Ultrastable-Stealth Large Gold Nanoparticles with ... - ACS Publications

Article pubs.acs.org/Langmuir

Ultrastable-Stealth Large Gold Nanoparticles with DNA Directed Biological Functionality Jun Hyuk Heo,† Kyung-Il Kim,† Hui Hun Cho,‡ Jin Woong Lee,† Byoung Sang Lee,† Seokyoung Yoon,‡ Kyung Jin Park,‡ Seungwoo Lee,‡,§ Jaeyun Kim,§ Dongmok Whang,†,‡ and Jung Heon Lee*,†,‡ †

School of Advanced Materials Science and Engineering, ‡SKKU Advanced Institute of Nanotechnology (SAINT), and §School of Chemical Engineering, Sungkyunkwan University (SKKU), Suwon, South Korea, 16419 S Supporting Information *

ABSTRACT: The stability of gold nanoparticles (AuNPs) in biological samples is very important for their biomedical applications. Although various molecules such as polystyrenesulfonate (PSS), phosphine, DNA, and polyethylene glycol (PEG) have been used to stabilize AuNPs, it is still very difficult to stabilize large AuNPs. As a result, biomedical applications of large (30−100 nm) AuNPs are limited, even though they possess more favorable optical properties and are easier to be taken up by cells than smaller AuNPs. To overcome this limitation, we herein report a novel method of preparing large (30−100 nm) AuNPs with a high colloidal stability and facile chemical or biological functionality, via surface passivation with an amphiphilic polymer polyvinylpyrrolidone (PVP). This PVP passivation results in an extraordinary colloidal stability for 13, 30, 50, 70, and 100 nm AuNPs to be stabilized in PBS for at least 3 months. More importantly, the PVP capped AuNPs (AuNP-PVP) were also resistant to protein adsorption in the presence of serum containing media and exhibit a negligible cytotoxicity. The AuNP-PVPs functionalized with a DNA aptamer AS1411 remain biologically active, resulting in significant increase in the uptake of the AuNPs (∼12 200 AuNPs per cell) in comparison with AuNPs capped by a control DNA of the same length. The novel method developed in this study to stabilize large AuNPs with high colloidal stability and biological activity will allow much wider applications of these large AuNPs for biomedical applications, such as cellular imaging, molecular diagnosis, and targeted therapy.

■

INTRODUCTION Gold nanoparticles (AuNPs) are one of the most commonly used nanomaterials for biological,1−5 environmental,6−8 analytical,9−12 and energy applications,13 because they are benign in terms of toxicity and have a simple but powerful surface chemistry and unique optical and electrical properties,14,15 that often change substantially according to changes in their size and shape.16,17 In particular, large AuNPs with a size between 30 and 100 nm exhibit very interesting properties. Through Mie’s theory, large AuNPs scatter more light, because they have larger optical cross sections.18 These strong scatterings allow large AuNPs to have significant surface enhanced spectroscopy (SES) effects such as surface enhanced Raman scattering (SERS) and surface enhanced fluorescence (SEF) that can be used to detect analytes with ultrahigh sensitivity.19,20 For biological imaging, large sized AuNPs are also preferred because AuNPs within those ranges (30−80 nm) have been reported to be the most efficiently taken up by cells.21 Therefore, AuNPs should be sufficiently large for a wide range of biological and biomedical applications such as cellular imaging, molecular diagnosis, and targeted therapy. Despite the promise of these large AuNPs, their applications have so far been quite limited, because they tend to be much © 2015 American Chemical Society

less stable than smaller AuNPs. A major factor in the stability of the NPs is the repulsive force that occurs between NPs, which is mainly induced by the electrostatic repulsion, steric effect, or formation of a hydration layer on the surface of the NPs.22,23 For small NPs, because their surface area-to-volume ratio is high, they remain stable in the presence of a few repulsively charged molecules that are functionalized on the surface. In contrast, stability becomes a critical issue for larger NPs because their large surface area induces high total surface energy, which is thermodynamically unstable. Consequently, the large NPs tend to aggregate to reduce their surface energy. To address this issue, stabilizers such as polyethylene glycol (PEG),24 polystyrenesulfonate (PSS),25 phosphine,26 or DNA23,27−29 are introduced to the NPs surface. However, some strategies have not worked well for large NPs, because it still is very difficult to overcome the dipole−induced dipole forces that occur between NPs with a repulsive interaction that is generated from molecules attached on the surface of the NPs. Therefore, Received: September 21, 2015 Revised: December 4, 2015 Published: December 6, 2015 13773

DOI: 10.1021/acs.langmuir.5b03534 Langmuir 2015, 31, 13773−13782

Article

Langmuir

were purchased from Bio-Rad (Hercules, CA, USA). Sodium dodecyl sulfate (SDS, electrophoresis grade) was obtained from Biopure Technology (Waterlooville, UK). Modification of Gold Nanoparticles with PVP. Ten milliliters of each AuNP solution (Au13NPs: 6 nM, Au30NPs: 0.6 nM, Au50NPs: 0.6 nM, Au70NPs: 0.04 nM, Au100NPs: 0.036 nM) was placed into a vial, pretreated with aqua regia. Next, 10 mL of polyvinylpyrrolidone (average Mw: 10 000 and 40 000 Da) aqueous solution (from 0.2 μM to 2 mM) was added to each AuNP solution. This solution was then stored in a dark room for 1 day before use. DNA Functionalization on AuNPs Capped with PVP. First, 1 mL of each PVP-capped AuNP (Au13NPs, Au30NPs, Au50NPs, Au70NPs, Au100NPs) solution was placed in a vial, which was pretreated with 1 M sodium hydroxide solution and rinsed several times with deionized water. 9 μL of 1 mM DNA stock solution was mixed with 1 μL of 500 mM acetate buffer solution (pH = 5.2) and a 3 μL aliquot of 10 mM trisodium citrate dihydrate (TCEP) and the mixed sample was incubated for 1 h to activate thiol-tagged ssDNA. Subsequently, thiol-tagged ssDNA stock solution (9, 2.33, 1.66, 1, and 0.5 μL of 0.69 mM DNA solution) was transferred to 1 mL of PVPcapped AuNP solution (Au13NP-PVP10k: 3 nM, Au30NP-PVP10k: 0.3 nM, Au50NP-PVP10k: 0.03 nM, Au70NP-PVP40K: 0.02 nM, Au100NPPVP40K: 0.018 nM), respectively. After 16 h of incubation, we added 100 μL of 1 M NaCl solution to the sample and sonicated it for 10 s after gentle shaking to minimize the physical adsorption and increase the chemical conjugation of the DNA on the AuNP surface. This solution was stored in a dark room for 1 day before use. Complementary Reaction of ssDNA Functionalized on AuNP-PVP. One milliliter of AuNP-PVP solution (Au13NP-PVP, Au30NP-PVP, Au50NP-PVP, Au70NP-PVP, Au100NP-PVP) treated with thiol-tagged ssDNA (DNA1) was centrifuged (3 times at 15 000 rcf, for 10 min) to remove free DNA in the solution and then redispersed in 500 μL of PBS solution. AuNP-PVP solutions functionalized with the complementary DNA2 (DNA2-AuNP-PVP solution) and a noncomplementary DNA3 (DNA3-AuNP-PVP solution) were prepared in a similar manner. Subsequently, 100 μL of DNA1-AuNP-PVP solution (DNA-Au13NP-PVP10k: 6 nM, DNA-Au30NP-PVP10k: 0.6 nM, DNA-Au50NP-PVP10k: 0.06 nM, DNA-Au70NP-PVP40K: 0.04 nM, DNA-Au100NP-PBP40K: 0.036 nM) was mixed with same amount of either DNA2-AuNP-PVP solution or DNA3-AuNP-PVP solution and the extinction spectra of the solution was measured using a UV−vis spectrometer. Fluorescence of 6-FAM Tagged ssDNA in the Presence of DNA-AuNP-PVP. For purification, AuNP-PVP (Au 13NP-PVP, Au30NP-PVP, Au50NP-PVP, Au70NP-PVP, Au100NP-PVP) treated with ssDNA was centrifuged 3 times at 15 000 rcf for 10 min. After discarding the supernatant, the precipitated DNA-AuNP-PVP was dispersed in PBS. A concentrated DNA-AuNP-PVP solution (DNAAu13NP-PVP10k: 3 nM × 10 μL of solution, Au30NP10k: 2 nM × 10 μL of solution, Au50NP10k: 3 nM × 10 μL of solution, Au70NP-PVP40K: 10 nM × 10 μL of solution, Au100NP40K: 15 nM × 10 μL of solution) was added to the 90 μL × 1 μM of 6-FAM tagged ssDNA solution, and the variation in the fluorescence signal of FAM was measured using a fluorimeter (BioTek, Winooski, VT, USA). Cell Uptake of DNA-AuNPs-PVP and AuNP-PVP. Ten thousand MDA-MB-231 cells were plated in 48 wells of a 96-well plate for 24 h. After removal of the growth media, RPMI-1640 media containing 300 μL × 0.3 nM of Au50NP-PVP10k (or Apt-Au50NP-PVP10k/cAptAu50NP-PVP10k) were added to each well and were incubated for another 24 h. The MDA-MB-231 cells incubated with Au50NP-PVP10k (or Apt-Au50NP-PVP10k/cApt-Au50NP-PVP10k) were then treated with trypsin and were centrifuged at 8000 rcf for 5 min to form a cell pellet. Each pellet was resuspended and washed three times with PBS. The amount of Au50NPs that were taken up into each cell were measured by using an inductively coupled plasma mass spectrometer (ICP-MS, Agilent 7500), and images of the Au50NPs taken up by cells were taken using a dark field microscope (Nikon, Eclipse Ni).

developing novel methods to stabilize large AuNPs is critical to their successful application. Polyvinylpyrrolidone (PVP) is a U.S. Food and Drug Administration (FDA) approved polymer that is extensively used in pharmaceutical applications, personal products, cosmetics, and so forth.30,31 PVP is similar to PEG because it is a flexible, highly water-soluble molecule due to its electronegative O and N in the pyrrolidone side chain. Its amphiphilic structure allows PVP to interact with nanomaterials and to change their surface properties. Therefore, a few examples of the use of PVP as a stabilizing agent for gold nanoparticles (AuNPs), silver nanoparticles (AgNPs), magnetic NPs, and palladium NPs have been already reported.32−35 PVP is also used as a reducing agent in the synthesis of nanoparticles, including AuNSs36 and AgNPs.37 However, the studies have only shown limited use of PVP as a passivating agent, because it has been mostly used on small nanoparticles that do not exhibit significant issues in terms of the colloidal stability.32,33 In this work, we demonstrate using PVP as a versatile passivating agent to make large colloidal AuNPs that are highly functional with outstanding colloidal stability in complex biological media. The PVP-capped AuNPs (AuNP-PVP) exhibited not only long-term stability in PBS (>3 months), regardless of their size, but also had negligible binding with proteins in serum containing media. More importantly, the AuNPs that were passivated with PVP could further undergo a direct chemical reaction with thiolate moieties, such as DNA, while preserving their colloidal stability in PBS. The DNA strands attach directly to the AuNP surface of the AuNP-PVP complex and maintain their biologically activity, which is very difficult to achieve when the AuNPs are passivated with thioltagged PEG. Finally, a significant uptake of the AuNPs into MDA-MB-231 cancer cells occurs when the cells were treated with AuNP-PVP functionalized with a cancer cell-targeting molecule, AS1411 aptamer. To the best of our knowledge, this work is the first to investigate the stability issues of large (30−100 nm) AuNPs in complex biological media and demonstrate a novel method of using PVP as a powerful NP passivating agent that ensures ultrahigh stability of large AuNPs in biological media, allows chemical conjugation and biological functionalization, while preventing nonspecific interactions with proteins and cells.

■

EXPERIMENTAL SECTION

Materials. Gold(III) chloride hydrate (HAuCl4·3H2O), sodium citrate dihydrate (C6H5Na3O7·2H2O), trisodium citrate dihydrate, polyvinylpyrrolidone (average Mw 10 and 40 kDa), poly(ethylene glycol) methyl ether thiol (average Mw 6 kDa), and all other materials were purchased from Sigma-Aldrich (St. Louis, MO, USA). 100 nm AuNPs (Product code: EM.GC100) were purchased from BBI Solution (Madison, WI, USA). Deionized water (18.2 MΩ·cm) was prepared with a Sartorius Arium pro Ultrapure water system. The oligonucleotides used in this work were purchased from Integrated DNA Technologies Inc. (Coralville, IA, USA) and Bioneer Corp. (Daejeon, Korea) and were used after purification via high performance liquid chromatography. RPMI-1640 cell culture medium, penicillin/streptomycin, and trypsin/EDTA were purchased from Lonza (Walkersville, MD, USA). Fetal bovine serum (FBS) was purchased from RMBIO (Missoula, MT, USA). DMSO was purchased from Samchun Chemical (Pyeongtaek, Korea). 3-(4,5-dimethylthiazol2-yl)-2,5-diphenyltetrazolium bromide (MTT) and 1,4-Dithio-DLthreitol (DTT) were purchased from Alfa Aesar (Ward Hill, MA, USA). 4× Laemmli sample buffer, Coomassie Brilliant Blue R-250 staining buffer and Coomassie Brilliant Blue R-250 destaining solution 13774


Article

Langmuir

Figure 1. Stability of sub-100 nm AuNPs capped with 0.5 mM of thiol tagged PEG (Mn = 6 kDa). (A,B) The 13 nm AuNPs capped with PEG (Au13NP-PEG) were very stable in PBS with no change in color and minimal red-shift of the extinction spectrum after 4 days of incubation in PBS. The color of the Au30NP-PEG did not change significantly in the presence of PBS. However, the absorbance of Au30NP-PEG decreased after incubation in PBS. The color and extinction spectra of the 50 nm, 70 nm, 100 nm AuNPs (Au50NPs, Au70NPs, Au100NPs) capped with PEG changed substantially after incubation in PBS. (C) The normalized extinction ratio of each AuNP-PEG in PBS (Au13NP-PEG: E523/E700, Au30NP-PEG: E528/ E700, Au50NP-PEG: E535/E700, Au70NP-PEG: E552/E750, Au100NP-PEG: E578/E800).

■

RESULTS AND DISCUSSION Stability of PEG-Capped AuNPs in Biological Media. Polyethylene glycol (PEG) has been widely used to help stabilize many kinds of nanoparticles including AuNPs, because it has a neutral charge and forms strong hydrogen bonds with water molecules; such a property prevents the NPs from electrostatically binding with charged molecules. In addition, as the size of the PEGs becomes larger than the London interaction range, the steric hindrance or depletion force provided by the PEG can help stabilize various NPs.23,38,39 The PEG tagged with thiol group has also found wide use in surface modification of metal NPs due to their low stability, toxicity, and resistance of protein adsorption. However, most studies have only shown the use of thiol-tagged PEG on small NPs (≤20 nm) that do not find serious problems with colloidal stability. So, we became curious whether thiol-tagged PEG can also efficiently stabilize the large NPs in biological buffer. To confirm this, we treated AuNPs of five different sizes (13, 30, 50, 70, and 100 nm) with 0.5 mM of thiol-tagged PEG (AuNPPEG, Mn = 6 kDa), based on previous studies, and then incubated them in PBS.40−43 As we expected, 13 nm AuNP capped thiolate PEG (Au13NP-PEG) remained red color in PBS even after 4 days, and this insignificant color change was also monitored by the UV−vis spectrometer, which showed little change in the surface plasmon resonance peak. This result indicates that Au13NP-PEG remains stable in PBS even after 4 days. In the case of Au30NP-PEG, we did not observe any color change of sample, when we dispersed them in PBS. After 4 days of incubation in PBS, we still did not observe a significant

change in the color of Au30NP-PEG with the naked eye, but the UV−vis measurements revealed a slight decrease in the absorbance at 523 nm, suggesting that there is little aggregation of Au30NP-PEG in PBS. The color of the Au50NP-PEG solution remained pink when they were dispersed in PBS for a few hours. On the other hand, the solution became light purple after 1 day of incubation in PBS. This change of color and redshift of the plasmon resonance peak also occurred from 70 and 100 nm AuNPs treated with 0.5 mM thiol-tagged PEG. In order to quantify the stability of AuNP-PEGs, we monitored the normalized extinction ratio between the wavelength with maximum extinction and the wavelength in the scattering region (700−800 nm) after incubating each AuNP-PEG in PBS for a week (see Figure 1C). So, the normalized extinction ratio of AuNPs will decrease from 1 as they become aggregated. The normalized extinction ratio of Au13NP-PEG decreased from 1 to ∼0.8 within a week, while the AuNPs larger than 50 nm became substantially aggregated with extinction ratio lower than 0.5 in a few days. This result suggests that thiol-tagged PEG (Mn = 6 kDa) provides stability on AuNPs but its effect may be limited for a few hours or days for large AuNPs. We also carried out another experiment to passivate the AuNPs with other reported stabilizers, polystyrene soulfonate (PSS) and bis(p-sulfonatophenyl)phenylphosphine (BSPP). However, we observed an instantaneous aggregation of all AuNPs, when we dispersed them in PBS (see Figure S1). So, we thought that it is necessary to find another method for providing stability of large NPs in order to use them in biological buffer. 13775


Article

Langmuir

Figure 2. Effect of PVP (MW = 10 kDa) on the stability of 50 nm AuNPs (Au50NPs) in PBS. (A) PVP is a neutral but amphipathic molecule. Therefore, once PVP interacts with the AuNPs, its hydrophobic carbon chain will bind to the AuNP surface while its highly polar part will remain exposed on the AuNP surface. (B) Color of Au50NP capped with different concentrations of PVP10k. A PVP concentration of 500 μM is enough to maintain the stability of Au50NPs in PBS for a long-term (>3 months). (C) Extinction spectra of Au50NPs capped with different concentrations of PVP10k. The extinction spectrum of Au50NP-PVP10k taken in PBS was similar to the spectrum of bare Au50NP taken in DI water. (D) The normalized extinction ratio of each AuNP-PVP in PBS (Au13NP-PVP10k: E523/E700, Au30NP-PVP10k: E528/E700, Au50NP-PVP10k: E535/E700, Au70NP-PVP40k: E552/ E750, Au100NP-PVP40k: E578/E800).

Significantly Enhanced Stability of PVP-Capped AuNPs. After we confirmed that it was difficult to stabilize large AuNPs in PBS, we searched for a candidate molecule to stabilize large NPs in biological buffer. Polyvinylpyrrolidone (PVP) is water-soluble polymer with no charge, like PEG.44 Since PEG has a linear structure with no particular chemical affinity for the surface of the AuNP, it is mostly attached to the surface of the AuNPs through the thiol-gold chemistry. In contrast, PVP is an amphiphilic molecule that has a hydrophobic chain and highly polar pyrrolidone side groups. Therefore, once the hydrophobic carbon chain of the PVP binds to the surface of the AuNPs, the highly polar side groups can be densely exposed on the surface of the AuNP (see Figure 2A). We thus wondered whether this amphiphilic structure of PVP can help passivate AuNPs in PBS more effectively. To test this hypothesis, we first treated Au13NP with different amounts of 10 kDa PVP (PVP10k) to form PVP-capped 13 nm AuNPs (Au13NP-PVP10k). The samples were washed three times by centrifugation to remove the free PVP molecules from the Au13NP-PVP10k, which was then transferred into PBS. Its stability was determined by monitoring both the color changes and their extinction spectra via UV−vis. The Au13NP treated with 0.2 μM or lower concentration of PVP exhibited a change in color from red to blue in PBS. The red-shift of the surface plasmon resonance peak of Au13NP-PVP10k was also observed in the extinction spectra. In contrast, the Au13NP-PVP10k prepared with a 0.5 μM or higher concentration of PVP did not exhibit any color changes in PBS for more than a year (see Figure S2A). This negligible change in color for the Au13NPPVP10k was also verified by the insignificant change in the extinction spectra. These results thus show that the PVPs can

successfully passivate and provide long-term stability of the Au13NPs in PBS. Encouraged by the result of PVP stabilization of Au13NP, we investigated whether the same strategy could also stabilize AuNPs larger than 13 nm. We carried out parallel experiments by capping Au30NPs, Au50NPs, Au70NPs, and Au100NPs with PVP10k in a similar manner and evaluated their changes in color and the extinction spectra from UV−vis measurements after placing them in PBS. The Au30NPs became stable when a 50 μM or higher concentration of PVP10k was used (see Figure S2B). For the Au50NPs, a 500 μM or higher concentration of PVP10k was needed to achieve stability in PBS (see Figure 2B). Excitingly, both the Au30NP-PVP10k and Au50NP-PVP10k remained stable in PBS with no color change and no shift in extinction spectra for at least 3 months. These results indicate that the PVP10k can render both Au30NPs and Au50NPs stable in PBS. However, the PVP10k could not make Au70NPs and Au100NPs stable since both Au70NP-PVP10k and Au100NPPVP10k aggregated within 3 days in PBS (data not shown). To provide stability for these larger AuNPs, we employed PVPs with larger molecular weight than PVP10k, and found out that 40 kDa PVP (PVP40k) could provide protection of the Au70NPs and Au100NPs. The Au70NP-PVP40k remained stable in PBS when 2 μM or higher concentration of PVP40k was used to cap the AuNPs (see Figure S3A). For the Au100NPs, 7 μM of PVP40K was sufficient to keep them stable in PBS (see Figure S3B). As a result, both Au70NP-PVP40k and Au100NP-PVP40k maintained their colloidal stability for at least 3 months in PBS. These results indicate that an optimal molecular weight and amount of PVP are required to make AuNPs stable and these parameters depend on many factors, including the size of the 13776


Article

Langmuir

Figure 3. Chemical functionalization and biological activity of the Au50NP-PVP10k. (A) In the presence of thiol-tagged ssDNA, PVP physically capped on AuNP surface could be partially replaced with ssDNA and can form DNA-AuNP-PVP. (B) The biological activity of the ssDNA functionalized on AuNP-PVP was investigated by monitoring the fluorescence signal of a fluorescent dye (6-FAM) tagged on the complementary ssDNA. The hybridization of 6-FAM tagged ssDNA (DNA1-FAM) with complementary ssDNA functionalized on AuNP-PVP (DNA2-AuNP-PVP) will place FAM close the surface of the AuNPs, resulting in the quenching of the fluorescent signal. (C) Au50NP-PVP10k samples functionalized with a pair of complementary ssDNA (DNA1-Au50NP-PVP10k and DNA2-Au50NP-PVP10k) remained pink with a minimal peak shift in the extinction spectra. When DNA1-Au50NP-PVP10k and DNA2-Au50NP-PVP10k were mixed together, a color change of the AuNP from pink to purple occurred with a red-shift of the surface plasmon resonance peak. This color change and shift of the extinction spectra was reversible after heating the sample beyond the melting temperature of the ssDNA. On the other hand, none of these changes were observed when Au50NP-PVP10k functionalized with two noncomplementary ssDNA (DNA1-Au50NP-PVP10k and DNA3-Au50NP-PVP10k) were mixed together. (D) Fluorescence signal of 6-FAM tagged ssDNA (DNA1-FAM) decreased up to 80.5% in the presence of DNA2-Au50NP-PVP10k. On the other hand, the fluorescence intensity of the DNA1-FAM showed a minimal decrease in the presence of Au50NP-PVP10k functionalized with noncomplementary ssDNA (DNA3-Au50NPPVP10k).

Biological Activity of ssDNA Functionalized on AuNPPVP. After we determined that the PVP can stabilize large AuNPs in PBS for at least a few months, we were then curious about whether or not PVP-capped AuNPs could also undergo chemical conjugation with biological or chemical moieties, while maintaining their stability in biological media. The chemical conjugation of NPs with high stability is very important for many applications, including drug delivery, cell and animal imaging, diagnostics, and environmental monitoring. However, to the best of our knowledge, no method has been reported thus far to provide both colloidal stability and functional flexibility for large AuNPs.34−36 In order to investigate this possibility, we chose a 3′-thioltagged single-stranded DNA (ssDNA) as a model molecule for chemical conjugation. Since the covalent gold−thiol interaction is stronger than the physisorption of PVP on the surface of the AuNPs, we anticipated that the 3′-thiol-tagged ssDNA can still bind to the surface of the PVP-capped AuNPs. Functionalization of only a small number of DNA molecules might not cause serious changes in the stability of the AuNP-PVP. However, a significant amount of ssDNA functionalization on the AuNPPVP can seriously influence their stability because PVP that is physically absorbed on the AuNP surface can be replaced with thiol-tagged ssDNA. Therefore, we anticipated that there will be a maximum amount of ssDNA that can be functionalized on a stable DNA-AuNP-PVP complex. To verify this assumption, we first prepared stable Au50NP-PVP10k in PBS and functionalized them with a different amount of thiol-tagged ssDNA (3′thiol-tagged ssDNA) to find the conditions under which DNAAuNP-PVP remained stable. An optimal condition was found

AuNPs and the ionic strength of media. In order to quantify the optimal amount of PVP that are necessary to make each AuNP stable, we prepared 500 μL of AuNP-PVP samples (Au13NPPVP10k, Au30NP-PVP10k, Au50NP-PVP10k, Au70NP-PVP40k, Au100NP-PVP40k) which were rigorously washed by three times of centrifugation but remained highly stable in PBS. Subsequently, 100 μL of DTT (1 mM) was added to the each sample, and the mixture was then incubated for 12 h at 25 °C to isolate the surface-capped PVP. After taking the PVP supernatant by centrifugation (3 times at 18 000 rcf, 4 °C), we measured the absorbance of supernatant by using the UV−vis spectrometer. Due to the presence of carbonyl group in ring structure, PVP shows only one peak (207 nm) in the UV−vis spectrum. The intensity of 207 nm peak increased with increasing the size of AuNPs. Therefore, we could construct a calibration curve of the peak, and then calculate the number of PVP attached on the each AuNP. The number of PVP10k molecules attached on Au13NP, Au30NP and Au50NP turned out to be approximately 53, 416, and 876, while 225 and 500 molecules of PVP40k had been attached on the surface of Au70NP and Au100NP, respectively (see Figure S4). We also monitored the normalized extinction ratios of PVP passivated AuNPs of different sizes in order to quantify the stability of AuNP-PVPs in PBS. As shown in Figure 2D, most ratios of AuNP-PVP samples remained above 0.8 for 3 months in PBS, indicating that PVP can help to maintain the colloidal stability of most AuNPs in biological media. Therefore, we could conclude that PVP is an appropriate molecule that can make large AuNPs extremely stable in biological media, such as PBS. 13777


Article

Langmuir when 1.66 μL of thiol-tagged ssDNA (0.69 mM) were treated with 1 mL of Au50NP-PVP10k (0.03 nM). The number of ssDNA that was actually attached on Au50NP-PVP10k was estimated by functionalizing 6-FAM (carboxyfluorescenin) tagged ssDNA on the Au50 NP-PVP10k under identical conditions and measuring the fluorescence of the ssDNA after detaching them from the AuNP surface.45 This method revealed that approximately 364 DNA strands (ssDNA) had been attached on the surface of the Au50NP-PVP10k. On the other hand, we found that in the absence of PVP passivation, Au50NP could be fully attached with approximately 723 DNA strands (ssDNA), which is nearly twice the number of ssDNA functionalized on the Au50NP-PVP10k. We also carried out parallel experiments to determine the number of ssDNA that was functionalized on the Au13NP-PVP10k, Au30NP-PVP10k, Au70NP-PVP40k, and Au100NP-PVP40k samples. The amount of ssDNA that was functionalized on each was approximately 65, 150, 738, and 1238, which is also approximately half of the ssDNA that was functionalized on Au13NP, Au30NP, Au70NP, and Au100NP without PVP (see Figure S5). To determine whether the ssDNA that was chemically conjugated on the surface of the AuNP-PVP has biological activity, we prepared two samples of Au50NP-PVP10k functionalized with a pair of complementary 5′-thiol-tagged ssDNA (DNA1-Au50NP-PVP10k and DNA2-Au50NP-PVP10k). Both samples remained pink with a minimal peak shift in the extinction spectra in PBS (see Figure 3C), indicating that the DNA-functionalized Au50NP-PVP10k can remain stable in PBS. On the other hand, when we mixed the DNA1-Au50NP-PVP10k and DNA2-Au50NP-PVP10k, a change occurred in the color of Au50NP from pink to purple. This color change was also observed in the red-shift of the extinction spectra. When we heated the mixed sample up to 70 °C, beyond the melting temperature of the two complementary ssDNA strands, we observed the color of the sample changing from purple to pink with a blue-shift in the extinction spectra. These results indicate that the change in color occurred for the mixed sample through the hybridization of two complementary ssDNA strands (DNA1 and DNA2) functionalized on the AuNP surface. When we mixed two noncomplementary DNA-Au50NP-PVP10k complexes (DNA1 and DNA3) to conduct a control experiment, we did not observe a change in color for the mixed sample. This indicates that DNA-Au50NP-PVP10k aggregation occurs selectively, based on the sequence of the ssDNA functionalized on the Au50NP-PVP10k surface. To ensure that the ssDNA functionalized on other AuNPs was also active, we conducted parallel experiments with Au13NP-PVP10k, Au30NPPVP10k, Au70NP-PVP40k, and Au100NP-PVP40k and found similar aggregation from all complementary paired DNAAuNP-PVP (see Figures S6 and 7). These results suggest that the ssDNA that was functionalized on the AuNPs was all active and induced the aggregation of AuNPs regardless of the size of the AuNPs. Although ssDNA-induced AuNP aggregation is a general way to determine the activity of ssDNA functionalized on AuNPs, it might not be the best way to quantitatively determine the biological activity of the ssDNA since AuNP aggregation can only happen when multiple ssDNA strands act together. Therefore, we designed another experiment to determine the activity of individual ssDNA functionalized on the Au50NPPVP10k surface (see Figure 3B). The fluorescent signal of a fluorescent dye has been reported to be quenched due to electron transfer close to the AuNP surface.46,47 Therefore, we

carried out an experiment to place a FAM dye close to the AuNP surface through the biological activity of the ssDNA functionalized on the Au50NP-PVP10k surface. We first prepared an 18-mer ssDNA tagged with 6-FAM (6-FAM tagged DNA1) on its 3′ end. Then we prepared DNA2-Au50NP-PVP10k with a 5′-thiol tagged ssDNA (DNA2) complementary to the 6-FAM tagged ssDNA1. In order to ensure that FAM quenching is maximized after hybridization, we placed a thiol modification on the 5′ end of DNA2 so that FAM would be located close to the AuNP surface after hybridization of the two strands. Subsequently, we mixed 6-FAM tagged ssDNA1 with DNA2Au50NP-PVP10k and monitored the variation in the intensity of the fluorescence by using a fluorimeter. As shown in Figure 3D, we observed a decrease of approximately 80.5% in the fluorescence signal of the 6-FAM tagged ssDNA1 in the presence of DNA2-Au50NP-PVP10k. We conducted two control experiments to ensure that a significant decrease had occurred in the fluorescence due to the sequence-specific hybridization of the two complementary ssDNA. First, when we substituted the ssDNA of DNA2-Au50NP-PVP10k with a noncomplementary 5′-thiol tagged ssDNA (DNA3), we observed a relatively small decrease (13.1%) in the fluorescence intensity. We also observed a similar decrease (13.2%) in the fluorescence intensity when we replaced DNA2-Au50NP-PVP10k with Au50NP-PVP10k with no ssDNA functionalized on the surface. We attribute this slight decrease in the fluorescence intensities of the two control samples to the energy transfer of light emitted from the FAM dyes to the AuNP.46 These results thus verify that the 5′-thiol tagged ssDNA functionalized on Au50NP-PVP10k is biologically active with sequence specificity and that the PVP functionalized on the AuNP surface does not hinder the activity of the functionalized ssDNA. When we carried out analogous experiments with Au13NP-PVP10k, Au30NP-PVP10k, Au70NP-PVP40k, and Au100NP-PVP40k, we observed similar quenching of the fluorescence (see Figures S8 and 9). These results indicate that the ssDNA strands that was functionalized on the AuNP-PVP maintain their biological activity regardless of the size of the AuNPs and of the molecular weight of the PVP. Although we could obtain facile chemical functionality of AuNPs by passivating them with PVP, a similar chemical functionality could not be achieved by replacing the PVP with thiol-tagged PEG (see Figure S10). Since PEG molecules are already attached on the AuNP-PEG surface through the Authiol chemistry, it was very difficult to replace PEG with a thioltagged DNA. So, thiol-tagged PEGs with carboxylic and amine modification placed at the other end are generally used for additional functionalization.48,49 Liu’s group reported direct chemical functionalization of DNA on NP surface by passivating them with PEG with no thiol termination through depletion stabilization.23 Passivation of AuNPs with PVP allows direct conjugation of DNA molecules on the surface of AuNP with excellent colloidal stability due to the hydrophobic interaction of the PVP and AuNP surface. Negligible Serum Protein Adsorption on PVP Capped AuNP. After demonstrating the stability of the PVP capped AuNPs in buffer, we investigated whether PVP could help to maintain the colloidal stability of the AuNPs in protein-rich biological media with various components including amino acids, salts, glucose, and so forth. To assess their stability, we incubated Au50NP-PVP10k in a standard cell culture medium (Dulbecco’s modified Eagle’s medium, DMEM) supplemented with 10% fetal bovine serum (FBS) and monitored the 13778


Article

Langmuir

Figure 4. (A) SDS-PAGE protein binding pattern of Au50NP with citrate, PVP10k, and thiol tagged PEG6k, incubated with 10% FBS in PBS. The protein molecular weight marker (lane 1) is shown on the left for reference. Citrate-capped Au50NP showed blue bands in the region of 55−100 kDa, indicating protein adsorption on the surface of the Au50NP. In contrast, Au50NP capped with PVP10k and PEG6k showed an insignificant blue band in the range from 55 to 70 kDa (minimal protein adsorption of Au50NPs surface). (B) Quantification of the adsorbed proteins on Au50NP, Au50NP-PVP10k, and Au50NP-PEG incubated with 10% FBS in PBS.

Au50NP-PVP10k, and Au50NP-PEG, we first scanned gels (8-bit grayscale), and the digitized images were analyzed to confirm the relative amount of proteins of each lane using Matlab program.50 When we treated Au50NPs with 10% FBS in PBS, we calculated that relatively 53.5% of proteins were bounded on the surface of Au50NPs as shown in Figure 4B. However, we observed a significant decrease of protein adsorption on Au50NP-PVP10k (3.8%) and Au50NP-PEG (3.8%). These results indicate that PVP can help inhibit the absorption of proteins on the AuNP surface significantly, like PEG. It has been proposed that preventing the nonspecific binding of protein on the NP surface by PEG is a result of its neutral charge and strong hydrogen interaction with water. We think that PVP helps prevent protein binding on the AuNP surface in a similar way. Although PVP is an amphipathic molecule, its hydrophobic chain binds to the surface of the AuNPs, exposing highly electronegative oxygen of the carbonyl group after the formation of AuNP-PVP. This was proven by the zeta potential measurement verifying the slightly negative charge of PVP capped AuNPs, similar to the PEG capped AuNPs (see Table S2). Since PVP is a molecule that is generally considered to be safe and has been approved by the FDA for a broad number of applications in biological,51 medical,52 and environmental fields,53 we reasoned that they will not exhibit significant cytotoxicity when used with AuNPs. Therefore, we tested the cytotoxicity by conducting a MTT cell viability assay after treating MDA-MB 231 breast cancer cells with Au50NP-PVP10k. The viabilities of the cells treated with 0.4−40 nM (5−500 μM of Au) of Au50NP-PVP10k for 24 h were similar to those of the control sample with no Au50NP-PVP10k treatment (see Figure S12). In addition, we did not observe a significant variation in the cell viability after treating the cells with 8 nM (100 μM of Au) Au50NP-PVP10k for up to 3 days (see Figure S13). These results indicate that the PVP-capped AuNP does not exhibit significant cytotoxicity under these conditions. We compared the cytotoxicity of PVP with that of PEG by conducting control experiments where Au50NP-PVP was replaced with Au50NPPEG. The results showed that the viability of cells treated with Au50NP-PVP was comparable to that of cells treated with

variation in the extinction spectra of the AuNP with UV−vis. The extinction spectra of the Au50NP-PVP10k did not exhibit a significant change after 24 h of incubation in DMEM (see Figure S11). In order to compare the stability of the PVPcapped AuNPs to those with PEG, we also conducted a control experiment with Au50NP-PEG and took their extinction spectra after 24 h of incubation in DMEM with FBS as well. The extinction spectrum of Au50NP-PVP10k was shown to be very similar to that of Au50NP-PEG, which indicates that PVP is comparable to PEG in terms of stabilizing Au50NPs in proteinrich media. On the other hand, when we carried out a control experiment with citrate-capped (bare) Au50NP, the extinction spectra significantly decreased with a red-shift of the plasmon resonance peak. This result shows that the AuNPs become unstable in protein-rich media in the absence of a passivating molecule, such as PVP or PEG. After we found that PVP provides excellent colloidal stability for AuNPs in protein-rich media, we next wondered whether PVP could also prevent nonspecific protein binding on the AuNP surface. To determine the amount of protein absorbed on the AuNP surface, we incubated Au50NP-PVP10k (1.8 × 1013 particles/mL) in dilute FBS (10% in PBS) solution for 3 h at 37 °C. We subsequently separated the Au50NPs from nonadsorbed proteins by centrifugation and then desorbed the molecules that were attached to the AuNP surface by treating the AuNPs with 500 mM dithiolthreitol (DTT) with 10% sodium dodecyl sulfate (SDS). Finally, we ran SDS polyacrylamide gel electrophoresis (SDS-PAGE) with a supernatant after separating the AuNPs through centrifugation (see Figure 4A). As controls, we also carried out parallel experiments with Au50NPPEG and bare Au50NP to compare their properties to those of PVP-capped Au50NP. As shown in lane 3, a thick band was observed within the range from 50 to 100 kDa for the sample prepared from citrate capped (bare) Au50NP, which indicates that a certain amount of FBS proteins are absorbed on the surface of the bare Au50NP. In contrast, the sample prepared from Au50NP-PVP produced a single but very faint band in the range from 50 to 70 kDa (see lane 4). This band was similar to the one prepared from Au50NP-PEG, as shown in lane 5. To quantitatively compare the protein adsorption of Au50NP, 13779


Article

Langmuir

Figure 5. Cellular uptake of 50 nm AuNPs (Au50NPs) by MDA-MB-231 cells. (A) Dark field microscopy images of MDA-MB-231 cells (control) and those treated with Au50NP-PVP10k, control aptamer (cApt) functionalized Au50NP-PVP10k, and AS1411 (Apt) functionalized Au50NP-PVP10k. The bright yellow dots are Au50NPs in MDA-MB-231 cells. (B) Quantification of the amount of gold present in MDA-MB-231 cells. After 48 h incubation of cells treated with Apt-Au50NP-PVP10k and cApt-Au50NP-PVP10K, 12 200 and 2912 AuNPs were taken up into each cell, respectively. However, without DNA functionalization with Au50NP-PVP10k, the amount of gold taken up into each cell was less than the detection limit of ICPMS equipment that we used (