) and TiSiTrans (no. PNR-EST-2010/2/79)grants. We acknowledge the confocal microscope platform in theInstitute Jacques Monod, Paris, France. Authors would also like tothank to Dr Emmanuel Lopez (Cochin Hospital, Paris) for helping us toobtain the surfactant used in this study.

References

Blaaderen A, Vrij A (1992) Synthesis and characterization of colloidaldispersions of fluorescent. Monodisperse Silica Spheres,Langmuir, pp 2921–2931

Casals E, Pfaller T, Duschl A, Oostingh GJ, Puntes V (2010) Timeevolution of the nanoparticle protein corona. ACSNano 4:3623–3632

Chroneos ZC, Sever-Chroneos Z, Shepherd VL (2010) Pulmonarysurfactant: an immunological perspective. Cell Physiol Biochem25:13–26

Geiser M, Rothen-Rutishauser B, Kapp N, Schürch S, Kreyling W,Schulz H, Semmler M, Im Hof V, Heyder J, Gehr P (2005)Ultrafine particles cross cellular membranes by nonphagocyticmechanisms in lungs and in cultured cells. Environ HealthPerspect 113:1555–1560

Geiser M, Casaulta M, Kupferschmid B, Schulz H, Semmler-BehnkeM, Kreyling W (2008) The role of macrophages in the clearanceof inhaled ultrafine titanium dioxide particles. Am J Respir CellMol Biol 38:371–376

Geiser M (2010) Update on macrophage clearance of inhaled micro-and nanoparticles. J Aerosol Med Pulm Drug Deliv 23:207–217

Kapra lov AA, Feng WH, Amosca to AA, Yanamala N,Balasubramanian K, Winnica DE, Kisin ER, Kotchey GP, GouP, Sparvero LJ, Ray P, Mallampalli RK, Klein-Seetharaman J,Fadeel B, Star A, Shvedova AA, Kagan VE (2012) Adsorption ofsurfactant lipids by single-walled carbon nanotubes in mouse lungupon pharyngeal aspiration. ACS Nano 6:4147–4156

Kruth HS (2002) Sequestration of agglomerated low-density lipopro-teins by macrophages. Curr Opin Lipidol 13:483–488

Lesniak A, Fenaroli F, Monopoli MP, Aberg C, Dawson KA, Salvati A(2012) Effects of the presence or absence of a protein corona onsilica nanoparticle uptake and impact on cells. ACS Nano6:5845–5857

Lundqvist M, Stigler J, Cedervall T, Berggård T, Flanagan MB, LynchI, Elia G, Dawson K (2011) The evolution of the protein coronaaround nanoparticles: a test study. ACS Nano 5:7503–7509

Lynch I, Cedervall T, Lundqvist M, Cabaleiro-Lago C, Linse S,Dawson KA (2007) The nanoparticle–protein complex as a bio-logical entity; a complex fluids and surface science challenge forthe 21st century. Adv Colloid Interface Sci 134–135:167–174

Lynch I, Salvati A, Dawson KA (2009) Protein–nanoparticle interac-tions: what does the cell see? Nat Nanotechnol 4:546–547

Mohamed BM, Verma NK, Prina-Mello A, Williams Y, Davies AM,Bakos G, Tormey L, Edwards C, Hanrahan J, Salvati A, Lynch I,DawsonK, Kelleher D, Volkov Y (2011) Activation of stress-relatedsignalling pathway in human cells upon SiO2 nanoparticles expo-sure as an early indicator of cytotoxicity. J Nanobiotechnology 9:29

Environ Sci Pollut Res

Author's personal copy

Mu Q, Hondow NS, Krzemi Ski L, Brown AP, Jeuken LJ, RoutledgeMN (2012) Mechanism of cellular uptake of genotoxic silicananoparticles. Part Fibre Toxicol 9:29

Nel A, Xia T, Mädler L, Li N (2006) Toxic potential of materials at thenanolevel. Science 311:622–627

Oberdörster G, Oberdörster E, Oberdörster J (2005) Nanotoxicology:an emerging discipline evolving from studies of ultrafine par-ticles. Environ Health Perspect 113:823–839

Rothen-Rutishauser BM, Schürch S, Haenni B, Kapp N, Gehr P (2006)Interaction of fine particles and nanoparticles with red blood cellsvisualized with advanced microscopic techniques. Environ SciTechnol 40:4353–4359

Roy I, Ohulchanskyy TY, Bharali DJ, Pudavar HE,Mistretta RA, Kaur N,Prasad PN (2005) Optical tracking of organically modified silicananoparticles as DNA carriers: a nonviral, nanomedicine approachfor gene delivery. Proc Natl Acad Sci U S A 102:279–284

Ruge CA, Schaefer UF, Herrmann J, Kirch J, Cañadas O, Echaide M,Pérez-Gil J, Casals C, Müller R, Lehr CM (2012) The interplay oflung surfactant proteins and lipids assimilates the macrophageclearance of nanoparticles. PLoS One 7:e40775

Schleh C, Mühlfeld C, Pulskamp K, Schmiedl A, Nassimi M,Lauenstein HD, Braun A, Krug N, Erpenbeck VJ, Hohlfeld JM(2009) The effect of titanium dioxide nanoparticles on pulmonarysurfactant function and ultrastructure. Respir Res 10:90

Schleh C, Semmler-Behnke M, Lipka J, Wenk A, Hirn S, Schäffler M,Schmid G, Simon U, Kreyling WG (2012) Size and surface

charge of gold nanoparticles determine absorption across intesti-nal barriers and accumulation in secondary target organs after oraladministration. Nanotoxicology 6:36–46

Semmler-Behnke M, Takenaka S, Fertsch S, Wenk A, Seitz J, Mayer P,Oberdörster G, Kreyling WG (2007) Efficient elimination ofinhaled nanoparticles from the alveolar region: evidence for in-terstitial uptake and subsequent reentrainment onto airways epi-thelium. Environ Health Perspect 115:728–733

Semmler-Behnke M, Kreyling WG, Lipka J, Fertsch S, Wenk A,Takenaka S, Schmid G, Brandau W (2008) Biodistribution of1.4- and 18-nm gold particles in rats. Small 4:2108–2111

Sergent JA, Paget V, Chevillard S (2012) Toxicity and genotoxicity ofNano-SiO2 on human epithelial intestinal HT-29 cell line. AnnOccup Hyg 56(5):622–630

Wemhöner A, Frick M, Dietl P, Jennings P, Haller T (2006) A fluores-cent microplate assay for exocytosis in alveolar type II cells. JBiomol Screen 11:286–295

Wu W, Samet JM, Peden DB, Bromberg PA (2010) Phosphorylation ofp65 is required for zinc oxide nanoparticle-induced interleukin8 expression in human bronchial epithelial cells. Environ HealthPerspect 118:982–987

Xia T, Kovochich M, Liong M, Mädler L, Gilbert B, Shi H, YehJI, Zink JI, Nel AE (2008) Comparison of the mechanism oftoxicity of zinc oxide and cerium oxide nanoparticles basedon dissolution and oxidative stress properties. ACS Nano2:2121–2134

Environ Sci Pollut Res

Author's personal copy