Proteomic Analysis of Cellular Response Induced by Multi-Walled Carbon Nanotubes Exposure in A549 Cells

The wide application of multi-walled carbon nanotubes (MWCNT) has raised serious concerns about their safety on human health and the environment. However, the potential harmful effects of MWCNT remain unclear and contradictory. To clarify the potentially toxic effects of MWCNT and to elucidate the associated underlying mechanisms, the effects of MWCNT on human lung adenocarcinoma A549 cells were examined at both the cellular and the protein level. Cytotoxicity and genotoxicity were examined, followed by a proteomic analysis (2-DE coupled with LC-MS/MS) of the cellular response to MWCNT. Our results demonstrate that MWCNT induces cytotoxicity in A549 cells only at relatively high concentrations and longer exposure time. Within a relatively low dosage range (30 µg/ml) and short time period (24 h), MWCNT treatment does not induce significant cytotoxicity, cell cycle changes, apoptosis, or DNA damage. However, at these low doses and times, MWCNT treatment causes significant changes in protein expression. A total of 106 proteins show altered expression at various time points and dosages, and of these, 52 proteins were further identified by MS. Identified proteins are involved in several cellular processes including proliferation, stress, and cellular skeleton organization. In particular, MWCNT treatment causes increases in actin expression. This increase has the potential to contribute to increased migration capacity and may be mediated by reactive oxygen species (ROS).


Introduction
Nanomaterials, with sizes ranging from 1 to 100 nm in one or more dimensions, are the core of an emerging technological revolution [1,2]. Multi-walled carbon nanotubes (MWCNT), discovered by Iijima in 1991, is one of the most common nanomaterials in use [3]. Structurally, MWCNT consists of several concentric graphene sheets which can be produced in laboratories, and can even be found in the particulate matter from ordinary combustion of fuel gases [4,5]. Due to its outstanding physicochemical and mechanical properties such as high tensile strength, ultra-light weight, thermal and chemical stability, as well as excellent semi-conductive electronic properties, MWCNT has been a highly desirable material in various sectors including electronics, aerospace, chemicals, construction and pharmaceuticals [6]. MWCNT has also being developed for a range of biomedical applications such as miniaturized biosensors, or for targeted drug delivery and tissue engineering [7,8]. However, the wide application of MWCNT has raised serious concerns about their possible impact on safety for human health and the environment. Human may be exposed to MWCNT through inhalation, ingestion, or skin uptake, and when MWCNT interacts with biological systems, adverse biological effects might be generated.
Many studies have been conducted over the past several years to evaluate the toxicological effects of MWCNT. However, existing data are frequently contradictory. For example, MWCNT was able to induce the time-and dose-dependent cytotoxicity in several cell lines, leading to the release of proinflammatory cytokines, cell cycle arrest and apoptosis [9][10][11][12][13][14]. Due to its fibrous-or microtubelike structure, genotoxic damage such as chromosomal aberrations, DNA strand breakages and micronuclei were also found in cells after MWCNT treatment [9,[15][16][17][18]. Several studies have suggested that reactive oxygen species (ROS) may be responsible for the cytotoxicity and genotoxicity of MWCNT [19][20][21], and signaling pathways such as NF-kB, AP-1, p38/ERK-MAPK cascades have also been implicated [14,22]. However, some other reports reveal no cytotoxicity following MWCNT treatment [20,23]. For instance, MWCNT demonstrates no sign of acute toxicity on cell viability and could not induce any inflammatory mediators, such as NO, TNF-alpha and IL-8 in either rat macrophages (NR8383) or in human lung adenocarcinoma A549 cells [23]. Although the exact reason for the different biological effects of MWCNT is still unknown, it is believed that the size and shape of the nanomaterials, the presence of trace amounts of metals, and the cell types examined may all contribute to the observed differences [23][24][25].
Similar contradictory results were also reported in in vivo studies. For example, it has been shown that MWCNT induces pulmonary inflammation, granuloma, fibrosis, and mesothelioma in experimental models [26][27][28][29][30]. MWCNT could also alter the systemic immune function in mice [31][32][33][34][35], and individuals with preexisting allergic inflammation may be susceptible to airway fibrosis from inhaled MWCNT [36]. Furthermore, recent observations suggest that the nervous system is vulnerable to MWCNT as well [37][38][39]. On the other hand, there are also reports showing no inflammation or cancer occurrence after MWCNT exposure in rats [40,41].
As noted above, the potential large-scale exposure of humans to the biological effects of MWCNT requires a much better understanding of the risks and mechanisms involved, as well as a clarification of the origin of these contradictory results. To this end, the cytotoxic and genotoxic effects of MWCNT on A549 cells were examined, followed by application of a proteomics-based approach. By screening and identifying the differentially expressed proteins, we further investigated the possible roles of such proteins in MWCNT-induced toxicity. As reported here, MWCNT induces cytotoxicity in A549 cells only at relatively high concentrations and longer exposure time. Within a relatively low dosage range (30 mg/ml) and short time period (24 h), MWCNT treatment of A549 cells does not induce significant cytotoxicity, cell cycle arrest, apoptosis, or DNA damage. However, at these low doses and times, proteomic analysis reveals that MWCNT causes significant protein expression changes, and that differentially expressed proteins are involved in cellular processes such as proliferation, metabolism, and organization of the cellular skeleton.

MWCNT preparation
MWCNT was provided by Dr. F. Chen (Lawrence Berkeley National Laboratory, Berkeley, CA), and it was synthesized with a chemical vapor deposition (CVD) method [14]. The detailed physicochemical characterizations of MWCNT were described in our previous study [12]. The sterile raw material was suspended in 1640 culture medium (Gibco, Grand Island, NY, USA) containing 10% fetal calf serum and then the suspension was sonicated at 180 W for 30 cycles, with 10 s ultrasonication and 5 s pause using an ultrasonic disrupter (JY92-IIN, Scientz, Ningbo, China). The suspensions were always prepared fresh prior to use.

Cell culture and Cytotoxicity analysis
Human lung adenocarcinoma A549 cells, obtained from the ATCC (CCL-185), were routinely subcultured in 1640 culture medium containing 10% newborn calf serum, 100 U/ml penicillin, 125 mg/ml streptomycin and 0.03% glutamine at 37uC in 5% humidified CO 2 . MWCNT was added to cell media at different concentrations (0, 0.3, 3, 30 and 300 mg/ml) for various times (0, 2, 12 and 24 h). The effects of MWCNT on cell viability were examined by trypan blue exclusion as described previously [42]. Briefly, A549 cells were cultured in 24-well plates and subjected to various treatments. Then the cells were harvested and mixed with an equal volume of 0.4% (w/v) trypan blue solution prepared in PBS. The number of trypan blue-excluding cells was determined using a hemocytometer, and cell viability was calculated as the  Flow cytometry analysis of apoptosis and the cell cycle A549 cells were collected after MWCNT treatment, and washed twice with PBS. For apoptosis detection, the cells were suspended in 400 ml binding buffer and incubated with 5 ml Annexin V-FITC (MultiSciences, Hangzhou, China) and 5 ml PI in the dark at 37uC for 15 min. Cells were put on ice until analysis. The ratio of apoptotic cells was measured using a Beckman Coulter Epics XL-MCL device (Fullerton, CA, USA).
For cell cycle detection, cells were fixed in 70% ethanol for at least 2 h. After fixation, cells were centrifuged at 200 g for 5 min, and the cell pellet suspended in 500 ml propidium iodide (PI)/ Triton X-100 staining solution containing RNase A for 15 min at 37uC. Cell cycle was assessed using a Beckman Coulter Epics XL-MCL device.
Immunofluorescence microscopy to detect genotoxicity of MWCNT Immunofluorescence microscopy to observe the formation of cH2AX foci and distribution of actin was conducted essentially as described previously with slight modification [42]. Briefly, 1-2610 5 cells were seeded into glass-bottom 6-well plates. After treatment, cells were fixed in 4% paraformaldehyde for 15 min, washed with PBS once, and permeabilized in 0.2% Triton X-100 for 5 min. After blocking for 1 h, samples were incubated with a mouse monoclonal anti-cH2AX antibody (1:3000) (Upstate

Comet assay
The neutral comet assay was conducted as described previously [42]. First, the fully frosted microscope slides were covered with 100 ml of 0.65% normal melting point agarose and immediately covered with a coverslip. Slides were placed on ice for 8 min to allow the agarose to solidify. Next, the coverslips were removed, and the first agarose layer was covered with the cell suspension (1610 6 cells in 15 ml PBS were mixed with 75 ml of 0.65% low melting point agarose). After replacing the coverslips, the slide was allowed to solidify on ice for 8 min. Another layer of agarose (75 ml of 0.65% low melting point agarose) was then added as described above. Finally, the coverslips were removed, and the slides were immersed in the lysis buffer (2 M NaCl, 30 mM EDTA, 10 mM Tris, with 1% Triton X-100 and 10% DMSO added just before use, pH 8.2-8.5) for 2 h at 4uC. The slides were removed from the lysis buffer, washed for 10 min in 0.56TBE, and transferred to an electrophoresis chamber. After equilibration in 0.56TBE for 20 min, electrophoresis was conducted at 25 V for 20 min. The slides were then washed in a neutralization buffer (0.4 M Tris, pH 7.5) 3 times for 5 min. The slides were drained and stained with gel-red, and observed with a fluorescent microscope. Single cell images were captured and analyzed using an Olympus AX70 (Olympus, Japan) immunofluorescent microscope, and tail moment was scored using CometScore software (TriTek Corporation, Southern California, USA).

2-Dimentional electrophoresis (2-DE)
2-DE was conducted as previously described [44]. Briefly, 200 mg total cellular proteins were loaded onto 24 cm, pH 3-10 linear immobilized pH gradient strips (Bio-Rad, CA, USA) for isoelectric focusing (IEF). After 12 h of rehydration, the strips were transferred to the IEF cell. The parameters were set as follows: 250 V for 30 min, step; 500 V for 1 h, step; 1000 V for 3 h, gradient; 10,000 V for 6 h, gradient; 10,000 V until 80,000 Vhr, step. After IEF was completed, the strips were equilibrated and the second dimension was performed by vertical 12% SDS-PAGE. Gels were stained using silver staining and scanned with a Bio-Rad GS-800 scanner. Images were analyzed by PDQuest software Version 7.4.0 (Bio-Rad, Hercules, USA). A statistical analysis was performed using the Student's t-test. Proteins with significant differences (1.5 fold change, P,0.05) were selected for MS identification.

LC-MS/MS identification
Differentially expressed protein spots were manually cut from the silver-stained gels and were destained with a solution of 30 mM potassium ferricyanide and 100 mM sodium thiosulfate,

Measurement of intracellular reactive oxygen species (ROS)
The ROS production was first observed by Immunofluorescence microscopy using MitoSOX TM Red mitochondrial superoxide indicator (lot number: M36008, Invitrogen, USA) as described before [45]. Briefly, 1-2610 5 cells were seeded into glass-bottom 6-well plates. After treatment, cells were incubated with MitoSOX TM reagent working solution (5 mM) for 15 min and Hoechst33258 for 15 min in the dark. The glass-bottom 6-   well plates were observed using a LSM710 Laser Scanning Confocal Microscope (Zeiss, Jena, Germany). The quantitative analysis of ROS was also measured using 2,7dichlorofluorescin diacetate (DCFH-DA) as described previously [42]. Briefly, 10 mM DCFH-DA stock solution (in methanol) was diluted 500-fold in PBS to yield a 20 mM working solution. After MWCNT treatment, cells in each 96-well plate were washed twice with PBS and then incubated in 100 ml working solution of DCFH-DA at 37uC for 30 min. Fluorescence was determined at 485 nm excitation and 520 nm emission wavelength using an Infinite M200 microplate reader (Tecan, USA). To determine the role of ROS in the actin expression and cell migration, cells were incubated with a ROS scavenger, N-acetylcysteine (NAC) (Sigma, USA) for 2 h at 10 mM followed by MWCNT treatment. ROS level was represented as the absorbance of treated group/ absorbance of control group.
Cell scratch assay A549 cells plated in 35-mm dishes were exposed to MWCNT at 30 mg/ml for 24 h. When cells grew to confluence, the cell monolayer was scratched to form a 100-mm ''wound'' using sterile pipette tips and washed gently once with PBS. A549 cells were then incubated with normal medium for another 24 h. The wound was photographed at 0 and 24 h using a TS100F microscope (Nikon, Tokyo, Japan). The cell migration distance was measured by Image J software (National Institute of Mental Health, Bethesda, USA). Data are presented as mean6standard deviation (SD) of three independent experiments. (A) Cells were exposed to 0.3, 3 and 30 mg/ml MWCNT for 2, 12 and 24 h, and ROS generation was measured using MitoSOX TM Red mitochondrial superoxide indicator by Immunofluorescence microscopy. (B) Cells were exposed to the indicated concentrations of MWCNT for various times, and ROS generation was measured using 29,79-dichlorofluorescin diacetate (DCFH-DA) using a spectrofluorometer. ROS generation was expressed as fold increase of fluorescence compared to the control; * p,0.05, compared to PBS treated control. (C) Cells were pretreated with NAC (10 mM) for 2 h, and then exposed to 30 mg/ml of MWNCT for 24 h. * p,0.05, compared with the control. # p,0.05, compared to MWCNT treatment without NAC pre-incubation. doi:10.1371/journal.pone.0084974.g007

Cytokine assay
A commercial BD TM Cytometric Bead Array (CBA) Human Th1/Th2 Cytokine Kit II (lot number: 551809, BD Biosciences Pharmingen, San Diego, USA) was used to determine the levels (pg/ml) of Th1 and Th2 cytokines, including Interferon-gamma (IFN-c), tumor necrosis factor (TNF), interleukin (IL)-10, IL-6, IL-4 and IL-2 in A549 supernatant samples following the manufacturer's instructions, after treatment of MWCNT at 0.3, 3, 30 mg/ ml for 2, 12 and 24 h. A standard calibration curve was established for each kit. The maximum and minimum limits of detection for all six cytokines were 1.0 and 5000 pg/ml, respectively. Fluorescence was analyzed using a flow cytometer (FACS Calibur, Becton-Dickinson Biosciences, Heidelberg, Germany) and cytokine level was determined using a BD CBA Software. Statistical analysis was performed using a Kruskal-Wallis test (nonparametric analogue of one-way ANOVA) for the data that did not meet the assumptions of parametric tests. The differences were considered significant at p,0.05.

Statistics analysis
All experiments were conducted at least three times. Statistical analysis was performed using one-way ANOVA and Student's ttest. Numerical values are represented by mean6SD. A statistical probability of p,0.05 was considered significant.

MWCNT causes cytotoxicity following treatment for high doses and times
A549 cells were treated with various concentrations (0.3, 3, 30 and 300 mg/ml) of MWCNT for the indicated times (2, 12 and 24 h), and cytotoxicity was evaluated using the trypan blue assay. As shown in Fig. 1, as compared with untreated cells, there is no significant cytotoxic effect observed for MWCNT at 0.3, 3 and 30 mg/ml throughout the 24 h period. On the other hand, at the highest concentration (300 mg/ml), the cell proliferation is significantly suppressed after treatment with MWCNT at 12 and 24 h (Fig. 1).

MWCNT causes apoptosis and cell cycle perturbation
To determine whether apoptosis and/or cell cycle arrest had occurred, cells were examined by flow cytometry following MWCNT treatment. The results (Figure 2A) demonstrate that MWCNT fails to induce apoptosis at doses of 0.3, 3 and 30 mg/ml at 24 and 48 h. However, when treated with 300 mg/ml MWCNT for 24 h, the percentage of early apoptotic cells increases from 3% to 7% (Fig. 2A), and with prolonged exposure to 48 h, there is an approximately 2-fold increase in the percentage of both early apoptotic and apoptotic cells ( Fig. 2A).
Cell cycle perturbation was also examined for cells exposed to MWCNT at 30 mg/ml for various time periods (6, 8, 24, 48 and 72 h). At shorter time periods (6, 8 and 24 h), the cell cycle distribution is almost the same for both control and treated cells (Fig. 2B). In contrast, after exposure for 48 h, the percentage of cells in G1/G0 increases significantly from 66% to 86%, while percentage of cells decreases from 9% to 4% for G2/M and 25% to 11% for S phase, respectively. Cells exposed to MWCNT for 72 h have a similar distribution throughout the cell cycle (Fig. 2B).

MWCNT fails to demonstrate genotoxicity in A549 cells
To determine whether MWCNT is genotoxic to A549 cells, the effect of MWCNT on cH2AX foci formation, a sensitive indicator for DNA damage [46], was examined. Figure 3A shows representative immunofluorescent images of cells, and the results demonstrate that MWCNT treatment at 0.3, 3 and 30 mg/ml for 2, 12 or 24 h does not induce significant cH2AX foci formation.
To verify further that MWCNT indeed could not induce DNA damage, as implied by the results of our cH2AX foci formation assay, MWCNT treated cells were further subjected to the neutral comet assay. Shown in Fig. 3B are representative images from this comet assay for cells exposed to MWCNT at 0.3, 3 and 30 mg/ml for 24 h. No significant changes in tail moment or in the percentage of tail DNA, i.e., DNA damage, are observed (Table 1).

2-DE analysis and MS identification of cellular proteins in A549 cells exposed to MWCNT
Based on the results described above, A549 cells were treated with 0.3, 3 and 30 mg/ml MWCNT for 2, 12 or 24 h, and total cellular proteins were extracted and separated by 2-DE (Fig. 4A- C). After silver staining, the images of each treatment group and control were compared using PDQuest 7.4.0. In total, 106 spots show changed expression after MWCNT treatment. Of these, 62 spots were selected for LC-MS/MS identification, of which 10 spots could not be identified due to missing MS signals or low abundance. Of the 52 successfully identified spots, 29 are upregulated and 23 are down-regulated. 50% of these proteins are located in the cytoplasm, 24.19% are located in the nucleus, and 6.45% are found on the cellular membrane, while others are mitochondrial or endoplasmic reticulum proteins (Fig. 4D). The identified proteins are involved in cellular processes such as cell cycle regulation and apoptosis, cell metabolism, and regulation of cellular skeleton, etc. (Fig. 4E and Table 2).

Validation of identified proteins
To verify the proteomic results, Western blot was used to confirm the expression changes of three identified proteins, namely, 14-3-3e, HSP27 and actin. The results support the proteomic results, showing that MWCNT treatment induces a significant decrease in 14-3-3e (Fig. 5, left panel) and HSP27 (Fig. 5, right panel) at the indicated concentrations. On the other hand, actin is up-regulated after 24 h of treatment (Fig. 6A-C). Each of these results is consistent with the proteomic analysis.
To further confirm these results, confocal microscopy imaging analysis was applied to investigate the effects of MWCNT on actin expression in A549 cells. Cells were exposed to 0.3, 3 and 30 mg/ ml MWCNT for the indicated times and the expression level of actin filaments was determined by phalloidin staining. As shown in Fig. 6D-E, the fluorescent intensity of actin is significantly increased after exposure to MWCNT, consistent with our Western blot results.

ROS modulates the changes of protein expression induced by MWCNT
It has been reported that many of the bioeffects of MWCNT can be mediated through the changes of cellular oxidative status [14,19,20]. To determine whether MWCNT could affect ROS production in A549 cells, MitoSOX, a novel fluoroprobe, was introduced for selective detection of superoxide in the mitochondria of live cells. Shown in Fig. 7A are the representative images of cells after MWCNT treatment, which indicated that mitochondrial ROS was increased in a time-and dose-dependent manner. The intracellular ROS levels were also quantitatively measured by DCFH-DA staining. The results show that the ROS levels are significantly increased after exposure to MWCNT at 12 and 24 h (Fig. 7B), which is consistent with the confocal microscopy results.
Next, to determine whether ROS is involved in MWCNTinduced increases in actin expression, A549 cells were pretreated with an antioxidant NAC (10 mM, 2 h) before exposure to MWCNT. As anticipated, pre-treatment with NAC dramatically decreases the level of ROS (Fig. 7C). Consistent with the hypothesis that MWCNT-mediated increases in actin expression are mediated through ROS, we find that actin expression is significantly decreased after NAC-pretreatment as compared to MWCNT exposure only (Fig. 8). MWNCT promotes cell migration As described above, MWCNT exposure induces ROS generation and actin expression, and both are known to be directly related to the ability of cells to migrate [21,47]. Therefore, the effects of MWCNT on cell migration were measured by cell scratch analysis. It is found that cell migration ability is indeed increased by MWCNT exposure (Fig. 9A-B). Moreover, the pretreatment by NAC, which inhibits the production of ROS, also leads to decreased cell migration (Fig. 9C), indicating the involvement of ROS during this process.
The effects of MWCNT on Th1/Th2 cytokine production A549 cells were treated with various concentrations (0.3, 3 and 30 mg/ml) of MWCNT for the indicated times (2, 12 and 24 h), and the secretion of Th1/Th2 cytokines were meausred in the supernatant. TNF is not detected, probably due to the detection limit of the kit. For IFN-c, IL-2, IL-10, and IL-4 secretion, as shown in Fig. 10A, no significant changes are found after MWCNT treatment. In contrast, IL-6 secretion is affected by MWCNT treatment, which is increased in a time-and dosedependent manner (Fig. 10B).

Discussion
The increasing use of MWCNT in consumer products and medical applications underlines the importance of understanding its potential toxic effects to human and the environment. However, the toxicity and related molecular mechanisms of MWCNT have not been fully elucidated, with many contradictive results reported over the years. Thus, in the present study, we examined the effects of MWCNT on A549 cells. In particular, we employed a proteomic approach in an effort to provide a systematic view of the cellular response to MWCNT exposure.
We first find that MWCNT causes cytotoxic injuries to A549 cells only at relatively high concentrations and longer exposure time, which appears to be relatively more resistant to MWCNT exposure as compared to other cell types. In our study, after exposure to 30 mg/ml MWCNT for 24 h, no significant changes in cell cycle, cell apoptosis, or DNA damage are detected. This is consistent with previous reports showing that no significant suppression of A549 cell proliferation could be detected after exposure to 5-100 mg/ml MWCNT for 24 h [21,24]. On the other hand, in a previous study we had shown that exposure of the same MWCNT to human umbilical vein endothelial cell (HUVEC) significantly increases the percentage of apoptotic cells and cH2AX-positive cells in a dose-dependent manner (0.5-20 mg/ ml) [12]. Human skin fibroblast cells (HSF) are even more sensitive, as exposure to the exact same MWCNT at a dose of only 0.06 mg/ml for 48 h could result in 50% reduction in proliferation, a 2-fold increase in apoptosis/necrosis and G2/M block [14]. Still, Ursini et al. reports that significant cell death and DNA damage occur after 20 mg/ml MWCNT treatment of A549 cells for 24 h  [48]. These contradictory observations could be the results of different MWCNT materials, cell lines, detection methods and protocols in individual laboratories. Thus, a more comprehensive and standardized evaluation of the cytotoxic effects of MWCNT is warranted in order to obtain a clearer answer.
High-throughput and systems-based technologies, such as proteomics, can reveal complex interactions in biological systems and thereby provide new leads for mechanistic study [49]. Therefore, we examined the cellular response to MWCNT at dosages of 0.3, 3 and 30 mg/ml for 2, 12 and 24 h, for which no significant cytotoxicity, apoptosis, cell cycle perturbation, and DNA damage are induced. Eventually, 106 proteins with altered expression are detected, of which 52 are successfully identified. These 52 proteins are involved in various biological processes, including apoptosis, cell cycle arrest, cell metabolization and cellular skeleton regulation. Some of the identified proteins, such as HSP27, Annexin A2, 14-3-3e and proteasome components, match those reported by Haniu, et al, who analyzed the proteome response by exposing U937 cells to MWCNTs at non-cytotoxic dosages [50]. Similar functional perturbation of cellular processes like apoptosis, stress, metabolism, and cytoskeleton are also reported following MWCNT treatment of RAW264.7 macrophage cells and human epidermal keratinocytes (HEKs) [51,52], and in human hepatoma HepG2 cells treated with single-walled carbon nanotubes (SWCNT) [53].
Among the identified proteins, 21.15% are cellular skeleton proteins such as actin, confilin and profilin. Actin is the major component of microfilaments, and its structural modulation is the basis of molecular adhesion, cellular communication, cell permeability changes and cell movement [54][55][56]. Alterations in actin can be triggered by many external events, including HIV infection [57]. In one previous study, exposure to 1.5-4.5 mg/ml of MWCNT causes actin filament disruption and reduced tubule formation in human aortic endothelial cells [58]. In another study, MWCNT induces actin filament remodeling to form peripheral motile structures, lamellipodia and filopodia, and central actin filament bundles in human microvascular endothelial cells (HMVEC), and the cells are pulled apart to form small gaps in the HMVEC monolayer [59]. Mechanistically, reports have revealed that the elevation of ROS levels might be responsible for changes in the actin structure [59][60][61][62]. In our study, MWCNT causes significant increases in actin expression under all treatments, along with elevated ROS levels. Following pre-treatment with NAC for 2 h, ROS levels are decreased, followed by a significant decrease of actin expression in A549 cells. These results indicate that ROS is likely involved in MWCNT-induced actin alterations, consistent with previous reports.
Cell migration is an important process under several physiological conditions like development and wound healing, and also under the pathological conditions such as cancer cell invasion and metastasis [63]. Migration involves a wide array of cellular changes including alterations in cellular structure by regulation of cytoskeleton dynamics and expression of adhesion molecules [64]. Therefore, changes in actin structure have the ability to directly affect cell migration [47,65,66]. ROS are also known to actively participate in each of the above events [67][68][69]. Since MWCNT significantly induces the generation of ROS and actin expression, it is of interest to evaluate the effects of MWCNT on cell migration. The cell scratch assay results clearly show that MWCNT exposure can increase cell migration, while pretreatment of NAC abolishes this effect, which are consistent with other studies regarding the effects of MWCNT on cell migration [59,70].
It has been reported that MWCNT could trigger inflammatory response in mice or cell lines [31,33], and the increased ROS production is at least partially responsible for it [71,72], although contradictive results also exist [23,41]. In our study, several proteins involved in inflammation and infection processes, such as Annexin A1, Complement component 1 Q subcomponentbinding protein (C1qBP), Poly(rC)-binding protein 2 (hnRNP E2) are indentified (Table 2). Additionally, we also find an elevated ROS production. Therefore, it is of interest to know whether MWCNT could influence the inflammatory response in A549 cells. Among the Th1/Th2 cytokines examined, only IL-6 secretion is affected by MWCNT treatment. Thus, the IL-6mediated inflammatory response might be a major target of MWCNT.
Several of the other identified proteins are also of interest for further study. 14-3-3e is the most conserved member of the 14-3-3 family, and is involved in a wide range of physiological processes. For example, homocysteine induces apoptosis of rat hippocampal neurons by inhibiting 14-3-3e expression [73], and downregulation of 14-3-3e causes partial meiotic resumption of the mouse oocyte [74]. Moreover, a significant reduction of 14-3-3e protein expression is detected in gastric cancer [75]. Given the prevalence of reduced 14-3-3e expression in such processes, the importance of reduced 14-3-3e expression after MWCNT exposure merits further investigation. Another interesting protein is Hsp27, a member of the small heat shock protein (HSP) family, which protects against apoptotic cell death induced by a variety of stimuli including elevated temperature, heavy metals, oxidative stress and cytotoxic agents [76,77]. Our initial prediction is that under stress conditions, such as MWCNT exposure, HSP expression would have been induced for its protective function, however, we find instead that HSP27 expression is decreased. The physiological significance of such changes requires further clarification.

Conclusions
In the present study, we demonstrate that MWCNT induces cytotoxicity in A549 cells only at relatively high concentrations and longer exposure time. Within a relatively low dosage range (30 mg/ml) and short time period (24 h), MWCNT treatment of A549 cells does not induce significant cytotoxicity, cell cycle arrest, cell apoptosis, or DNA damage. However, under the same treatment condition, MWCNT causes significant changes in protein expression. The differentially expressed proteins could provide new leads for deciphering the cellular response to MWCNT. As one example, MWCNT treatment causes increased actin expression, likely mediated by ROS and leading to increased migration capacity.