Size- and Coating-Dependent Uptake of Polymer-Coated Gold

Article pubs.acs.org/Biomac

Size- and Coating-Dependent Uptake of Polymer-Coated Gold Nanoparticles in Primary Human Dermal Microvascular Endothelial Cells Christian Freese,*,†,∥ Matthew I. Gibson,‡,§,∥ Harm-Anton Klok,‡ Ronald E. Unger,† and C. James Kirkpatrick† †

REPAIR-lab, Institute of Pathology, University Medical Center of the Johannes Gutenberg University, Langenbeckstrasse 1, D-55101 Mainz, Germany ‡ ́ Ecole Polytechnique Fédérale de Lausanne (EPFL), Institut des Matériaux and Institut des Sciences et Ingénierie Chimiques, Laboratoire des Polymères, Bâtiment MXD, Station 12, CH-1015 Lausanne, Switzerland § University of Warwick, Institute of Chemistry, Coventry, United Kingdom S Supporting Information *

ABSTRACT: A library-orientated approach is used to gain understanding of the interactions of well-defined nanoparticles with primary human endothelial cells, which are a key component of the vasculature. Fifteen sequentially modified gold nanoparticles (AuNPs) based on three different core sizes (18, 35, 65 nm) and five polymeric coatings were prepared. The synthetic methodology ensured homogeneity across each series of particles to allow sequential investigation of the chemical features on cellular interactions. The toxicity of these nanoparticles, their uptake behavior in primary human dermal microvascular endothelial cells (HDMECs), and quantification of uptake were all investigated. The results of our studies indicated that high concentrations of gold nanoparticles (250 μg/mL) were nontoxic and that the number of internalized nanoparticles was related to nanoparticle size and surface chemistry. In summary, the positive-charged ethanediamine-coated AuNPs were internalized to a greater extent than the negative- or neutral-charged AuNPs. Moreover, differences in the amounts of internalized AuNPs could be shown for the three neutral-charged AuNPs, whereas the uptake of hydroxypropylamine-coated particles was preferred compared with glucosamine-coated or PEGylated AuNPs. Hydroxypropylamine-coated AuNPs were found to be the most efficient neutral-charged particles in overcoming the endothelial cell barrier and entering the cell.

■

surface.7,8 These are properties that make gold nanoparticles so attractive.9 To exploit nanoparticles in biological systems, we must determine the exclusion of contaminants such as endotoxins10 and also their cytotoxicological profile. Several studies have suggested that gold nanoparticles are nontoxic, but Pan and coworkers demonstrated that 1.4 nm gold nanoparticles induce cytotoxicity in HeLa cells to a greater extent than the corresponding 15 nm particles.11 Toxicity is also a function of the surface chemistry, however, so it is essential to control precisely the functional groups on the surface as well as remove residual contaminants, especially citrate, arising from the particle synthesis.12 A useful method to modify the surface properties of nanoparticles involves coating with polymers, which allows the introduction of a wide range of functionalities, composition, and charge (neutral, positive, and negative). For example, the addition of poly(ethylene glycol) (PEG) tends to

INTRODUCTION Progress in nanotechnology, especially in nanoscience, is a benefit for medical applications. Nanomedicine has become an expanding field with the potential to solve many problems of today’s medicine. Nanodimensions and the corresponding surface/volume ratio as well as the uptake potential in various cell types are just some properties of nanoparticles that make them suitable for applications in medicine and biology. The use of inorganic nanoparticles, especially gold nanoparticles, has been described as being of great interest for many biomedical applications, such as drug delivery, imaging, and diagnostics, due to their biocompatibility in vitro as well as in vivo.1−3 In addition to the broad spectrum of applications in biology and medicine, gold nanoparticles are easy to produce, and because of the unique electronic properties, they are easily detectable by various methods such as transmission electron microscopy (TEM) and spectroscopy (i.e., surface-enhanced Raman spectroscopy (SERS) and inductively coupled plasma atomic emission spectroscopy (ICP-AES)). Furthermore, gold is easily modified by the use of thiols4−6 and has been used to incorporated targeting groups or therapeutics onto their © XXXX American Chemical Society

Received: February 16, 2012 Revised: March 20, 2012

A

dx.doi.org/10.1021/bm300248u | Biomacromolecules XXXX, XXX, XXX−XXX

Biomacromolecules

Article

25 °C) was obtained from a Millipore Milli-Q gradient machine fitted with a 0.22 μm filter. All other chemicals were used as received unless otherwise indicated. Physical and Analytical Methods. UV−visible absorbance spectra were obtained on a Varian Cary 100 Bio spectrophotometer operating at 25 °C using 10 mm path length cuvettes with a total volume of 2.5 mL. The method of Fernig and coworkers was used to estimate the diameter of citrate-stabilized gold nanoparticle using UV−visible spectroscopy.20 Dynamic light scattering measurements were carried out on a Brookhaven Instruments system consisting of a BI-200SM goniometer and a BI-9000AT autocorrelator. A 100mW Ar+ ion laser (Lexel Lasers) operating at 488 nm was used. All measurements were performed at a temperature of 25.0 (±0.2) °C and at a scattering angle of 90°. Borosilicate cuvettes were used with a minimum of 3 mL of the nanoparticle solution. The nanoparticles were used at a concentration of ∼0.05 mg/mL (total gold mass) in Milli-Q water and passed through a 0.4 μm filter before measurement. Synthesis and characterization of the polymers was previously described.21 See the Supporting Information for full details of the polymers used. Synthesis of Citrate-Stabilized Gold Nanoparticles. As a typical example, the synthesis of 18 nm diameter gold nanoparticles is described. First, 600 mL of a 0.85 mmol/L (0.33 mg/mL) aqueous solution of HAuCl4 was heated to reflux in a scratch-free roundbottomed flask. After that, 10.5 mL of a 0.5 mol/L aqueous solution of sodium citrate was added in a single portion to the HAuCl4 solution to give a Au/citrate ratio 1:3.5. The temperature was maintained at reflux for 30 min, during which time a deep-red coloration formed. The reaction mixture was then allowed to cool to room temperature over a period of 3 h. Gold nanoparticles with diameters of 35 and 68 nm were prepared using the same protocol but by adjusting the concentration of the sodium citrate solution to the yield Au/citrate ratios of 1:2 and 1:1, respectively. Assuming complete reduction of the HAuCl4 into the particles, the total gold concentration in the final solution was 0.85 mmol/L (0.17 mg/mL). General Procedure for the Synthesis of Polymer-Coated Gold Nanoparticles. Approximately 10 mg of the desired thiol-terminated polymer was dissolved in 0.5 mL of high-purity water with the pH adjusted to ∼10 using 0.1 M NaCl solution. To this tube, 120 mL of the citrate-stabilized gold nanoparticle solution (0.85 mmol/L total gold concentration) was added and then agitated overnight in the absence of light. To remove excess polymer, the particles were centrifuged for 3 h at 8000 rpm. Following careful decantation of the supernatant, the particles were then redispersed in 20 mL of highquality water, and the centrifugation−resuspension process was repeated for a total of three cycles. After the final cycle, the particles were dispersed in 10 mL of high-quality water for future use. Assuming complete incorporation of the citrate-coated gold particles into the final polymer-coated particles, the total concentration of gold in the final solution was 10.2 mmol/L (2 mg/mL). Isolation and Cell Culture of Primary Human Dermal Microvascular Endothelial Cells. Human dermal microvascular endothelial cells (HDMECs) were isolated from juvenile foreskin, as previously described.22 In brief, cells were isolated by cutting the foreskin into small pieces. After enzymatic digestion in 0.4% collagenase (Gibco, Carlsbad, CA) for 16 h, the epidermis of the foreskin was manually separated from the dermis. After a second incubation with versene (Gibco) and 80 μL of 2.5% trypsin (Gibco) for 2 h and after a mechanical treatment of the foreskin and filtering of the digested tissue, the cells were seeded onto 0.2% gelatin-coated culture flasks and cultured in endothelial cell basal medium (Customer Formulation) supplemented with supplement mix (both PromoCell, Heidelberg, Germany) and penicillin (10 000 units/mL)/streptomycin (10 000 μg/mL; both Gibco). To separate the endothelial cells from other contaminating cells such as fibroblasts, we performed two separating steps with magnetic CD-31 beads after the cultures became confluent. Following the first separation step, cells were grown in ECBM supplemented with 15% fetal bovine serum, 2.5 ng/mL basal fibroblast growth factor and 10 μg/mL sodium heparin (all SigmaAldrich, St. Louis, MO), and penicillin (10 000 U/mL) and

prolong the circulation time of nanoparticles in the blood compared with NPs without PEG.13 Positive-charged nanoparticles have been reported to be internalized in mammalian cells more efficiently than neutral- or negative-charged particles because of electrostatic interactions with the anionic surface of most cell membranes.14 Whereas it is widely understood that surface, size, and shape will all play a key role in the interactions between nanoparticles and biological membranes,15 there is still a significant requirement to investigate these in a systematic manner, where only a single feature is changed at a time. In particular, the role of these on cellular internalization and processing is critical.4 For many applications in vivo, the nanoparticles are injected intravenously. In addition to the immune cells in blood,16 endothelial cells, which line the inner side of the vasculature, are the first cells that the nanoparticles come into contact with. Therefore, understanding how nanoparticles of different sizes, charges, or surface modification react with microvascular endothelial cells making up the capillaries is essential to develop new drug delivery systems capable of specifically targeting diseased tissue. A specific advantage of gold nanoparticles compared with, for example, self-assembled polymer micelles/vesicles is their ease of detection. For example, because of the high electron density of gold nanoparticles, direct visualization by optical or TEM is possible. Other nanoparticle classes require the addition of fluorescence dyes that require further synthetic steps and the challenge of ensuring that each particle has an equal number of dye molecules. Quantification of cellular uptake by microscopic methods is, however, very challenging because only a small number of cells are evaluated delivering qualitative information. A relatively new approach to determine the amount of gold within cells is ICP-AES of acid-digested cells.17 This method ensures a highly sensitive analysis and comparison of gold nanoparticles taken up by all exposed cells in the range of parts per billion (ppb) and has been successfully used to quantify the uptake of unmodified gold nanoparticles into cells.18,19 Despite the huge diversity of nanoparticles that has been synthesized and biologically evaluated in the literature, there have still been very few systematic studies into the role of nanoparticle surface chemistry and size on their cellular interactions. A key challenge with most synthetic methods is ensuring that only one variable is changed between particles (e.g., size, shape, charge). Therefore, we have developed a new method to obtain libraries of sequentially modified, polymerstabilized gold nanoparticles with precise control over size and coating.6 Here we have investigated a 15-member polymercoated gold nanoparticle library based on three different gold cores (18, 35, 65 nm) and coated with five different polymers: PEG and glucosamine-, hydroxypropylamine-, ethanediamine-, and taurine-functional poly(methacrylates) that were conjugated to the gold nanoparticle surfaces via thiol-gold bonds. The toxicity and uptake of the library into primary human endothelial cells was evaluated to provide insight into the behavior of nanoparticles in the circulatory system.

■

MATERIALS AND METHODS

Synthesis and Characterization of Polymer-Coated Gold Nanoparticles. Materials. Gold(III) chloride trihydrate (HAuCl4; > 49% Au, ACS grade) and thiol-terminated PEG (Mn = 3000 g/mol) were purchased from Sigma-Aldrich. Trisodium citrate (99.8%) was purchased from Acros Organics. Ultra-high-quality water with a resistance of 18.2 MΩ × cm2 (at B

dx.doi.org/10.1021/bm300248u | Biomacromolecules XXXX, XXX, XXX−XXX

Biomacromolecules

Article

Scheme 1. Synthesis of Polymer-Coated Gold Nanoparticle Librariesa

a (i) 4-Cyanopentanoic acid dithiobenzoate and 4,4-azobis(4-cyanovaleric acid) and initiator, dioxane, 90 °C; (ii) 2 equiv of R-NH2/NEt3, DMF, 50 °C, 16 h; (iii) sodium citrate, H2O, reflux, 30 min; and (iv) thiol-terminated polymer (2 mass equiv), H2O 12 h.

streptomycin (10 000 μg/mL). In the following text, this medium is termed ECBM culture medium. Cells were used up to passage three. Cells were maintained in an atmosphere of 5% CO2 and 95% humidified condition at 37 °C. All experiments were carried out with at least three different donors. Stimulation with Gold Nanoparticles, Staining and Microscopy. Cells were seeded onto fibronectin-coated LabTek chamber slides (Nunc, Roskilde, Denmark) in ECGM culture medium. After 48 h, cells were incubated with a 100 μg/mL nanoparticle suspension. After the incubation period, cells were washed twice with HEPES buffer including 0.2% BSA and then fixed with 3.7% paraformaldehyde at room temperature for 15 min. Afterward, cells were washed and incubated with mouse antihuman CD31 antibody (DakoCytomation, Glostrup, Denmark) and the corresponding secondary antibody (goat antimouse Alexa Fluor 488; Molecular Probes, Carlsbad, CA) at room temperature for 1 h each. The nuclei were stained with Hoechst 33342 (Sigma-Aldrich). The LabTeks were embedded with GelMount (Biomeda, Natutec, Germany) and analyzed via light/fluorescence microscopy (Olympus IX71 with Delta Vision system, Applied Precision, USA). For analyzing the uptake of gold nanoparticles by TEM, cells were seeded onto fibronectin-coated Thermanox coverslips (Nunc, Roskilde, Denmark), treated with gold nanoparticles (10 μg/ mL) as described above, and then fixed with cacodylate-buffered glutaraldehyde (Serva, Heidelberg, Germany) (pH 7.2) for 20 min. This was followed by a fixation step in 1% (w/v) osmium tetroxide for 2 h and dehydration in ethanol. Cells were transferred through propylene oxide. Afterward, the samples were embedded in agar-100 resin (PLANO, Germany) and polymerized at 60 °C for 48 h. Ultrathin sections were cut with an ultramicrotome (Leica Microsystems, Germany), placed onto copper grids and stained with 1% (w/ v) uranyl acetate in alcoholic solution and lead citrate. Ultrastructural analysis was performed with a TEM, EM 410 (Philips; Eindhoven, The Netherlands). Quantification of Internalized Gold Nanoparticles by ICPAES. Cells were seeded onto fibronectin-coated 24-well plates. After reaching confluence, the medium was replaced by the nanoparticle suspension (10 μg/mL). After treatment for 4 and 24 h the cells were washed with HEPES + 0.2% BSA, detached by trypsin incubation, and transferred after the addition of 0.9 mL of PBS to microcentrifuge tubes. The cell suspension was stored at −20 °C until analysis. To the cell lysate solution was added 0.15 mL of aqua regia (3:1 hydrochloric acid/nitric acid (both purchased from Fisher Scientific, U.K.)). Following incubation overnight, the samples were then further diluted to 5 mL using Milli-Q water to give a total sample volume of 5 mL. These samples were then analyzed for total gold content by ICP-AES, and the measurement was repeated three times. Cell Viability Assay (MTS Assay). Cells were seeded onto fibronectin-coated 96-well plates with ECGM culture medium and cultured to confluence for 48 h. Cells were exposed to different

concentrations of gold nanoparticles for 24−48 h. Cell viability was measured using the CellTiter 96 AQueous nonradioactive assay (Promega, Mannheim, Germany), as recommended by the manufacturer. Data were analyzed using GraphPad Prism version 5.00 for Windows (GraphPad Software, San Diego, CA www.graphpad.com).

■

RESULTS Synthesis and Characterization of Polymer-Coated Gold Nanoparticles. The first stage of the nanoparticle synthesis is a well-defined reactive polymer precursor, poly(pentafluorophenyl methacrylate), PPFMA, which is obtained by RAFT (reversible addition−fragmentation chain transfer) polymerization.21,23 The obtained polymer was well-defined, with a number-average degree of polymerization of 91. (See the Supporting Information for full details.) Because of the activated ester side chain, PPFMA can be modified; postpolymerization with amines gives a library of functional poly(methacrylamides).24 Furthermore, during this process, the RAFT agent end group is removed, liberating a ω-terminal thiol group on each chain, which is available for further conjugation, as shown in Scheme 1A, and polymer structures are shown in Table 1. The well-established citrate reduction method was used to obtain gold nanoparticles of predetermined diameter, and particle sizes of 18, 35, and 65 nm were obtained, as determined by UV−visible and DLS analysis, shown in Table 2. These particles were subsequently coated by simple addition of Table 1. Polymers Used to Coat Gold Nanoparticles Used in This Studya

a R indicates the side-chain chemical functionality. Polymer A: ethanediamine; Polymer B: glucosamine; Polymer C: hydroxypropylamine; Polymer D: taurine; and Polymer E: linear PEG. The degree of polymerization for polymers A, B, C, and D = 91 (poly(methacrylamides) and 144 for polymer E (which is poly(ethylene glycol)).

C

dx.doi.org/10.1021/bm300248u | Biomacromolecules XXXX, XXX, XXX−XXX

Biomacromolecules

Article

24 and 48 h. None of the synthesized gold nanoparticles induced cell toxicity after 24 h of exposure (Figure 1A). Even after 48 h neither the different sizes nor the various coatings of the gold nanoparticles decreased the cell viability compared with the untreated control (Figure 1B) in the concentration range tested. Internalization of Polymer-Coated Gold Nanoparticles in Human Dermal Endothelial Cells. To determine if the gold nanoparticles were internalized by HDMECs, we treated cells with 100 μg/mL gold nanoparticles for 24 h. Cell membranes were visualized by staining of the CD31 membrane protein by using a specific antibody against CD31 and the corresponding fluorescently labeled secondary antibody. In some cases, sufficient gold particles were internalized to allow direct visualization (black dots) using optical microscopy, as shown in Figure 2. Most of the internalized gold nanoparticles were located in the perinuclear region after 24 h of exposure. The gold nanoparticles did not influence the cell morphology of HDMEC, even though in some cases high amounts of gold were detected within the cells. The particles were present as agglomerates not individual particles and therefore only gave a qualitative indication of cellular uptake, which allowed for some general observations. Less uptake of PEGylated gold nanoparticles (E) was observed, whereas the uptake of ethandiamine-modified particles (A), which exhibited a positive surface charge, was extremely high. A moderate uptake was detected after exposure to glucosamine (B)- and hydroxypropylamine (C)-modified gold nanoparticles, whereas the uptake of the smallest and the largest taurine (D)-modified gold nanoparticles (negative surface charge) was lower compared with the other gold nanoparticles. The internalization and intracellular distribution of the gold nanoparticles in HDMEC was confirmed by TEM analysis: because of the high density of the gold nanoparticles, these are visible without additional staining. The uptake of 35 nm gold nanoparticles with various surface modifications is presented in Figure 3. No gold nanoparticles were detected in the untreated control cells. Moreover, the same trend with respect to the amount of internalized particles, which was demonstrated by optical microscopy (Figure 2), was also observed by TEM analysis. High amounts of positive-charged ethanediaminecoated gold nanoparticles were found in HDMEC sections, whereas the neutral- (B,C) and negative-charged (D) particles were internalized in lower amounts. In addition, none of the PEGylated gold nanoparticles (E) were detected in the HDMEC sections. Because of the high resolution of the TEM, single particles were observed to be enclosed in vesicles. After increasing the incubation time, the particles were mostly located in the perinuclear region. However, it was not possible to determine

Table 2. Gold Nanoparticle Characteristics particle citrate/Au ratioa UVmax (nm) core diameter UV (nm)b core diameter DLS (nm)c

Au18

Au35

Au65

3.5:1 520 20 18

2:1 527 39 35

1:1 537 63 65

a

Ratio of reactants used during the synthesis of citrate-stabilized nanoparticles. bDiameter of the citrate-stabilized particles from UV− vis spectroscopy. cDiameter of the citrate-stabilized particles determined from dynamic light scattering.

the thiol-terminated polymers (Table 1), which form a selfassembled monolayer on the nanoparticle surface (Scheme 1 B). Centrifugation−resuspension cycles allowed for easy removal of excess polymer, and the particles were again characterized by DLS and UV−visible spectroscopy (Table 3). In all cases, there was an increase in the hydrodynamic diameter of the particles following the addition of the polymer, consistent with immobilization on the surface. There was also a small shift (2−10 nm) in the surface plasmon resonance band of the gold particles following functionalization, which is consistent with a change in the surface chemistry, and its magnitude does not suggest aggregation. Within this manuscript, the quoted size of the nanoparticles refers to the diameter of the core not the hydrodynamic diameter of the polymer-coated nanoparticle. Our previous studies, which developed this synthetic methodology,6 indicated a grafting density of ∼0.5 chains·nm−2 for amino, carboxyl, or PEG functional polymers, and corresponding shifts in their zeta potential were observed. In a previous study, we investigated the role of solution conditions on nanoparticle aggregation.6 Importantly, we demonstrated that in the endothelial cell culture medium containing 15% serum proteins all of the nanoparticles were well-dispersed. Conversely, in the absence of serum proteins, charged particles tended to aggregate. This is a key observation, as the interactions between nanoparticles and cells are expected to be diameter-dependent. The reasons for the dispersal of particles in the presence of serum proteins is not fully understood but is hypothesized to be due to absorption of the proteins on the nanoparticle surface. Considering these prior observations, all cell culture experiments were conducted in the presence of serum proteins to (i) ensure that particles were well-dispersed and (ii) to present conditions closer to those found in vivo in which blood plasma contains very high levels of dissolved serum proteins. Evaluation of Cytotoxicity. Primary HDMECs were incubated with increasing amounts of gold nanoparticles for

Table 3. Polymer-Coated Gold Nanoparticles Used in This Studya Au18 code

diameter (nm)

A@Au18 B@Au18 C@Au18 D@Au18 E@Au18

30 31 23 28 26

Au35 b

UVmax (nm) 528 523 522 519 522

c

code

diameter (nm)

A@Au35 B@Au35 C@Au35 D@Au35 E@Au35

50 46 46 44 42

Au65 b

UVmax (nm) 529 529 536 526 528

c

code

diameter (nm)b

UVmax (nm)c

A@Au65 B@Au65 C@Au65 D@Au65 E@Au65

76 75 73 75 71

546 559 555 535 540

a Letter refers to the polymer coated from Table. bObtained in water by dynamic light scattering. cRepresents the wavelength with maximum absorbance in the UV−visible spectrum and is the SPR peak (surface plasmon resonance).

D

dx.doi.org/10.1021/bm300248u | Biomacromolecules XXXX, XXX, XXX−XXX

Biomacromolecules

Article

Figure 1. Cell viability of HDMEC after exposure to various gold nanoparticles for 24 and 48 h. The endothelial cells HDMEC were exposed to different concentrations of gold nanoparticles (10, 50, 100, and 250 μg/mL) for 24 (A) and 48 h (B). The cell viability was measured by the MTS assay (reduction of a substrate solution to formazan by viable cells), with untreated cells (Ctrl) being set to 100% (highlighted by red horizontal line). Each column represents the mean ± standard deviation (SD) of three independent experiments (three different donors); each of these was performed at least in triplicate.

the mechanism of uptake and internalization of the gold nanoparticles in the TEM images. Quantification of the Amount of Internalized Gold Nanoparticles. Microscopic analysis is useful for obtaining qualitative information on uptake and for intracellular distribution but only evaluates a relatively small population of the cells. Therefore, a quantitative measurement was necessary. Because gold is not found in measurable concentration in mammalian cells, ICP-AES can be used to quantify the total amount of gold in a large number of cells and to determine cellular uptake and has sub −ppb detection limits. Following incubation of the cells with particles, the cell extracts were digested with aqua regia to dissolve the particles and the gold concentration determined by ICP-AES. The results of this analysis are shown in Figure 4. The quantification data were analyzed in two different but also complementary ways. First, the total number of particles per cells was determined. Because the diameter of the nanoparticles was known, it was possible to calculate the number of gold atoms per particle and hence the number of particles per cell, as the total number of cells/well was known. The second measurement was to determine the percentage uptake of the particles into the cells, as a function of the total amount applied. It is important to determine both measurements because for a therapeutic application a certain number of particles must enter the cell to deliver a sufficient quantity of a drug molecule. Conversely, as endothelial cells line the vasculature, the percentage of particles taken up by the cells is critical to understand how this affects biodistribution. In general, the quantification data presented in Figure 4 illustrated that the uptake by cells was time-dependent. Nearly all of the gold nanoparticles examined were taken up after 4 h, although the number increased after prolongation of the

incubation time to 24 h. In general, the smallest gold nanoparticles (18 nm) were internalized in the highest amount compared with the 35 nm and 65 nm-sized gold nanoparticles, whereas the internalized proportion of the initial concentration was always the lowest. In addition to the images presented in Figure 2, the quantification of the internalized gold nanoparticles showed that the positive-charged gold nanoparticles were internalized to the highest degree by HDMEC. The number of the smallest gold nanoparticles with an ethanediamine surface modification (A@Au18) was extremely high and significantly increased compared with A@Au35 and A@Au65, although only 25% of the initial nanoparticles in the suspension were internalized (Figure 4A,B). This amount was significantly decreased compared with the medium- and the largest-sized gold nanoparticles with the same surface modification after exposure for 4 and 24 h. However, after 24 h, most of the positive-charged nanoparticles in suspension were internalized by HDMEC. In contrast with the ethanediamine-coated gold nanoparticles (A), the PEGylated gold nanoparticles (E) were taken up to a very low extent. The PEGylated 18 nm-sized gold nanoparticles (E@Au18) were internalized best compared with their counterparts with diameters of 35 and 65 nm. However, even after 24 h of exposure, the amount of PEGylated gold nanoparticles (E@ Au18) within the cells is