Physicochemical and biological characterization of single-walled and double-walled carbon nanotubes in biological media

Journal of Hazardous Materials 280 (2014) 216–225

Contents lists available at ScienceDirect

Journal of Hazardous Materials journal homepage: www.elsevier.com/locate/jhazmat

Physicochemical and biological characterization of single-walled and double-walled carbon nanotubes in biological media Wen-Te Liu a,b,1 , Mauo-Ying Bien a,c,1 , Kai-Jen Chuang d,e , Ta-Yuan Chang f , Tim Jones g , Kelly BéruBé h , Georgi Lalev i , Dai-Hua Tsai j , Hsiao-Chi Chuang a,b,∗∗ , Tsun-Jen Cheng k,l,∗ , on behalf of the Taiwan CardioPulmonary Research (T-CPR) Group a

School of Respiratory Therapy, College of Medicine, Taipei Medical University, Taipei, Taiwan Division of Pulmonary Medicine, Department of Internal Medicine, Shuang Ho Hospital, Taipei Medical University, Taipei, Taiwan c Division of Pulmonary Medicine, Department of Internal Medicine, Taipei Medical University Hospital, Taipei, Taiwan d Department of Public Health, School of Medicine, College of Medicine, Taipei Medical University, Taipei, Taiwan e School of Public Health, College of Public Health and Nutrition, Taipei Medical University, Taipei, Taiwan f Department of Occupational Safety and Health, College of Public Health, China Medical University, Taichung, Taiwan g School of Earth and Ocean Sciences, Cardiff University, Cardiff, Wales, UK h School of Biosciences, Cardiff University, Cardiff, Wales, UK i School of Chemistry, Cardiff University, Cardiff, Wales, UK j Institute of Social and Preventive Medicine, Lausanne University Hospital, Lausanne, Switzerland k Institute of Occupational Medicine and Industrial Hygiene, College of Public Health, National Taiwan University, Taipei, Taiwan l School of Public Health, College of Public Health, National Taiwan University, Taipei, Taiwan b

h i g h l i g h t s • • • • •

Investigation of current dispersal method for nanotoxicology. BSA and BALF were used to investigate protein and CNT interaction. Alteration in physicochemistry of CNT was induced by BSA. Protein binding to CNT could result in misinterpretation of in vitro results. Protein-to-CNT interactions were associated with the coagulation pathways.

a r t i c l e

i n f o

Article history: Received 3 March 2014 Received in revised form 14 July 2014 Accepted 23 July 2014 Available online 10 August 2014 Keywords: Carbon nanotube Inflammation Nanomaterials Nanotoxicology Proteomics

a b s t r a c t To study the toxicity of nanoparticles under relevant conditions, it is important to reproducibly disperse nanoparticles in biological media in in vitro and in vivo studies. Here, single-walled nanotubes (SWNTs) and double-walled nanotubes (DWNTs) were physicochemically and biologically characterized when dispersed in phosphate-buffered saline (PBS) and bovine serum albumin (BSA). BSA-SWNT/DWNT interaction resulted in a reduction of aggregation and an increase in particle stabilization. Based on the protein sequence coverage and protein binding results, DWNTs exhibited higher protein binding than SWNTs. SWNT and DWNT suspensions in the presence of BSA increased interleukin-6 (IL-6) levels and reduced tumor necrosis factor-alpha (TNF-␣) levels in A549 cells as compared to corresponding samples in the absence of BSA. We next determined the effects of SWNTs and DWNTs on pulmonary protein modification using bronchoalveolar lavage fluid (BALF) as a surrogate collected form BALB/c mice. The BALF proteins bound to SWNTs (13 proteins) and DWNTs (11 proteins), suggesting that these proteins were

Abbreviations: BSA, bovine serum albumin; BALF, bronchoalveolar lavage fluid; DWNTs, double-walled nanotubes; DLS, dynamic light scattering; EDX, energy-dispersive X-ray; HRTEM, high-resolution transmission electron microscopy; IFN-␥, interferon-gamma; IL-6, interleukin-6; LC–MS, liquid chromatography–mass spectrometry; NTs, nanotubes; PANTHER, Protein ANalysis THrough Evolutionary Relationships; PBS, phosphate-buffered saline; SEM, scanning electron microscope; SWNTs, single-walled nanotubes; SDS-PAGE, sodium dodecyl sulphate-polyacrylamide gel electrophoresis; TFA, trifluoroacetic acid; TNF-␣, tumor necrosis factor-alpha. ∗ Corresponding author at: Institute of Occupational Medicine and Industrial Hygiene, College of Public Health, National Taiwan University, Taipei, Taiwan. Tel.: +886 2 33668090; fax: +886 2 23957845. ∗∗ Corresponding author at: School of Respiratory Therapy, College of Medicine, Taipei Medical University, Taipei, Taiwan. Tel.: +886 2 27361661x3515; fax: +886 2 27391143. E-mail addresses: [email protected] (W.-T. Liu), [email protected] (M.-Y. Bien), [email protected] (K.-J. Chuang), [email protected] (T.-Y. Chang), [email protected] (T. Jones), [email protected] (K. BéruBé), [email protected] (G. Lalev), [email protected] (D.-H. Tsai), [email protected] (H.-C. Chuang), [email protected] (T.-J. Cheng). 1 These authors contributed equally to this study. http://dx.doi.org/10.1016/j.jhazmat.2014.07.069 0304-3894/Crown Copyright © 2014 Published by Elsevier B.V. All rights reserved.

W.-T. Liu et al. / Journal of Hazardous Materials 280 (2014) 216–225

217

associated with blood coagulation pathways. Lastly, we demonstrated the importance of physicochemical and biological alterations of SWNTs and DWNTs when dispersed in biological media, since protein binding may result in the misinterpretation of in vitro results and the activation of protein-regulated biological responses. Crown Copyright © 2014 Published by Elsevier B.V. All rights reserved.

1. Introduction Single-walled nanotubes (SWNTs) and double-walled nanotubes (DWNTs) are hollow tubes comprising a single wall or two walls, respectively, of crystalline graphene. The possible biomedical applications of NTs have attracted a great deal of interest as novel drug delivery systems [1], imaging contrast agents [2] and detection devices for tumor cells in the blood [3]. Increasing evidence suggests that NTs induce oxidative stress [4], inflammation [5] and immunotoxicity [6]. Xu et al. [7] also demonstrated that multi-walled nanotubes translocated into the pleural cavity and induced visceral mesothelial proliferation in rats. Therefore, the potential adverse human health impact of the biomedical use of NTs must be thoroughly evaluated prior to its widespread application. Aggregation and the more dynamic process of agglomeration are important issues when examining nanoparticle toxicity. Aggregation indicates strongly bonded or fused nanoparticles, whereas agglomeration indicates more weakly bonded nanoparticles [8]. Agglomeration and aggregation are both associated with nanoparticle toxicity due to their regulation of cellular uptake mechanisms [9]. Therefore, increasing evidence has suggested that physicochemical characterization is required to explore the toxicity of nanoparticles [10]. Previous toxicological studies have physically investigated nanoparticle agglomeration in solutions in the presence and absence of protein addition. For example, Zhang and colleagues [11] illustrated that serum-containing medium thoroughly disaggregated titanium dioxide nanoparticles after 120 h. Vippola et al. [12] investigated the addition of bovine serum albumin (BSA) to cell culture media, which reduced agglomerates and increased stability. Surfactants are commonly used to disperse nanoparticles for toxicological investigations [11–13]. For sample, BSA, a type of surfactant, is a globular serum protein that is commonly used in biochemical studies. BSA consists of 583 amino acid residues and its molecular weight is 66,463 Da. The structure of BSA is composed of 67% alpha helices and 17 disulphide bridges, thereby increasing its stability [14]. The advantages of serum-containing media, such as media containing BSA, are that the addition of BSA can disaggregate and stabilize the size of nanoparticles, allowing for a better understanding of nanoparticle toxicity under consistent physicochemical conditions. However, the interactions between proteins and NTs and the consequent biological effects remain unclear. Nanoparticle surface chemistry is a considerable factor in controlling the binding of proteins to specific ligands, resulting in inflammation and immunity [6,15]. Furthermore, protein modification may regulate cell morphology, cell communication and enzymatic processes, leading to abnormal cellular homeostasis. However, the interaction of NTs with proteins and the consequent responses of biological systems are less well understood and are important for the evaluation of nanoparticle toxicity. To perform this investigation, we physicochemically characterized the interaction of SWNTs and DWNTs with BSA. Next, we examined BSA modification by SWNTs and DWNTs [using liquid chromatography–mass spectrometry (LC–MS)] and consequent biological responses [indicated by interleukin-6 (IL-6), interferongamma (IFN-␥) and tumor necrosis factor-alpha (TNF-␣) levels] in vitro in the presence and absence of BSA. Proteins in murine bronchoalveolar lavage fluid (BALF) were allowed to interact with

SWNTs and DWNTs in order to examine the bindability of the protein–particle conjugates. Lastly, SWNT- and DWNT-specific proteins were identified, and the biological processes, molecular functions and pathways affected by the protein–particle conjugates were analyzed. 2. Materials and methods 2.1. Sample preparation Two near-pure, manufactured, chemical-less carbon NTs were used: SWNTs and DWNTs. The SWNTs and DWNTs were purchased from Nanostructured & Amorphous Materials Inc. (Houston, USA). The detailed parameters were as follows: Outer diameters of 1–2 nm for SWNTs and 5 nm for DWNTs, lengths of 1–3 ␮m for SWNTs and 5–15 ␮m for DWNTs, purities >95% and special surface areas of 360–400 m2 /g for SWNTs and 400 m2 /g for DWNTs. The preparation of SWNTs or DWNTs in protein solution was based on protocols outlined in previous studies [16–18]. Briefly, filtered BSA at 1 mg/ml was prepared in sterile phosphate-buffered saline (PBS). Following incubation at 37 ◦ C for 2 h under constant shaking at 500 rpm, SWNTs or DWNTs were suspended and sonicated in two solutions (BSA and BSA-free) at final concentrations of 0, 50, 150 and 1000 ␮g/ml. In the present study, three SWNT and DWNT fractions were used, as follows: (1) supernatant (unbound protein), (2) pelleted particles (protein bound to particles) and (3) suspension. The supernatant fraction (1) was defined as the filtered supernatant collected from the centrifuged particle–protein mixture. The pelleted fraction (2) was denoted as the pelleted particles obtained after centrifugation without the supernatant fraction. The suspension fraction (3) was described as the well-mixing particle–protein samples. To obtain the supernatant and pelleted fractions, the samples (0, 50, 150 and 1000 ␮g/ml of SWNTs or DWNTs) were then separated into two fractions using centrifugation (3500 rpm, or approximately 2100 × g) and filtration following a PBS wash. In the present study, the three SWNT and DWNT fractions were used in different bioreactive investigations. First, protein modification in the SWNT and DWNT supernatant fractions and the corresponding pelleted fractions was determined. Second, inflammation and immune responses that were caused by the SWNT and DWNT suspension fractions and their corresponding supernatant fractions were investigated in vitro. NT-free samples prepared in BSA or a BSA-free solution served as negative controls with the corresponding treatment. All chemicals that were used in the present study were of reagent grade (Sigma–Aldrich, UK), unless otherwise stated. 2.2. NT characterization The BSA-containing and BSA-free pelleted fractions were lyophilized for analysis using a scanning electron microscope (SEM). The preparation of SEM samples has been previously reported [19]. Briefly, the samples were fixed onto 13 mm aluminum SEM stubs after platinum coating (at an average thickness of 10 nm). An InspectTM SEM (FEI, USA) was used to investigate the morphologies of samples at a voltage of 15 kV and a spot size of 2.5. The energy-dispersive X-ray (EDX) Genesis Microanalysis

218

W.-T. Liu et al. / Journal of Hazardous Materials 280 (2014) 216–225

System was used to determine alterations in elemental content. Additionally, the microstructures of SWNTs and DWNTs suspended in BSA were investigated using high-resolution transmission electron microscopy (HRTEM; JEOL 2100, Jeol, Japan) operated at 200 keV. The hydrodynamic diameters and zeta potentials of 150 ␮g/ml SWNT and DWNT suspensions were determined using a Zetasizer (Malvern Zetasizer Nano-ZS, UK). 2.3. Determination of BSA modification in NT supernatants To determine BSA modification by NTs, a BSA control and 1000 ␮g/ml of the SWNT and DWNT supernatant fractions were prepared in a 5% acetonitrile (J.T. Baker, USA)/0.1% formic acid (Fluka, USA) solution [20]. A Q Exactive MS (Thermo Fisher Scientific, Bremen, Germany) that was connected to an UltiMate 3000 Rapid Separation LC (RSLC) system (Dionex, Sunnyvale, USA) was used. Peptide separation was conducted using LC with a C4 100 mm × 2.1 mm, 5 ␮m particle size column (BioBasic, Thermo Fisher Scientific, Germany). The conditions were listed in Supplementary Information Table S1. The mass spectra, ranging from 800 to 3500 m/z, were deconvoluted using Protein Deconvolution v2.0 (Thermo, Germany). 2.4. Characterization of BSA-NT conjugates Following alkylation using 10 mM iodoacetamide, the BSA control and the 1000 ␮g/ml SWNTs and DWNTs pelleted fractions were incubated with 6.5 mM dithiothreitol [20]. Tryptic digestion (25 mM ammonium bicarbonate) and acidification (0.1% formic acid) was conducted prior to analysis using Q Exactive MS coupled to an UltiMate 3000 RSLC system. An acclaim PepMap RSLC C18 column (75 ␮m × 150 mm, 2 ␮m, Dionex) was used to separate peptides under the conditions listed in Supplementary Information Table S2. Full scans of mass spectra ranging from 300 to 2000 m/z were determined. The 10 most intense ions were then subjected to fragmentation, producing MS/MS spectra. Proteome Discoverer v1.4 (Mascot database searches, http://www.matrixscience.com) was used. 2.5. Cell culture Human alveolar epithelial A549 cells were obtained from the American Type Culture Collection (ATCC) and cultured in RPMI 1640 medium that had been supplemented with 10% heat-inactivated fetal bovine serum (FBS) and 1% penicillin–streptomycin at 37 ◦ C in a humidified atmosphere containing 5% CO2 . 2.6. Cell treatment A549 cells were seeded onto surface-treated Transwell® inserts (BD Biosciences, UK) and incubated for 24 h (1 × 105 cells/well). The cells were subjected to various treatment conditions (SWNT and DWNT suspension fractions and SWNT and DWNT supernatant fractions in BSA or BSA-free solution at 0, 50 and 150 ␮g/ml). The supernatants were collected for IL-6, IFN-␥ and TNF-␣ determinations after 4 h of treatment. Each experiment was conducted in quadruplicate. The concentrations were chosen based on their ability to induce inflammatory and immune responses with >80% cell viability [21,22]. 2.7. Determination of IL-6, IFN- and TNF-˛ levels Levels of IL-6, IFN-␥ and TNF-␣ were examined using an enzyme-linked immunosorbent assay (BD OptEIATM set, BD

Biosciences, USA). The experimental processes were conducted according to the manufacturer’s recommended instructions. 2.8. Animals To investigate the impact of NTs on the lining fluid in the lungs, fresh BALF was used to evaluate the binding ability of BALF proteins and NTs. In addition, the types of BALF proteins that bound to NTs were identified. Eight-week-old healthy female BALB/c mice obtained from BioLASCO (Taipei, Taiwan) were maintained at a consistent temperature (22 ± 2 ◦ C) and relative humidity (55 ± 10%) on a 12 h light/dark cycle. BALF samples were collected after 1-week acclimation. All animal experiments were performed in compliance with the animal and ethics review committee of the Laboratory Animal Center of National Taiwan University, Taiwan (IACUC Approval No. 20120074). 2.9. NT and protein binding assay BALF-containing SWNT and DWNT supernatant fractions and pelleted fractions (1000 ␮g/ml) were prepared according to the method described above. Particle-free BALF was used as a negative control. The supernatant fractions and pelleted fractions were heated at 95 ◦ C for 10 min in sample buffer containing 80 mM sodium dodecyl sulfate (SDS), 75 ␮M SDS, 1.25% ␤mercaptoethanol, 10% glycerol and 62.5 mM Tris–HCl (pH 6.8). SDS-polyacrylamide gel electrophoresis (SDS-PAGE, 12%) was conducted to separate the proteins prior to Coomassie Brilliant Blue staining. The gel images were acquired using VisionWorks software (Ultra-Violet Products Ltd., UK) and subjected to densitometric analyses using ImageJ software [23]. 2.10. In-gel protein digestion Proteins of interest (bands 1 and 2) were excised from the gel prior to destaining (with 40% acetonitrile containing 25 mM ammonium bicarbonate) [24]. The samples were dried using a SpeedVac concentrator. Digestion of the gel pieces was performed in 30 ␮l of 25 mM ammonium bicarbonate buffer containing 0.03 mg/ml of modified sequencing-grade trypsin (Promega, USA). Peptides were extracted with 0.1% trifluoroacetic acid (TFA) and 0.1% TFA containing 60% acetonitrile. 2.11. MS and protein identification A QSTARTM XL Q-TOF mass spectrometer (Applied Biosystems, USA) was used to analyze the BALF proteins of interest. Peptides were separated using Dionex Ultimate® 3000 nanoflow liquid chromatography (Thermo Scientific, USA), as previously reported [25]. Full MS scans were conducted in the range of 400–1600 m/z, followed by the selection of the 10 most intense peaks. MASCOT search engine (http://www.matrixscience.com), National Center for Biotechnology Information and UniProt database prediction were performed for protein identification. 2.12. Protein functional analyses The proteins identified in the BALF-NT pelleted fractions were analyzed using the protein annotation through evolutionary relationships (PANTHER) classification system (http://www.pantherdb.org/) against the Mus musculus reference dataset [26,27]. The biological processes and molecular functions related to the proteins of interest were obtained based on functional annotation associated with individual genes/proteins or groups of genes/proteins.

W.-T. Liu et al. / Journal of Hazardous Materials 280 (2014) 216–225

2.13. Statistical analysis For comparisons between groups, the Mann–Whitney U-test was used [28]. The statistical analyses were performed using GraphPad Version 5 for Windows. The level of significance was set to p < 0.05. 3. Results 3.1. Particle characterization No significant differences were observed in the long-length morphology of NTs and protein–nanoparticle conjugates, which retained their morphological features (SWNTs, Fig. 1a and b; DWNTs, Fig. 1d and e). The HRTEM images clearly revealed the single-walled (Fig. 1c) and double-walled morphologies (Fig. 1f) of these samples. To determine elemental alterations post-BSA treatment, a ternary diagram of the C–O–N ratio was used to evaluate changes in SWNTs and DWNTs, pre- and post-BSA treatment (Fig. 1g). BSA served as the NT-free protein control. The C–O–N ratios of the SWNTs/DWNTs and BSA–SWNT/DWNT samples revealed that the O–N ratios increased in NTs after the addition of BSA. To investigate the dispersibility of SWNTs and DWNTs in BSA, the hydrodynamic diameters were determined in SWNT and DWNT suspension fractions in the presence and absence of BSA (Fig. 1h). A slight decrease in the hydrodynamic diameter of the SWNTs was observed when BSA was added (from 246 ± 28 nm to 235 ± 11 nm), whereas a significant decrease in the hydrodynamic diameter of the DWNTs was observed following the addition of BSA (from 241 ± 10 nm to 198 ± 19 nm; p < 0.05). The addition of BSA resulted in NT dispersal, which decreased the hydrodynamic diameters of SWNTs and DWNTs 1.05–1.23 times. The zeta potentials of SWNTs and DWNTs alone and the BSA–SWNT/DWNT conjugates were determined (Fig. 1i). The zeta potentials of SWNTs and DWNTs suspended in deionized water (−0.4 mV) were −8.7 and 8.2 mV, respectively, and were significantly reduced to −19.3 and −20.4 mV, respectively, following the addition of −15.4 mV BSA (p < 0.05). 3.2. Investigation of BSA modification and binding using LC–MS To investigate BSA modification following SWNT or DWNT exposure, the supernatant fractions were first analyzed. Fig. 2 shows the mass spectra of the BSA control and SWNT and DWNT supernatant fractions, the molecular weights of which were 66,427 Da, 66,427 Da and 66,428 Da, respectively. Next, the tryptic peptides from the BSA control and SWNT and DWNT pellet fractions were determined (Fig. 2). The protein sequence coverage of BSA was 73%; however, only 4 and 13% protein sequence coverage values were detected for the protein–SWNT and protein–DWNT conjugates, respectively. These results suggested that protein binding to DWNTs was higher than protein binding to SWNTs. 3.3. Effects of BSA and SWNT or DWNT interaction on IL-6, IFN- and TNF-˛ production Dose-dependent alterations in IL-6, IFN-␥ and TNF-␣ levels were observed after SWNT or DWNT suspension exposure (in the presence and absence of BSA; Fig. 3). However, no significant effects on IL-6, IFN-␥ and TNF-␣ production post-exposure to SWNT or DWNT supernatant were observed. In a comparison between suspension samples in the presence and absence of BSA, the IL-6 levels were significantly elevated to 4.8- (for 50 ␮g/ml) and 6.2-times higher (for 150 ␮g/ml) for SWNTs in the presence of BSA (p < 0.05 as compared to the corresponding samples in the absence of BSA), whereas the

219

Table 1 BALF proteins unbound (supernatant) and bound (pellet) to SWNTs and DWNTs. The identified proteins were divided into two groups (i.e., bands 1 and 2) based on their molecular weights, as analyzed by SDS-PAGE. Protein name

Band 1 Carboxylesterase Contrapsin mCG9583, isoform CRA-a Serum albumin Serum albumin precursor Transferrin Band 2 Alpha-1 antitrypsin precursor Alpha-1 protease inhibitor 2 Alpha-1-antitrypin 1–1 precursor Alpha-1-antitrypsin Alpha-1-antitrypsin 1–4 precursor Alpha-1-antitrypsin 1–5 precursor Immunoglobulin heavy chain variable region Serpinalb protein Serum albumin Serum albumin precursor

SWNT

DWNT

Unbound

Bound

Unbound

Bound

V V V V V V

V V V V

V V V V V V

V V V V V V

V V V V

V

V V V V V

V

V V V V V

V

V V V V

V V V

V V V

V V V

IL-6 levels were significantly increased to 1.5- (for 50 ␮g/ml) and 4-times higher (for 150 ␮g/ml) for DWNTs in the presence of BSA (p < 0.05 as compared to the corresponding samples in the absence of BSA). No significant effects of SWNT or DWNT suspensions (in the presence or absence of BSA) on IFN-␥ production were observed, with the exception of DWNTs (in the presence or absence of BSA) at 150 ␮g/ml (p < 0.05). When comparing the effects of the SWNT and DWNT suspensions on TNF-␣ levels, lower TNF-␣ production was observed for suspension samples in the presence of BSA than in the absence of BSA.

3.4. Identification of BALF protein binding to SWNTs and DWNTs The biological interactions between BALF and SWNTs or DWNTs were studied based on the protein mass ratio. BALF samples interacting with SWNTs or DWNTs displayed two significant bands of approximately 70 (band 1) and 55 kDa (band 2) (Fig. 4a). A significant difference in protein bindability between SWNTs and DWNTs was observed based on band 1: Ten percent for SWNTs and 47% for DWNTs (Fig. 4b). Consistently, 13% and 46% of BALF protein in band 2 was bound to SWNTs and DWNTs, respectively. Together, the results revealed that DWNTs exhibited higher protein bindability than SWNTs. BALF proteins that were unbound or bound to SWNTs or DWNTs were primarily found in bands 1 and 2 (Table 1). Carboxylesterase, contrapsin, mCG9583 isoform CRA-a, serum albumin, serum albumin precursor and transferrin were identified among the unbound and bound proteins in band 1, with the exception of serum albumin precursor bound to SWNTs. Alpha-1 protease inhibitor 2, alpha-1-antitrypsin 1–4 precursor, serpina lb, serum albumin and serum albumin precursor were observed among the unbound and bound proteins in band 2. Alpha-1 antitrypsin precursor and the immunoglobulin heavy-chain variable region were only observed among unbound proteins after SWNT and DWNT treatment, respectively. Alpha-1-antitrypsin 1-1 precursor was observed among unbound and bound SWNT and DWNT samples, with the exception of bound DWNTs. Alpha-1-antitrypsin was found among bound SWNTs and unbound DWNTs, whereas alpha-1-antitrypsin 1–5 precursor was found among unbound and bound SWNTs.

220

W.-T. Liu et al. / Journal of Hazardous Materials 280 (2014) 216–225

Fig. 1. SEM photographs of long-length carbon NTs for (a) SWNTs alone, (b) BSA–SWNT conjugates, (d) DWNTs alone and (e) BSA–DWNT conjugates. The microstructures of (c) BSA–SWNT conjugates and (f) BSA–DWNT conjugates were investigated using HRTEM. (g) Ternary diagrams of C–O–N EDX data for SWNTs and DWNTs (n = 4) with (+) and without (−) BSA treatment, as determined using SEM under consistent conditions. The results indicated that the O–N ratios increased after the particles were treated with BSA. Alterations in (h) hydrodynamic diameter and (i) zeta potentials with (+) and without (−) BSA treatment (n = 5). The values are the means ± SD; *p < 0.05.

3.5. Functional pathway analyses of BALF proteins bound to SWNTs or DWNTs The biological processes, molecular functions and pathways were summarized using PANTHER analysis. These proteins were associated with eight biological processes (Fig. 4c): Cell communication (9%), cellular processes (9%), transport (4%), system processes (11%), the response to stimuli (9%), developmental processes (9%), metabolic processes (33%), cell adhesion (9%) and immune system processes (9%). Four molecular functions were involved in BALF–SWNT/DWNT interaction (Fig. 4c): Binding (45%),

receptor activity (3%), enzyme regulatory activity (31%) and catalytic activity (21%). Blood coagulation was identified to be an important pathway in response to BALF protein binding to SWNTs and DWNTs. 4. Discussion and conclusions NTs often present as either an “aggregation” or an “agglomeration” in air and liquid phases. Increasing evidence indicates that proteins bind to nanoparticles and alter the physical characteristics and bioreactivity of the particles [9,17,23,29,30]. However,

W.-T. Liu et al. / Journal of Hazardous Materials 280 (2014) 216–225

221

Fig. 2. Mass spectra of the BSA control, SWNT supernatant with BSA and DWNT supernatant with BSA.

few studies have investigated protein–nanoparticle interactions on the macromolecular level and the biological consequences of these interactions. Furthermore, the effects of current protein dispersal methods on toxicological examinations remain unclear. For example, Kocbach and colleagues [31] have reported that protein-bound particles altered data obtained in vitro. Therefore, understanding the physicochemical interactions between NTs and biomolecules is important for the future application of NTs. The present study revealed that SWNTs and DWNTs were well dispersed by BSA, without modification of their morphology, suggesting that the BSA–SWNT/DWNT interaction resulted in a reduction of aggregation (reduced hydrodynamic diameter) and an increase in particle stabilization (increased absolute zeta-potential values). No significant BSA modification by SWNTs or DWNTs in the supernatant fraction was observed. Based on the protein sequence coverage and SDS-PAGE results, DWNTs exhibited higher protein binding than SWNTs. BSA in SWNT and DWNT suspensions increased cellular IL-6 levels and reduced TNF-␣ levels as compared to the corresponding samples in the absence of BSA. We next determined the effects of SWNTs and DWNTs on pulmonary protein modification using murine BALF as a surrogate. There were 13 and 11 BALF proteins bound to SWNTs and DWNTs, respectively. These

proteins were related to blood coagulation pathways. The main findings of the present study indicate that BSA is a dispersant that reduces particle aggregation/agglomeration and stabilizes particles for toxicological preparation; however, protein binding may result in misinterpretation of in vitro results and activation of proteinregulated biological responses. A detailed analysis of the physicochemical characteristics of SWNT and DWNT samples under experimental conditions is an important step in understanding their toxicological potential [32]. In the present study, hydrophobic carbon-based SWNTs and DWNTs were used. First, we investigated the morphology of SWNTs and DWNTs in the presence and absence of BSA and found that BSA did not induce significant morphological and structural modifications, but did induce increases in the oxygen and nitrogen ratios. Alterations in the carbon, oxygen and nitrogen proportions implied that BSA interacted with SWNT and DWNT samples, but their bindability requires further investigation (as discussed below). NTs are unstable when dispersed in water. We observed that SWNT and DWNT aggregates were slightly dispersed into smaller sizes in BSA, with higher absolute zeta-potential values. Previous work has consistently indicated that BSA is a good dispersant for disaggregating nanoparticles for toxicological testing [8,11]. The hydrodynamic

222

W.-T. Liu et al. / Journal of Hazardous Materials 280 (2014) 216–225

Fig. 3. A549 cells were exposed to SWNT and DWNT supernatants and suspensions with and without BSA. The endpoints were assessed after exposure for 4 h. IL-6, IFN-␥ and TNF-␣ levels were determined using suspensions with 0, 50 and 150 ␮g/ml NTs. The values are the means ± SD (n = 4); *p < 0.05.

diameter of the SWNT and DWNT suspensions was measured using dynamic light scattering (DLS). However, due to the morphological nature of tubes/fibers (i.e., defined aspect ratios of length to width), the size-distribution produced by DLS measurements may not be accurate. A more appropriate parameter was derived by calculating the ‘hydrodynamic equivalent size-distributions’. The hydrodynamic diameter of the SWNTs and DWNTs is determined by more than one of the dimensions of the particles. In the future, a more detailed understanding of SWNT and DWNT morphological organization would be better addressed through observations

using transmission electron microscopy. A previous study revealed that the hydrophobicity and surface charge of nanoparticles play important roles in regulating protein binding and stability [33]. The possible mechanisms underlying the stability of SWNT and DWNT dispersions in BSA may be associated with electrostatic repulsion [34]. However, the effects of an NT protein coating on alterations in SWNT and DWNT toxicity remain unclear. To clarify the biological effects of BSA–SWNT/DWNT conjugates, it is important to understand the biological role of dispersion solutions. The supernatant fractions were first used to investigate

W.-T. Liu et al. / Journal of Hazardous Materials 280 (2014) 216–225

223

Fig. 4. Protein expression profiles with BALF and particle interactions. (a) 1-D gel patterns of BALF-SWNT/-DWNT supernatants (unbound) and pellets (bound) were determined. (b) Protein bindability is presented as the percentages of BALF in bands 1 and 2. Values are the means ± SD; n = 5; *p < 0.05. (c) Biological process and molecular function associated with the BALF–SWNT/DWNT conjugates in BALF, which were obtained using PANTHER analysis.

unbound BSA modifications by NTs. No significant BSA modifications were induced by the SWNTs or DWNTs. This finding suggested that the interaction of BSA with NTs did not modify the protein structure. Next, the pelleted fractions from the SWNT and DWNT suspensions were used to investigate the interaction of BSA with SWNTs and DWNTs via LC–MS, and the results suggested that DWNTs exhibited higher bindability than SWNTs. The present study was the first study to attempt to investigate relative protein binding to SWNTs and DWNTs using LC–MS. However, further work is required to quantify the binding results. We then added the SWNT or DWNT suspension fractions and their corresponding supernatant fractions to A549 cells to evaluate the effects of pro-inflammatory and immune responses. Alveolar epithelial cells actively participate in host defense to protect underlying tissues from desiccation, toxic challenge and physical trauma. Furthermore, these cells produce and respond to a variety of cellular signaling pathways, including those involving eicosanoids, cytokines and growth factors, that have been commonly used for toxicological screening [35–38]. The SWNT and DWNT supernatant results revealed that there were no significant alterations in IL-6, IFN-␥ or TNF-␣ levels in the presence or absence of BSA. These results were confirmed by our LC–MS findings, suggesting that no protein modification occurred in SWNT or DWNT supernatants. Notably, we found that BSA-coated SWNTs and DWNTs both increased IL-6 levels and reduced TNF-␣ levels as compared to the corresponding BSA-free samples. This finding may have resulted from the activation of pro-inflammatory cytokines, such as IL-6. We observed that SWNT and DWNT suspensions containing BSA displayed reduced TNF-␣ levels as compared to the

corresponding samples in the absence of BSA. A previous study revealed that the addition of BSA significantly inhibited TNF-␣induced mRNA expression in human arterial endothelial cells [39]. This finding may partially explain the observed reductions in TNF␣ levels following treatment with BSA-coated SWNTs and DWNTs; however, the underlying mechanisms remain unclear. The evaluation of the pulmonary toxicity of respirable materials commonly involves animal exposure via inhalation. The fluid surfactant layer lining the alveoli, which contains proteins, surfactant lipids and antioxidants, is the first layer to interact with inhaled nanoparticles. Particle bioreactivity may be mitigated after inhalation as a result of the composition of the lining fluid [20]. Therefore, it is critical to understand the interaction between protein and particles. Consistent with the LC–MS results (interaction with BSA), the SDS-PAGE results revealed that DWNTs exhibited higher binding to BALF proteins than SWNTs in the two bands. We observed that 13 and 11 BALF proteins interacted with SWNTs and DWNTs, respectively. Protein binding to SWNTs and DWNTs is highly selective due to the multitude of proteins in BALF, and very few proteins bound to the NT samples. A similar observation was also reported for plasma protein combined with NTs [40]. We found that carboxylesterase, contrapsin, mCG9583 isoform CRAa, serum albumin, transferrin, alpha-1-antitrypsin and serpina lb were the proteins that commonly bound to SWNTs and DWNTs. PANTHER analyses were used to understand the roles of the identified proteins in biological processes and molecular functions. We found that among the biological processes examined, metabolic processes predominated, whereas binding and enzyme regulatory activities were the predominant molecular functions identified.

224

W.-T. Liu et al. / Journal of Hazardous Materials 280 (2014) 216–225

These biological processes and molecular functions are associated with blood coagulation pathways. A previous study revealed that SWNT exposure increased procoagulant activity and reduced fibrinolytic activity in rats [5]. Meng et al. [41] reported that NTs affected blood coagulation and the mechanical properties of blood clots. Similar systemic inflammation and coagulatory disturbances were observed for SWNTs and DWNTs in mice [42]. Our results are consistent with previous findings that blood coagulation after exposure is the critical mechanism underlying the response to SWNTs and DWNTs [5,41,42]. In the present study, we investigated the application of BSA in assessing NT toxicity in vitro and the potential interactions of pulmonary proteins with NTs. Physicochemically, particle–protein interactions may reduce particle aggregation and increase particle stability, without morphological modification. Biologically, no alterations in BSA structure in the supernatant fractions were observed, although BSA–SWNT/DWNT conjugates may result in misinterpretation of NT toxicity in vitro. Oxidative stress and inflammation have been linked to protein–particle conjugates [15,29,43]. Together with our results, we suggest that the interaction of NTs and proteins should be considered in future risk assessment studies. Furthermore, we identified murine BALF proteins that were associated with blood coagulation pathways. Authors’ contribution All authors have contributed substantially to the concept and design of the study, drafting of the article, and critically revising the manuscript for important intellectual content. All authors have read and approved the final version of the manuscript for publication. Conflict of interest The authors declare that they have no conflicts of interest. Funding This study was founded by the Ministry of Science and Technology of Taiwan (grant numbers: 100-2621-M-002-005 and 103-2314-B-038-018) and the Taipei Medical University and Shuang Ho Hospital (TMU101-AE1-B53, TMU101-AE1-B58 and 102TMU-SHH-09). Acknowledgements The authors wish to thank the Prof. Mong-Hsun Tsai, Ms. ShuTsen Yeh, Ms. Chun-Jung Lin and Mr. Ya Chien Chang-Chien for the technical assistance of this research. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.jhazmat.2014.07.069. References [1] W. Wu, X. Jiang, A practical strategy for constructing nanodrugs using carbon nanotubes as carriers, Methods Mol. Biol. 751 (2011) 565–582. [2] N.W. Kam, Z. Liu, H. Dai, Functionalization of carbon nanotubes via cleavable disulfide bonds for efficient intracellular delivery of siRNA and potent gene silencing, J. Am. Chem. Soc. 127 (2005) 12492–12493. [3] S. Kolhe, K. Parikh, Application of nanotechnology in cancer: a review, Int. J. Bioinform. Res. Appl. 8 (2012) 112–125. [4] A.A. Shvedova, A. Pietroiusti, B. Fadeel, V.E. Kagan, Mechanisms of carbon nanotube-induced toxicity: focus on oxidative stress, Toxicol. Appl. Pharmacol. 261 (2012) 121–133.

[5] L. Tian, Z.Q. Lin, B.C. Lin, H.L. Liu, J. Yan, et al., Single wall carbon nanotube induced inflammation in cruor-fibrinolysis system, Biomed. Environ. Sci. 26 (2013) 338–345. [6] G. Laverny, A. Casset, A. Purohit, E. Schaeffer, C. Spiegelhalter, et al., Immunomodulatory properties of multi-walled carbon nanotubes in peripheral blood mononuclear cells from healthy subjects and allergic patients, Toxicol. Lett. 217 (2013) 91–101. [7] J. Xu, M. Futakuchi, H. Shimizu, D.B. Alexander, K. Yanagihara, et al., Multiwalled carbon nanotubes translocate into the pleural cavity and induce visceral mesothelial proliferation in rats, Cancer Sci. 103 (2012) 2045–2050. [8] J.M. Zook, R.I. Maccuspie, L.E. Locascio, M.D. Halter, J.T. Elliott, Stable nanoparticle aggregates/agglomerates of different sizes and the effect of their size on hemolytic cytotoxicity, Nanotoxicology 5 (2011) 517–530. [9] G. Baier, C. Costa, A. Zeller, D. Baumann, C. Sayer, et al., BSA adsorption on differently charged polystyrene nanoparticles using isothermal titration calorimetry and the influence on cellular uptake, Macromol. Biosci. 11 (2011) 628–638. [10] G. Oberdörster, A. Maynard, K. Donaldson, V. Castranova, J. Fitzpatrick, et al., Principles for characterizing the potential human health effects from exposure to nanomaterials: elements of a screening strategy, Part Fibre Toxicol. 2 (2005) 8. [11] X. Zhang, L. Yin, M. Tang, Y. Pu, Optimized method for preparation of TiO2 nanoparticles dispersion for biological study, J. Nanosci. Nanotechnol. 10 (2010) 5213–5219. [12] M. Vippola, G.C. Falck, H.K. Lindberg, S. Suhonen, E. Vanhala, et al., Preparation of nanoparticle dispersions for in vitro toxicity testing, Hum. Exp. Toxicol. 28 (2009) 377–385. [13] P. Bihari, M. Vippola, S. Schultes, M. Praetner, A.G. Khandoga, et al., Optimized dispersion of nanoparticles for biological in vitro and in vivo studies, Part Fibre Toxicol. 5 (2008) 14. [14] D.C. Carter, J.X. Ho, Structure of serum albumin, Adv. Protein Chem. 45 (1994) 153–203. [15] M. Gasser, P. Wick, M.J. Clift, F. Blank, L. Diener, et al., Pulmonary surfactant coating of multi-walled carbon nanotubes (MWCNTs) influences their oxidative and pro-inflammatory potential in vitro, Part Fibre Toxicol. 9 (2012) 17. [16] M.S. Ehrenberg, A.E. Friedman, J.N. Finkelstein, G. Oberdörster, J.L. McGrath, The influence of protein adsorption on nanoparticle association with cultured endothelial cells, Biomaterials 30 (2009) 603–610. [17] H. Ruh, B. Kuhl, G. Brenner-Weiss, C. Hopf, S. Diabate, et al., Identification of serum proteins bound to industrial nanomaterials, Toxicol. Lett. 208 (2012) 41–50. [18] T. Jones, K. Berube, The bioreactivity of the sub-10 mum component of volcanic ash: Soufriere Hills volcano, Montserrat, J. Hazard. Mater. 194 (2011) 128–134. [19] T. Moreno, W. Gibbons, T. Jones, R. Richards, The geology of ambient aerosols: characterising urban and rural/coastal silicate PM10–2.5 and PM2.5 using highvolume cascade collection and scanning electron microscopy, Atmos. Environ. 37 (2003) 4265–4276. [20] L.L. Chiang, H.C. Chen, C.N. Lee, K.J. Chuang, T.T. Chen, et al., Serum protein oxidation by diesel exhaust particles: effects on oxidative stress and inflammatory response in vitro, Chem. Biol. Interact. 206 (2013) 385–393. [21] M.R. Wilson, J.H. Lightbody, K. Donaldson, J. Sales, V. Stone, Interactions between ultrafine particles and transition metals in vivo and in vitro, Toxicol. Appl. Pharmacol. 184 (2002) 172–179. [22] H.C. Chuang, C.W. Fan, K.Y. Chen, G.P. Chang-Chien, C.C. Chan, Vasoactive alteration and inflammation induced by polycyclic aromatic hydrocarbons and trace metals of vehicle exhaust particles, Toxicol. Lett. 214 (2012) 131–136. [23] S. Fertsch-Gapp, M. Semmler-Behnke, A. Wenk, W.G. Kreyling, Binding of polystyrene and carbon black nanoparticles to blood serum proteins, Inhal. Toxicol. 23 (2011) 468–475. [24] C.L. Wen, K.Y. Chen, C.T. Chen, J.G. Chuang, P.C. Yang, et al., Development of an AlphaLISA assay to quantify serum core-fucosylated E-cadherin as a metastatic lung adenocarcinoma biomarker, J. Proteomics 75 (2012) 3963–3976. [25] T.-T. Chen, K.-J. Chuang, L.-L. Chiang, C.-C. Chen, C.-T. Yeh, et al., Characterization of the interactions between protein and carbon black, J. Hazard. Mater. 264 (2014) 127–135. [26] D.W. Huang, B.T. Sherman, Q. Tan, J. Kir, D. Liu, et al., DAVID Bioinformatics Resources: expanded annotation database and novel algorithms to better extract biology from large gene lists, Nucleic Acids Res. 35 (2007) W169–W175. [27] C.L. Su, T.T. Chen, C.C. Chang, K.J. Chuang, C.K. Wu, et al., Comparative proteomics of inhaled silver nanoparticles in healthy and allergen provoked mice, Int. J. Nanomed. 8 (2013) 2783–2799. [28] B. Sitkauskiene, M. Radinger, A. Bossios, A.K. Johansson, R. Sakalauskas, et al., Airway allergen exposure stimulates bone marrow eosinophilia partly via IL-9, Respir. Res. 6 (2005) 33. [29] C. Schulze, U.F. Schaefer, C.A. Ruge, W. Wohlleben, C.M. Lehr, Interaction of metal oxide nanoparticles with lung surfactant protein A, Eur. J. Pharm. Biopharm. 77 (2011) 376–383. [30] P. Aggarwal, J.B. Hall, C.B. McLeland, M.A. Dobrovolskaia, S.E. McNeil, Nanoparticle interaction with plasma proteins as it relates to particle biodistribution, biocompatibility and therapeutic efficacy, Adv Drug Deliv. Rev. 61 (2009) 428–437. [31] A. Kocbach, A.I. Totlandsdal, M. Lag, M. Refsnes, P.E. Schwarze, Differential binding of cytokines to environmentally relevant particles: a possible source for misinterpretation of in vitro results? Toxicol. Lett. 176 (2008) 131–137. [32] K. Donaldson, L. Tran, L.A. Jimenez, R. Duffin, D.E. Newby, et al., Combustionderived nanoparticles: a review of their toxicology following inhalation exposure, Part Fibre Toxicol. 2 (2005) 10.

W.-T. Liu et al. / Journal of Hazardous Materials 280 (2014) 216–225 [33] A.A. Shemetov, I. Nabiev, A. Sukhanova, Molecular interaction of proteins and peptides with nanoparticles, ACS Nano 6 (2012) 4585–4602. [34] P. Nithiyasri, K. Balaji, P. Brindha, M. Parthasarathy, Programmable selfassembly of carbon nanotubes assisted by reversible denaturation of a protein, Nanotechnology 23 (2012) 465603. [35] L.-Y. Lin, I.J. Liu, H.-C. Chuang, H.-Y. Lin, K.-J. Chuang, Size and composition effects of household particles on inflammation and endothelial dysfunction of human coronary artery endothelial cells, Atmos. Environ. 77 (2013) 490–495. [36] H.C. Chuang, T. Jones, T.T. Chen, K. Berube, Cytotoxic effects of incense particles in relation to oxidative stress, the cell cycle and F-actin assembly, Toxicol. Lett. 220 (2013) 229–237. [37] W.S. Cho, R. Duffin, C.A. Poland, A. Duschl, G.J. Oostingh, et al., Differential pro-inflammatory effects of metal oxide nanoparticles and their soluble ions in vitro and in vivo; zinc and copper nanoparticles, but not their ions, recruit eosinophils to the lungs, Nanotoxicology 6 (2012) 22–35. [38] E. Herzog, H.J. Byrne, M. Davoren, A. Casey, A. Duschl, et al., Dispersion medium modulates oxidative stress response of human lung epithelial cells upon exposure to carbon nanomaterial samples, Toxicol. Appl. Pharmacol. 236 (2009) 276–281.

225

[39] W.J. Zhang, B. Frei, Albumin selectively inhibits TNF alpha-induced expression of vascular cell adhesion molecule-1 in human aortic endothelial cells, Cardiovasc. Res. 55 (2002) 820–829. [40] C. Salvador-Morales, E. Flahaut, E. Sim, J. Sloan, M.L. Green, et al., Complement activation and protein adsorption by carbon nanotubes, Mol. Immunol. 43 (2006) 193–201. [41] J. Meng, X. Cheng, J. Liu, W. Zhang, X. Li, et al., Effects of long and short carboxylated or aminated multiwalled carbon nanotubes on blood coagulation, PLoS ONE 7 (2012) e38995. [42] K. Inoue, H. Takano, E. Koike, R. Yanagisawa, M. Sakurai, et al., Effects of pulmonary exposure to carbon nanotubes on lung and systemic inflammation with coagulatory disturbance induced by lipopolysaccharide in mice, Exp. Biol. Med. (Maywood) 233 (2008) 1583–1590. [43] S. Mandal, M. Hossain, P.S. Devi, G.S. Kumar, K. Chaudhuri, Interaction of carbon nanoparticles to serum albumin: elucidation of the extent of perturbation of serum albumin conformations and thermodynamical parameters, J. Hazard. Mater. 248/249 (2013) 238–245.