In Vitro Toxicity of Silver Nanoparticles at Noncytotoxic Doses to

Jul 9, 2009 - We also used polystyrene nanoparticles (PS-NPs) and silver carbonate (Ag2CO3) as test materials to compare the toxic effects with respec...
6 downloads 9 Views 397KB Size
Download PDF
Environ. Sci. Technol. 2009, 43, 6046–6051

In Vitro Toxicity of Silver Nanoparticles at Noncytotoxic Doses to HepG2 Human Hepatoma Cells KOJI KAWATA, MASATO OSAWA, AND SATOSHI OKABE* Department of Urban and Environmental Engineering, Graduate School of Engineering, Hokkaido University, North-13, West-8, Kita-ku, Sapporo 060-8628, Japan

Received March 11, 2009. Revised manuscript received June 22, 2009. Accepted June 26, 2009.

Although it has been reported that silver nanoparticles (AgNPs) have strong acute toxic effects to various cultured cells, the toxic effects at noncytotoxic doses are still unknown. We, therefore, evaluated in vitro toxicity of Ag-NPs at noncytotoxic doses in human hepatoma cell line, HepG2, based on cell viability assay, micronucleus test, and DNA microarray analysis. We also used polystyrene nanoparticles (PS-NPs) and silver carbonate (Ag2CO3) as test materials to compare the toxic effects with respect to different raw chemical composition and form of silver. The cell viability assay demonstrated that Ag-NPs accelerated cell proliferation at low doses (1.0 mg/L) and induced abnormal cellular morphology, displaying cellular shrinkage and acquisition of an irregular shape. In addition, only Ag-NPs exposure increased the frequency of micronucleus formation up to 47.9 ( 3.2% of binucleated cells, suggesting that Ag-NPs appear to cause much stronger damages to chromosome than PS-NPs and ionic Ag+. Cysteine, a strong ionic Ag+ ligand, only partially abolished the formation of micronuclei mediated by Ag-NPs and changed the gene expression, indicating that ionic Ag+ derived from Ag-NPs could not fully explain these biological actions. Based on these discussions, it is concluded that both “nanosized particle of Ag” as well as “ionic Ag+” contribute to the toxic effects of Ag-NPs.

Introduction Nanomaterials are increasingly being manufactured and used for commercial purposes because of their novel and unique physicochemical properties. These novel properties differ substantially from those bulk materials of the same composition. There are, however, rising concerns about the adverse effects of these materials on human health and environments. Some nanomaterials have been reported to produce reactive oxygen species (ROS) and exert cytotoxicity in vitro (1). Recent studies have shown that nanoparticles can readily pass through cell membranes (2, 3) and even * Corresponding author phone: +81-11-706-6266; fax: +81-11707-6266; e-mail: [email protected]. 6046

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 15, 2009

biological barriers such as the blood-brain barrier and bloodtestis barrier (4), deposit in target organs, and interact with biological systems, which may create toxicity to living cells. Therefore, the establishment of principles and test procedures to ensure the safety of nanomaterials is urgently required. Of various nanomaterials, silver nanoparticles (Ag-NPs) are used most commonly in numerous consumer products including textiles, cosmetics, and health care products for exploiting its strong antimicrobial activity. Information on the toxicity of silver metals and silver salts, which have been used hitherto as silver-impregnated dressing and pharmaceuticals, is available (5), and they are considered to have no adverse effects on the human body when they used in reasonable amounts. However, despite widespread use, there is a serious lack of information concerning the toxicity of Ag-NPs to humans at the cellular and molecular level. There is growing evidence that Ag-NPs are highly toxic to various cultured cells. It has been reported that Ag-NPs exposures decreased viability, increased lactate dehydrogenase (LDH)leakage, or inhibited mitochondrial function in rat liver cells (6), mouse germline stem cells (7), human fibroblasts (8), and rat adrenal cells (9). Furthermore, Kim et al. demonstrated the dose-dependent changes of alkaline phosphatase and cholesterol values, which might be as a consequence of liver damage, in either the male or female rats following 28-day oral exposures to Ag-NPs (10). However, most of the studies evaluated the acute toxic effects of Ag-NPs at relatively high doses. It is required to evaluate the chronic toxicity at low doses, which could be developed by the prolonged internal exposure because Ag-NPs may remain in target organs for a long time. The genotoxicity such as mutagenicity and carcinogenicity of Ag-NPs is still largely unknown. In addition, the toxicity of Ag-NPs at molecular level has not been reported yet so far. In this study, we therefore investigated toxic effects of Ag-NPs to human hepatoma derived cell line HepG2 that were exposed to Ag-NPs at low doses. For toxicity evaluations, cellular morphology, cell viability, and micronucleus formation were assessed under exposed conditions. Furthermore, DNA microarray analysis, which enables the examination of the expression of thousands of genes simultaneously and has been used in in vitro toxicogenomics, was performed to understand the cellular responses at molecular level. We evaluated the potential toxicity of Ag-NPs with emphasis on DNA damaging action or carcinogenicity by correlating cellular responses to gene expression patterns, which could provide a mechanistic understanding of the toxicity of Ag-NPs.

Materials and Methods Test Materials. Silver nanoparticles (Ag-NPs; 7-10 nm, stabilized with polyethylenimine) were purchased from Kyoto Nano Chemical Co., Ltd. (Kyoto, Japan). Polystyrene nanoparticles (micromerR, PS-NPs; 15 nm) was used as a control material because polystyrene itself has no apparent toxicity to human cells, but the PS-PNs are nanoscale particles that have nanoparticle properties, which displays the toxicity. To evaluate the toxic effect of ionic silver (Ag+) that could be released from the Ag-NPs, silver carbonate (Ag2CO3) was used as a control material. The Ag-NPs and Ag2CO3 were purchased from COREFRONT Co., Ltd. (Tokyo, Japan) and Wako Pure Chemical Industries (Osaka, Japan), respectively. Cell Culture and Treatments. In this study, we used human hepatoma HepG2 cells, which retain normal cell functions and have been used in a number of toxicological studies. The cells obtained from the Riken Cell Bank (Tsukuba, 10.1021/es900754q CCC: $40.75

 2009 American Chemical Society

Published on Web 07/09/2009

Japan) were cultured in Eagle’s minimal essential medium (MEM) (Nissui, Tokyo, Japan) supplemented with 1% nonessential amino acid (Invitrogen, Carlsbad, CA), 10% fetal bovine serum, and 60 mg/mL kanamycin at 37 °C and 5% CO2. In addition, to evaluate the contribution of Ag+ to the toxicity of Ag-NPs, 5 mM N-acetyl-L-cysteine was used as a strong Ag+ ligand. Neutral Red (NR) Uptake Assay. Viability of HepG2 cells after exposure to each material was determined by neutral red (NR) uptake assay. This assay was performed as described by Borenfreund and Puerner (11) with slight modification. Briefly, the cells were seeded in 96-well cell culture plate at a density of ca. 5.5 × 104 cells per well and incubated overnight. Following exposure to three test materials, the medium was replaced with MEM containing 0.005% neutral red dye. The plates were then incubated in a 5% CO2 incubator for 3 h at 37 °C. After incubation, the dye-containing medium was discarded. After washing with PBS, extractant solution (50% ethanol and 1% acetic acid) was added to each well. The microplates were shaken for few minutes, and the absorbance of solutions was measured at 540 nm using a microplate reader. All absorbance values were corrected against blank wells which contained growth media alone. Care was taken to ensure that the neutral red-containing plates and solutions were completely protected from light throughout the experimental procedure. Each assay involved eight wells per condition. Micronucleus Test. Cells were grown to 70% confluency in 60 mm culture dishes and were exposed to 1 mg/L of Ag-NPs for 24 h. The exposure doses were chosen as maximum concentrations, at which significant cytotoxicity was not observed by the neutral red uptake assay mentioned above. In addition, the cells were exposed to 1 mg/L of PSNPs and 1.3 mg/L of Ag2CO3 (corresponding to 1 mg/L of Ag+) for comparison purposes. The cells exposed to the tested materials were trypsinized and incubated in a cold hypotonic solution (KCl 5.6 g/L) for 20 min and spread onto glass slides. After air-drying, the cells were fixed with methanol for 10 min and stained with 5% Giemsa. A total of 1000 binucleated cells were scored for the evaluation of the frequencies of micronucleus formation. Microarray Experiment. Cells were grown to 70% confluency in 60 mm culture dishes and were exposed to 1 mg/L of Ag-NPs, 1 mg/L of PS-NPs, and 1.3 mg/L of Ag2CO3 for 24 h. Following chemical exposure, the cells were washed with PBS and immediately subjected to RNA extraction. Three independent cultures were prepared for each treatment or control group. The cells were lysed directly on culture dishes, and total RNA was extracted using RNeasy Mini Kit (Qiagen, Hilden, Germany). Target preparation and hybridization were performed according to one-cycle eukaryotic target labeling assay protocols described in Affymetrix technical manual (Affymetrix, Santa Clara, CA). cDNA was synthesized from the total RNA by using One-Cycle cDNA Synthesis kit (Invitrogen, Carlsbad, CA) with a T7-(dT)24 primer incorporating a T7 RNA polymerase promoter. cRNA was synthesized from the cDNA and biotin-labeled by in vitro transcription using IVT Labeling kit (Affymetrix, Santa Clara, CA). Labeled cRNA was fragmented by incubation at 94 °C for 35 min in the presence of 40 mM Tris acetate, pH 8.1, 100 mM potassium acetate, and 30 mM magnesium acetate. Ten µg of fragmented cRNA was hybridized to a human genome focus array (Affymetrix, Santa Clara, CA) containing probes for 8795 human genes for 16 h at 45 °C. After hybridization, the microarrays were automatically washed and stained with streptavidin-phycoerythrin by using a fluidics station (Af-

FIGURE 1. Effects of three test material exposures on cell viability of HepG2 cells. Cells were exposed to different concentrations of Ag-nanoparticles, PS-nanoparticles and Ag2CO3 for 24 h. Cell viabilities were determined by the neutral red uptake assay. The data are expressed as mean (SD of eight wells of a cell culture plate. The concentrations of Ag2CO3 are expressed as the amounts of ionic silver (Ag+). fymetrix, Santa Clara, CA). Finally, probe arrays were scanned with the Genechip System confocal scanner (Affymetrix, Santa Clara, CA). Microarray Data Analysis. Expression data of 12 samples (four treatments, n ) 3) stored as “CEL file” in the Gene Chip Operating Software (GCOS) (Affymetrix, Santa Clara, CA) were transferred into the Avadis 4.3 prophetic (Strand Genomics, Redwood City, CA). Signal intensity of probes were scaled and normalized by MAS5 algorism. These summarized data have been deposited to the National Center for Biotechnology Information (NBCI) Gene Expression Omnibus (GEO; http:// www.ncbi.nlm.nih.gov/geo), and are accessible through GEO series accession number GSE14452. From the results of detection call analyses, the genes with “Present calls” in three replications were selected and used in the subsequent steps. To identify differentially expressed genes, the unpaired t-test for control and respective treatment groups (n ) 3) was performed for each gene. From the results of these analyses, the genes with p < 0.05 and g2.0 fold change in either direction were identified as being differentially expressed. The genes that were differentially expressed by the test material treatments were functionally categorized based on gene ontology categories at level 6 and KEGG biological pathways by using web based gene ontology program Fatigo (http://fatigo.bioinfo.cipf.es).

Results Cytotoxicity of Test Materials to HepG2 Cells. As the preliminary experiment, we assessed the cytotoxicity of three test materials (Ag-NPs, PS-NPs, and AgCO3) by measuring cell viabilities. The cell viabilities after 24 h exposure to the test materials were shown in Figure 1. Up to 0.5 mg/L of Ag-NPs, PS-NPs, and Ag2CO3, no significant cytotoxicity was observed, instead the cell viabilities increased up to around 120% relative to the nonexposed control. These results suggest that low dose of NPs accelerate cell proliferation. However, Ag-NPs exposure exhibited a significant cytotoxicity at higher doses (>1.0 mg/L), whereas PS-NPs and Ag2CO3 exposures did not produce a significant cytotoxicity. In this study, we chose 1.0 mg/L as an exposure dose for the following experiments, at which a significant cytotoxicity was not observed in the Ag-NPs exposure. VOL. 43, NO. 15, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6047

TABLE 1. Inductions of Micronuclei in HepG2 Cells Treated with Three Test Materialsa

frequencies of micrometeruclei (%)

control

Ag-nanoparticle

PS-nanoparticle

Ag2CO3

Ag-nanoparticle + Cysteine

2.1 ( 0.40

47.9 ( 3.2

2.5 ( 0.50

2.6 ( 0.36

29.3 ( 2.5

a

For each treatment, triplicate biological samples were prepared, and 1,000 cells were analysed for each biological sample. Values are means ( SD.

FIGURE 2. Venn diagrams illustrating shared gene expression in HepG2 cells under the exposures of three test materials (Ag-NPs, PS-NPs, and Ag2CO3). Supporting Information (SI) Figure S1 shows the general morphology of the HepG2 cells that were exposed to each test material at 1.0 mg/mL. There was no distinct change in cellular morphology after 24 h exposure to PS-NPs and Ag2CO3 as compared with the control (nonexposed) cells. However, Ag-NPs exposed cells became abnormal in shape, displaying the widened intercellular spaces (cellular shrinkage) and pseudopodic form (acquisition of an irregular shape) (SI Figure S1B and E). Micronucleus Test for the NPs Exposed Cells. The frequencies of micrometerucleus formation in Ag-NPs, PS-NPs, or Ag2CO3 exposed cells are shown in Table 1. In the nonexposed (control) cells, micronuclei were found in 2.1 ( 0.40% of binucleated cells. Ag-NPs exposure remarkably increased the frequency of micronucleus formation up to 47.9 ( 3.2% of binucleated cells, indicating DNA damage and chromosome aberrations, while formations of micrometeruclei in PS-NPs and Ag2CO3 exposed cells were not significant (2.5 ( 0.50% and 2.6 ( 0.36% of binucleated cells, respectively). Genes Altered by Ag-NPs Exposure. Ag-NPs exposure altered the expression levels of 529 (induction: 236 and repression: 293) genes (Figure 2). To assess the effect of AgNPs exposure on cellular functions, we classified these altered genes functionally based on gene ontology (GO) categories of “biological process” (level 6). Figure 3 shows the major biological process, which assign functional characteristics, and the percentage of classified genes to total altered genes. From the results of this analysis, 521 genes could be annotated, and 255 biological processes were found. An important finding was remarkable inductions of genes classified in “M phase” (31 genes), “microtuble-based process” (19 genes), “DNA repair” (16 genes), “DNA replication” (24 genes) and “intracellular transport” (32 genes). The individual genes classified in these biological processes are shown in SI Table S1. Most of the genes classified in the “M phase”, “microtuble-based process” and “intracellular transport”, were involved in chromosome segregation, cell division, and proliferation. Furthermore, the genes categorized as “DNA repair” and “DNA replication” were involved in DNA biosynthesis and restoration of DNA after DNA damage. In this study, inductions of some well-known stressinducible genes were observed. Three metallothionein genes (MT1H; 4.5 fold, MT1X; 3.4 fold, and MT2A; 4.1 fold) and three heat shock protein genes (HSPA4L; 2.2 fold, HSPB1; 2.1 6048

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 15, 2009

fold, and HSPH1; 2.1 fold) were significantly up-regulated in the cells exposed to Ag-NPs. Comparison of Ag-NPs-, PS-NPs-, and Ag2CO3-Altered Gene Profiles. In this study, we also performed DNA microarray analyses for the HepG2 cells exposed to PS-NPs and Ag2CO3 in the same manner to understand the cellular responses at molecular level. The exposure to PS-NPs and Ag2CO3 altered the expression levels of 189 (induction: 98 and repression: 91) and 304 (induction: 162 and repression: 142) genes, respectively. The overlaps of gene expressions among these two test materials and Ag-NPs are shown in Figure 2. In particular, 35 up-regulated genes and 27 downregulated genes were altered commonly in the same direction among all three chemical treatments. For further analyses, we compared the results of functional classification of altered genes among treatments with three test materials (Figure 3). In PS-NPs exposure, 188 genes were annotatable and 161 biological processes (level 6) were found. Of these, the up-regulations of the genes classified in “M phase” (14 genes), “microtuble-based process” (6 genes), “DNA replication” (4 genes) and “intracellular transport” (7 genes) were observed (SI Table S1). These classification patterns were similar to those in Ag-NPs exposure mentioned above. In particular, 13 of 14 up-regulated genes classified in “M-phase” in the PS-NPs exposure overlapped with those in the Ag-NPs exposure, while the genes classified in “DNA repair” were not significantly overlapped (induction: 2 and repression: 5). In Ag2CO3 exposure, 298 genes were annotatable and 210 biological processes (level 6) were found. As shown in Figure 3 and SI Table S1, classification patterns of the altered genes were similar to those of Ag-NPs exposure. The inductions of genes classified in “M phase” (28 genes), “microtuble-based process” (17 genes), “DNA repair” (17 genes), “DNA replication” (24 genes), and “intracellular transport” (14 genes) were observed. Furthermore, we focused on the stress inducible genes that were remarkably induced by Ag-NPs exposure. Except for HSPB1, the genes up-regulated by Ag-NPs exposure mentioned above (MT1H, MT1X, MT2A, HSPA4L, and HSPH1) were not up-regulated by PS-NPs and Ag2CO3 exposure. Effects of Cysteine on Toxicity of Ag-NPs. In this study, 5 mM N-acetyl-L-cysteine was used as an Ag+ ligand to assess the contribution of Ag+ to the toxicity of Ag-NPs. In the presence of cysteine, no significant cytotoxicity was observed even at high concentrations of Ag-NPs (Figure 1). However, the cellular morphological change mentioned above was observed even in the presence of cysteine (SI Figure S1E). Furthermore, the formation of micronuclei by Ag-NPs was partially counteracted by the addition of cysteine (formation frequency; 29.3 ( 2.5% of binucleated cells) (Table 1). The Ag-NPs exposure altered less number of total genes (induction: 213 and repression: 179) in the presence of cysteine. In these genes, only 165 (induction: 81 and repression: 84) genes were altered in the same direction in the absence of cysteine. The results of functional classification analysis of these genes were shown in Table S1 and Figure 3. The cysteine addition decreased the number of Ag-NPs induced specific genes that were classified into the “M phase”

FIGURE 3. Comparison of the gene ontology (GO) categories among the annotatable genes altered by Ag-nanoparticle (521 genes), PS-nanoparticle (188 genes), and Ag2CO3 (298 genes). GO terms of biological processes were represented at level 6 using the web based gene ontology program Fatigo (http://fatigo.bioinfo.cipf.es). Each bar describes the percentage of genes classified in each biological process to total altered genes. (22/31 genes: cysteine-treated/ nontreated cells) and “DNArepair” (11/16 genes: cysteine-treated/ nontreated cells) (SI Table S1). Furthermore, five stress inducible genes, which were significantly induced by Ag-NPs exposure, were downregulated (MT1H: -3.1 fold and MT1X: -3.1 fold) or did not exhibit significant expression level alteration (MT2A, HSPA4L, and HSPH1) in the presence of cysteine.

Discussion To date, some studies have evaluated the acute toxic effects of Ag-NPs and demonstrated that Ag-NPs were highly cytotoxic to mammalian cells based on the assessment on mitochondrial function, membrane leakage of lactate dehydrogenase (LDH), abnormal cell morphologies (6, 7, 9). However, biological effects at noncytotoxic doses of Ag-NPs such as carcinogenesis are still unknown. The aim of this study was to evaluate potential toxicity of Ag-NPs at noncytotoxic doses and the general mechanism involved in the toxicity of Ag-NPs. For this purpose, we performed the neutral red uptake assay, micronucleus test and DNA microarray analysis for the HepG2 cells under noncytotoxic (100% cell viability detected by the neutral red uptake assay) exposure conditions of Ag-NPs. Furthermore, we used two test materials, PS-NPs and Ag2CO3, to evaluate the contribution of “nanosized particle” and “ionic Ag+” to the toxicity of Ag-NPs. The cytotoxicity of three test materials was assessed at various doses. Ag-NPs exhibited a significant cytotoxicity at high dose exposures (>1.0 mg/L), whereas PS-NPs and Ag2CO3 had no measurable effects at the doses tested (Figure 1). In addition, abnormal cellular morphology was observed only in the Ag-NPs exposed cells (SI Figures S1B and E). Hussain et al. have shown that 5-50 mg/L of Ag-NPs exhibited a significant cytotoxicity in BRL 3A rat river cells. The cytotoxic doses determined in this study were on similar level with one reported in their report. Furthermore, it have been shown that Ag-NPs (5-10 mg/L, diameter: 15 nm) reduced mitochondrial function drastically and increased LDH leakage in the mammalian germline stem cells, whereas a significant cytotoxicity was not observed in Ag2CO3 exposed cells (7). Our experimental results agree with the result of this report. Intriguingly, noncytotoxic dose of Ag-NPs drastically increased cell viability. This may be a subsequence of hormesis, namely, stimulatory effects caused by low levels of potentially

toxic agents. Furthermore, Ag-NPs and Ag2CO3, which did not produce a significant toxicity, also induced cell proliferation. It is obvious that the test materials used in this study would have some stimulant effects on human cells. In general, information on the genotoxicity of metal nanoparticles is limited. Exceptionally, it has been reported that ultrafine titanium dioxide (TiO2) induced micronuclei in Syrian hamster embryo fibroblasts (12). Micronuclei can be expressed in dividing cells as a result of chromosome breaks. At telophase, these fragments that did not reach the spindle poles during mitosis, form a separate and smaller nucleus. Micronuclei represent therefore a measure of DNA and chromosome breakage. In this study, the only Ag-NPs significantly increased the frequency of micronucleus formation (Table 1), suggesting that Ag-NPs have a potential to cause damage to chromosome. In contrast, PS-NPs and Ag2CO3 have no significant effect. Thus, the chromosome aberrations as well as cytotoxicty of Ag-NPs are likely to be mediated through a combined effect of “nanosized particle” and “raw chemical composition of silver”. From the results of DNA microarray analysis, Ag-NPs induced larger number of genes than the other two test materials, indicating that As-NPs would affect various biological functions (Figure 2 and SI Table S1). In particular, only Ag-NPs induced well-known stress associated genes coding metallothionein (MT1H, MT1X, and MT2A) and heat shock protein (HSPA4L and HSPH1), which have been reported to be induced by cellular stresses such as heavy metal and various cytotoxic agent exposures (13, 14). From the results of functional classification of the altered genes, we highlighted the genes classified in biological process “M-phase”. These genes are associated with cell cycle progression through mitotic (M) phase, and most of these genes are included in the “cell cycle” pathway based on the KEGG pathway mapping. In particular, we observed the increases in expression levels of checkpoint related genes (BIRC5, BUB1B, CCNA2, CDC25B, CDC20, and CKS2) (15-19) in the Ag-NPs, PS-NP, or Ag2CO3 exposed cells (SI Table S1). Abnormal expression of these genes would cause dysregulated cellular proliferation and play a critical role in carcinogenesis and tumor progression. Furthermore, induction of these genes has been observed by exposing to nongenotoxic carcinogens such as 12-O-tetradecanoylphorbol-13acetate and tetrachloroethylene (20). In this study, the induction of these genes was found in all the cells exposed VOL. 43, NO. 15, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6049

to three test materials including PS-NPs and would reflect the abnormal cell proliferation as shown in the cell viability assay (Figure 1). This suggests that the HepG2 cells likely respond to the nanosized particles regardless of their raw materials. In addition, induction of these genes by Ag2CO3 demonstrated that ionic Ag+ also promotes the cell proliferation. Therefore, the abnormal cell proliferating action of Ag-NPs would be mediated by both the nanosized particle and ionic Ag+. The genes classified in “DNA repair” were induced by only Ag-NPs and Ag2CO3, but not by PS-NPs. These Aginduced specific genes were involved in various DNA repair pathways activated by DNA damage, and its induction would be closely related to carcinogenesis. The previous report demonstrated that Ag-NPs induce an expression of RAD51 protein involved in DNA damage repair (21). In this study, induction of RAD51C gene, a member of the RAD51 family, was observed (SI Table S1). The RAD51 proteins including RAD51C are thought to promote DNA strand exchange and be involved in recombinational repair of damaged DNA (22, 23). Abnormal expression of RAD51 proteins has been reported in various tumor cells (24, 25). It has been reported that Ag+ binds with nucleobase covalently and increases DNA damage (26). The induction of DNA repair-associated genes by Ag2CO3 might reflect this interaction between Ag+ and DNA. However, the micronucleus formation was not significant in the Ag2CO3 exposed cells (Table 1). Thus, Ag+ binds with DNA, but does not cause damages to chromosome. Ag-NPs appeared to cause much more damages to chromosome than Ag+. Based on these discussions, it is concluded that both “nanosized particle of Ag” as well as “ionic Ag+” contribute to the DNA damaging action of Ag-NPs. In this study, we used cysteine, a strong Ag+ ligand, to assess the contribution of Ag+ to toxicity of Ag-NPs. Navarro et al. (27) have reported that cysteine abolished the inhibitory effects of Ag-NPs on photosynthesis in algae, Chlamydomonas reinhardtii and concluded that Ag-NPs contributed to the toxicity as a source of ionic Ag+. In this study, the addition of cysteine effectively inhibited the cell death (Figure 1) and the induction of stress-associated genes caused by Ag-NPs. These results suggest that ionic Ag+ contribute mainly to cytotoxic and stress associated effects of Ag-NPs. In addition, remarkable induction of cell proliferation by addition of cysteine (Figure 1) would be a consequence of growth stimulatory effects of NPs, which was also observed in PSNPs exposure. This result was also supported by the induction of the genes classified in “M-phase” (Figure 3), which was only partially inhibited by cysteine. DNA damaging effect demonstrated by the micronucleus test was not completely counteracted (Table S1). Furthermore, the induction of the genes classified in “DNA repair” was only partially inhibited (Figure 3). These results suggest that the DNA damaging effect of Ag-NPs cannot be explained solely by the contribution of ionic Ag+ that is released from Ag-NPs. Thus, the nanosized particles of Ag alone have unique toxic effects to the cells. Metal ions including silver act as a catalyst and exhibit the ability to produce reactive oxygen species (ROS) in the presence of oxygen species, which is thought to be a mechanism of toxicity. The recent studies have indicated that Ag-NPs increased production of intracellular ROS (6). In addition, PS-NPs produced ROS in cell free medium (28). The ROS can act as signal molecules that promote cell cycle progression by affecting growth factor receptors, AP-1, NFkB, and so on (29-31) and induce the oxidative DNA damage. These mechanisms have been speculated to play important roles in carcinogenesis and tumor progressing actions of carcinogenic chemicals. Coincidentally, it has been reported that 2,3-dimethoxy-1,4-naphthoquinone (DMNQ), which has been known as a ROS generating chemical and frequently 6050

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 15, 2009

used as a model chemical for oxidative stress, induced the gene alteration patterns similar to one induced by Ag-NPs in this study (20, 32). Although little is known about relationship between carcinogenesis and ROS production by nanoparticles, it has been reported that nanoparticle carbon black induced DNA damage by ROS production, activating p53, proteins involved in DNA repair and regulation of cell growth and apotosis (33). Hence, it is speculated that the ROS production induces the genes associated with cell proliferation and DNA damage as shown in SI Table S1 and Figure 3. However, significant inductions of oxidative stress associated genes were not observed in this study, and a significant increase in intracellular ROS was not detected in the cells exposed to 1.0 mg/L of Ag-NPs by using a fluorescent probe 2′,7′-dichlorofluorescein diacetate (DCFH-DA) (34). (SI Figure S2). This is probably because the Ag-NPs concentration was too low to detect measurable ROS generation in the cells. It should be noted that such low dose of Ag-NPs caused a significant damage to chromosome (Table 1), reflecting a unique toxic effect of Ag-NPs. According to the criteria of the United States Environmental Protection Agency (EPA), silver is not classifiable as to human carcinogenicity (group D). Silver powder and colloidal silver do not induce cancer in animals, and silver chloride is considered nonmutagenic in rec-assay. Thus, silver compounds have been generally considered not to have carcinogenicity in humans and animals. No evidence of the carcinogenicity of Ag-NPs has so far been reported despite the growing commercialization of Ag-NPs. In this study, however, the up-regulation of a number of the genes associated with DNA repair and the increase in micronuclei in the Ag-NPs exposed cells at relatively low doses (