toxicological sciences 138(1), 104–116 2014 doi:10.1093/toxsci/kft328 Advance Access publication January 15, 2014

Vascular Effects of Multiwalled Carbon Nanotubes in Dyslipidemic ApoE−/− Mice and Cultured Endothelial Cells Yi Cao,* Nicklas Raun Jacobsen,† Pernille Høgh Danielsen,* Anke G. Lenz,‡ Tobias Stoeger,‡ Steffen Loft,* Håkan Wallin,†,* Martin Roursgaard,* Lone Mikkelsen,* and Peter Møller*,1  *Department of Public Health, Section of Environmental Health, University of Copenhagen, Øster Farimagsgade 5A, DK-1014 Copenhagen K, Denmark; †The National Research Centre for the Working Environment, Lersø Parkalle 105, DK-2100 Copenhagen, Denmark; and ‡Comprehensive Pneumology Center, Institute of Lung Biology and Disease, Helmholtz Zentrum München, German Research Center for Environmental Health, 85764 Neuherberg, Germany 1

Received May 30, 2013; accepted December 11, 2013

Accumulating evidences indicate that pulmonary exposure to carbon nanotubes (CNTs) is associated with increased risk of lung diseases, whereas the effect on the vascular system is less studied. We investigated vascular effects of 2 types of multiwalled CNTs (MWCNTs) in apolipoprotein E−/− mice, wild-type mice, and cultured cells. The ApoE−/− mice had accelerated plaque progression in aorta after 5 intracheal instillations of MWCNT (25.6 μg/mouse weekly for 5 weeks). The exposure was associated with pulmonary inflammation, lipid peroxidation, and increased expression of inflammatory, oxidative stress, DNA repair, and vascular activation response genes. The level of oxidatively damaged DNA in lung tissue was unaltered, probably due to increased DNA repair capacities. Despite upregulation of inflammatory genes in the liver, effects on systemic cytokines and lipid peroxidation were minimal. The exposure to MWCNTs in cultured human endothelial cells increased the expression of cell adhesion molecules (ICAM1 and VCAM1). In cocultures, there was increased adhesion of monocytes to endothelial cells after exposure to MWCNT. The exposure to both types of MWCNT was also associated with increased lipid accumulation in monocytic-derived foam cells, which was dependent on concomitant oxidative stress because the antioxidant N-acetylcysteine inhibited the lipid accumulation. Collectively, our results indicate that exposure to MWCNT is associated with accelerated progression of atherosclerosis, which could be related to both increased adherence of monocytes onto the endothelium and oxidative stress-mediated transformation of monocytes to foam cells. Key Words:  8-isoprostanes; multiwalled carbon nanotubes; oxidative stress; comet assay; atherosclerosis; lipid accumulation.

Substantial evidence from both human studies and animal experimental models indicates that airway exposure to particles is associated with vasomotor dysfunction and progression of atherosclerosis (Møller et al., 2011). The exposure to high aspect ratio materials has mainly been linked to pulmonary

diseases such as fibrosis and cancers, although asbestosexposed workers also have increased risk of cardiovascular disease (Harding et al., 2012). Carbon nanotubes (CNTs) are high aspect ratio nanomaterials, which are being produced for industrial and medical uses. It has been shown that pulmonary exposure to single-walled CNTs (SWCNTs) promotes plaque progression in atherosclerosis-prone ApoE knockout (ApoE−/−) mice (Li et  al., 2007). To the best of our knowledge, there are no publications showing association between pulmonary exposure to multiwalled CNTs (MWCNTs) and atherosclerosis, whereas it has recently been reported that IV injection of MWCNT promoted plaque progression in rats (Xu et  al., 2012). The mechanisms linking airway exposure to atherosclerosis and inflammation are unknown, although pulmonary and/ or systemic inflammation and oxidative stress are considered to be important steps for CNT-mediated cardiovascular diseases (Shvedova et al., 2012). Even a single oropharyngeal aspiration of MWCNT is associated with pulmonary inflammation (Wang et al., 2011). Subchronic inhalation studies with MWCNT have shown little effect on hematology parameters suggested in the OECD test guideline 413, including total and differential leukocyte counts and clotting potential (Ma-Hock et  al., 2009; Pauluhn, 2010). However, it has been described that inhalation exposure to MWCNT increased the systemic immune response, evidenced as increased interleukin 10 (IL10) levels in the spleen, whereas the same doses did not cause pulmonary inflammation (Mitchell et al., 2007; Pauluhn, 2010). Atherosclerosis occurs in arteries, where a gradual accumulation of lipids in the intima leads to narrowing of the vessel. The first step in this process occurs when endothelial cells express cell adhesion molecules on their plasma membrane, which facilitates adherence of monocytes to the endothelium. The monocytes may subsequently migrate through the endothelium to the intima where they transform to foam cells, accumulate

© The Author 2014. Published by Oxford University Press on behalf of the Society of Toxicology. All rights reserved. For permissions, please email: [email protected].

Downloaded from http://toxsci.oxfordjournals.org/ at Universidade de Vigo on May 7, 2014

To whom correspondence should be addressed at Department of Public Health, Section of Environmental Health, University of Copenhagen, Øster Farimagsgade 5A, Building 5B, 2nd Floor, DK-1014 Copenhagen, Denmark. Fax: +45 3532 7686. E-mail: [email protected].

Vascular Effects of Carbon Nanotubes



Materials and Methods Particle Preparations The MWCNTs were NM400 (Nanocyl; entangled MWCNT, diameter 30 nm) and NM402 (Arkema Graphistrength C100; entangled MWCNT, diameter 30 nm) from the European Commission Joint Research Centre

Nanomaterials Repository (http://ihcp.jrc.ec.europa.eu/our_activities/nanotechnology/nanomaterials-repository). They were selected as materials with short and long entangled fibers, based on information from the supplier. The materials had similar characteristics in protein rich cell media and they did not differ in regard to cytotoxicity and inflammatory response in hepatocytes (Kermanizadeh et al., 2013). Physicochemical characterization by transmission electron microscopy and Brunauer, Emmett, and Teller (BET) surface area has been reported by Kermanizadeh et al. (2013). The BET surface area of NM400 and NM402 was 298 and 225 m2/g, respectively. The length and diameter of NM400 was 700–3000 and 5–35 nm, respectively. The length and diameter of NM402 was 400–4000 and 6–20 nm, respectively. NM400 contained Fe/Co catalysts (6–9 nm, average length 7.5 nm) inside the tubes, whereas NM402 contained amorphous carbon structures with Fe (5–20 nm) as impurities. Vietti et al. (2013) have determined the level of impurities in NM400 and NM402 to 5.4% and 3.2% of the mass. According to the particle characterization by the joint action Nanogenotox project, the most abundant impurities in NM400 and NM402 were aluminum and iron as determined by inductive coupled plasma with mass spectrometry. The content of aluminum was 1.0% and 1.3% in NM400 and NM402, respectively. The content of iron was 0.2% and 1.6% in NM400 and NM402, respectively (http://www.nanogenotox.eu). In the present study, we have investigated the structure of the MWCNTs in suspension before and after sonication by scanning electron microscopy. The MWCNTs were sonicated in nanopure water with 2% mouse serum and 5  μl of the suspension was applied onto a carbon conductive double-coated 12-mm spectro tab (Canemco-Marivac, Quebec, Canada) and fixed on a specimen mount. The structure of nonsonicated MWCNTs was determined by softly touching the surface of a conductive carbon tape on the raw particle material that had been taking straight out of the container. All samples were coated with a conductive layer of Au of 10–20 nm thickness before imaging on a Zeiss Ultra 55 SEM equipped with a field emission electron source. High vacuum conditions were applied and a secondary electron detector was used for image acquisition. The particle size of MWCNT agglomerates in the vehicle was measured by Nanosight LM20 with NanoSight Tracking Analysis (NTA) software version 2.1 (NanoSight Ltd, Amesbury, UK). We used the same dispersion procedure for the NTA, cell cultures, and animal experiments, although it was necessary to dilute the samples to 1 μg/ml for the measurement of particle size. The particle size was measured in water, saline, or RPMI 1640 medium with 10% serum. The measured particle size of the MWCNTs by NTA is an indication of spherical size of agglomerates in the suspension, which is not identical to the true particle size. Nevertheless, the NTA measurement is relevant as a measure of the difference in agglomerate size between NM400 and NM402 in suspension. MWCNT stock solution was prepared by sonicating a 2.56 mg/ml suspension of particles in double-distilled water (Sigma-Aldrich, St Louis, Missouri) containing either 2% mouse serum (homemade from sibling C57 mice) or 2% bovine serum (for cultured cells; Sigma B9433), using a Branson Sonifier S-450D (Branson Ultrasonics Corp, Danbury, Connecticut) equipped with a disrupter horn (Model number: 101-147-037) before each experiment. The suspension was sonicated continuously for 16 min at amplitude of 10% and cooling on ice/water to avoid sample heating and evaporation. After sonication, the suspension was diluted and used only once. In Vivo Experiments Exposure of mice to MWCNT.  Female wild-type (WT) C57BL/6N Tac and ApoE−/− (C57BL/6-Apoe tm1) mice were purchased from Taconic MB (Ejby, Denmark) at the age of 5–8 weeks. They were housed in cages with standard 12-h light and 12-h dark cycle conditions in a temperature and humidity controlled environment with access to Western-type diet and tap water ad libitum. The WT and ApoE−/− mice were exposed by intracheal (i.t.) instillation to a total dose of 32 or 128 μg/mouse of MWCNT from the age of 9–10 weeks. The mice were exposed once a week (6.4 or 25.6 per instillation) for 5 weeks. In an auxiliary experiment for assessment of pulmonary inflammation, WT mice were sacrificed 24 h after a single i.t. instillation of 32 or 128  μg/ mouse. The content of cells in BALF was assessed in all groups of mice.

Downloaded from http://toxsci.oxfordjournals.org/ at Universidade de Vigo on May 7, 2014

lipids, and mediate inflammatory responses (Moore and Tabas, 2011). It has been shown that exposure of cultured endothelial cells to MWCNT was associated with increased production of reactive oxygen species (ROS) and promoted the expression of vascular adhesion molecule 1 (VCAM1) and release of proinflammatory cytokines (Vidanapathirana et  al., 2012). Furthermore, exposure to nanoparticles can downregulate lipid oxidation and thus induce lipid droplet formation in cultured cells (Przybytkowski et al., 2009). Cultured macrophages also take up SWCNT and the same type of particles accumulates in macrophages in the neointima of the carotid artery after IV injection of SWCNT (Kosuge et al., 2012). The purpose of this study was to investigate the association between pulmonary exposure to MWCNT and plaque progression, including the underlying mechanisms of action. We investigated MWCNTs of different size because the fiber length is important for pulmonary toxicity of high aspect ratio nanomaterials. We fed the dyslipidemic ApoE−/− mice a high-fat diet to accelerate the plaque progression, associated with endothelial dysfunction and aortic stiffness. Mechanisms in terms of oxidative stress and inflammation were assessed in bronchiolar lavage fluid (BALF), lung tissue, serum, and liver tissue. In liver and lung tissues, gene expression responses included oxidative stress (Hmox1: heme oxygenase (decycling) 1), inflammation (Ccl2: chemokine (C-C motif) ligand 2 and Nos2: nitric oxide synthase 2-inducible), vascular activation (Vcam1: vascular adhesion molecule 1 and Vegfa: vascular endothelial growth factor A), and repair of oxidatively damaged DNA (Ogg1: oxoguanine DNA glycosylase 1). The latter was included because we also assessed oxidative stress effects in terms of oxidatively damaged DNA, which has been found to be induced by several carbon-based nanoparticles in lungs in vivo and in vitro along with ROS production, which we assessed here in the cell cultures (Jacobsen et al., 2008; Møller et al., 2013). We investigated the effect of exposure to MWCNTs in cultured cells with biomarkers reflecting activation of endothelial cells and monocytes, their interaction, and lipid accumulation in foam cells. The effect of MWCNT was assessed in human umbilical vein endothelial cells (HUVEC), which express intercellular adhesion molecule 1 (ICAM1) and VCAM1 after exposure to particles and interact with monocytes in terms of THP-1 cells (Forchhammer et al., 2012). The formation of foam cells in vitro was assessed in THP-1 cells, which were transformed to macrophages prior to the exposure to MWCNT. These in vitro assays mimic the steps in the development of atherosclerosis, whereas the cell culture exposure system does not fully model the pathway from pulmonary MWCNT exposure to development of atherosclerosis.

105

106

Cao et al.

Pulmonary inflammation.  We assessed the inflammation response in WT and ApoE−/− mice because it has previously been observed that ApoE−/− mice responded with stronger influx of neutrophils than WT mice to a pulmonary particle exposure (Jacobsen et al., 2009). The number of cells in BALF was determined by visual scoring of 200 cells according to the methods described by Jacobsen et al. (2009). In addition, we measured the level of 11 cytokines in BALF, including IL1β, IL4, IL6, IL12p40, granulocyte colony-stimulating factor (G-CSF), chemokine (C-X-C motif) ligand 1 (CXCL1, also known as keratinocyte-derived chemokine [KC]), tumor necrosis factor (TNF), CCL2 (also known as macrophage chemotactic protein 1 [MCP1]), chemokine (C-C motif) ligand 4 (CCL4, also known as macrophage inflammatory protein 1β [MIP1β]), and chemokine (C-C motif) ligand 5 (CCL5, also known as regulated on activation, normal T cell expressed and secreted [RANTES]). We measured 10 cytokines in serum, including IL1a, IL1β, IL2, IL4, IL5, IL6, IL10, IL12p40, G-CSF, KC/CXCL1, and TNF was determined by multiplexed immunoassays (Bio-Plex Pro Mouse Cytokine 11-Plex Panel No. M60006RDAY; Bio-Rad, Munich, Germany), as described by Götz et al. (2011). DNA damage in pulmonary tissue.  The nuclear DNA was isolated as described by Vesterdal et al. (2012). The level of DNA damage was measured as DNA strand breaks and FPG sensitive sites by the alkaline comet assay as described by Jantzen et al. (2012) and further described in the Supplementary Information. The level of DNA migration was scored by visual classification according to the 5-class scoring system and transformed to lesions per 106 base pair according to the calibration curve from the European Comet Assay Validation Group (Forchhammer et al., 2010). Isoprostanes.  The level of 8-isoprostanes was determined in lung tissue and serum by the enzyme immunoassay (No. 516351, Cayman Chemical Company, Ann Arbor, Michigan) as described by Beck-Speier et al. (2012). mRNA expression in lung and liver tissue.  The gene expression level of Icam1 (GeneID: 15894), Vcam1 (GeneID: 22329), Vegfa (Gene ID: 22339), Hmox1 (GeneID: 15368), Nos2 (GeneID: 18126), Ccl2 (Gene ID: 20296), and Ogg1 (GeneID: 18294)  was determined by reverse transcriptase-PCR (RT-PCR) as described previously (Vesterdal et al., 2010). We have not measured gene expression in aorta because the preparation of the vessels for plaque progression analysis could be associated with degradation of mRNA. Plaque area assessment.  The area of atherosclerotic plaques was determined as described by Mikkelsen et al. (2011). The entire aorta from the junction with the heart to the iliac bifurcation was placed in a flatbed scanner. The plaques were observed as whitish spots against a black background. The fraction of area of the total aortic surface that contained plaques was determined on a digital microscope image (Digital Imaging Solutions; analySIS getIT!) by means of the computer software Image J and calculated as % plaques proportional to total aorta area.

In Vitro Experiments Cell lines.  The monocytic THP-1 cell line (American Type Culture Collection, Manassas, Virginia) was cultured in suspension in RPMI cell medium with 10% serum as previously described (Danielsen et al., 2009). HUVEC and their growth medium (ie, Endothelial Cell Growth Medium Kit containing 2% serum) were purchased from Cell Applications (San Diego, California). The HUVEC were used at passage 2 to maintain equivalent morphologic and phenotypic characteristics (Mikkelsen et al., 2011). HUVEC are considered to be arterial endothelial cells because the umbilical cord vein delivers oxygenated blood to the fetus and it has a higher blood pressure than the central venous blood vessels. We exposed cells to 2, 4, 8, 16, 32, 64, 128, and 256 μg/ml in the initial screening for cytotoxicity, ROS production, and surface adhesion molecules ICAM1 and VCAM1. Subsequent experiments on gene expression included fewer concentrations (4, 16, 64, 128, and 256  μg/ml). The assessment of adhesion of THP-1 cells on HUVEC in coculture was limited to concentrations with no or modest cytotoxicity (2, 4, 8, 16, 32, and 64 μg/ml). As the cultures contain adherent (HUVEC) and suspension (THP-1) cells, we have reported the concentration in μg/ml. The conversion to μg/cm2 cell surface area for the cultures with HUVEC corresponds to a concentration range between 0.63 μg/cm2 (equal to 2 μg/ ml) and 80 μg/cm2 (equal to 256 μg/ml). The experiments on cytotoxicity, proliferation, cellular ROS production, THP-1 attachment assay, and lipid accumulation were repeated on 3 different days and were all done in triplicates per day in 96-well plates. The mRNA expression was assessed in 6-well plates because it requires more cells; in this assay, 250  μg/ml corresponds to 110 μg/cm2. WST-1 assay.  The cytotoxicity was examined by the WST-1 assay (Roche Diagnostics GmbH, Mannheim, Germany), which measures mitochondrial succinate dehydrogenase activity in living cells. HUVEC (2 × 104 cells/well) or THP-1 monocytes (5 × 104 cells/well) were exposed to various concentrations of MWCNT for 24 h in 96-well plates. After exposure, the cells were rinsed and cultured in 100 μl of fresh media containing 10% WST-1 reagent for 2 h. The THP-1 macrophages (5 × 104 cells/well) were exposed to various concentrations of MWCNT for 18 h, rinsed once, followed by 3-h exposure to 200 μl of fresh media containing 10% WST-1 reagent, with or without 0.5mM free fatty acid (oleic/palmitic acid, 2:1). The enzymatic conversion of WST-1 by succinate dehydrogenase was measured spectrometrically at 450 nm with 630 nm as reference. Proliferation assay.  The proliferation ability was measured by a BrdU incorporation assay in HUVEC and THP-1 cells according to the manufacturer’s instructions (Roche Diagnostics GmbH). Briefly, 5 × 103/well of HUVEC or 2 × 104/well of THP-1 cells were exposed to MWCNT for 24 h, and the BrdU incorporation was measured. As we observed increased BrdU incorporation in 32 μg/ml MWCNT-exposed THP-1 cells, we also counted the numbers of THP-1 cells from 0- to 48-h exposure. ROS measurement.  The ROS generation in cells and in acellular condition was measured by the 2',7'-dihydrofluorescin diacetate (DCFH-DA) assay as previously described (Hemmingsen et  al., 2011). Briefly, 2 × 104/ well HUVEC or 5 × 104/well THP-1 cells were preloaded with 2μM DCFH-DA for 15 min, washed, and exposed to various concentrations of MWCNT. The fluorescence was continuously measured for 3 h and the accumulated ROS production was calculated. ICAM1 and VCAM1 measurement.  The expression of ICAM1 and VCAM1 was measured with a modified ELISA procedure as previously described (Frikke-Schmidt et  al., 2011). Briefly, HUVEC were exposed to MWCNT for 24 h and the subsequently incubated with anti-ICAM1 or antiVCAM1 (R&D Systems, Abingdon, UK) for 1 h. The cells were rinsed 3 times with 1% bovine serum albumin media and further incubated with anti-goat IgG peroxidase-coupled antibody (Sigma, St Louis, Missouri) for another 1 h. After 5 times washes, the substrate solution containing 0.024% H2O2 and 0.4 mg/ml o-phenylenediamine (Sigma, St Louis, Missouri) was added for 30 min, and the product was read at 450 nm.

Downloaded from http://toxsci.oxfordjournals.org/ at Universidade de Vigo on May 7, 2014

The other biomarkers were assessed in groups of mice that had been exposed to 128 μg/mouse, except the results on the aortic plaque area in ApoE−/− mice at day 28 after the last exposure that included both 32 and 128  μg/mouse. Supplementary Figure 1 shows a flow chart of the exposure and biomarkers at different time points and experimental models. It has been described that workplace exposure to MWCNT can be up to 400 μg/m3, although one-tenth is a more likely exposure (Mercer et al., 2013). The weekly exposure was approximately 1 mg/kg body weight in our study (25.6  μg/week and approximately 25 g body weight for mice). Thus, the human exposure can be as high as 68 μg/kg body weight per week, assuming that humans inhale 8 m3 per working day, 70 kg body weight, and a 30% deposition of MWCNT. The highest dose (128 μg per week) in our study corresponds to approximately 3 weeks of exposure at 400  μg/m3 in humans. All animal procedures followed the guidelines for the care and handling of laboratory animals established by the Danish government, and the Animal Experiment Inspectorate, Ministry of Justice, approved the study. Further information on the housing of animals, i.t. instillation, preparation of mouse serum, and isolation of organs are available in the Supplementary Information.



107

Vascular Effects of Carbon Nanotubes

THP-1 attachment assay.  The attachment of THP-1 cells onto HUVEC was done with the BrdU assay as previously described (Forchhammer et al., 2012). Briefly, cocultures of HUVEC (2 × 104 cells/well) and BrdU-labeled THP-1 cells (5 × 103 cells/well) were exposed to various concentrations of MWCNT for 24 h. After exposure, the media were removed and BrdU content both in the coculture and supernatant was determined. Data were expressed as the percentage of THP-1 cells remaining in cocultures.

Gene expression in THP-1 cells.  The expressions of CCL2 (Gene ID: 6347), IL8 (Gene ID: 3576), TNF (Gene ID: 7124), and HMOX1 (Gene ID: 3162)  were assessed by real-time RT-PCR in THP-1 cells. Briefly, 2 × 106/well of THP-1 cells were exposed to 4, 16, or 64 μg/ml of MWCNT for 24 h. After exposure, the cells were washed twice, the mRNA was extracted, and the gene expression was measured as described by Vesterdal et al. (2010). Statistics.  The results were analyzed by parametric 1- or 2-factor ANOVA if there was homogeneity of variance between the groups. Otherwise, they were analysed by nonparametric ANOVA. A  detailed description of the statistical analysis is outlined in the Supplementary Information. Statistically significant results were accepted on p < .05 in the overall ANOVA. The p values in the text correspond to post hoc least significant difference tests. The statistical analysis was carried out in Statistica 5.5 from StatSoft, Inc, Tulsa, Oklahoma. All data are presented as mean and SEM values.

Characterization of MWCNTs Figure  1 shows scanning electron microscopy images of NM400 and NM402 as raw material or in dispersion vehicle. Both of the MWCNT samples appeared to consist of entangled, irregular, and bend nanotubes. Table 1 displays the particle size of MWCNT agglomerates in water, saline, or RPMI medium after sonication. The distribution of the size by NTA is shown in Supplementary Figure 2. A collective assessment of 3 different approaches to characterize NM400 and NM402 indicates that they differ slightly by diameter, impurity content, and agglomeration size (Kermanizadeh et  al., 2013; Vietti et  al., 2013; present study). There are more characterization properties than samples of MWCNT, indicating that it is not possible to establish associations between specific characterization parameters and biomarker measurements. In Vivo Results Plaque progression.  The mice that had been exposed to 5 i.t. instillations of NM400 (p < .01) or NM402 (p < .01) had accelerated plaque progression in the aorta tissue at 24 h after the last of 5 i.t. instillations (Fig.  2). An additional recovery period of 4 weeks after the last exposure showed that the difference in plaque progression between the MWCNT-exposed mice and controls was decreased, although the highest dose of NM402 was statistically nonsignificantly increased by 1.4-fold (95% CI: 0.9–2.2 fold) compared with the controls (Supplementary Figure 3). Pulmonary inflammation.  The cell counts in BALF at 24 h after the last of 5 i.t. instillations in WT and ApoE−/− mice are shown in Table 2. There was increased influx of neutrophils in BALF in ApoE−/− mice after exposure to 128  μg/mouse (p < .05), whereas there was no difference in the influx of neutrophils between the 2 types of MWCNT (p > .05). The exposure

Fig. 1.  Scanning electron microscopy images of NM400 and NM402 without (A, B, E, and F) or with (C, D, G, and H) sonication. Images (A–D) and (E–H) are NM400 and NM402, respectively.

Downloaded from http://toxsci.oxfordjournals.org/ at Universidade de Vigo on May 7, 2014

Lipid accumulation in foam cells.  THP-1 cells were transformed to macrophages by 24-h treatment with phorbol 12-myristate 13-acetate. These cells were subsequently seeded in 96-well black plates (5 × 104 cells/well) and exposed to various concentrations of MWCNT for 18 h, with or without the presence of 1mM of the antioxidant N-acetylcysteine (NAC; Sigma). The cells were subsequently rinsed once, and then incubated with either medium (for control) or 0.5mM free fatty acid (oleic/palmitic acid, 2:1) for another 3 h. The cells were rinsed once with Hanks solution and stained in Hanks solution containing 0.5 μg/ml Nile red and 0.01% Pluronic F127 (Sigma-Aldrich) for 15 min in the dark. The cells were rinsed twice with Hanks solution and the fluorescence was measured in a fluorescence spectrophotometer (excitation: 544 nm, emission: 590 nm). The formation of intracellular lipid droplets was assessed visually by microscopy. The THP-1 macrophages were seeded onto precoated 8-well microscopy chamber (Ibidi, Munich, Germany), exposed to MWCNT for 18 h, and subsequently stained with Nile red in medium for 5 min. The nuclei were stained with Hoechst 33342. After staining, the cells were inspected in a Leica FA6000 inverted wide-field microscope with ×100 magnification (Leica Microsystems GmbH, Wetzlar, Germany).

Results

108

Cao et al.

Table 1 Size of Particle Agglomerates in Suspensions of MWCNT Measured by NanoSight Tracking Analysis Method

NM402

198 ± 85 nm 186 (95–315) nm 186 nm

206 ± 108 nm 176 (85–355) nm 106 nm

124 ± 50 nm 116 (71–179) nm 110 nm

153 ± 54 nm 147 (92–214) nm 145 nm

143 ± 59 nm 132 (79–217) nm 112 nm

150 ± 64 nm 140 (81–231) nm 147 nm

Fig. 2.  Plaque progression in aorta of ApoE−/− mice at 24 h after the last exposure to MWCNT by i.t. instillation once a week for 5 weeks. The total dose was 128 μg/mouse. The results are means and SEM from 5–6 mice/group. *p < .01 compared with the control. Abbreviation: MWCNT, multiwalled carbon nanotube.

to 128 μg/mouse of NM400 was associated with decreased level of lymphocytes in WT and ApoE−/− mice (p < .05), whereas the NM402 decreased only the number of lymphocytes in WT mice (p < .05). The number of eosinophils decreased in the WT mice that were exposed to 128 μg/mouse of MWCNT (p < .05). The same tendency toward decreased number of eosinophils was also observed in ApoE−/− mice, albeit it was not statistically significant (p > .05). The number of macrophages (p > .05) and epithelial cells (p > .05) were unaltered by the MWCNT exposure in WT and ApoE−/− mice at 24 h after the last of 5 i.t. instillations. The cell counts in BALF at day 28 after the last of 5 i.t. instillations in WT and ApoE−/− mice are shown in Table  3. The levels of neutrophils increased dose dependently in WT and ApoE−/− mice after exposure to 32 and 128 μg/mouse (p < .05). The exposure was associated with modest increases in the levels of lymphocytes, albeit with some inconsistency in regard to the dose and strain. The level of macrophages (p > .05) and

Cytokines in BALF and serum.  Table 4 shows the levels of cytokines in BALF from WT and ApoE−/− mice at 24 h after the last of 5 i.t. instillations with a total dose of 128 μg/mouse. The concentrations of cytokines in BALF of the control mice were higher at 24 h after the last of the 5 i.t. instillations compared with values in the mice that were sacrificed at day 28 after the last exposure or 24 h after a single i.t. instillation. It indicates that the multiple i.t. instillations have evoked an inflammation reaction, which could be related to serum components in the instillation vehicle. The exposure to NM400 generated higher induction of IL6 in WT and ApoE−/− mice compared with NM402 (p < .05). In WT mice, NM400 also generated higher levels of G-CSF, KC, and CCL2 compared with NM402 (p < .05). The cytokine levels in BALF at day 28 after the last i.t. instillation are shown in Table 5. The levels of IL1β, IL6, KC, CCL2, MIP1β, and RANTES were more increased in ApoE−/− mice that were exposed to NM400 compared with NM402 (p < .05). This was also pronounced in WT mice where all cytokines were elevated in mice exposed to NM400 compared with NM402 (p < .05). Also, in WT mice instilled only a single time with 128 μg/mouse of MWCNT, the cytokine levels were generally elevated (Supplementary Table 2). The cytokine profile in serum is reported in Supplementary Table  3. There were no differences in serum concentrations of cytokines in WT mice at 24 h after a single i.t. instillation to 128  μg/mouse of MWCNT. The exposure to NM400 or

Downloaded from http://toxsci.oxfordjournals.org/ at Universidade de Vigo on May 7, 2014

Size in water   Mean ± SD   Median (10–90 percentile)  Mode Size in saline   Mean ± SD   Median (10–90 percentile)  Mode Size in RPMI medium   Mean ± SD   Median (10–90 percentile)  Mode

NM400

epithelial cells (p > .05) were unaltered, whereas the levels of eosinophils were decreased in WT mice exposed to 128  μg/ mouse of NM400 (p < .05) or NM402 (p .05 for interaction between strain and dose). The 2 types of MWCNT elicited the same level of influx of neutrophils at 24 h (p > .05) and day 28 (p > .05) after the last i.t. instillation. In a separate experiment, we assessed the effect of a single dose of MWCNT in WT mice with the same cumulated dose as the repeated experiment study (Supplementary Table 1). The influx of neutrophils increased dose dependently for both NM400 and NM402 (p < .05), whereas there was no difference in responses between the MWCNT (p > .05). The 5 i.t. instillations with 2% mouse serum (from sibling mice) were associated with elevated levels of eosinophils in the BALF (Table 2), which was not observed after single i.t. instillation (Supplementary Table 1) or in BALF from mice at day 28 after the last i.t. instillation (Table 3). The effect on eosinophil response is possibly evoked by multiple instillations of small amounts of organic material in the lung, whereas it is unlikely that it was caused by infections because the single exposure study was conducted at the same time as the repeated exposure studies.

109

Vascular Effects of Carbon Nanotubes



Table 2 Cells (× 105) in BALF at 24 h After i.t. Instillation of MWCNT Once a Week for 5 Weeks Group (n)

Dose (μg)

Neutrophils

Macrophages

Lymphocytes

Eosinophils

Epithelial Cells

0 32 128 32 128

1.37 (0.24) 3.2 (0.7) 1.9 (0.3) 4.8 (0.8) 3.2 (0.7)*

4.1 (11.2) 1.9 (0.5) 2.0 (0.3) 4.1 (1.4) 2.7 (0.3)

1.25 (0.29) 0.7 (0.3) 0.3 (0.1)* 1.7 (0.7) 0.4 (0.2)*

16.3 (2.4) 4.9 (2.3) 0.3 (0.1)* 14.3 (4.2) 4.3 (3.2)*

0.6 (0.2) 0.3 (0.1) 0.4 (0.1) 1.0 (0.2) 0.5 (0.1)

0 128 128

2.99 (1.42) 4.58 (0.65)* 4.77 (0.78)*

5.2 (1.3) 3.7 (1.1) 2.9 (0.4)

2.2 (0.65) 0.93 (0.37)* 1.67 (0.48)

19.7 (5.4) 9.6 (1.6) 8.1 (1.6)

0.62 (0.30) 0.59 (0.20) 0.55 (0.16)

Wild type   Control (11)   NM400 (5)   NM400 (6)   NM402 (6)   NM402 (6) ApoE−/−   Control (6)   NM400 (6)   NM402 (6)

Table 3 Cells (× 105) in BALF at Day 28 After 5 i.t. Instillations of MWCNT Dose (n)

Dose (μg)

Neutrophils

Macrophages

Lymphocytes

Eosinophils

Epithelial Cells

0 32 128 0 32 128

0.07 (0.03) 0.54 (0.15)* 1.42 (0.16)* 0.06 (0.01) 1.13 (0.19)* 1.59 (0.41)*

2.00 (0.42) 1.54 (0.13) 3.03 (0.31) 1.77 (0.13) 2.47 (0.36) 2.28 (0.36)

0.30 (0.11) 0.14 (0.04) 0.63 (0.13)* 0.28 (0.05) 1.26 (0.16)* 0.73 (0.15)*

0.20 (0.11) 0.06 (0.01) 0.02 (0.02)* 0.03 (0.01) 0.31 (0.21) 0.16 (0.06)

0.26 (0.03) 0.32 (0.04) 0.39 (0.06) 0.28 (0.02) 0.35 (0.06) 0.52 (0.10)

0 32 128 0 32 128

0.18 (0.07) 0.84 (0.28)* 1.17 (0.16)* 0.07 (0.02) 0.73 (0.10)* 2.00 (0.31)*

2.30 (0.24) 2.30 (0.29) 2.00 (0.42) 2.08 (0.39) 2.85 (0.29) 1.89 (0.20)

0.33 (0.08) 0.45 (0.15) 1.00 (0.38) 0.49 (0.13) 0.88 (0.12)* 0.67 (0.10)

0.09 (0.02) 0.09 (0.02) 0.01 (0.01)* 0.06 (0.03) 0.10 (0.05) 0.06 (0.03)

0.33 (0.06) 0.25 (0.02) 0.34 (0.09) 0.38 (0.09) 0.39 (0.07) 0.27 (0.04)

NM400   WT (6)   WT (6)   WT (6)  ApoE−/− (10)  ApoE−/− (10)  ApoE−/− (9) NM402   WT (6)   WT (5)   WT (6)  ApoE−/− (10)  ApoE−/− (10)  ApoE−/− (9)

The results are reported as mean and SEM. *p < .05 compared with control (ANOVA).

Table 4 Cytokines in BALF at 24 h After Repeated i.t. Instillations of MWCNT Once a Week for 5 Weeks in WT or ApoE−/− mice Cytokine

IL1β IL4 IL6 IL12 IL13 G-CSF KC CCL2 MIP1β RANTES TNF

WT

WT

WT

ApoE−/−

ApoE−/−

ApoE−/−

Control

NM400

NM402

Control

NM400

NM402

291 (24.8) 62.5 (30.1) 66.9 (44.9) 831 (217) 1180 (417) 18.1 (2.2) 209 (74) 228 (25) 77.8 (1.9) 22.1 (4.3) 190 (8.4)

696 (38.1)* 95.0 (28.5) 997 (190)*,§ 4487 (981)* 618 (218) 404 (47)*,§ 3655 (421)*,§ 7850 (918)*,§ 632 (142)* 133 (21)* 644 (225)*

579 (74.4)* 98.6 (59.5) 560 (352)* 6563 (1542)* 1422 (661) 153 (39)* 2160 (644)* 4285 (1365)* 446 (133)* 113 (29)* 398 (164)*

160 (16.5)# 74.4 (29.0) 106 (42.3) 1186 (329) 1790 (336) 34.9 (4.7) 125 (33) 371 (112) 107 (18) 26.2 (10) 34.6 (6.7)

313 (20.3)* 137 (17.1) 2154 (458)*,§ 6765 (582)* 2053 (458) 204 (48)* 738 (125)* 10199 (968)* 221 (24)* 129 (19)* 152 (24)*

274 (20.3)* 102 (32.4) 451 (118)* 7257 (528)* 743 (102) 109 (29)* 437 (115)* 5141 (982)* 180 (38) 93.8 (17)* 95.7 (21)*

The total dose was 128 μg/mouse. The cytokines are reported in pg/ml. *p < .05 compared with control in the same strain. §p < .05 compared with NM402 in the same strain. #p < .05 compared with WT mice.

Downloaded from http://toxsci.oxfordjournals.org/ at Universidade de Vigo on May 7, 2014

The statistical analysis included mice that were exposed to 128 μg in order to assess interactions between particle exposure and strain. The results are reported as mean and SEM. *p < .05 compared with controls (ANOVA).

110

Cao et al.

Table 5 Cytokines in BALF at Day 28 After Repeated i.t. Instillations of MWCNT Once a Week for 5 Weeks in WT or ApoE−/− mice Cytokine

WT

WT

ApoE−/−

ApoE−/−

ApoE−/−

Control

NM400

NM402

Control

NM400

NM402

549 (33)*,§ 39.4 (7.5)*,§ 272 (56)*,§ 7608 (406)*,§ 679 (74)*,§ 124 (16)*,§ 4936 (613)*,§ 19682 (5354)*,§ 478 (70)*,§ 273 (51)*,§ 486 (69)*,§

373 (53) 33.7 (8.4) 13.0 (5.5) 3730 (928)* 343 (144)* 51.0 (23)* 1850 (516)* 3394 (1362)* 252 (147) 74.8 (27)* 203 (139)*

  378 (40) < 25.3   < 5.0   473 (103)   118 (18)   17.5 (2.1)   61.7 (25) 236 (56)   88.9 (6.3)   15.0 (0.4)   13.5 (5.0)

397 (16) < 25.3   < 5.0   445 (87)   150 (26) 14.7 (0.3) 29.1 (1.1)   248 (22) 89.2 (6.2) 15.6 (0.6) 20.4 (2.5)  

529 (54)*,§ < 25.3 293 (68)*,§ 6133 (820)* 445 (50)* 103 (30)* 2974 (423)*,§ 10826 (2266)*,§ 141 (22)*,§ 265 (61)*,§ 149 (40)*

  403 (46) < 25.3 23.7 (5.3)   4902 (402)*   384 (47)* 50.8 (6.1)*   1907 (205)*   3628 (703)* 98.2 (11)   136 (15)* 90.3 (17)*

The total dose was 128 μg/mouse. The cytokines are reported in pg/ml. *p < .05 compared with control in the same strain. §p < .05 compared with NM400 in the same strain.

NM402 was associated with increased levels of KC in serum at 24 h after the last of 5 i.t. instillations in ApoE−/− mice, whereas there were no significant differences in WT mice (p < .05 for interaction between strain of mice and particle exposure). There were also increased serum concentrations of IL1β, IL4, IL6, G-CSF, CCL2, RANTES, and TNF, whereas the levels of IL12, IL13, and MIP1β were decreased in MWCNT-exposed ApoE−/− mice, although none of these differences were statistically significant. The cytokine concentrations of IL4, CCL2, and MIP1β were higher in ApoE−/− mice compared with WT mice (p < .05 for single-factor effect of the strain). These differences were not observed in the samples from mice that were sacrificed at day 28 after the exposure where only the levels of CCL2 and RANTES were increased in the ApoE−/− mice compared with the controls (p < .05, single-factor effect of strain) and there were no effects of the exposure to MWCNT (p > .05). 8-Isoprostanes in lung tissue, BALF, and serum.  The levels of 8-isoprostanes in lung tissue were increased at 24 h after the last of 5 i.t. instillations to NM400 and NM402 in WT and ApoE−/− mice (p < .05; Fig.  3). In contrast, the levels of 8-isoprostanes were the same between MWCNT-exposed mice and controls at day 28 after the 5 i.t. instillations (p > .05). Similarly, in the WT mice, the levels of 8-isoprostanes after a single i.t. instillation of NM400 (1368 ± 463 pg/ml, n = 6) and NM402 (800 ± 69 pg/ml, n = 6) were not statistically significantly different from the levels in the controls (1034 ± 146 pg/ml, n = 6; p > .05). The concentrations of 8-isoprostanes were also measured in BALF and serum after a single i.t. instillation (Supplementary Table 4). The levels were typically several orders of magnitude lower than the values in lung tissue. In BALF, there seemed to be decreased levels of 8-isoprostanes after exposure to MWCNT, but this was mainly driven by substantially increased protein levels causing the ratio to decline. Overall, there was not convincing evidence for increased level of 8-isoprostanes in

BALF. The levels in serum were unaltered in the groups of mice that were sacrificed at 24 h after 1 or repeated i.t. instillations. DNA damage.  The level of DNA strand breaks increased in the lung tissue at 24 h after the last exposure to 5 i.t. instillations of MWCNT (p < .01), whereas the level of FPG sensitive sites was unaltered in the MWCNT-exposed mice compared with the controls (Fig. 4). Gene expression.  The expression of vascular adhesion molecules (Vcam1), growth factor (Vegfa), inflammation (Ccl2 and Nos2), oxidative stress (Hmox1), and DNA repair (Ogg1) was assessed in lung and liver tissue of ApoE−/− mice that had been sacrificed at 24 h after the last of 5 i.t. instillations (Table 6). There was increased expression of all the genes in the lungs of MWCNT-exposed mice. In addition, the exposure to NM402 was associated with higher gene expression in the lungs compared with NM400. The gene expression in the liver was modest compared with the responses in the lungs. The exposure to NM400 and NM402 increased the inflammatory response in the liver in terms of Nos2 and Ccl2. There were also modest but significant increases in the Ogg1 and Vcam1 expression in the liver of mice that had been exposed to NM402. In Vitro Results Cytotoxicity and cell proliferation.  We observed a concentration-dependent decrease in viability in HUVEC that were exposed to MWCNT for 24 h (Supplementary Figure  4). The lowest concentrations that were associated with decreased succinate dehydrogenase activity in the HUVEC were 32 and 64  μg/ml for the NM402 (p < .001) and NM400 (p < .01), respectively. There was also a gradual decline in the succinate dehydrogenase activity in THP-1 cells, which reached statistically significantly effect at low concentrations (2 and 4  μg/ml for NM400 and NM402,

Downloaded from http://toxsci.oxfordjournals.org/ at Universidade de Vigo on May 7, 2014

IL1β IL4 IL6 IL12 IL13 G-CSF KC CCL2 MIP1β RANTES TNF

WT

Vascular Effects of Carbon Nanotubes



111

Table 6 Levels of mRNA in the Liver and Lungs of ApoE−/− Mice at 24 h After i.t. Instillation of MWCNT Once a Week for 5 Weeks Gene Expression Levels Lung  Ccl2  Hmox1  Nos2  Ogg1  Vcam1  Vegfa Liver  Ccl2  Hmox1  Nos2  Ogg1  Vcam1  Vegfa

Control

NM400

NM402

32.2 (6.0) 2.5 (0.12) 42.0 (8.4) 0.26 (0.06) 1.14 (0.23) 13.6 (5.4)

781 (316)* 8.0 (1.29)* 120 (38)* 0.71 (0.10)* 2.1 (0.31)* 33.3 (14.1)*

1489 (452)* 15.3 (1.9)*,# 260 (45)*,# 1.60 (0.19)*,# 4.9 (0.87)*,# 68.3 (18.3)*,#

0.90 (0.15) 37.9 (5.7) 73.9 (15.7) 1.00 (0.08) 8.32 (0.77) 20.2 (3.2)

5.77 (0.26)* 42.5 (4.5) 149.5 (21.0)* 1.21 (0.16) 8.68 (4.00) 16.0 (7.2)

7.48 (0.50)* 58.8 (11.5) 146.2 (25.8)* 1.63 (0.30)* 10.1 (1.56)* 13.2 (5.78)

The total dose was 128 μg/ml. The p values correspond to ANOVA with post hoc LSD test. The results are expressed relative to 18S per 106, except for Ccl2 in the liver and Nos2 in the liver and lungs that are expressed relative to 18S per 109. The results are mean and SEM, n = 5–6 mice/group. Abbreviation: LSD, least significant difference. *p < .05 compared with control. #p < .05 compared with NM400.

Fig. 4.  Levels of DNA strand breaks (A) and FPGss (B) in lung tissue at 24 h after the last exposure in ApoE−/− mice that have been exposed to MWCNT by i.t. instillation once a week for 5 weeks. The total dose was 128 μg/mouse. The results are means and SEM (n = 4–6 mice/group). *p < .01 compared with the control. Abbreviations: FPGss, FPG sensitive sites; MWCNT, multiwalled carbon nanotube.

respectively). The succinate dehydrogenase activity was reduced with 20% and 30% in THP-1 cells in the concentration range of 4–64 μg/ml. The assessment of succinate dehydrogense activity in THP-1a cells that had been exposed to free fatty acids and MWCNT showed no synergistic effect on cytotoxicity (p > .05 for the interaction between particles and free fatty acids), whereas the MWCNT induced a slightly increased cytotoxicity. The exposure to NM400 and NM402 was associated with 20% (p < .01) and 14% (p < .05) decreased succinate dehydrogense activity in THP-1a cells (Supplementary Figure 5). The exposure to MWCNT did not affect the proliferation of HUVEC, assessed as incorporation of BrdU (p > .05), whereas the exposure to 32 μg/ml of MWCNT was associated with increased BrdU incorporation in THP-1 cells (p < .01; Supplementary Figure  6). However, the incubation of THP-1 cells with 32  μg/ml MWCNT did not affect number of cells during a 48-h exposure period (Supplementary Figure 7).

Downloaded from http://toxsci.oxfordjournals.org/ at Universidade de Vigo on May 7, 2014

Fig. 3.  Levels of 8-isoprostanes in lung tissue from WT and ApoE−/− mice that have been exposed to MWCNT by i.t. instillation once a week for 5 weeks and sacrificed at 24 h (A) or 28 days (B) after the last exposure. The total dose was 128 μg/mouse. The results are means and SEM (n = 6–10 mice/group). *p < .05 compared with the control. Abbreviations: MWCNT, multiwalled carbon nanotube; WT, wild type.

112

Cao et al.

A collected analysis of the responses in HUVEC and THP-1 cells indicated that the exposure to MWCNT in the concentration range of 0–64 μg/ml was not associated with cytotoxic responses. ROS production.  The ROS production ability of the MWCNT in cell-free dispersions was clearly bell shaped with increased generation in the dose range 4–32  μg/ml (Supplementary Figure  8). The same effect was also observed for the cellular ROS production in THP-1 cells and HUVEC (Fig. 5). There was increased ROS production in HUVEC after exposure to NM400 (2–8  μg/ml, p < .05) and NM402 (2–16  μg/ml, p < .05). The exposure to NM400 (2–8 μg/ml, p < .05) and NM402 (2–16 μg/ ml, p < .05) also increased the ROS production in THP-1 cells.

Expression of THP-1 cells.  The exposure of THP-1 cells to NM400 or NM402 for 3 h did not alter the gene expression level of CCL2, IL8, TNF, or HMOX1 (Supplementary Figure 9).

Fig. 5.  ROS production in THP-1 cells (A) and HUVEC (B) after exposure to MWCNT for 3 h. p < .05 compared with controls. Abbreviations: HUVEC, human umbilical vein endothelial cell; MWCNT, multiwalled carbon nanotube; ROS, reactive oxygen species.

Attachment of THP-1 cells onto HUVEC.  The results on the attachment of MWCNT into HUVEC are shown in Figure 7. This analysis indicated that the exposure to MWCNT caused increased adherence of THP-1 cells onto HUVEC at 16 (p < .01) and 32 μg/ ml (p < .01), whereas there was no difference between the potency of the particles (p > .05, single-factor effect of particles).

As an independent confirmatory assay, we also determined the lipid accumulation by wide-field microscopy. This indicated that the lipids accumulated in the cytoplasm of the cells (Supplementary Figure 11).

Lipid accumulation in THP-1a cells.  The results on the lipid accumulation in THP-1a cells are shown in Figure  8. There was a 3-factor interaction between the concentration of particles, presence of free fatty acids, and NAC (p < .01). The level of lipid accumulation was observed in the cells that were treated with free fatty acids and MWCNT. These cultures had increased lipid levels after exposure to 2–16 μg/ml of MWCNT compared with the controls. The cotreatment with NAC reduced this accumulation of lipids in THP-1a cells. In addition, there was a single-factor effect of the type of particle (p < .001); on average, the cultures that were exposed to NM402 and free fatty acids had 22% (95% CI: 3%–43%) higher lipid content than the cultures that were exposed to NM400 and free fatty acids. The treatment with free fatty acids was not associated with lipid accumulation in this experimental setting; this was because we had selected an exposure period that did not increase the lipid level, whereas a 24-h incubation period increased the lipid load in THP-1a cells (Supplementary Figure 10).

Discussion

There is accumulating evidence that pulmonary exposure to MWCNT is associated with lung inflammation, whereas the effect on the vascular system still remains to be investigated beyond the first observations of effect. We firstly showed that pulmonary exposure to 2 different types of MWCNT caused progression of atherosclerosis with assessment of mechanistic endpoints in relation to oxidative stress and inflammation and subsequently explored the effect in vitro of MWCNT exposure in cultured endothelial and monocytic cells. The in vivo findings indicated that the 2 MWCNTs evoked pulmonary neutrophilic influx in WT and ApoE−/− mice and in different dosing regimes. Curiously, the repeated instillation was associated with influx of eosinophils and high levels of IL4 and IL13 in BALF. This may have been caused by repeated exposure to the small amounts of organic material in the serum. The lung response to MWCNT was thus driven from the classical

Downloaded from http://toxsci.oxfordjournals.org/ at Universidade de Vigo on May 7, 2014

Expression of cell adhesion molecules.  The results on the expression of cell adhesion molecules are shown in Figure 6. The exposure to 64  μg/ml of NM400 was associated with increased expression of ICAM1 (p < .05) and VCAM1 (p < .001). The exposure to NM402 was associated with increased expression of ICAM1 at concentrations of 16 (p < .05) and 32 μg/ml (p < .01) of NM402. The VCAM1 was increased at concentrations of 16 (p < .001), 32 (p < .001), and 64 μg/ml (p < .001) of NM402. In addition, the NM402 evoked higher expression of VCAM1 in HUVEC at 32 (p < .05) and 64 μg/ml (p < .001) than did the NM400.



Vascular Effects of Carbon Nanotubes

Fig. 7.  Attachment of THP-1 cells onto HUVEC after 24-h exposure to MWCNT. p < .05 compared with controls. *p < .001 compared with control. Abbreviations: HUVEC, human umbilical vein endothelial cell; MWCNT, multiwalled carbon nanotube.

Th1-dominated aseptic inflammation to a Th2 driven one. The impact of the eosinophilic response on the MWCNT-mediated pulmonary oxidative stress response and systemic effects is uncertain but deserve further investigation. Nevertheless, the levels of cytokines in the BALF were highest for NM400 exposed mice, whereas NM402 elicited the strongest effect on the gene expression related to oxidative stress and vascular activation

in the lung. The exposure to NM402 was also associated with increased level of DNA strand breaks in the lung tissue. This is in keeping with oxidative stress because some types of ROS (eg, H2O2) generate strand breaks. The levels of FPG sensitive sites were unaltered, which could be related to OGG1-mediated repair of oxidized purine lesions in DNA as supported by the increased expression level of Ogg1. This is in keeping with earlier studies showing inverse associations between oxidatively damaged DNA and OGG1-mediated DNA repair in the lungs after pulmonary exposure to diesel exhaust particles (Risom et al., 2003, 2007). A similar level of daily exposure to SWCNT (0.5 mg/kg administered at 26 and 2 h before sacrifice) was not associated with increased levels of FPG sensitive sites in the lungs (Vesterdal et  al., 2012). Nevertheless, there was increased level of 8-isoprostanes in lung tissue at 24 h after the last of 5 MWCNT exposures in the WT and ApoE−/− mice. The level of 8-isoprostanes and inflammation were unaltered at day 28 after the last exposure. The MWCNTs are most likely still present in the lungs at the late time point (Ryman-Rasmussen et al., 2009), although it can be speculated that they are less hazardous because of protein coating. Alternatively, it has been shown that MWCNTs are predominantly located in alveolar macrophages at day 1 after a single pharyngeal aspiration of 80 μg/mouse (Mercer et al., 2010). Collectively, the i.t. instillation exposure to both MWCNT was associated with inflammation and oxidative stress in the lungs with a possible stronger oxidative stress response to NM402. The exposure to MWCNT was associated with almost a doubling of the aortic lesion area, which is essentially the same level of response as seen after the exposure to SWCNT in ApoE−/− mice by i.t. instillation (Li et al., 2007). Still, despite persistent pulmonary inflammation, the cessation of exposure to NM400 seemed to normalize aortic plaque progression, whereas 28 days after the exposure to NM402, there was statistically nonsignificantly elevated plaque area level. The results of the MWCNT-mediated plaque progression in the present study are supported by findings from IV administration of MWCNT in rats where 0.2 mg/kg twice weekly for 4 months increased the level of atherosclerosis and calcification in aorta, whereas lower doses (0.05 or 0.1 mg/kg) had no effect on plaque progression (Xu et al., 2012). This dose corresponds to a plasma concentration of 3 μg/ml, assuming that male Sprague Dawley rats have a plasma volume of 7 ml/100 g. This is similar to the concentration that produced the highest level of lipid accumulation in foam cells in our in vitro study. It is yet unresolved how pulmonary exposure to particles generates vascular effects. Direct translocation of particles is low and systemic effects may thus come about by inflammation and oxidative stress in the lung (Møller et al., 2011). We observed systemic effects in terms of increased expression of genes related to inflammation responses and oxidative stress in the liver in particular related to NM402. The mainly unaltered serum levels of cytokines and 8-isoprostanes suggest that these biomarkers may not be sufficiently sensitive to detect inflammation and oxidative stress in the circulation of MWCNT-exposed mice.

Downloaded from http://toxsci.oxfordjournals.org/ at Universidade de Vigo on May 7, 2014

Fig. 6.  Surface expression of ICAM1 (A) and VCAM1 (B) on HUVEC after 24-h exposure to MWCNT. *p < .05 compared with controls, #p < .05 for difference between particles. Abbreviations: HUVEC, human umbilical vein endothelial cell; ICAM1, intercellular adhesion molecule 1; MWCNT, multiwalled carbon nanotubes; VCAM1, vascular adhesion molecule 1.

113

114

Cao et al.

It has been shown that the exposure to MWCNT stimulated platelet aggregation in vitro and thrombosis in the carotid artery of rats after IV injection of 25 μg/rat in the femoral vein (Radomski et al., 2005). Pulmonary exposure to MWCNT was associated with impairment of coronary arteriolar vasodilatation response (Stapleton et al., 2012). Another study showed that a single oropharyngeal aspiration of MWCNT at doses of 0.1–100 μg/mouse was associated with pulmonary inflammation and increased susceptibility to cardiac ischemia/reperfusion injury in mice, whereas there was no systemic inflammation (Urankar et  al., 2012). This is in keeping with our observations that plaque progression occurred without substantially increased inflammation

in the circulation. The same researchers have also looked into the effect of ischemia/reperfusion injury in rats, showing increased infarct size at a dose of 100 μg/rat, whereas there were no effects at doses that were 1 or 2 orders of magnitude lower (Thompson et  al., 2012). Collectively, the results indicate that exposure to MWCNT is associated with cardiovascular diseases in animals by mechanisms encompassing atherosclerosis and thrombosis, conferring increased risk of atherothrombosis and susceptibility to ischemic/reperfusion injury. Accelerated progression of plaques may be due to increased atherogenesis or faster accumulation of lipids in foam cells of fatty streaks or atheromas. We used cell culture models

Downloaded from http://toxsci.oxfordjournals.org/ at Universidade de Vigo on May 7, 2014

Fig. 8.  Accumulation of lipids in THP-1a cells that have been exposed to MWCNT for 18 h with or without the presence of NAC and subsequently incubated with free fatty acids (black columns) or media (white columns) for 3 h. *p < .05 compared with the controls with the same treatment. #p < .05 compared with cells that have not been exposed to NAC. Abbreviations: MWCNT, multiwalled carbon nanotube; NAC, N-acetylcysteine.



115

Vascular Effects of Carbon Nanotubes

endothelial cells and oxidative stress-mediated transformation of monocytes to foam cells. The declared longer of the 2 MWCNTs (NM402) appeared to give stronger responses in terms of oxidative stress and gene expression in lung and liver tissue, adhesion molecule expression and lipid uptake in foam cells supporting a possible more persistent accelerated plaque progression. Supplementary Data

Supplementary data are available online at http://toxsci. oxfordjournals.org/. Funding

The European Union supported FP7 ENPRA (NMP4-SL-2009-228789); the Danish Research Councils (2052-03-0016); the Center for Pharmaceutical Nanotechnology and Nanotoxicology (Strategic Research Council 2106-080081); the Lundbeck Foundation Center for Biomembranes in Nanomedicine and Danish Centre for Nanosafety (20110092173/3) from the Danish Working Environment Research Foundation. References Beck-Speier, I., Karg, E., Behrendt, H., Stoeger, T., and Alessandrini, F. (2012). Ultrafine particles affect the balance of endogenous pro- and anti-inflammatory lipid mediators in the lung: In-vitro and in-vivo studies. Part. Fibre Toxicol. 9, 27. Danielsen, P. H., Loft, S., Kocbach, A., Schwarze, P. E., and Møller, P. (2009). Oxidative damage to DNA and repair induced by Norwegian wood smoke particles in human A549 and THP-1 cell lines. Mutat. Res. 674, 116–122. Forchhammer, L., Johansson, C., Loft, S., Möller, L., Godschalk, R. W., Langie, S. A., Jones, G. D., Kwok, R. W., Collins, A. R., Azqueta, A., et al. (2010). Variation in the measurement of DNA damage by comet assay measured by the ECVAG inter-laboratory validation trial. Mutagenesis 25, 113–123. Forchhammer, L., Loft, S., Roursgaard, M., Cao, Y., Riddervold, I. S., Sigsgaard, T., and Møller, P. (2012). Expression of adhesion molecules, monocyte interactions and oxidative stress in human endothelial cells exposed to wood smoke and diesel exhaust particulate matter. Toxicol. Lett. 209, 121–128. Frikke-Schmidt, H., Roursgaard, M., Lykkesfeldt, J., Loft, S., Nøjgaard, J. K., and Møller, P. (2011). Effect of vitamin C and iron chelation on diesel exhaust particle and carbon black induced oxidative damage and cell adhesion molecule expression in human endothelial cells. Toxicol. Lett. 203, 181–189. Fubini, B., Ghiazza, M., and Fenoglio, I. (2010). Physico-chemical features of engineered nanoparticles relevant to their toxicity. Nanotoxicology 4, 347–363. Götz, A. A., Rozman, J., Rödel, H. G., Fuchs, H., Gailus-Durner, V., Hrabě de Angelis, M., Klingenspor, M., and Stoeger, T. (2011). Comparison of particle-exposure triggered pulmonary and systemic inflammation in mice fed with three different diets. Part. Fibre Toxicol. 8, 30. Harding, A. H., Darnton, A., and Osman, J. (2012). Cardiovascular disease mortality among British asbestos workers (1971-2005). Occup. Environ. Med. 69, 417–421. Hemmingsen, J. G., Møller, P., Nøjgaard, J. K., Roursgaard, M., and Loft, S. (2011). Oxidative stress, genotoxicity, and vascular cell adhesion molecule expression in cells exposed to particulate matter from combustion of

Downloaded from http://toxsci.oxfordjournals.org/ at Universidade de Vigo on May 7, 2014

to investigate the effect of direct exposure of MWCNT to endothelial and foam cells. The MWCNT exposures of the cells were associated with cytotoxicity at concentrations above 64 μg/ml. Below this concentration, there was increased ROS production in acellular conditions and within cells with peak levels at 2–8 μg/ml of MWCNT. The bell-shaped curves for the ROS production at high concentration may be due to destruction of the probe or quenching of its fluorescence, although it has also been described that some types of MWCNT have free radical scavenging capacity (Fubini et al., 2010). Importantly, the exposure to MWCNT, in particular NM402, rendered foam cells more apt to accumulate microvesicular lipid droplets in the cytoplasma. The concentration curve for the lipid accumulation resembled that of the ROS production with peak levels at concentrations of 2–4 μg/ml. This accumulation of lipids was dependent on oxidative stress because a concomitant treatment with NAC ameliorated the lipid accumulation. We also investigated the effect of MWCNT exposure on the attachment of monocytes onto endothelial cells by binding to cell adhesion proteins. The exposure to both types of MWCNT was associated with increased attachment of THP-1 cells on HUVEC. This was supported by increased expression of ICAM1 and VCAM1 on HUVEC in particular for NM402. However, there was no effect on THP-1 cells, assessed as gene expression of IL8, TNF, CCL2, and HMOX1, despite increased ROS production. It has been speculated that CNTs function as molecular bridges whereby they stimulate aggregation of platelets (Radomski et  al., 2005). Similarly, MWCNT might facilitate the binding of monocytes onto activated endothelial, thus promoting atherogenesis. We did not pursue experiments of antioxidant treatments on the expression of ICAM1 and VCAM1 on HUVEC because we previously have observed that maintenance of intracellular ascorbate levels in HUVEC did not affect the particle-mediated expression of these adhesion molecules (Frikke-Schmidt et  al., 2011). This suggests that particle-induced expression of cell adhesion molecules is not directly linked to oxidative stress level. It has been shown that 1% of the applied dose of MWCNT translocated systemically at 24 h after a 12-day period, predominantly to lymph nodes (Mercer et al., 2013). It indicates that the MWCNT concentrations in the culture medium were well above the concentrations that are likely to occur in the circulation by translocation of fibers after airway exposure. Nevertheless, the cell culture experiments are useful as proof of principle, showing that MWCNTs have the ability to activate endothelial cells and stimulate accumulation of lipid in foam cells. A recent study has extended this concept in cocultures, showing that MWCNTexposed alveolar cells on Transwell membranes mediated activation of endothelial cells, which were not directly exposed to MWCNT (Snyder-Talkington et al., 2013). In summary, our results show that pulmonary exposure to MWCNT is associated with oxidative stress and inflammation in lungs and liver and progression of atherosclerosis that could be related to increased adherence of monocytes onto activated

116

Cao et al.

conventional diesel and methyl ester biodiesel blends. Environ. Sci. Technol. 45, 8545–8551. Jacobsen, N. R., Pojana, G., White, P., Møller, P., Cohn, C. A., Korsholm, K. S., Vogel, U., Marcomini, A., Loft, S., and Wallin, H. (2008). Genotoxicity, cytotoxicity, and reactive oxygen species induced by single-walled carbon nanotubes and C(60) fullerenes in the FE1-Mutatrade markMouse lung epithelial cells. Environ. Mol. Mutagen. 49, 476–487. Jacobsen, N. R., Møller, P., Jensen, K. A., Vogel, U., Ladefoged, O., Loft, S., and Wallin, H. (2009). Lung inflammation and genotoxicity following pulmonary exposure to nanoparticles in ApoE-/- mice. Part. Fibre Toxicol. 6, 2. Jantzen, K., Roursgaard, M., Desler, C., Loft, S., Rasmussen, L. J., and Møller, P. (2012). Oxidative damage to DNA by diesel exhaust particle exposure in co-cultures of human lung epithelial cells and macrophages. Mutagenesis 27, 693–701.

Li, Z., Hulderman, T., Salmen, R., Chapman, R., Leonard, S. S., Young, S. H., Shvedova, A., Luster, M. I., and Simeonova, P. P. (2007). Cardiovascular effects of pulmonary exposure to single-wall carbon nanotubes. Environ. Health Perspect. 115, 377–382. Ma-Hock, L., Treumann, S., Strauss, V., Brill, S., Luizi, F., Mertler, M., Wiench, K., Gamer, A. O., van Ravenzwaay, B., and Landsiedel, R. (2009). Inhalation toxicity of multiwall carbon nanotubes in rats exposed for 3 months. Toxicol. Sci. 112, 468–481. Mercer, R. R., Hubbs, A. F., Scabilloni, J. F., Wang, L., Battelli, L. A., Schwegler-Berry, D., Castranova, V., and Porter, D. W. (2010). Distribution and persistence of pleural penetrations by multi-walled carbon nanotubes. Part. Fibre Toxicol. 7, 28. Mercer, R. R., Scabilloni, J. F., Hubbs, A. F., Wang, L., Battelli, L. A., McKinney, W., Castranova, V., and Porter, D. W. (2013). Extrapulmonary transport of MWCNT following inhalation exposure. Part. Fibre Toxicol. 10, 38. Mikkelsen, L., Sheykhzade, M., Jensen, K. A., Saber, A. T., Jacobsen, N. R., Vogel, U., Wallin, H., Loft, S., and Møller, P. (2011). Modest effect on plaque progression and vasodilatory function in atherosclerosis-prone mice exposed to nanosized TiO(2). Part. Fibre Toxicol. 8, 32. Mitchell, L. A., Gao, J., Wal, R. V., Gigliotti, A., Burchiel, S. W., and McDonald, J. D. (2007). Pulmonary and systemic immune response to inhaled multiwalled carbon nanotubes. Toxicol. Sci. 100, 203–214. Møller, P., Danielsen, P. H., Jantzen, K., Roursgaard, M., and Loft, S. (2013). Oxidatively damaged DNA in animals exposed to particles. Crit. Rev. Toxicol. 43, 96–118. Møller, P., Mikkelsen, L., Vesterdal, L. K., Folkmann, J. K., Forchhammer, L., Roursgaard, M., Danielsen, P. H., and Loft, S. (2011). Hazard identification of particulate matter on vasomotor dysfunction and progression of atherosclerosis. Crit. Rev. Toxicol. 41, 339–368. Moore, K. J., and Tabas, I. (2011). Macrophages in the pathogenesis of atherosclerosis. Cell 145, 341–355. Pauluhn, J. (2010). Subchronic 13-week inhalation exposure of rats to multiwalled carbon nanotubes: Toxic effects are determined by density of agglomerate structures, not fibrillar structures. Toxicol. Sci. 113, 226–242. Przybytkowski, E., Behrendt, M., Dubois, D., and Maysinger, D. (2009). Nanoparticles can induce changes in the intracellular metabolism of lipids without compromising cellular viability. FEBS J. 276, 6204–6217.

Risom, L., Dybdahl, M., Bornholdt, J., Vogel, U., Wallin, H., Møller, P., and Loft, S. (2003). Oxidative DNA damage and defence gene expression in the mouse lung after short-term exposure to diesel exhaust particles by inhalation. Carcinogenesis 24, 1847–1852. Risom, L., Dybdahl, M., Møller, P., Wallin, H., Haug, T., Vogel, U., Klungland, A., and Loft, S. (2007). Repeated inhalations of diesel exhaust particles and oxidatively damaged DNA in young oxoguanine DNA glycosylase (OGG1) deficient mice. Free Radic. Res. 41, 172–181. Ryman-Rasmussen, J. P., Cesta, M. F., Brody, A. R., Shipley-Phillips, J. K., Everitt, J. I., Tewksbury, E. W., Moss, O. R., Wong, B. A., Dodd, D. E., Andersen, M. E., et  al. (2009). Inhaled carbon nanotubes reach the subpleural tissue in mice. Nat. Nanotechnol. 4, 747–751. Shvedova, A. A., Pietroiusti, A., Fadeel, B., and Kagan, V. E. (2012). Mechanisms of carbon nanotube-induced toxicity: Focus on oxidative stress. Toxicol. Appl. Pharmacol. 261, 121–133. Snyder-Talkington, B. N., Schwegler-Berry, D., Castranova, V., Qian, Y., and Guo, N. L. (2013). Multi-walled carbon nanotubes induce human microvascular endothelial cellular effects in an alveolar-capillary co-culture with small airway epithelial cells. Part. Fibre Toxicol. 10, 35. Stapleton, P. A., Minarchick, V. C., Cumpston, A. M., McKinney, W., Chen, B. T., Sager, T. M., Frazer, D. G., Mercer, R. R., Scabilloni, J., Andrew, M. E., et al. (2012). Impairment of coronary arteriolar endothelium-dependent dilation after multi-walled carbon nanotube inhalation: A time-course study. Int. J. Mol. Sci. 13, 13781–13803. Thompson, L. C., Frasier, C. R., Sloan, R. C., Mann, E. E., Harrison, B. S., Brown, J. M., Brown, D. A., and Wingard, C. J. (2014). Pulmonary instillation of multi-walled carbon nanotubes promotes coronary vasoconstriction and exacerbates injury in isolated hearts. Nanotoxicology. 8, 38–49. Urankar, R. N., Lust, R. M., Mann, E., Katwa, P., Wang, X., Podila, R., Hilderbrand, S. C., Harrison, B. S., Chen, P., Ke, P. C., et  al. (2012). Expansion of cardiac ischemia/reperfusion injury after instillation of three forms of multi-walled carbon nanotubes. Part. Fibre Toxicol. 9, 38. Vesterdal, L. K., Folkmann, J. K., Jacobsen, N. R., Sheykhzade, M., Wallin, H., Loft, S., and Møller, P. (2010). Pulmonary exposure to carbon black nanoparticles and vascular effects. Part. Fibre Toxicol. 7, 33. Vesterdal, L. K., Jantzen, K., Sheykhzade, M., Roursgaard, M., Folkmann, J. K., Loft, S., and Møller, P. (2014). Pulmonary exposure to particles from diesel exhaust, urban dust or single-walled carbon nanotubes and oxidatively damaged DNA and vascular function in apoE-/-mice. 8, 61–71. Vidanapathirana, A. K., Lai, X., Hilderbrand, S. C., Pitzer, J. E., Podila, R., Sumner, S. J., Fennell, T. R., Wingard, C. J., Witzmann, F. A., and Brown, J. M. (2012). Multi-walled carbon nanotube directed gene and protein expression in cultured human aortic endothelial cells is influenced by suspension medium. Toxicology 302, 114–122. Vietti, G., Ibouraadaten, S., Palmai-Pallag, M., Yakoub, Y., Bailly, C., Fenoglio, I., Marbaix, E., Lison, D., and van den Brule, S. (2013). Towards predicting the lung fibrogenic activity of nanomaterials: Experimental validation of an in vitro fibroblast proliferation assay. Part. Fibre Toxicol. 10, 52. Wang, X., Katwa, P., Podila, R., Chen, P., Ke, P. C., Rao, A. M., Walters, D. M., Wingard, C. J., and Brown, J. M. (2011). Multi-walled carbon nanotube instillation impairs pulmonary function in C57BL/6 mice. Part. Fibre Toxicol. 8, 24. Xu, Y. Y., Yang, J., Shen, T., Zhou, F., Xia, Y., Fu, J. Y., Meng, J., Zhang, J., Zheng, Y. F., Yang, J., et  al. (2012). Intravenous administration of multiwalled carbon nanotubes affects the formation of atherosclerosis in SpragueDawley rats. J. Occup. Health 54, 361–369.

Downloaded from http://toxsci.oxfordjournals.org/ at Universidade de Vigo on May 7, 2014

Kermanizadeh, A., Pojana, G., Gaiser, B. K., Birkedal, R., Bilanicova, D., Wallin, H., Jensen, K. A., Sellergren, B., Hutchison, G. R., Marcomini, A., et al. (2013). In vitro assessment of engineered nanomaterials using a hepatocyte cell line: Cytotoxicity, pro-inflammatory cytokines and functional markers. Nanotoxicology, 7, 693–701. Kosuge, H., Sherlock, S. P., Kitagawa, T., Dash, R., Robinson, J. T., Dai, H., and McConnell, M. V. (2012). Near infrared imaging and photothermal ablation of vascular inflammation using single-walled carbon nanotubes. J. Am. Heart Assoc. 1, e002568.

Radomski, A., Jurasz, P., Alonso-Escolano, D., Drews, M., Morandi, M., Malinski, T., and Radomski, M. W. (2005). Nanoparticle-induced platelet aggregation and vascular thrombosis. Br. J.  Pharmacol. 146, 882–893.