Article pubs.acs.org/est
Size-Dependent Toxicity of Nano-C60 Aggregates: More Sensitive Indication by Apoptosis-Related Bax Translocation in Cultured Human Cells Maoyong Song,† Shaopeng Yuan,‡ Junfa Yin,† Xiaoli Wang,† Zihui Meng,§ Hailin Wang,*,† and Guibin Jiang† †
State Key Laboratory of Environmental Chemistry and Ecotoxicology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing, 100085, China ‡ Department of Pharmacology, Institute of Materia Medica, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, 100050, China § School of Chemical Engineering and Environment, Beijing Institute of Technology, Beijing, 100081, China ABSTRACT: The toxicity of NPs is not well characterized in terms of their size. In particular, the size-based toxicity of fullerene (C60) remains an issue because of a lack of C60 NPs with a well-controlled size. In this work, six fractions of the nano-C60 aggregates (nC60) with different size distribution were prepared by a simple differential centrifugation. By using these nC60 fractions, we demonstrate the size-dependent inhibition of DNA polymerase and reduced-size enhanced cytotoxicity. Above all, we found that nC60 NPs with smaller size may have higher toxicity potency. These sizedependent effects were observed at the high exposure doses (4−6 mg/L). Interestingly, at 20-times lower and noncytotoxic doses, the size-dependent effect can be indicated by apoptosis-related fluorescent protein fused Bax translocation. Considering the toxicity of NPs is often ignored in the traditional end-point analysis for cytotoxicity when the exposure dose is low, the findings presented here will assist in the evaluation of the size-dependent cytotoxicity and dose−response relationships of toxicity mediated by nC60 NPs at low doses.
■
INTRODUCTION Fullerenes have attracted great attention in biological and medical applications,1−6 meanwhile, the safety of these nanomaterials is of great concern.7 Although pristine C60 is poorly soluble in water, several methods have been developed to prepare dispersible colloidal aggregates of C60 (nC60) in aqueous solutions.8,9 These nC60 particles are monthly or yearly stable. This implicates that nC60 could be chronically exposed to the biological and environmental systems. Because of the high stability of buckyball structure,10 increasing quantities in the lab,3−7 and industrial production in large scale, strong attention has been paid to the potential risk of nC60 NPs to human health and environmental impact. Previous toxicity tests of aqueous fullerene C60 suspensions yielded both positive and negative results. Jia et al. incubated alveolar macrophage (from adult pathogen-free healthy guinea pigs) with nC60 (up to 226 μg/cm2) and found no significant cytotoxicity.11 Fiorito et al. also found a very low toxicity to elicit an inflammatory response and low cellular uptake of pristine C60 against human macrophage cells.12 Conversely, Sayes et al. reported that nC60 was highly cytotoxic to human dermal fibroblast and human hepatocellular liver carcinoma (HepG2).13 Since it was reported that nC60 NPs induced oxidative stress on the brain of juvenile largemouth bass,14 later studies presented more experimental evidence that fullerenes © 2012 American Chemical Society
could migrate throughout the mice body and accumulate in the mouse liver,15,16 and accumulate in lysosomes, cytoplasm, along the plasma membrane, and inside the nucleus of human monocyte macrophages.17,18 However, a result of Shinohara et al. indicated that very few or no nC60 NPs reached the brain of C. carpio across the blood−brain barrier and no lipid peroxidation was induced in their brains.19 To date, the understanding of fullerenes’ toxicity and bioactivity remains limited, probably due to the diversity of experimental conditions and the complexity of nC60 NPs, which may vary in size, shape, and charge. The discoveries in the past decades have implicated that the particle size plays an important role in the cellular uptake and toxicity of artificial viruses (DNA coated glycocluster NPs),20 quantum dots,21,22 iron oxide NPs,23 titanium dioxide NPs,24 and gold NPs.25−27 Recently, Chan et al. reported that the binding and activation of membrane receptor and subsequent protein expression strongly depended on the size of the nanoparticles.28 For nC60 NPs, their reactivity may be influenced by the size and aggregation state of these particles.29 Received: Revised: Accepted: Published: 3457
November 2, 2011 February 21, 2012 February 21, 2012 February 21, 2012 dx.doi.org/10.1021/es2039008 | Environ. Sci. Technol. 2012, 46, 3457−3464
Environmental Science & Technology
Article
fractions for TEM analysis were prepared by depositing 5 μL nC60 suspensions on an ultrafine carbon support film on copper grid, and drying overnight in a dust-free box. UV/vis absorbance spectra of nC60 suspensions were obtained using a dual-beam, high-resolution UV/vis spectrophotometer (DU800, Beckman Coulter, USA). The scan was performed in the wavelength range of 200−800 nm. The slit width and sample interval were set 1 and 0.2 nm. Inhibition of Enzyme Activity in vitro. The enzyme activity assays was carried out based on our previous work.34 Briefly, the polymerase chain reaction can be inhibited by water-soluble fullerene via inhibition of the activity of polymerase. The inhibition of polymerase activity results from high affinity interaction between nC60 NPs and the polymerase. Here, different nC60 suspensions were added to DNA amplified system, and the activity of polymerase was measured by realtime PCR and gel electrophoresis. The sequence of ssDNA template (75 bases) is 5′CCGCCTGATTAG CGATACTTACGTGAGCGTGCTGTCCCCTAAAGGTGATACGTCACTTGAGCAAAATCACCTGCA-3′, which was synthesized by Sangon Biological Engineering Technology and Services (Shanghai, China). The real-time amplification and analysis of the template DNA was carried out using a Mx3005P QPCR Systems (Stratagene, USA). The real-time PCR reagents were from Brilliant SYBR QPCR Master Mix kit (Stratagene, USA). The subsequent temperature cycling program included 94 °C for 30 s, 55 °C for 30 s, and 72 °C for 30 s, 40 cycles, and final extension at 72 °C for 10 min. In vitro Cytotoxicity. Human lung adenocarcinoma cell line A549 was cultured in a 5% CO2 in Dulbecco’s modification of Eagles media (DMEM) containing 10% fetal bovine serum (Hyclone, South Logan, Australia), 100 U/mL penicillin, and 100 μg/mL streptomycin. Cells were grown to 70−80% confluency before exposure to nC60 NPs; each exposure plate was incubated in the dark at 37 °C for 16 h. Cytotoxicity was measured using a LIVE/DEAD Viability/Cytotoxicity Kit (Molecular Probes, Invitrogen, Eugene, OR). Live cells were determined with the polyanionic dye calcein AM, which is retained within live cells, and produces an intense uniform green fluorescence. Dead cells were identified by use of ethidium homodimer-1 (EthD-1) dye, which binds occult DNA when cells are lysed, and produces a bright red fluorescence. The percentage of dead cells can be calculated from the fluorescence readings and defined as
However, different from gold NPs, there is no method available for the size-controlled synthesis of nC60 NPs. Lyon et al. separated nC60 NPs in suspension into two size fractions by centrifugation (large and small particle sizes).30 The smaller nC60 NPs tend to have a higher level of antibacterial activity than the larger NPs. The suspension of fullerene was also fractionated by successive filtration with membranes having nominal pore sizes from 50 to 800 nm.31 It reported that the production of singlet oxygen in the smallest fraction (filtered with 50-nm membrane) was approximately 10 times than that of the bulk nC60 suspension. In spite of what has been achieved so far, little is known about the size effect of nC60 NPs on toxicity and bioactivity, primarily because of a lack of nC60 with well-controlled size and the size variance caused by aggregation in environment media. Herein, this research has two aims: the first is to separate nC60 aggregates into a series of fractions with a different size distribution by differential centrifugation, and the second is to investigate whether the size of nC60 NPs affects their toxicity and bioactivity using the obtained size-different fractions. To obtain a reliable result, the size of fractionated nC60 aggregates was characterized using a laser particle sizer, UV/vis spectrometer, and TEM; meanwhile, several biological assays were also combined to evaluate the cytotoxicity and bioactivity: enzyme activity inhibition, live/dead cell staining, and GFP-Bax translocation.
■
MATERIALS AND METHODS Size Separation of nC60 NPs. Fullerene (C60, 99.9%) was purchased from Sigma-Aldrich (St. Lous, MO). Suspension of nC60 was produced using the method described by Anderievsky.32 This technique is based on the transfer of fullerenes from nonpolar solvent (toluene) to a polar one (water) under ultrasonic treatment. A solution of 1 g/L C60 in toluene was prepared, and then 100 mL of this solution was added to 1 L of Milli-Q (Millipore) water. After mixing, this mixture was sonicated until all of the toluene had evaporated. The yellow suspension was ultimately filtered through a 0.45-μm microfilter, resulting in a suspension with concentration up to 60 mg/ L. The nC60 NPs in suspension were separated into different particle size fractions by differential centrifugation in a CR22GII centrifuge (Hitachi, Japan). The supernates were removed with a pipette from the bottles while the depositions were still in the bottom of bottles. The original nC60 suspension that was not separated with centrifugation is named as S0, and the supernates obtained by centrifugation at 4000, 8000, and 10000 rpm for 5 min were correspondingly named as S4, S8, and S10. The nC60 suspension (S0) was also further centrifuged at 15 000 rpm for 20 min in order to obtain more “smaller” fraction, and the supernate was named S15. One deposition obtained by the centrifugation of nC60 suspension at 4000 rpm for 5 min was added by the same volume of water then shaken to form suspension. The reformed suspension was named SR. Size Characterization of nC60 Suspensions. Concentration of C60 in all suspensions was determined by the mass quantification method for C60 described in a separate publication.33 The hydrodynamic diameter of particle was obtained with a Zetasizer Nano S90 (Malvern Instruments Ltd. UK). All measurements were performed at 25 °C. Six measurements (12 runs per measurement) were acquired for each fraction. Morphology and structure of nC60 NPs were analyzed by TEM (H-7500, Hitachi, Japan) imaging. The
%Dead cells =
F(645)sam − F(645)min × 100% F(645)max − F(645)min
where F(645)sam is fluorescence at 645 nm in the experimental cell sample, labeled with calcein AM and EthD-1; F(645)max is fluorescence at 645 nm in a sample where all the cells are dead (treatment with 0.1% saponin for 10 min), labeled with EthD-1 only; F(645)min is fluorescence at 645 nm in a sample where all the cells are dead, labeled with calcein AM only. Bax Translocation Analysis. GFP-Bax stable MCF7 cells were plated onto 96-well plates. The cells were then treated with different concentrations of nC60 suspensions for 24 h. The percentages of GFP-Bax punctuate cells were determined by fluorescence microscopy as previously described.35,36 3458
dx.doi.org/10.1021/es2039008 | Environ. Sci. Technol. 2012, 46, 3457−3464
Environmental Science & Technology
Article
Figure 1. (a) Digital image of centrifuge vessels containing nC60 suspensions; (b) mean hydrodynamic diameter of nC60 suspensions; (c) size distributions by intensity of nC60 suspensions.
Figure 2. TEM micrographs of nC60 NPs in suspensions.
■
RESULTS AND DISCUSSION Separation and Characterization of nC60 Suspensions. Because there is no procedure exploited for the sizecontrollable synthesis of colloidal suspensions of nC60, postsynthesis separation is of choice to prepare particle fractions with different size distribution. Here we developed a differential centrifugation method to separate nC60 suspension into a series of fractions with different mean hydrodynamic diameter. It is known that the sedimentation speed of the particles increases as the centrifugal force increases. Thus, we performed a rapid separation by simply changing centrifugation speed to generate differential centrifugal force. As shown in
Figure 1 a, with the same centrifugation time, the color of nC60 suspension or supernatant is fading gradually with the increasing centrifugation speed, indicating the precipitation of more nC 60 from the suspension with the increasing centrifugation speed. The mean hydrodynamic diameter of each nC60 suspension gradually decreased with the increasing of centrifugation speed (Figure 1b and 1c). This is consistent with the centrifugation-based separation principle, by which the isolated nC60 NPs suspended in water can be fractionized by size. After being centrifuged at 10 000 rpm for 5 min, the mean hydrodynamic diameter of nC60 NPs was decreased from 179.3 ± 3.4 nm (S0) to 123.4 ± 4.9 nm (S10). S15 has the smallest 3459
dx.doi.org/10.1021/es2039008 | Environ. Sci. Technol. 2012, 46, 3457−3464
Environmental Science & Technology
Article
mean hydrodynamic diameter (108.6 ± 3.5) among all nC60 fractions. The redissolved deposition (SR) has a larger mean hydrodynamic diameter (228.2 ± 7.5) than other fractions. The proportion of “smaller” nC60 NPs accounted in each fraction increased with the increasing centrifugation speed. These results indicate that the “larger” nC60 NPs will be preferentially precipitated from the suspension by centrifugation. For the SR fraction, it should be emphasized that only a part of nC60 NPs could be deposited again at the same centrifugation condition. This means the resuspended nC60 NPs, at least a part of them, have smaller size than that before centrifugation, indicating that the larger aggregates can be sheared into smaller aggregates by centrifugal force. We further characterized the prepared fractions of nC60 using TEM analysis. As shown by TEM images in Figure 2, the uncentrifuged nC60 suspension was typically present as wide scale aggregates with dimensions on the order of 50−450 nm (S0). After being centrifuged at 4000 rpm for 5 min, more particles with smaller size (d < 200 nm) were observed (Figure 2, S4). When the centrifugation speed increased to 8000 rpm, almost all the particles observed in the TEM image were less than 200 nm (Figure 2, S8). And the proportion of “smaller” particles (d < 100 nm) and “larger” particles (d > 100 nm) was close to 1:1 (determined from TEM images, n = 3). When the centrifugation speed was set at 10 000 rpm, almost only “smaller” ones (d < 100 nm) could be observed (Figure 2, S10). For the further centrifuged fraction of S15, the size of nC60 NPs was less than 50 nm. Resuspended nC60 NPs (SR fraction) were characterized as larger elongated aggregates (d > 200 nm) and very few smaller round aggregates (d < 200 nm). In the TEM images, the shape appeared to be relatively smooth and round for the smaller nC60 NPs, while the larger ones were more amorphous and angular. We also recorded UV/vis absorption spectra of nC60 suspensions. As shown in Figure 3, a blue shift was observed
Size-Dependent Inhibition of DNA Polymerase Activity by nC60. Recently, we reported that the activity of Taq DNA polymerase can be inhibited by nC60.34 Because the polymerase activity is associated with biologically important DNA replication and repair, the C60 NPs induced inhibition of the amplification activity may affect DNA replication, DNA repair, and gene expression, and further change the cellular functions. By taking advantage of this known enzyme system, here we further examined whether the inhibition is size dependent by using the obtained size-different fractions of nC60. The inhibition of polymerase activity by nC60 NPs was determined by real-time PCR and gel electrophoresis analysis after incubation of Taq polymerase with nC60 NPs. The amplified efficiency of PCR decreased upon addition of any investigated fraction of nC60 NPs, showing the inhibition of Taq DNA polymerase activity by nC60. This is consistent with our recent work.34 Moreover, the inhibition is dependent on the size of nC60 (Figure 4a and 4b). The fraction S10 exhibited the strongest inhibitory potency (about 60% inhibition), meanwhile, the fraction S0 exhibited only about 30% inhibition. The fraction SR, with the largest size, displayed the least inhibition, and the residual activity of the polymerase was over >80%. The results indicate that nC60 NPs with smaller size can more strongly block the enzyme activity than bigger ones at the same concentration. Cytotoxicity of nC60 NPs. We measured the cytotoxicity of nC60 NPs on human lung adenocarcinoma cell line A549. After incubating the cell culture with nC60 NPs for 16 h, the cytotoxicity was determined by staining with calcein AM and ethidium homodimer (Ethd-1). The polyanionic dye calcein is well retained within live cells, producing an intense uniform green fluorescence in live cells. On the other hand, the stained impermeable Ethd-1 can only enter dead cells, in which the cellular membranes are damaged and become permeable to the positively charged dye, and intercalated into dsDNA in chromosome. The intercalated ethidium homodimer can emit bright red fluorescence. Each strong red fluorescence dot indicates one dead cell. Given the values of the relative number of live and dead cells (expressed in terms of percentages of whole cells) measured by microplate reader were much lower than 50% of control at available highest concentration of SR, S0, and S4, we did not quantify the cytotoxicity of nC60 NPs by determining the IC50 values. By imaging analysis of live/dead cells, the fractions containing smaller nC60 NPs exhibited stronger cytotoxicity than the fractions containing the larger nC60 NPs (Figure 5). The live and dead cells could be obviously distinguished by colors after exposure to nC60 NPs of the fractions of S10, S8, S4, and S0 at the same concentration (6 mg/L). For the fraction SR, only very few dead cells were observed. The cytotoxicity of nC60 NPs prepared in this work was far less than that in a previous report (the nano-C60 aggregates are toxic to HDF cells at a LC50 value of 20 μg/L).13 This is consistent with the work of Spohn et al.,39 which indicated that water-soluble side products in THF-nC60 suspension rather than fullerenes themselves caused acute toxic effects. The low cytotoxicity of our prepared nC60 indicates the reliability of the preparation without introducing toxic ingredients. nC60 NPs Induce Bax Translocation to Mitochondria. The Bcl-2-associated X protein, Bax, is a protein of the Bcl-2 gene family, and regulates and contributes to cell apoptosis. Because Bax translocation from the cytoplasm to mitochondria represents an important step in apoptosis, we examined the
Figure 3. Changes in UV/vis absorption spectra of nC60 suspensions after centrifugation.
from 356 nm (S0) to 344 nm (S10) and 334 nm (S15) as the average particle size is smaller. For SR, an obvious red shift (372 nm) was observed due to the average particle size increased.37,38 All the analysis obtained by three analytical techniques (dynamic laser scattering, TEM, and UV/vis absorption spectra consistently show that the prepared fractions have distinct size distribution, providing the possibility to evaluate size-dependent effects of nC60 NPs. 3460
dx.doi.org/10.1021/es2039008 | Environ. Sci. Technol. 2012, 46, 3457−3464
Environmental Science & Technology
Article
Figure 4. (a) Real-time PCR curves by plotting SYBR Green I fluorescence intensity versus cycle number in the presence of nC60 NPs with different centrifugation speed. (b) Activity of polymerase after incubation with nC60 NPs. The fluorescence signal at the 25th cycle was estimated as the percentage of that of the positive control. The insert is an electropherogram from polyacrylamide gel electrophoresis analysis of routine PCR products (25 cycles) in the presence of nC60 NPs. The concentrations of nC60 suspensions were 4 mg/L.
Figure 5. (a) Differences in cytotoxicity of nC60 NPs with different size. The live cell is stained with calcein AM (green), and dead cell is stained with ethidium homodimer (red). (b) Percentage of dead cells calculated from the fluorescence readings. Concentrations of nC60 suspensions were 6 mg/ L.
magnified picture in white frame). These results indicated nC60 NPs might potentially trigger apoptotic process. To quantify the distribution of GFP-Bax, we calculated the number of GFP-Bax clusters appearing in the images (Figure 6c). The number of GFP-Bax clusters induced by S15 was about three times that induced by S10 and more than six times that induced by S0. The results indicated that nC60 NPs induced Bax redistribution in a size-dependent manner and smaller nC60 NPs had stronger induction of Bax redistribution than larger ones. The products of nC60 NPs may vary in size, structure, and surface chemistry. These chemistry-similar products, which have been used for the studies of their environmental
effect of nC60 NPs on the subcellular distribution of this protein as the indicator of the cellular signal transduction of apoptosis. An MCF7 cell line stably expressing GFP-Bax for a rapid assessment of Bax subcellular localization was used. As described in previous work, Bax was diffusely distributed in the cytosol initially; some GFP-Bax molecules started to translocate to mitochondria to show a partially filamentous structure afterward; later, more and more GFP-Bax molecules were transloacted to mitochondria to form large clusters.40 As show in Figure 6, there was an increase in the fluorescence pattern (white arrow) of GFP-Bax when these cells were subjected to the exposure of nC60 NPs. Stage of dynamic redistribution of GFP-Bax was also observed (Figure 6, 3461
dx.doi.org/10.1021/es2039008 | Environ. Sci. Technol. 2012, 46, 3457−3464
Environmental Science & Technology
Article
Figure 6. Measurement of GFP-Bax distribution during nC60 NPs induced apoptosis in live MCF-7 cells. (a) TEM images of nC60 NPs (right) and GFP-Bax cluster (left, indicated with arrows) induced by nC60 NPs. (b) Blank of live MCF-7 cells. (c) Results of measurements on the number of GFP-Bax clusters from images (n = 4). The concentrations of nC60 suspensions were 200 μg/L.
Meanwhile, the dose−response relationships obtained in toxicological study are important because they are used for determining appropriate dosages of acceptable limits for exposure to nC60 NPs. Although these results indicate a significant size dependency of enzyme inhibition and the cytotoxicity of nC60 NPs, there is no significant change in both enzyme activity inhibition and cytotoxicity tests for all the fractions when the exposure dose is as low as 200 μg/L. However, at this exposure dose, significant difference in Bax translocation induced by S0, S10, and S15 could be observed, especially S15 (Figure 6). The results indicate that nC60 NPs with smaller size may have higher toxicity potency leading to cell apoptosis at lower exposure dosage or may play an active role in mediating biological effects below the lethal doses. The toxicity of NPs is often ignored due to insensitivity of the traditional end point analysis to the potential cytotoxicity when the exposure dose is low. So the finding present here will assist in the validation of the size-dependent cytotoxicity and dose− response relationships of toxicity mediated by nC60 NPs. Based on our study, we can speculate the mechanisms that govern size-dependent bioactivity of nC60 NPs. Smaller size equals exponentially high numbers per exposure volume at the same mass dose. It is proposed that particle surface area for particles of different sizes is a better dosemetric than particle mass or particle number in dose−response curve.16,43 In this work, the smaller size of nC60 in solution ensured a significantly larger surface area of particles, which was expected to enhance the extent of molecular contact. By coating or nonspecific absorption of proteins, the NPs with smaller size tended to the binding and activation of membrane receptors.25,28 NPs with smaller size (
