- Email: [email protected]
The effect of gold nanoparticle size on osteogenic differentiation of adipose-derived stem cells
Journal of Colloid and Interface Science 438 (2015) 68–76
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
The effect of gold nanoparticle size on osteogenic differentiation of adipose-derived stem cells Wan-Kyu Ko a,1, Dong Nyoung Heo a,1, Ho-Jin Moon a, Sang Jin Lee a, Min Soo Bae a, Jung Bok Lee a, In-Cheol Sun b, Hoon Bong Jeon c, Hun Kuk Park c, Il Keun Kwon a,⇑ a Department of Maxillofacial Biomedical Engineering and Institute of Oral Biology, School of Dentistry, Kyung Hee University, 26, Kyungheedae-ro, Dongdaemun-gu, Seoul 130-701, Republic of Korea b Biomedical Research Center, Korea Institute of Science and Technology, 39-1 Hawolgok-dong, Seongbuk-gu, Seoul 136-791, Republic of Korea c Department of Biomedical Engineering, College of Medicine, Kyung Hee University, 26, Kyungheedae-ro, Dongdaemun-gu, Seoul 130-701, Republic of Korea
a r t i c l e
i n f o
Article history: Received 26 June 2014 Accepted 28 August 2014 Available online 16 September 2014 Keywords: Gold nanoparticles Human adipose-derived stem cell Uptake Osteogenesis Osteogenic differentiation
a b s t r a c t There have been many medical applications based on gold nanoparticles (GNPs) over the past several centuries. Recently, researchers have focused on bone tissue engineering applications utilizing GNPs. The effect of various sizes of gold nanoparticles on the differentiation of human adipose-derived stem cells (ADSCs) into osteoblasts was investigated. The concentration of gold nanoparticles was fixed at 1 lM and varying sizes of 15, 30, 50, 75 and 100 nm (spherical GNPs) were used. The lack of cytotoxicity was confirmed by establishing viability of ADSCs using cell counting kit-8 (CCK-8) and live/dead assays. The results showed that each size of GNPs had no significant toxicity on ADSCs during 1 week of incubation. Osteogenic differentiation of ADSCs was confirmed by alkaline phosphatase (ALP) staining, ALP activity, calcium deposition, and real time PCR experiments. It was found, through dark field assays and microscope cell images, that 30 nm and 50 nm GNPs were preferentially up taken into the ADSCs. As expected, all sizes of gold nanoparticles promoted the differentiation of ADSCs toward osteoblasts more than control. Among all sizes, 30 and 50 nm GNPs appeared to have the highest differentiation rates. The data consistently demonstrated that 30 and 50 nm GNPs are the most effective in promoting osteogenic differentiation of ADSCs. Ó 2014 Elsevier Inc. All rights reserved.
1. Introduction In recent years, nanoparticle-based biomaterials have become one of the most popular research fields in the areas of tissue engineering, biochemistry, biotechnology, and biomedicine [1–3]. One major reason for this interest is that nanoparticles show a new possibility as compared to traditional medicine, which opens up a great potential for new technological applications. In particular, many medical applications [4–8] based on gold nanoparticles (GNPs) have been reported for centuries. GNPs have many advantages due to their unique chemical and physical properties, superior biocompatibility, and well-established strategies for surface modification. Due to their unique properties, GNPs are widely used in various research and practical applications. In biology they are being applied for use as imaging/diagnostic tools, targeted drug
⇑ Corresponding author. Fax: +82 2 967 1457. 1
E-mail address: [email protected] (I.K. Kwon). These authors contributed equally to this work.
http://dx.doi.org/10.1016/j.jcis.2014.08.058 0021-9797/Ó 2014 Elsevier Inc. All rights reserved.
and gene delivery, photo thermal therapy, biosensors, and tissue engineering [4–7]. More recently, many researchers have focused on bone tissue engineering research using GNPs. In a recent report, Yi et al. [9] found that GNPs could promote osteogenic differentiation of mesenchymal stem cells (MSCs) after endocytosis into the cytoplasm. The authors purported that the GNPs stimulated the MSCs by interacting with the cell membrane and binding with proteins in the cytoplasm. This leads to the promotion of osteogenic differentiation by activating the p38 mitogen-activated protein kinase (MAPK) pathway. Additionally, the GNPs inhibited the adipogenic differentiation of MSCs. Dan et al. [10] have also previously observed that GNPs with diameters of both 20 and 40 nm led to osteogenic differentiation of the osteoblast-like cells named MC3T3-E1. Furthermore, GNPs do not only affect osteogenic differentiation, but osteoclast formation as well. In a recent report, Sul et al. [11] demonstrated that GNPs inhibited the differentiation toward osteoclasts from bone marrow-derived macrophages (BMMs) by reducing the receptor activator of nuclear factor-jB ligand (RANKL).
W.-K. Ko et al. / Journal of Colloid and Interface Science 438 (2015) 68–76 Table 1 GNPs average size according to citrate/HAuCl4 volume, and maximum absorption wavelength of each GNPs group.
69
counting kit (CCK-8) was purchased from Dojindo (Kumamoto, Japan).
GNPs size (nm)
Maximum absorption wavelength (nm)
Citrate amount (ll/100 ml)
2.2. Cell culture
30 50 75 100
522 533 541.5 546.5
600 460 400 360
The cells were cultured in MPRO or osteogenic medium in a humidified 5% CO2 and 37 °C incubator. All the mediums were replaced every 3 days. 2.3. Gold nanoparticle synthesis and stability in medium
As noted above, GNPs are a very attractive material for bone tissue regeneration. However, there is a serious lack of information concerning the biological activity in human tissue cells as it relates to the size of GNPs. Understanding the intracellular activity between GNPs and cells is important to regulate the response. Therefore, detailed analysis of the cellular effects of varying sizes of GNPs needs to be performed in order to select the effective candidates for bone tissue regeneration. Adipose-derived stem cells (ADSCs) are excellent multipotent stem cells which can be readily induced to several types of cells such as adipocytes, chondrocytes, myoblasts, and osteoblasts [12]. ADSCs are easy to handle and widely used in bone tissue engineering [13–15]. In this study, the potential effects of the various sizes of GNPs on differentiation of ADSCs toward osteoblasts were analyzed. In order to compare the osteogenic differentiation levels among GNPs groups, a positive control was established by mixing 3 ng/mL of BMP-2 in the medium [16]. Recent reports have shown that BMP-2 plays a crucial role in the regulation of osteogenic differentiation of ADSCs [16–19]. The following groups were used in this study: medium alone (negative control), BMP-2 (positive control), and each size of spherical GNPs (15, 30, 50, 75, and 100 nm). This is the first study to examine the osteogenic differentiation from ADSCs using highly defined sizes of gold nanoparticles. Therefore, we focused on GNP synthesis optimized to form appropriately sized particles with a spherical shape. Transmission electron microscopy (TEM), dynamic light scattering (DLS), and UV–Visible spectroscopy (UV–Vis) were used to measure the size of each GNPs. Cytotoxicity was established using cell counting kits-8 (CCK-8) and live/dead kits were used. To compare the intracellular uptake properties, dark field assays were conducted. Differentiation of ADSCs was evaluated by alkaline phosphatase (ALP) staining, ALP activity, calcium deposition assay, and real time polymerase chain reaction (RT-PCR). 2. Materials and methods 2.1. Chemicals Chloroauric acid (HAuCl4) and trisodium citrate were purchased from Sigma Aldrich. StemProÒ Human ADSCs obtained from Invitrogen (Carlsbad, CA, USA) were cultured in MesenPRO RS™ medium (MPRO medium, Invitrogen) and three passages of ADSCs (Invitrogen) were used. MPRO medium was supplemented with MesenPRO RS™ Growth Supplement (2% FBS, 2 mM glutamine). Osteogenic medium consisted of dulbecco’s modified eagle medium (DMEM) containing 10% fetal bovine serum (FBS, GIBCO, Gran Island, NY), 100 U/ml penicillin (GIBCO), 100 lg/ml streptomycin (GIBCO), 10 mM b-glycerol phosphate disodium salt hydrate (Sigma–Aldrich, St. Louis, MO), 300 lM ascorbic acid (Sigma– Aldrich) and 0.1 lM dexamethasone (Sigma–Aldrich) [20]. Trypsin–EDTA was purchased from GIBCO. Bone morphogenetic protein-2 was purchased from PeproTeck (Rockhill, NJ) and mixed with 0.1% bovine serum albumin (BSA, Sigma–Aldrich) for long usage. Live/dead assay kit was obtained from Invitrogen. Cell
The aqueous dispersions of each size of GNPs were synthesized by citrate reduction of HAuCl4 [21–23]. To synthesize 15 nm GNPs [21], 100 mL of 1 mM HAuCl4 solution (0.04 g HAuCl43H2O in 100 mL distilled water (D.W.)) was refluxed, and then 10 ml of 38.8 mM trisodium citrate solution (0.1141 g trisodium citrate in 10 mL D.W.) was quickly added. The color of the solution changed from yellow to black and to deep red. After the color changed, the solution was refluxed for an additional 15 min, and then the solution was stirred until it reached room temperature. Synthesized spherical GNPs had an average size of 15.5 nm. Synthesis methods of GNPs which have 30–100 nm sizes are described by Frens [23]. 100 mL of 1 mM HAuCl4 solution are heated to boiling while being stirred in clean two-neck round bottle, and then different amounts of 1% trisodium citrate solution is quickly added to the auric solution (Table 1). GNPs solutions were then filtered (0.22 lm, Millipore Carrigtwohill, Ireland) prior to usage in all experiments. Poly ethylene glycol (PEG, MW = 1000, Sigma–Aldrich) was added to synthesized GNPs solutions in order to prevent GNPs [24]. Synthesized high concentration GNPs solution was centrifuged at 1.7 104 g to concentrate the particles. Then the supernatant liquid was replaced with suitable medium for each experiment. The concentration of GNPs in the medium was fixed at a concentration of 1 lM inviability, live/dead, ALP, calcium deposition, real time PCR assay. For the TEM, DLS, UV–Vis spectroscopy, and dark field assays, 500 lM of GNPs solution was used. Transmission electron microscopy (TEM) observations were carried out using an H7100 (Hitachi, Japan). Dynamic light scattering (DLS) measurements were carried out using a 90 PLUS (Brookhaven, USA). Ultraviolet/Vis (UV/Vis) absorption was measured using an UV-1650PC spectrophotometer (Shimadzu, Japan). 2.4. Cell viability test ADSCs with no treatment (negative control), BMP-2 treated, 15, 30, 50, 75 and 100 nm sized GNPs groups were seeded onto 24well culture plates at a seeding density of 5.0 104 cells (n = 3) including MPRO medium. At 1, 3, and 7 days, optical densities of cells were evaluated by using CCK-8. After 2 h of incubation, the intensity was measured by a microplate reader (BioRad, USA) at a wavelength of 450 nm. Cell viability was evaluated at these same conditions using calcein-AM/ethidium homodimer-1 (EthD-1) from the live/dead assay kit. At 1, 3 and 7 days, cell-seeded wells were rinsed with Dulbecco’s Phosphate Buffered Saline (DPBS) twice, and calcein AM and EthD-1 were diluted with DPBS and added to each well. After reacting for 15 min, cells in all groups were observed at 40 magnification using an inverted fluorescence microscope (Olympus IX71, JAPAN). 2.5. Dark field assay ADSCs (5.0 104 cells) were seeded on a 10 mm washed disk with MPRO medium, and incubated overnight to allow for cell attachment. After overnight incubation, each medium was changed to a new medium containing 500 lM of each sized GNPs and then incubated for 8 h. After removing the medium, the cells were
70
W.-K. Ko et al. / Journal of Colloid and Interface Science 438 (2015) 68–76 Table 2 Base sequences of human osteogenic primers for real time PCR. Primer name
Forward primer sequence (50 –30 )
Reverse primer sequence (50 –30 )
COL 1 Runx2 BSP OCN GAPDH
ATG ACT ATG AGT ATG GGG AAG CA AAC CCA CGA ATG CAC TAT CCA AAC GAA GAA AGC GAA GCA GAA AGC AAA GGT GCA GCC TTT GT ATG GGG AAG GTG AAG GTC G
TGG GTC CCT CTG TTA CAC TTT CGG ACA TAC CGA CGA GGG ACA TG TCT GCC TCT GTG CTG TTG GT GCG CCT GGG TCT CTT CAC T GGG GTC ATT GAT GGC AAC AAT A
Fig. 1. Size confirmation of each GNPs group. (A) TEM images, (B) size distribution, measured by DLS, and (C) surface plasmon adsorption, measured by UV–Vis spectroscopy.
rinsed with fresh DPBS twice, fixed in 3.7% formaldehyde at room temperature for 20 min, and were then washed twice again. The light scattering images of GNPs were obtained by using a 12-bit CCD (charge coupled device) camera (Kodak Image Station 4000 MM, New Haven, CT, USA) equipped with a special C-mount lens. The quantitative analysis of GNPs intracellular up take was analyzed by Image J program. The total white spots in the whole area were analyzed to calculate the amount of intracellular uptake. 2.6. Alkaline phosphatase (ALP) staining ADSCs (a density with 5.0 104 cells/well) were seeded one each well (n = 3, 24-well culture plate), and incubated in osteogenic medium alone, BMP-2, and each size of GNPs for 7, 14 and 21 days. At each predetermined time interval, the cell-seeded wells were fixed in 3.7% formaldehyde at room temperature for 20 min. After washing with DPBS twice, the cells were incubated with a mixture of solutions FRV-alkaline solution, sodium nitrite solution, and napthol AS-BI alkaline solution from the ALP staining kit (Sigma–Aldrich), following the manufacturer’s instruction. After staining, the cell-seeded wells were washed twice with DPBS and observed by optical microscopy (Olympus CKX41, JAPAN). 2.7. ALP activity ADSCs (a density with 5.0 104 cells/well) were seeded on each well (n = 3, 24-well culture plate), and incubated in osteogenic
medium containing no treatment, BMP-2, and each size of GNPs for 7, 14 and 21 days. At each predetermined time interval, the cell-seeded wells were washed with DPBS twice and then lysed using 1 RIPA buffer (50 mM Tri-HCL (pH 7.4), 150 mM NaCl, 0.25% deoxycholic acid, 1% NP-40 and 1 mM EDTA) with a protease inhibitor cocktail (Boehringer Mannheim GmbH, Germany) for 30 min on ice. Each of the lysates was centrifuged at 1.7 104 g at 4 °C for 15 min to remove the cell debris. After centrifugation, the supernatant was collected and then reacted with p-nitrophenyl phosphate solution (pNPP, Sigma) in a 5% CO2 humidified incubator at 37 °C for 30 min. The reaction with pNPP was then terminated by adding 50 ll of 1 M NaOH. The production of p-nitrophenol was measured by absorption at 405 nm using a microreader. A calibration curve was generated using standard p-nitrophenol solutions. The quantity of produced total p-nitrophenol from cultured cells was obtained by comparing their results against a calibration curve. Finally, the enzyme activity was expressed as lM of reaction product (p-nitrophenol) per minute per lg of total cellular protein. 2.8. Calcium deposition assay ADSCs (a density of 5.0 104 cells/well) were seeded on each wells (n = 3, 24-well culture plate), and incubated in osteogenic medium containing no treatment, BMP-2, and each size of GNPs for 7, 14 and 21 days. At each predetermined time interval, the cell-seeded wells were washed with DPBS twice, fixed for 20 min
W.-K. Ko et al. / Journal of Colloid and Interface Science 438 (2015) 68–76
71
Fig. 2. Evaluation of cytotoxicity on non-treated ADSCs (control), ADSCs containing BMP-2 and each size of GNPs for 1, 3, and 7 days as measured by (A) live/dead assay and (B) CCK-8 (⁄P < 0.05, ⁄⁄P < 0.01, n = 3).
using 3.7% formaldehyde, and then washed with DPBS again. All the cells were stained in an incubator at 37 °C under 5% CO2 for 30 min using 40 mM alizarin red staining solution (pH 4.2) [25]. The staining solution was removed after 30 min, washed with D.W. three times, and then was examined using an inverted fluorescence microscope. For a quantitative analysis, the stained cells were desorbed with 10% 1-hexadecylpyridinium chloride and the absorbance was measured by using a microplate reader at 540 nm.
The fold change of the control group (osteogenic medium only) at 7 days of culture was set at 1-fold and the ratio of the normalized fold change was calculated. The real-time PCR amplifications were carried out 10 s at 95 °C, 30 s at 53–56 °C (COL1: 53 °C, Runx2: 56 °C, BSP:54 °C, OCN:55 °C) and 30 s at 72 °C for 45 cycles after the initial denaturation step for 10 min at 95 °C. Each primer was designed for real time PCR (Table 2) and purchased from Bioneer. All the real time PCR results were normalized by glyceraldehyde 3-phosphate dehydrogenase (GAPDH) [26].
2.9. Real time PCR 2.10. Statistical analysis 4
ADSCs (a density of 5.0 10 cells/well) were seeded on each well (n = 3, 24-well culture plate), and incubated in osteogenic medium containing no treatment, BMP-2, and each size of GNPs for 7, 14 and 21 days. At each predetermined time interval, the cell-seeded wells were washed with DPBS twice and afterward the entire RNA from each group’s cells was isolated using an RNeasy Plus Mini Kit (Qiagen, CA, USA) according to manufacturer’s instructions. 1 lg of the entire RNA was extracted from all specimens, and subsequently transcribed into cDNA using an AccuPower Cycle Script RT Premix (Bioneer, Daejeon, Republic of Korea). Real-time PCR was analyzed by using iQ SYBR Green supermix (Bio-Rad, Hercules, CA, USA). Threshold cycle values were calculated by using a comparative cycle threshold (CT) method.
Statistical analysis was performed using PASW Statistics 18 software (SPSS, Inc.). All values were expressed as means ± standard deviations, and differences with p-values (⁄P < 0.05, ⁄⁄ P < 0.01) were considered statistically significant. 3. Results 3.1. Synthesis of gold nanoparticles and size confirmation GNPs have a different maximum absorbance wavelength in accordance with each GNPs size [21]. All spherical GNPs were synthesized according to the Huang report [21]. The size and
72
W.-K. Ko et al. / Journal of Colloid and Interface Science 438 (2015) 68–76
Fig. 3. Dark-field microscopy analysis of ADSCs involved with no treatment (control) and each size of GNPs after 8 h: (A) the scattering images, and (B) quantitative analysis according to various sizes of GNPs (⁄P < 0.05, n = 3).
distribution were confirmed by TEM, DLS, and UV–Vis spectroscopy (Fig. 1). For visual confirmation, GNPs were observed by TEM. Each GNPs solution was dropped on carbon and photographed (Fig. 1A). All synthesized GNPs had spherical shape, and the sizes were distributed well. For accurate confirmation, DLS was conducted and the following results were elicited. GNPs have the maximum frequency in a size distribution graph at 15.5 (15 nm target), 30.2 (30 nm target), 50.3 (50 nm target), 75.8 (75 nm target), and 100.3 nm (100 nm target), respectively (Fig. 1B). At last, synthesized GNPs were analyzed by UV–Vis spectroscopy. The measured range was from 400 to 800 nm (Fig. 1C). All GNP groups displayed a smooth curve from 400 to 800 nm range and the GNPs of each size have maximum absorptions at 517.5 (15 nm), 522 (30 nm), 533 (50 nm), 541.5 (75 nm), and 546.5 nm (100 nm), respectively. The color of each solution changed from yellow to black and then to red or purple depending on the size of the nanoparticles (data not shown). The approximate amounts of citrate acid and the corresponding sizes of nanoparticles are listed Table 1. This shows that smaller nanoparticles require more citrate solutions, and larger nanoparticles require less citrate solution. The peak intensity and peak position of GNP solutions are dependent on the size, shape, and concentration of GNPs [21]. In conclusion, the above results match Huang’s results indicating successful formation of spherical GNPs with different sizes. 3.2. Cell viability test We conducted a cell viability test on ADSCs which had been cultured for 1, 3, and 7 days with BMP-2 and each size of GNPs, and then these results were compared with that of non-treated group (control). There were almost no dead cells after 7 days of incubation by the live/dead assay (Fig. 2A). In order to quantify, we measured the cell viability percentage among groups using the CCK-8 assay (Fig. 2B). Generally, all the GNP treated groups displayed a higher cell viability than control group during 7 days of incubation. These results indicate that 1 lM of GNPs has no significant cytotoxicity toward ADSCs.
3.3. Dark field microscopy analysis To confirm the intracellular uptake properties of each size of GNPs, dark-field microscopy observations were carried out (Fig. 3A). Seeded cells were incubated for 1 day in order to attach to the culture plate. Subsequently these were treated with each size of GNPs for 8 h [27]. The ADSCs in each group were observed under bright or dark-field imaging conditions. In dark-field imaging conditions, the GNPs in cells were observed as white spots against a dark background [7,28]. The results showed that more white spots were observed in cells treated with 30 and 50 nm sized GNPs as compared to other sizes. As shown in Fig. 3B, the 50 nm group showed the highest uptake ratio of GNPs of all groups. All differences were significant except for those between 75 and 100 nm groups (⁄P < 0.05). In summary, it was found that the uptake amount of GNPs during 8 h increased in the order of 100, 75, 15, 30, and 50 nm groups. 3.4. ALP staining and activity assay The ALP staining/activity levels indicating differentiation of ADSCs toward osteoblasts were examined. This was done with cells cultured on plate wells for 7, 14 and 21 days with osteogenic medium containing no treatment, BMP-2, and each size of GNPs. The ALP staining was observed by an inverted fluorescence microscope (Fig. 4A). Generally, the maximum level of ALP is expressed at 2 weeks during osteogenesis [29]. The quantity of ALP staining had reached a maximum by 2 weeks, and then decreased at 3 weeks in all groups. The 30 and 50 nm groups had more ALP staining than others during the 3 week study. In order to quantify, the ALP activity assay was conducted under the same conditions and quantitatively measured (Fig. 4B). The ALP activity results showed strong similarities with the ALP staining results. At every predetermined time interval, the 30 and 50 nm groups were observed to have higher ALP levels than others. The BMP-2 positive control group’s level was similar to the level of 75 or 100 nm GNPs. These results indicate that GNPs can positively
W.-K. Ko et al. / Journal of Colloid and Interface Science 438 (2015) 68–76
73
Fig. 4. ALP staining and activity level during the process of differentiation from ADSCs toward osteoblasts for 3 weeks: (A) ALP stained cells, which were treated with osteogenic medium alone (control), BMP-2, and each size of GNPs for 3 weeks as observed by an optical microscope, and (B) ALP activity levels of cells which were cultured at the same conditions for 3 weeks as observed by a microplate reader (⁄P < 0.05, ⁄⁄P < 0.01, n = 3).
accelerate the osteogenic differentiation of ADSCs more so than BMP-2. The 50 nm GNPs group expressed the highest ALP level. In general the ALP activity level increased in the order of 100, 75, 15, 30, and 50 nm GNPs groups. 3.5. Calcium deposition staining and activity assay Calcium deposition was measured for cells incubated for 7, 14 and 21 days with osteogenic medium only, osteogenic medium containing BMP-2, and osteogenic medium containing each size of GNPs. The quantity of deposited calcium was compared to control group after staining with alizarin red solution (Fig. 5A). The deposited calcium degree of all groups increased during the 3 weeks. The 30 and 50 nm GNPs groups displayed strong deposition, which matched the ALP assay results. In order to quantify these results, the deposited calcium from each group was dissolved using 10% 1-hexadecylpyridinium chloride. Then the optical density (O.D.) was measured by a microplate reader (Fig. 5B). There was no great difference among all groups at 1 week, but 30 and 50 nm GNPs groups had higher levels at 2 and 3 weeks. This indicates that 30 and 50 nm of GNPs strongly promoted mineralization of osteoblasts. The BMP-2 group had a higher O.D. than the control group, but did not exceed the O.D. of the GNP groups.
3.6. Real-time PCR We examined mRNA expression of osteoblast gene markers, such as collagen type I (COL1), runt-related transcription factor 2 (Runx2), bone sialoprotein (BSP) and osteocalcin (OCN) of cells at the varying incubation conditions (Fig. 6). As the culture time increased, the amount of synthesized COL1 decreased in all specimens (Fig. 6A). Generally, COL1 is regarded as an early marker during osteogenic differentiation of MSCs [29] which is in agreement with our results. High levels of COL1 were observed with 30 and 50 nm GNPs groups at each time interval. Another osteogenic transcription factor, Runx2, was also expressed at its highest levels after 2 weeks of incubation in all specimens (Fig. 6B). The highest intensity of Runx2 was recorded with the 50 nm group which is similar to the COL1 results. Fig. 6C shows the BSP mRNA expression levels during 21 days of culture. It was found that the BSP mRNA expression levels increased in all groups as culture time increased and the 50 nm group’s level was higher during the 3 weeks than the rest. The mRNA expression of OCN was similar to BSP (Fig. 6D). For all markers, it was found that the mRNA expression levels of GNPs groups were higher than the control groups. The 30 and 50 nm GNPs groups showed the highest levels among all GNPs groups. The BMP-2 group showed similar levels to the 75 and 100 nm group for all primers.
74
W.-K. Ko et al. / Journal of Colloid and Interface Science 438 (2015) 68–76
Fig. 5. Alizarin red staining during the process of differentiation from ADSCs toward osteoblasts for 3 weeks: (A) Stained cells with alizarin red solutions, which were treated with osteogenic medium containing none (control), BMP-2, and each size of GNPs for 3 weeks as observed by confocal laser scanning microscopy, and (B) deposition degrees of lysed calcium, as measured by a microplate reader (⁄P < 0.05, ⁄⁄P < 0.01, n = 3).
4. Discussion GNPs are widely used in many applications such as biomedical imaging, clinical diagnostics, and therapeutics [1–8]. These areas are associated with nanomaterials that have different sizes and several types of shape [30–35]. GNPs are an attractive material in and of themselves. Here, we used GNPs as an accelerator for differentiation of ADSCs toward osteoblasts, and investigated the osteogenic expression of ADSCs as they were influenced by GNPs. GNPs were synthesized using the citrate acid reduction method of HAuCl4 and their sizes were confirmed by TEM, DLS, and UV–Vis spectroscopy (Fig. 1). There exists a specific oscillation depending on the size and shape of GNPs [21]. GNPs possess strong distance-dependent optical density in accordance with their plasmon resonance. The experimental results showed quite a similar graph as Huang’s [21], but did not show exactly the same maximum wavelength. Huang’s result had shown an increasing maximum wavelength with increasing size of GNPs. Therefore, it was necessary to confirm the sizes of the synthesized GNPs. DLS is used routinely to analyze the size and size distribution of polymers, proteins, colloids, and nanoparticles [36]. In this study, DLS was conducted for quantitative detection and analysis of GNPs size. As a result, we were able to prepare GNPs with the following
distribution: 15.5 ± 2.4, 30.2 ± 2.5, 50.3 ± 1.25, 75.8 ± 0.95 and 100.3 ± 5.05 nm, respectively (Fig. 1B). The distribution figures showed that the GNPs were synthesized appropriately for the experiments of this study. Prior to performing experimentation with ADSCs, the usable concentration of GNPs which does not cause toxicity had to be established. Too high of a concentration of GNPs will inevitably drive cells to death [38], too low concentrations of GNPs would have little influence on differentiation. It was established that 1 lM of GNPs displayed no toxicity toward cells in both viability test and live/dead assays over the course of 1 week of incubation (Fig. 2). This is despite the GNPs being gradually cumulated in the cells. Previous studies [37,38] have also supported that 1 lM of GNPs does not display toxicity toward cells. In addition, 1 lM of GNPs, demonstrated an obvious difference of each group toward the osteogenic differentiation process. To analyze these results, we measured the quantitative density of each size of GNPs up taken into the cell membranes. After a preset time of 8 h for each group, the uptake properties of the GNPs were measured via the dark field assay. This uptake time was selected based on a prior Devika et al. study [27] which demonstrated that maximum uptake was observed from 5 to 8 h. In Fig. 3, the 50 nm sized GNPs group expressed the highest uptake among the 5 sizes of GNPs groups. The uptake increased in the order of 100, 75, 15, 30, and
75
W.-K. Ko et al. / Journal of Colloid and Interface Science 438 (2015) 68–76
Fig. 6. The mRNA expression of osteogenic markers during the process of differentiation for 3 weeks: (A) COL1, (B) Runx2, (C) BSP, and (D) OCN (⁄P < 0.05,
50 nm sizes. We tested that this result might affect differentiation of ADSCs toward osteoblasts among groups through the following experiments. ALP and calcium deposition assays were used respectively as early and late markers of osteoblast differentiation. These results were consistent with our results as well (Figs. 4 and 5). ALP is a typical ecto-enzyme, which acts as a marker of osteogenic differentiation [29]. Therefore, we measured the ALP level by both quantitative and qualitative methods (Fig. 4). Fig. 4A and B both demonstrated the highest peak at 2 weeks in all specimens. The 50 nm GNPs group had the highest level among all groups and the 30 nm GNPs also demonstrated higher levels than other GNPs groups. These results corresponded well with the calcium deposition assay results (Fig. 5). Typically, after culturing for 2–3 weeks, osteoblast nodes were formed and reached a peak quantity when osteoblasts started to mineralize [39]. These results show that 30 and 50 nm GNPs groups especially lead to high osteoblast differentiation. For a more analytical approach, we evaluated the mRNA expressions of COL1, Runx2, BSP, and OCN by real time PCR (Fig. 6). At all tested time points for all osteogenic markers, the 30 and 50 nm GNPs had high levels of expression. COL1 is the most abundant protein in bone matrix and usually is formed during the proliferation of osteoblast cells [40,41]. In Fig. 6A, mRNA expressions in all groups were recorded at peak levels at 1 week during the 3 week study. The groups containing GNPs had higher levels than that of the group containing BMP-2, particularly the 50 nm group recorded the highest expression levels. We evaluated the typical osteogenic transcriptional factor Runx2 as a critical regulator of osteogenic differentiation [40]. The Runx2 mRNA expression of all specimens increased during 2 weeks, and then decreased at 3 weeks of culture (Fig. 6B). This result means that over expressed Runx2 presumably inhibits its own production to prevent the
⁄⁄
P < 0.01, n = 3).
differentiation of mature osteoblasts toward osteocytes [40]. During 3 weeks of culture time, the 30 and 50 nm GNPs groups had comparatively higher levels among the tested GNPs sizes. The expressions of BSP and OCN markers related to calcium deposition in the extracellular matrix during its mineralization stage [40] were increased over the course of 3 weeks and also showed the superiority of the 50 nm sized GNPs (Fig. 6C and D). The expressed Runx2 promoted BSP and OCN expression, which promotes the mineralization of osteoblasts (Fig. 5). These data show that the amount of uptake of GNPs increased in a similar order as osteogenic differentiation. The attached proteins around GNPs as well as uptake GNPs themselves might affect the osteoblast differentiation of ADSCs [42]. In summary, ADSCs incubated with 30 or 50 nm GNPs lead to more differentiation of the ADSCs as compared to other groups. The 50 nm GNPs group especially promoted osteogenic differentiation of ADSCs. However, the osteogenic differentiation levels of larger GNPs were lower than the 30 nm GNPs group. The above results indicate that the size of GNPs is quite an important factor in their impact on osteogenic differentiation. 5. Conclusion In this study, all the data consistently demonstrated that the 30 and 50 nm GNPs were effective at enhancing osteogenic differentiation. Although we could not confirm the exact signaling mechanism after endocytosis into the cell’s membrane, we could show the effectiveness of 30 and 50 nm GNPs through the expression of Runx2 genes and its linked osteogenic markers’ expression by several osteogenic experiments. All results show that the osteogenic differentiation was dependent on the size of GNPs. Therefore, we suggest that the nanoparticle size is important in the biological
76
W.-K. Ko et al. / Journal of Colloid and Interface Science 438 (2015) 68–76
environment. The results of this study might assist in designing nanostructures for future drug delivery and therapeutic systems. Acknowledgments This research was supported by the Pioneer Research Center Program through the National Research Foundation of Korea funded by the Ministry of Education, Science and Technology (2013-007552) and the Public welfare & Safety research program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (NRF-20120008610) and by Bio-industry Technology Development Program (312062-5) of iPET (Korea Institute of Planning and Evaluation for Technology in Food, Agriculture, Forestry and Fisheries), Ministry for Food, Agriculture, Forestry and Fisheries, Republic of Korea. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13]
M. De, P.S. Ghosh, V.M. Rotello, Adv. Mater. 20 (2008) 4225. J. Xie, S. Lee, X. Chen, Adv. Drug Deliv. Rev. 62 (2010) 1064. A. Tautzenberger, A. Kovtun, A. Ignatius, Int. J. Nanomed. 7 (2012) 4545. D.A. Giljohann, D.S. Seferos, W.L. Daniel, M.D. Massich, P.C. Patel, C.A. Mirkin, Angew. Chem. Int. Ed. 49 (2010) 3280. L. Dykman, N. Khlebtsov, Chem. Soc. Rev. 41 (2012) 2256. E. Boisselier, D. Astruc, Chem. Soc. Rev. 38 (2009) 1759. D.N. Heo, D.H. Yang, H.-J. Moon, J.B. Lee, M.S. Bae, S.C. Lee, W.J. Lee, I.-C. Sun, I.K. Kwon, Biomaterials 33 (2012) 856. Y.-I. Chung, K.-M. Ahn, S.-H. Jeon, S.-Y. Lee, J.-H. Lee, G. Tae, J. Control. Release 121 (2007) 91. C. Yi, D. Liu, C.-C. Fong, J. Zhang, M. Yang, ACS Nano 4 (2010) 6439. D. Liu, J. Zhang, C. Yi, M. Yang, Chin. Sci. Bull. 55 (2010) 1013. O.-J. Sul, J.-C. Kim, T.-W. Kyung, H.-J. Kim, Y.-Y. Kim, S.-H. Kim, J.-S. Kim, H.-S. Choi, Biosci. Biotechnol. Biochem. 74 (2010) 2209. P.A. Zuk, M. Zhu, P. Ashjian, D.A. De Ugarte, J.I. Huang, H. Mizuno, Z.C. Alfonso, J.K. Fraser, P. Benhaim, M.H. Hedrick, Mol. Biol. Cell 13 (2002) 4279. C.M. Cowan, Y.-Y. Shi, O.O. Aalami, Y.-F. Chou, C. Mari, R. Thomas, N. Quarto, C.H. Contag, B. Wu, M.T. Longaker, Nat. Biotechnol. 22 (2004) 560.
[14] M.D. Kwan, B.J. Slater, D.C. Wan, M.T. Longaker, Hum. Mol. Genet. 17 (2008) R93. [15] A.W. James, B. Levi, E.R. Nelson, M. Peng, G.W. Commons, M. Lee, B. Wu, M.T. Longaker, Stem Cells Dev. 20 (2010) 427. [16] M.S. Bae, J.E. Kim, J.B. Lee, D.N. Heo, D.H. Yang, J.-H. Kim, K.-R. Kwon, J.B. Bang, H. Bae, I.K. Kwon, Carbohydr. Polym. 92 (2013) 167. [17] J.L. Dragoo, J.Y. Choi, J.R. Lieberman, J. Huang, P.A. Zuk, J. Zhang, M.H. Hedrick, P. Benhaim, J. Orthop. Res. 21 (2003) 622. [18] J.L. Dragoo, J.R. Lieberman, R.S. Lee, D.A. Deugarte, Y. Lee, P.A. Zuk, M.H. Hedrick, P. Benhaim, Plast. Reconstr. Surg. 115 (2005) 1665. [19] M. Knippenberg, M. Helder, B. ZandiehDoulabi, P. Wuisman, J. Klein-Nulend, Biochem. Biophys. Res. Commun. 342 (2006) 902. [20] N. Jaiswal, S.E. Haynesworth, A.I. Caplan, S.P. Bruder, J. Cell. Biochem. 64 (1997) 295. [21] X. Huang, Ph.D. thesis, Georgia Tech, 2006. [22] P. CooperáStevenson, Discuss. Faraday Soc. 11 (1951) 55. [23] G. Frens, Nature 241 (1973) 20. [24] F.M. Veronese, Biomaterials 22 (2001) 405. [25] H. Anderson, Lab. Investig. J. Tech. Methods Pathol. 60 (1989) 320. [26] K. Yang, J.S. Lee, J. Kim, Y.B. Lee, H. Shin, S.H. Um, J.B. Kim, K.I. Park, H. Lee, S.W. Cho, Biomaterials 33 (2012) 6952. [27] B.D. Chithrani, A.A. Ghazani, W.C. Chan, Nano Lett. 6 (2006) 662. [28] I.-C. Sun, S. Lee, H. Koo, I.C. Kwon, K. Choi, C.-H. Ahn, K. Kim, Bioconjug. Chem. 21 (2010) 1939. [29] D. Liu, C. Yi, D. Zhang, J. Zhang, M. Yang, ACS Nano 4 (2010) 2185. [30] K.C. Grabar, R.G. Freeman, M.B. Hommer, M.J. Natan, Anal. Chem. 67 (1995) 735. [31] W.K. Leutwyler, S.L. Bürgi, H. Burgl, Science 271 (1996) 933. [32] R. Elghanian, J.J. Storhoff, R.C. Mucic, R.L. Letsinger, C.A. Mirkin, Science 277 (1997) 1078. [33] Y.C. Cao, R. Jin, C.A. Mirkin, Science 297 (2002) 1536. [34] J.M. Klostranec, W.C. Chan, Adv. Mater. 18 (2006) 1953. [35] L.R. Hirsch, R. Stafford, J. Bankson, S. Sershen, B. Rivera, R. Price, J. Hazle, N. Halas, J. West, Proc. Natl. Acad. Sci. 100 (2003) 13549. [36] B.J. Berne, R. Pecora, Dynamic light scattering: with applications to chemistry, biology, and physics, Courier Dover Publications, 2000. [37] Y. Pan, S. Neuss, A. Leifert, M. Fischler, F. Wen, U. Simon, G. Schmid, W. Brandau, W. Jahnen-Dechent, Small 3 (2007) 1941. [38] E.E. Connor, J. Mwamuka, A. Gole, C.J. Murphy, M.D. Wyatt, Small 1 (2005) 325. [39] G.S. Stein, J.B. Lian, Endocr. Rev. 14 (1993) 424. [40] T. Komori, J. Bone Miner. Metab. 21 (2003) 193. [41] K. Kadler, D. Holmes, J. Trotter, J. Chapman, Biochem. J. 316 (1996) 1. [42] M. Mahmoudi, I. Lynch, M.R. Ejtehadi, M.P. Monopoli, F.B. Bombelli, S. Laurent, Chem. Rev. 111 (2011) 5610.
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
The effect of gold nanoparticle size on osteogenic differentiation of adipose-derived stem cells Wan-Kyu Ko a,1, Dong Nyoung Heo a,1, Ho-Jin Moon a, Sang Jin Lee a, Min Soo Bae a, Jung Bok Lee a, In-Cheol Sun b, Hoon Bong Jeon c, Hun Kuk Park c, Il Keun Kwon a,⇑ a Department of Maxillofacial Biomedical Engineering and Institute of Oral Biology, School of Dentistry, Kyung Hee University, 26, Kyungheedae-ro, Dongdaemun-gu, Seoul 130-701, Republic of Korea b Biomedical Research Center, Korea Institute of Science and Technology, 39-1 Hawolgok-dong, Seongbuk-gu, Seoul 136-791, Republic of Korea c Department of Biomedical Engineering, College of Medicine, Kyung Hee University, 26, Kyungheedae-ro, Dongdaemun-gu, Seoul 130-701, Republic of Korea
a r t i c l e
i n f o
Article history: Received 26 June 2014 Accepted 28 August 2014 Available online 16 September 2014 Keywords: Gold nanoparticles Human adipose-derived stem cell Uptake Osteogenesis Osteogenic differentiation
a b s t r a c t There have been many medical applications based on gold nanoparticles (GNPs) over the past several centuries. Recently, researchers have focused on bone tissue engineering applications utilizing GNPs. The effect of various sizes of gold nanoparticles on the differentiation of human adipose-derived stem cells (ADSCs) into osteoblasts was investigated. The concentration of gold nanoparticles was fixed at 1 lM and varying sizes of 15, 30, 50, 75 and 100 nm (spherical GNPs) were used. The lack of cytotoxicity was confirmed by establishing viability of ADSCs using cell counting kit-8 (CCK-8) and live/dead assays. The results showed that each size of GNPs had no significant toxicity on ADSCs during 1 week of incubation. Osteogenic differentiation of ADSCs was confirmed by alkaline phosphatase (ALP) staining, ALP activity, calcium deposition, and real time PCR experiments. It was found, through dark field assays and microscope cell images, that 30 nm and 50 nm GNPs were preferentially up taken into the ADSCs. As expected, all sizes of gold nanoparticles promoted the differentiation of ADSCs toward osteoblasts more than control. Among all sizes, 30 and 50 nm GNPs appeared to have the highest differentiation rates. The data consistently demonstrated that 30 and 50 nm GNPs are the most effective in promoting osteogenic differentiation of ADSCs. Ó 2014 Elsevier Inc. All rights reserved.
1. Introduction In recent years, nanoparticle-based biomaterials have become one of the most popular research fields in the areas of tissue engineering, biochemistry, biotechnology, and biomedicine [1–3]. One major reason for this interest is that nanoparticles show a new possibility as compared to traditional medicine, which opens up a great potential for new technological applications. In particular, many medical applications [4–8] based on gold nanoparticles (GNPs) have been reported for centuries. GNPs have many advantages due to their unique chemical and physical properties, superior biocompatibility, and well-established strategies for surface modification. Due to their unique properties, GNPs are widely used in various research and practical applications. In biology they are being applied for use as imaging/diagnostic tools, targeted drug
⇑ Corresponding author. Fax: +82 2 967 1457. 1
E-mail address: [email protected] (I.K. Kwon). These authors contributed equally to this work.
http://dx.doi.org/10.1016/j.jcis.2014.08.058 0021-9797/Ó 2014 Elsevier Inc. All rights reserved.
and gene delivery, photo thermal therapy, biosensors, and tissue engineering [4–7]. More recently, many researchers have focused on bone tissue engineering research using GNPs. In a recent report, Yi et al. [9] found that GNPs could promote osteogenic differentiation of mesenchymal stem cells (MSCs) after endocytosis into the cytoplasm. The authors purported that the GNPs stimulated the MSCs by interacting with the cell membrane and binding with proteins in the cytoplasm. This leads to the promotion of osteogenic differentiation by activating the p38 mitogen-activated protein kinase (MAPK) pathway. Additionally, the GNPs inhibited the adipogenic differentiation of MSCs. Dan et al. [10] have also previously observed that GNPs with diameters of both 20 and 40 nm led to osteogenic differentiation of the osteoblast-like cells named MC3T3-E1. Furthermore, GNPs do not only affect osteogenic differentiation, but osteoclast formation as well. In a recent report, Sul et al. [11] demonstrated that GNPs inhibited the differentiation toward osteoclasts from bone marrow-derived macrophages (BMMs) by reducing the receptor activator of nuclear factor-jB ligand (RANKL).
W.-K. Ko et al. / Journal of Colloid and Interface Science 438 (2015) 68–76 Table 1 GNPs average size according to citrate/HAuCl4 volume, and maximum absorption wavelength of each GNPs group.
69
counting kit (CCK-8) was purchased from Dojindo (Kumamoto, Japan).
GNPs size (nm)
Maximum absorption wavelength (nm)
Citrate amount (ll/100 ml)
2.2. Cell culture
30 50 75 100
522 533 541.5 546.5
600 460 400 360
The cells were cultured in MPRO or osteogenic medium in a humidified 5% CO2 and 37 °C incubator. All the mediums were replaced every 3 days. 2.3. Gold nanoparticle synthesis and stability in medium
As noted above, GNPs are a very attractive material for bone tissue regeneration. However, there is a serious lack of information concerning the biological activity in human tissue cells as it relates to the size of GNPs. Understanding the intracellular activity between GNPs and cells is important to regulate the response. Therefore, detailed analysis of the cellular effects of varying sizes of GNPs needs to be performed in order to select the effective candidates for bone tissue regeneration. Adipose-derived stem cells (ADSCs) are excellent multipotent stem cells which can be readily induced to several types of cells such as adipocytes, chondrocytes, myoblasts, and osteoblasts [12]. ADSCs are easy to handle and widely used in bone tissue engineering [13–15]. In this study, the potential effects of the various sizes of GNPs on differentiation of ADSCs toward osteoblasts were analyzed. In order to compare the osteogenic differentiation levels among GNPs groups, a positive control was established by mixing 3 ng/mL of BMP-2 in the medium [16]. Recent reports have shown that BMP-2 plays a crucial role in the regulation of osteogenic differentiation of ADSCs [16–19]. The following groups were used in this study: medium alone (negative control), BMP-2 (positive control), and each size of spherical GNPs (15, 30, 50, 75, and 100 nm). This is the first study to examine the osteogenic differentiation from ADSCs using highly defined sizes of gold nanoparticles. Therefore, we focused on GNP synthesis optimized to form appropriately sized particles with a spherical shape. Transmission electron microscopy (TEM), dynamic light scattering (DLS), and UV–Visible spectroscopy (UV–Vis) were used to measure the size of each GNPs. Cytotoxicity was established using cell counting kits-8 (CCK-8) and live/dead kits were used. To compare the intracellular uptake properties, dark field assays were conducted. Differentiation of ADSCs was evaluated by alkaline phosphatase (ALP) staining, ALP activity, calcium deposition assay, and real time polymerase chain reaction (RT-PCR). 2. Materials and methods 2.1. Chemicals Chloroauric acid (HAuCl4) and trisodium citrate were purchased from Sigma Aldrich. StemProÒ Human ADSCs obtained from Invitrogen (Carlsbad, CA, USA) were cultured in MesenPRO RS™ medium (MPRO medium, Invitrogen) and three passages of ADSCs (Invitrogen) were used. MPRO medium was supplemented with MesenPRO RS™ Growth Supplement (2% FBS, 2 mM glutamine). Osteogenic medium consisted of dulbecco’s modified eagle medium (DMEM) containing 10% fetal bovine serum (FBS, GIBCO, Gran Island, NY), 100 U/ml penicillin (GIBCO), 100 lg/ml streptomycin (GIBCO), 10 mM b-glycerol phosphate disodium salt hydrate (Sigma–Aldrich, St. Louis, MO), 300 lM ascorbic acid (Sigma– Aldrich) and 0.1 lM dexamethasone (Sigma–Aldrich) [20]. Trypsin–EDTA was purchased from GIBCO. Bone morphogenetic protein-2 was purchased from PeproTeck (Rockhill, NJ) and mixed with 0.1% bovine serum albumin (BSA, Sigma–Aldrich) for long usage. Live/dead assay kit was obtained from Invitrogen. Cell
The aqueous dispersions of each size of GNPs were synthesized by citrate reduction of HAuCl4 [21–23]. To synthesize 15 nm GNPs [21], 100 mL of 1 mM HAuCl4 solution (0.04 g HAuCl43H2O in 100 mL distilled water (D.W.)) was refluxed, and then 10 ml of 38.8 mM trisodium citrate solution (0.1141 g trisodium citrate in 10 mL D.W.) was quickly added. The color of the solution changed from yellow to black and to deep red. After the color changed, the solution was refluxed for an additional 15 min, and then the solution was stirred until it reached room temperature. Synthesized spherical GNPs had an average size of 15.5 nm. Synthesis methods of GNPs which have 30–100 nm sizes are described by Frens [23]. 100 mL of 1 mM HAuCl4 solution are heated to boiling while being stirred in clean two-neck round bottle, and then different amounts of 1% trisodium citrate solution is quickly added to the auric solution (Table 1). GNPs solutions were then filtered (0.22 lm, Millipore Carrigtwohill, Ireland) prior to usage in all experiments. Poly ethylene glycol (PEG, MW = 1000, Sigma–Aldrich) was added to synthesized GNPs solutions in order to prevent GNPs [24]. Synthesized high concentration GNPs solution was centrifuged at 1.7 104 g to concentrate the particles. Then the supernatant liquid was replaced with suitable medium for each experiment. The concentration of GNPs in the medium was fixed at a concentration of 1 lM inviability, live/dead, ALP, calcium deposition, real time PCR assay. For the TEM, DLS, UV–Vis spectroscopy, and dark field assays, 500 lM of GNPs solution was used. Transmission electron microscopy (TEM) observations were carried out using an H7100 (Hitachi, Japan). Dynamic light scattering (DLS) measurements were carried out using a 90 PLUS (Brookhaven, USA). Ultraviolet/Vis (UV/Vis) absorption was measured using an UV-1650PC spectrophotometer (Shimadzu, Japan). 2.4. Cell viability test ADSCs with no treatment (negative control), BMP-2 treated, 15, 30, 50, 75 and 100 nm sized GNPs groups were seeded onto 24well culture plates at a seeding density of 5.0 104 cells (n = 3) including MPRO medium. At 1, 3, and 7 days, optical densities of cells were evaluated by using CCK-8. After 2 h of incubation, the intensity was measured by a microplate reader (BioRad, USA) at a wavelength of 450 nm. Cell viability was evaluated at these same conditions using calcein-AM/ethidium homodimer-1 (EthD-1) from the live/dead assay kit. At 1, 3 and 7 days, cell-seeded wells were rinsed with Dulbecco’s Phosphate Buffered Saline (DPBS) twice, and calcein AM and EthD-1 were diluted with DPBS and added to each well. After reacting for 15 min, cells in all groups were observed at 40 magnification using an inverted fluorescence microscope (Olympus IX71, JAPAN). 2.5. Dark field assay ADSCs (5.0 104 cells) were seeded on a 10 mm washed disk with MPRO medium, and incubated overnight to allow for cell attachment. After overnight incubation, each medium was changed to a new medium containing 500 lM of each sized GNPs and then incubated for 8 h. After removing the medium, the cells were
70
W.-K. Ko et al. / Journal of Colloid and Interface Science 438 (2015) 68–76 Table 2 Base sequences of human osteogenic primers for real time PCR. Primer name
Forward primer sequence (50 –30 )
Reverse primer sequence (50 –30 )
COL 1 Runx2 BSP OCN GAPDH
ATG ACT ATG AGT ATG GGG AAG CA AAC CCA CGA ATG CAC TAT CCA AAC GAA GAA AGC GAA GCA GAA AGC AAA GGT GCA GCC TTT GT ATG GGG AAG GTG AAG GTC G
TGG GTC CCT CTG TTA CAC TTT CGG ACA TAC CGA CGA GGG ACA TG TCT GCC TCT GTG CTG TTG GT GCG CCT GGG TCT CTT CAC T GGG GTC ATT GAT GGC AAC AAT A
Fig. 1. Size confirmation of each GNPs group. (A) TEM images, (B) size distribution, measured by DLS, and (C) surface plasmon adsorption, measured by UV–Vis spectroscopy.
rinsed with fresh DPBS twice, fixed in 3.7% formaldehyde at room temperature for 20 min, and were then washed twice again. The light scattering images of GNPs were obtained by using a 12-bit CCD (charge coupled device) camera (Kodak Image Station 4000 MM, New Haven, CT, USA) equipped with a special C-mount lens. The quantitative analysis of GNPs intracellular up take was analyzed by Image J program. The total white spots in the whole area were analyzed to calculate the amount of intracellular uptake. 2.6. Alkaline phosphatase (ALP) staining ADSCs (a density with 5.0 104 cells/well) were seeded one each well (n = 3, 24-well culture plate), and incubated in osteogenic medium alone, BMP-2, and each size of GNPs for 7, 14 and 21 days. At each predetermined time interval, the cell-seeded wells were fixed in 3.7% formaldehyde at room temperature for 20 min. After washing with DPBS twice, the cells were incubated with a mixture of solutions FRV-alkaline solution, sodium nitrite solution, and napthol AS-BI alkaline solution from the ALP staining kit (Sigma–Aldrich), following the manufacturer’s instruction. After staining, the cell-seeded wells were washed twice with DPBS and observed by optical microscopy (Olympus CKX41, JAPAN). 2.7. ALP activity ADSCs (a density with 5.0 104 cells/well) were seeded on each well (n = 3, 24-well culture plate), and incubated in osteogenic
medium containing no treatment, BMP-2, and each size of GNPs for 7, 14 and 21 days. At each predetermined time interval, the cell-seeded wells were washed with DPBS twice and then lysed using 1 RIPA buffer (50 mM Tri-HCL (pH 7.4), 150 mM NaCl, 0.25% deoxycholic acid, 1% NP-40 and 1 mM EDTA) with a protease inhibitor cocktail (Boehringer Mannheim GmbH, Germany) for 30 min on ice. Each of the lysates was centrifuged at 1.7 104 g at 4 °C for 15 min to remove the cell debris. After centrifugation, the supernatant was collected and then reacted with p-nitrophenyl phosphate solution (pNPP, Sigma) in a 5% CO2 humidified incubator at 37 °C for 30 min. The reaction with pNPP was then terminated by adding 50 ll of 1 M NaOH. The production of p-nitrophenol was measured by absorption at 405 nm using a microreader. A calibration curve was generated using standard p-nitrophenol solutions. The quantity of produced total p-nitrophenol from cultured cells was obtained by comparing their results against a calibration curve. Finally, the enzyme activity was expressed as lM of reaction product (p-nitrophenol) per minute per lg of total cellular protein. 2.8. Calcium deposition assay ADSCs (a density of 5.0 104 cells/well) were seeded on each wells (n = 3, 24-well culture plate), and incubated in osteogenic medium containing no treatment, BMP-2, and each size of GNPs for 7, 14 and 21 days. At each predetermined time interval, the cell-seeded wells were washed with DPBS twice, fixed for 20 min
W.-K. Ko et al. / Journal of Colloid and Interface Science 438 (2015) 68–76
71
Fig. 2. Evaluation of cytotoxicity on non-treated ADSCs (control), ADSCs containing BMP-2 and each size of GNPs for 1, 3, and 7 days as measured by (A) live/dead assay and (B) CCK-8 (⁄P < 0.05, ⁄⁄P < 0.01, n = 3).
using 3.7% formaldehyde, and then washed with DPBS again. All the cells were stained in an incubator at 37 °C under 5% CO2 for 30 min using 40 mM alizarin red staining solution (pH 4.2) [25]. The staining solution was removed after 30 min, washed with D.W. three times, and then was examined using an inverted fluorescence microscope. For a quantitative analysis, the stained cells were desorbed with 10% 1-hexadecylpyridinium chloride and the absorbance was measured by using a microplate reader at 540 nm.
The fold change of the control group (osteogenic medium only) at 7 days of culture was set at 1-fold and the ratio of the normalized fold change was calculated. The real-time PCR amplifications were carried out 10 s at 95 °C, 30 s at 53–56 °C (COL1: 53 °C, Runx2: 56 °C, BSP:54 °C, OCN:55 °C) and 30 s at 72 °C for 45 cycles after the initial denaturation step for 10 min at 95 °C. Each primer was designed for real time PCR (Table 2) and purchased from Bioneer. All the real time PCR results were normalized by glyceraldehyde 3-phosphate dehydrogenase (GAPDH) [26].
2.9. Real time PCR 2.10. Statistical analysis 4
ADSCs (a density of 5.0 10 cells/well) were seeded on each well (n = 3, 24-well culture plate), and incubated in osteogenic medium containing no treatment, BMP-2, and each size of GNPs for 7, 14 and 21 days. At each predetermined time interval, the cell-seeded wells were washed with DPBS twice and afterward the entire RNA from each group’s cells was isolated using an RNeasy Plus Mini Kit (Qiagen, CA, USA) according to manufacturer’s instructions. 1 lg of the entire RNA was extracted from all specimens, and subsequently transcribed into cDNA using an AccuPower Cycle Script RT Premix (Bioneer, Daejeon, Republic of Korea). Real-time PCR was analyzed by using iQ SYBR Green supermix (Bio-Rad, Hercules, CA, USA). Threshold cycle values were calculated by using a comparative cycle threshold (CT) method.
Statistical analysis was performed using PASW Statistics 18 software (SPSS, Inc.). All values were expressed as means ± standard deviations, and differences with p-values (⁄P < 0.05, ⁄⁄ P < 0.01) were considered statistically significant. 3. Results 3.1. Synthesis of gold nanoparticles and size confirmation GNPs have a different maximum absorbance wavelength in accordance with each GNPs size [21]. All spherical GNPs were synthesized according to the Huang report [21]. The size and
72
W.-K. Ko et al. / Journal of Colloid and Interface Science 438 (2015) 68–76
Fig. 3. Dark-field microscopy analysis of ADSCs involved with no treatment (control) and each size of GNPs after 8 h: (A) the scattering images, and (B) quantitative analysis according to various sizes of GNPs (⁄P < 0.05, n = 3).
distribution were confirmed by TEM, DLS, and UV–Vis spectroscopy (Fig. 1). For visual confirmation, GNPs were observed by TEM. Each GNPs solution was dropped on carbon and photographed (Fig. 1A). All synthesized GNPs had spherical shape, and the sizes were distributed well. For accurate confirmation, DLS was conducted and the following results were elicited. GNPs have the maximum frequency in a size distribution graph at 15.5 (15 nm target), 30.2 (30 nm target), 50.3 (50 nm target), 75.8 (75 nm target), and 100.3 nm (100 nm target), respectively (Fig. 1B). At last, synthesized GNPs were analyzed by UV–Vis spectroscopy. The measured range was from 400 to 800 nm (Fig. 1C). All GNP groups displayed a smooth curve from 400 to 800 nm range and the GNPs of each size have maximum absorptions at 517.5 (15 nm), 522 (30 nm), 533 (50 nm), 541.5 (75 nm), and 546.5 nm (100 nm), respectively. The color of each solution changed from yellow to black and then to red or purple depending on the size of the nanoparticles (data not shown). The approximate amounts of citrate acid and the corresponding sizes of nanoparticles are listed Table 1. This shows that smaller nanoparticles require more citrate solutions, and larger nanoparticles require less citrate solution. The peak intensity and peak position of GNP solutions are dependent on the size, shape, and concentration of GNPs [21]. In conclusion, the above results match Huang’s results indicating successful formation of spherical GNPs with different sizes. 3.2. Cell viability test We conducted a cell viability test on ADSCs which had been cultured for 1, 3, and 7 days with BMP-2 and each size of GNPs, and then these results were compared with that of non-treated group (control). There were almost no dead cells after 7 days of incubation by the live/dead assay (Fig. 2A). In order to quantify, we measured the cell viability percentage among groups using the CCK-8 assay (Fig. 2B). Generally, all the GNP treated groups displayed a higher cell viability than control group during 7 days of incubation. These results indicate that 1 lM of GNPs has no significant cytotoxicity toward ADSCs.
3.3. Dark field microscopy analysis To confirm the intracellular uptake properties of each size of GNPs, dark-field microscopy observations were carried out (Fig. 3A). Seeded cells were incubated for 1 day in order to attach to the culture plate. Subsequently these were treated with each size of GNPs for 8 h [27]. The ADSCs in each group were observed under bright or dark-field imaging conditions. In dark-field imaging conditions, the GNPs in cells were observed as white spots against a dark background [7,28]. The results showed that more white spots were observed in cells treated with 30 and 50 nm sized GNPs as compared to other sizes. As shown in Fig. 3B, the 50 nm group showed the highest uptake ratio of GNPs of all groups. All differences were significant except for those between 75 and 100 nm groups (⁄P < 0.05). In summary, it was found that the uptake amount of GNPs during 8 h increased in the order of 100, 75, 15, 30, and 50 nm groups. 3.4. ALP staining and activity assay The ALP staining/activity levels indicating differentiation of ADSCs toward osteoblasts were examined. This was done with cells cultured on plate wells for 7, 14 and 21 days with osteogenic medium containing no treatment, BMP-2, and each size of GNPs. The ALP staining was observed by an inverted fluorescence microscope (Fig. 4A). Generally, the maximum level of ALP is expressed at 2 weeks during osteogenesis [29]. The quantity of ALP staining had reached a maximum by 2 weeks, and then decreased at 3 weeks in all groups. The 30 and 50 nm groups had more ALP staining than others during the 3 week study. In order to quantify, the ALP activity assay was conducted under the same conditions and quantitatively measured (Fig. 4B). The ALP activity results showed strong similarities with the ALP staining results. At every predetermined time interval, the 30 and 50 nm groups were observed to have higher ALP levels than others. The BMP-2 positive control group’s level was similar to the level of 75 or 100 nm GNPs. These results indicate that GNPs can positively
W.-K. Ko et al. / Journal of Colloid and Interface Science 438 (2015) 68–76
73
Fig. 4. ALP staining and activity level during the process of differentiation from ADSCs toward osteoblasts for 3 weeks: (A) ALP stained cells, which were treated with osteogenic medium alone (control), BMP-2, and each size of GNPs for 3 weeks as observed by an optical microscope, and (B) ALP activity levels of cells which were cultured at the same conditions for 3 weeks as observed by a microplate reader (⁄P < 0.05, ⁄⁄P < 0.01, n = 3).
accelerate the osteogenic differentiation of ADSCs more so than BMP-2. The 50 nm GNPs group expressed the highest ALP level. In general the ALP activity level increased in the order of 100, 75, 15, 30, and 50 nm GNPs groups. 3.5. Calcium deposition staining and activity assay Calcium deposition was measured for cells incubated for 7, 14 and 21 days with osteogenic medium only, osteogenic medium containing BMP-2, and osteogenic medium containing each size of GNPs. The quantity of deposited calcium was compared to control group after staining with alizarin red solution (Fig. 5A). The deposited calcium degree of all groups increased during the 3 weeks. The 30 and 50 nm GNPs groups displayed strong deposition, which matched the ALP assay results. In order to quantify these results, the deposited calcium from each group was dissolved using 10% 1-hexadecylpyridinium chloride. Then the optical density (O.D.) was measured by a microplate reader (Fig. 5B). There was no great difference among all groups at 1 week, but 30 and 50 nm GNPs groups had higher levels at 2 and 3 weeks. This indicates that 30 and 50 nm of GNPs strongly promoted mineralization of osteoblasts. The BMP-2 group had a higher O.D. than the control group, but did not exceed the O.D. of the GNP groups.
3.6. Real-time PCR We examined mRNA expression of osteoblast gene markers, such as collagen type I (COL1), runt-related transcription factor 2 (Runx2), bone sialoprotein (BSP) and osteocalcin (OCN) of cells at the varying incubation conditions (Fig. 6). As the culture time increased, the amount of synthesized COL1 decreased in all specimens (Fig. 6A). Generally, COL1 is regarded as an early marker during osteogenic differentiation of MSCs [29] which is in agreement with our results. High levels of COL1 were observed with 30 and 50 nm GNPs groups at each time interval. Another osteogenic transcription factor, Runx2, was also expressed at its highest levels after 2 weeks of incubation in all specimens (Fig. 6B). The highest intensity of Runx2 was recorded with the 50 nm group which is similar to the COL1 results. Fig. 6C shows the BSP mRNA expression levels during 21 days of culture. It was found that the BSP mRNA expression levels increased in all groups as culture time increased and the 50 nm group’s level was higher during the 3 weeks than the rest. The mRNA expression of OCN was similar to BSP (Fig. 6D). For all markers, it was found that the mRNA expression levels of GNPs groups were higher than the control groups. The 30 and 50 nm GNPs groups showed the highest levels among all GNPs groups. The BMP-2 group showed similar levels to the 75 and 100 nm group for all primers.
74
W.-K. Ko et al. / Journal of Colloid and Interface Science 438 (2015) 68–76
Fig. 5. Alizarin red staining during the process of differentiation from ADSCs toward osteoblasts for 3 weeks: (A) Stained cells with alizarin red solutions, which were treated with osteogenic medium containing none (control), BMP-2, and each size of GNPs for 3 weeks as observed by confocal laser scanning microscopy, and (B) deposition degrees of lysed calcium, as measured by a microplate reader (⁄P < 0.05, ⁄⁄P < 0.01, n = 3).
4. Discussion GNPs are widely used in many applications such as biomedical imaging, clinical diagnostics, and therapeutics [1–8]. These areas are associated with nanomaterials that have different sizes and several types of shape [30–35]. GNPs are an attractive material in and of themselves. Here, we used GNPs as an accelerator for differentiation of ADSCs toward osteoblasts, and investigated the osteogenic expression of ADSCs as they were influenced by GNPs. GNPs were synthesized using the citrate acid reduction method of HAuCl4 and their sizes were confirmed by TEM, DLS, and UV–Vis spectroscopy (Fig. 1). There exists a specific oscillation depending on the size and shape of GNPs [21]. GNPs possess strong distance-dependent optical density in accordance with their plasmon resonance. The experimental results showed quite a similar graph as Huang’s [21], but did not show exactly the same maximum wavelength. Huang’s result had shown an increasing maximum wavelength with increasing size of GNPs. Therefore, it was necessary to confirm the sizes of the synthesized GNPs. DLS is used routinely to analyze the size and size distribution of polymers, proteins, colloids, and nanoparticles [36]. In this study, DLS was conducted for quantitative detection and analysis of GNPs size. As a result, we were able to prepare GNPs with the following
distribution: 15.5 ± 2.4, 30.2 ± 2.5, 50.3 ± 1.25, 75.8 ± 0.95 and 100.3 ± 5.05 nm, respectively (Fig. 1B). The distribution figures showed that the GNPs were synthesized appropriately for the experiments of this study. Prior to performing experimentation with ADSCs, the usable concentration of GNPs which does not cause toxicity had to be established. Too high of a concentration of GNPs will inevitably drive cells to death [38], too low concentrations of GNPs would have little influence on differentiation. It was established that 1 lM of GNPs displayed no toxicity toward cells in both viability test and live/dead assays over the course of 1 week of incubation (Fig. 2). This is despite the GNPs being gradually cumulated in the cells. Previous studies [37,38] have also supported that 1 lM of GNPs does not display toxicity toward cells. In addition, 1 lM of GNPs, demonstrated an obvious difference of each group toward the osteogenic differentiation process. To analyze these results, we measured the quantitative density of each size of GNPs up taken into the cell membranes. After a preset time of 8 h for each group, the uptake properties of the GNPs were measured via the dark field assay. This uptake time was selected based on a prior Devika et al. study [27] which demonstrated that maximum uptake was observed from 5 to 8 h. In Fig. 3, the 50 nm sized GNPs group expressed the highest uptake among the 5 sizes of GNPs groups. The uptake increased in the order of 100, 75, 15, 30, and
75
W.-K. Ko et al. / Journal of Colloid and Interface Science 438 (2015) 68–76
Fig. 6. The mRNA expression of osteogenic markers during the process of differentiation for 3 weeks: (A) COL1, (B) Runx2, (C) BSP, and (D) OCN (⁄P < 0.05,
50 nm sizes. We tested that this result might affect differentiation of ADSCs toward osteoblasts among groups through the following experiments. ALP and calcium deposition assays were used respectively as early and late markers of osteoblast differentiation. These results were consistent with our results as well (Figs. 4 and 5). ALP is a typical ecto-enzyme, which acts as a marker of osteogenic differentiation [29]. Therefore, we measured the ALP level by both quantitative and qualitative methods (Fig. 4). Fig. 4A and B both demonstrated the highest peak at 2 weeks in all specimens. The 50 nm GNPs group had the highest level among all groups and the 30 nm GNPs also demonstrated higher levels than other GNPs groups. These results corresponded well with the calcium deposition assay results (Fig. 5). Typically, after culturing for 2–3 weeks, osteoblast nodes were formed and reached a peak quantity when osteoblasts started to mineralize [39]. These results show that 30 and 50 nm GNPs groups especially lead to high osteoblast differentiation. For a more analytical approach, we evaluated the mRNA expressions of COL1, Runx2, BSP, and OCN by real time PCR (Fig. 6). At all tested time points for all osteogenic markers, the 30 and 50 nm GNPs had high levels of expression. COL1 is the most abundant protein in bone matrix and usually is formed during the proliferation of osteoblast cells [40,41]. In Fig. 6A, mRNA expressions in all groups were recorded at peak levels at 1 week during the 3 week study. The groups containing GNPs had higher levels than that of the group containing BMP-2, particularly the 50 nm group recorded the highest expression levels. We evaluated the typical osteogenic transcriptional factor Runx2 as a critical regulator of osteogenic differentiation [40]. The Runx2 mRNA expression of all specimens increased during 2 weeks, and then decreased at 3 weeks of culture (Fig. 6B). This result means that over expressed Runx2 presumably inhibits its own production to prevent the
⁄⁄
P < 0.01, n = 3).
differentiation of mature osteoblasts toward osteocytes [40]. During 3 weeks of culture time, the 30 and 50 nm GNPs groups had comparatively higher levels among the tested GNPs sizes. The expressions of BSP and OCN markers related to calcium deposition in the extracellular matrix during its mineralization stage [40] were increased over the course of 3 weeks and also showed the superiority of the 50 nm sized GNPs (Fig. 6C and D). The expressed Runx2 promoted BSP and OCN expression, which promotes the mineralization of osteoblasts (Fig. 5). These data show that the amount of uptake of GNPs increased in a similar order as osteogenic differentiation. The attached proteins around GNPs as well as uptake GNPs themselves might affect the osteoblast differentiation of ADSCs [42]. In summary, ADSCs incubated with 30 or 50 nm GNPs lead to more differentiation of the ADSCs as compared to other groups. The 50 nm GNPs group especially promoted osteogenic differentiation of ADSCs. However, the osteogenic differentiation levels of larger GNPs were lower than the 30 nm GNPs group. The above results indicate that the size of GNPs is quite an important factor in their impact on osteogenic differentiation. 5. Conclusion In this study, all the data consistently demonstrated that the 30 and 50 nm GNPs were effective at enhancing osteogenic differentiation. Although we could not confirm the exact signaling mechanism after endocytosis into the cell’s membrane, we could show the effectiveness of 30 and 50 nm GNPs through the expression of Runx2 genes and its linked osteogenic markers’ expression by several osteogenic experiments. All results show that the osteogenic differentiation was dependent on the size of GNPs. Therefore, we suggest that the nanoparticle size is important in the biological
76
W.-K. Ko et al. / Journal of Colloid and Interface Science 438 (2015) 68–76
environment. The results of this study might assist in designing nanostructures for future drug delivery and therapeutic systems. Acknowledgments This research was supported by the Pioneer Research Center Program through the National Research Foundation of Korea funded by the Ministry of Education, Science and Technology (2013-007552) and the Public welfare & Safety research program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (NRF-20120008610) and by Bio-industry Technology Development Program (312062-5) of iPET (Korea Institute of Planning and Evaluation for Technology in Food, Agriculture, Forestry and Fisheries), Ministry for Food, Agriculture, Forestry and Fisheries, Republic of Korea. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13]
M. De, P.S. Ghosh, V.M. Rotello, Adv. Mater. 20 (2008) 4225. J. Xie, S. Lee, X. Chen, Adv. Drug Deliv. Rev. 62 (2010) 1064. A. Tautzenberger, A. Kovtun, A. Ignatius, Int. J. Nanomed. 7 (2012) 4545. D.A. Giljohann, D.S. Seferos, W.L. Daniel, M.D. Massich, P.C. Patel, C.A. Mirkin, Angew. Chem. Int. Ed. 49 (2010) 3280. L. Dykman, N. Khlebtsov, Chem. Soc. Rev. 41 (2012) 2256. E. Boisselier, D. Astruc, Chem. Soc. Rev. 38 (2009) 1759. D.N. Heo, D.H. Yang, H.-J. Moon, J.B. Lee, M.S. Bae, S.C. Lee, W.J. Lee, I.-C. Sun, I.K. Kwon, Biomaterials 33 (2012) 856. Y.-I. Chung, K.-M. Ahn, S.-H. Jeon, S.-Y. Lee, J.-H. Lee, G. Tae, J. Control. Release 121 (2007) 91. C. Yi, D. Liu, C.-C. Fong, J. Zhang, M. Yang, ACS Nano 4 (2010) 6439. D. Liu, J. Zhang, C. Yi, M. Yang, Chin. Sci. Bull. 55 (2010) 1013. O.-J. Sul, J.-C. Kim, T.-W. Kyung, H.-J. Kim, Y.-Y. Kim, S.-H. Kim, J.-S. Kim, H.-S. Choi, Biosci. Biotechnol. Biochem. 74 (2010) 2209. P.A. Zuk, M. Zhu, P. Ashjian, D.A. De Ugarte, J.I. Huang, H. Mizuno, Z.C. Alfonso, J.K. Fraser, P. Benhaim, M.H. Hedrick, Mol. Biol. Cell 13 (2002) 4279. C.M. Cowan, Y.-Y. Shi, O.O. Aalami, Y.-F. Chou, C. Mari, R. Thomas, N. Quarto, C.H. Contag, B. Wu, M.T. Longaker, Nat. Biotechnol. 22 (2004) 560.
[14] M.D. Kwan, B.J. Slater, D.C. Wan, M.T. Longaker, Hum. Mol. Genet. 17 (2008) R93. [15] A.W. James, B. Levi, E.R. Nelson, M. Peng, G.W. Commons, M. Lee, B. Wu, M.T. Longaker, Stem Cells Dev. 20 (2010) 427. [16] M.S. Bae, J.E. Kim, J.B. Lee, D.N. Heo, D.H. Yang, J.-H. Kim, K.-R. Kwon, J.B. Bang, H. Bae, I.K. Kwon, Carbohydr. Polym. 92 (2013) 167. [17] J.L. Dragoo, J.Y. Choi, J.R. Lieberman, J. Huang, P.A. Zuk, J. Zhang, M.H. Hedrick, P. Benhaim, J. Orthop. Res. 21 (2003) 622. [18] J.L. Dragoo, J.R. Lieberman, R.S. Lee, D.A. Deugarte, Y. Lee, P.A. Zuk, M.H. Hedrick, P. Benhaim, Plast. Reconstr. Surg. 115 (2005) 1665. [19] M. Knippenberg, M. Helder, B. ZandiehDoulabi, P. Wuisman, J. Klein-Nulend, Biochem. Biophys. Res. Commun. 342 (2006) 902. [20] N. Jaiswal, S.E. Haynesworth, A.I. Caplan, S.P. Bruder, J. Cell. Biochem. 64 (1997) 295. [21] X. Huang, Ph.D. thesis, Georgia Tech, 2006. [22] P. CooperáStevenson, Discuss. Faraday Soc. 11 (1951) 55. [23] G. Frens, Nature 241 (1973) 20. [24] F.M. Veronese, Biomaterials 22 (2001) 405. [25] H. Anderson, Lab. Investig. J. Tech. Methods Pathol. 60 (1989) 320. [26] K. Yang, J.S. Lee, J. Kim, Y.B. Lee, H. Shin, S.H. Um, J.B. Kim, K.I. Park, H. Lee, S.W. Cho, Biomaterials 33 (2012) 6952. [27] B.D. Chithrani, A.A. Ghazani, W.C. Chan, Nano Lett. 6 (2006) 662. [28] I.-C. Sun, S. Lee, H. Koo, I.C. Kwon, K. Choi, C.-H. Ahn, K. Kim, Bioconjug. Chem. 21 (2010) 1939. [29] D. Liu, C. Yi, D. Zhang, J. Zhang, M. Yang, ACS Nano 4 (2010) 2185. [30] K.C. Grabar, R.G. Freeman, M.B. Hommer, M.J. Natan, Anal. Chem. 67 (1995) 735. [31] W.K. Leutwyler, S.L. Bürgi, H. Burgl, Science 271 (1996) 933. [32] R. Elghanian, J.J. Storhoff, R.C. Mucic, R.L. Letsinger, C.A. Mirkin, Science 277 (1997) 1078. [33] Y.C. Cao, R. Jin, C.A. Mirkin, Science 297 (2002) 1536. [34] J.M. Klostranec, W.C. Chan, Adv. Mater. 18 (2006) 1953. [35] L.R. Hirsch, R. Stafford, J. Bankson, S. Sershen, B. Rivera, R. Price, J. Hazle, N. Halas, J. West, Proc. Natl. Acad. Sci. 100 (2003) 13549. [36] B.J. Berne, R. Pecora, Dynamic light scattering: with applications to chemistry, biology, and physics, Courier Dover Publications, 2000. [37] Y. Pan, S. Neuss, A. Leifert, M. Fischler, F. Wen, U. Simon, G. Schmid, W. Brandau, W. Jahnen-Dechent, Small 3 (2007) 1941. [38] E.E. Connor, J. Mwamuka, A. Gole, C.J. Murphy, M.D. Wyatt, Small 1 (2005) 325. [39] G.S. Stein, J.B. Lian, Endocr. Rev. 14 (1993) 424. [40] T. Komori, J. Bone Miner. Metab. 21 (2003) 193. [41] K. Kadler, D. Holmes, J. Trotter, J. Chapman, Biochem. J. 316 (1996) 1. [42] M. Mahmoudi, I. Lynch, M.R. Ejtehadi, M.P. Monopoli, F.B. Bombelli, S. Laurent, Chem. Rev. 111 (2011) 5610.