Nanoscale materials are presently under development for diagnostic (nanomedicine) and electronic purposes. In contrast to the potential benefits of nanotechnology, the effects of nanomaterials on human health are poorly understood. Nanomaterials are known to translocate into the circulation and could thus directly affect vascular endothelial cells (ECs), causing vascular injury that might be responsible for the development of atherosclerosis. To explore the direct effects of nanomaterials on endothelial toxicity, human umbilical vein ECs were treated with 1–100 μg/ml hydroxyl fullerene [C60(OH)24; mean diameter, 7.1 ± 2.4 nm] for 24 h. C60(OH)24 induced cytotoxic morphological changes such as cytosolic vacuole formation and decreased cell density in a dose-dependent manner. Lactate dehydrogenase assay revealed that a maximal dose of C60(OH)24 (100 μg/ml) induced cytotoxic injury. Proliferation assay also showed that a maximal dose of C60(OH)24 inhibited EC growth. C60(OH)24 did not seem to induce apoptosis but caused the accumulation of polyubiquitinated proteins and facilitated autophagic cell death. Formation of autophagosomes was confirmed on the basis of Western blot analysis using a specific marker, light chain 3 antibody, and electron microscopy. Chronic treatment with low-dose C60(OH)24 (10 μg/ml for 8 days) inhibited cell attachment and delayed EC growth. In the present study, we have examined, for the first time, the toxicity of water-soluble fullerenes to ECs. Although fullerenes changed morphology in a dose-dependent manner, only maximal doses of fullerenes caused cytotoxic injury and/or death and inhibited cell growth. EC death seemed to be caused by activation of ubiquitin-autophagy cell death pathways. Although exposure to nanomaterials appears to represent a risk for cardiovascular disorders, further in vivo validations are necessary.
- ubiquitin proteasome
the advent of nanoscale materials seems to offer marvelous opportunities for biomedical applications such as therapeutic and diagnostic tools as well as benefits in the fields of engineering, electronics, and optics (1, 2, 12). Biomedical applications under development include targeted drug delivery systems to the brain and cancer tissues and intravascular nanosensor and nanorobotic devices for imaging and diagnosis. However, little is known yet regarding the potential adverse effects or humoral immune responses after the introduction of such devices or nanoscale particulates into the organism (7, 19, 25).
Several pathways have been proposed for potential exposure of humans to nanomaterials (19). Whereas inhalation may be the major route of exposure, ingestion and dermal exposure are also possible during the manufacture, use, and disposal of engineered nanomaterials. Furthermore, intravenous, subcutaneous, or intramuscular administration is needed for therapeutic and diagnostic applications of nanotechnological devices. After inhaled nanoparticles are deposited in the respiratory tract, their small size promotes uptake into cells and transcytosis into the vasculature and lymphatics. Because nanoparticles are barely recognized by phagocytosing cells such as lung macrophages compared with microsized particles (2), uptake seems likely to occur via epithelial or endothelial cells (ECs).
According to epidemiological studies conducted in the United States and Europe (4, 10), modest increases in the mass of particulate matter are associated with increased duration of hospitalization and mortality as a result of cardiovascular disorders. Traffic-derived nanosized particles are most likely responsible for these cardiovascular actions, because the larger surface area per mass potentially leads to enhanced biological toxicity (2, 19). Because nanoparticulate air pollution is known to translocate into the vasculature (17, 18), direct effects of nanoparticles on the cardiovascular system could be one possible mechanism explaining these epidemiological findings. In this context, vascular endothelium could represent a primary target for nanomaterials after translocation into the circulation. We thus hypothesized that nanomaterials may directly interact with ECs to induce endothelial injury or cell death, promote thrombosis, and destabilize atheromatous plaques. In the present study, we focused particularly on the direct effects of fullerenes, one of the major nanomaterials, on endothelial injury and toxicity, using cultured vascular ECs to explore the possibilities of cellular toxicity leading to cardiovascular disease.
MATERIALS AND METHODS
Hydroxyl fullerene [C60(OH)24; Tokyo Progress System, Tokyo, Japan] was suspended in culture medium by sonication and vortexing. Particle size was measured using a particle size analyzer (model UPA-EX150; Nikkiso, Tokyo, Japan), revealing a mean ± SD diameter of 7.1 ± 2.4 nm after filtration (50% mass accumulation). Antibody sources were as follows: total actin (Santa Cruz Biotechnology, Santa Cruz, CA), poly(ADP-ribose) polymerase (PARP; BD Biosciences, San Jose, CA), cleaved caspase-3 (Cell Signaling Technology, Beverly, MA), and ubiquitin (Chemicon International, Temecula, CA). Rabbit PAb against light chain (LC)3 was kindly provided by Dr. T. Yoshimori (National Institute of Genetics, Mishima, Japan).
Human umbilical vein ECs (HUVECs) were purchased from Cascade Biologics (Portland, OR) and cultured in Medium 200 supplemented with low-serum growth supplement (LSGS; Cascade Biologics) as described previously (30). Cells were used at passages 3–6 for experiments.
Samples were fixed in 0.1 M sodium cacodylate-buffered (pH 7.4) 2.0% glutaraldehyde solution at 4°C overnight and postfixed in 0.1 M sodium cacodylate-buffered (pH 7.4) 1% OsO4 solution at 4°C for 2 h. After dehydration in an ethanol gradient (50–100% for 10 min each), samples were embedded in EPON812 at 60°C for 2 days. Ultrathin sections (80 nm) were stained using uranyl acetate and lead citrate. Sections were examined using an electron microscope (model JEM2000EX; JEOL, Tokyo, Japan) at 100 kV.
LDH cytotoxicity assay.
Cytotoxicity assay was performed using a CytoTox 96 nonradioactive cytotoxicity assay kit (Promega, Madison, WI) in accordance with the instructions of the manufacturer. Briefly, after treating HUVECs at ∼90% confluence in six-well plates with C60(OH)24 (1–100 μg/ml) for 24 h, culture medium was collected. Lactate dehydrogenase (LDH), a stable cytosolic enzyme released during cell lysis, was measured at 490-nm absorbance using a standard 96-well plate reader. Cytotoxicity was expressed relative to basal LDH release in untreated control cells.
Proliferation assay was performed using a Cell Counting-8 kit (Dojindo Laboratories, Kumamoto, Japan) according to the instructions of the manufacturer. Briefly, after treating HUVECs at ∼30% confluence in 12-well plates with C60(OH)24 (1–100 μg/ml) for 24 h, water-soluble tetrazolium salt (WST-8) was added for 3 h and culture medium was collected. Conversion of WST-8 into formazan by living cells (active mitochondria) was measured using a standard 96-well plate reader at 450-nm absorbance. Total numbers of living cells were compared with untreated control samples.
Western blot analysis.
Western blot analysis was performed as described previously (29). Proteins were obtained by homogenizing HUVECs with Triton X-based lysis buffer (1% Triton X-100, 20 mM Tris, pH 7.4, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 2.5 mM sodium pyrophosphate, 1 mM β-glycerol phosphate, 1 mM Na3VO4, 1 μg/ml leupeptin, and 0.1% protease inhibitor mixture; Nacalai Tesque, Kyoto, Japan). Protein concentration was determined using the bicinchoninic acid method (Pierce, Rockford, IL). Equal amounts of proteins (15 μg) were separated by SDS-PAGE (7.5% and 14%) and transferred onto nitrocellulose membrane (Pall, Ann Arbor, MI). After being blocked with 3% BSA, membranes were incubated with primary antibody (1:1,000 dilution) at 4°C overnight and membrane-bound antibodies were visualized using horseradish peroxidase-conjugated secondary antibodies (1:10,000 dilution, 1 h) and the ECL system (Amersham Biosciences, Little Chalfont, UK). Experiments were performed three or more times, and equal loading of protein was ensured by measuring total actin expression.
Activity assay for 20S proteasome.
Activity of the 20S proteasome was determined using the 20S proteasome assay kit (Calbiochem, San Diego, CA) according to the instructions of the manufacturer. Cells were lysed in Triton X-based lysis buffer as described above. The assay mixture contained 178 μl of reaction buffer (25 mM HEPES and 0.5 mM EDTA, pH 7.6), 10 μl of substrate [10 μM Suc-Leu-Val-Tyr-7-amino-4-methylcoumarin (AMC)], 2 μl of SDS (0.03%), and 10 μl of cell lysate (10 μg of protein). After incubation for 30 min at 37°C, the fluorescence of liberated AMC was measured using excitation and emission wavelengths at 340 and 450 nm, respectively.
Total RNA was isolated from HUVECs treated with or without C60(OH)24 (100 μg/ml, 24 h) using an RNeasy Kit (Qiagen, Valencia, CA) according to the instructions of the manufacturer. Only samples with A260/A280 between 1.7 and 2.2 (measured in 10 mM Tris·HCl, pH 7.6) were considered suitable for use. Hybridization samples were prepared according to the GeneChip Expression Analysis Technical Manual (701021, Rev. 5, section 2, “Eukaryotic Sample and Array Processing,” chapt. 1, “Eukaryotic Target Preparation”; http://www.affymetrix.com/support/technical/manuals.affx). Total RNA (2 μg) was amplified for each sample. Next, cRNA (30 μg) was fragmented in 40 μl of 1× fragmentation buffer. Hybridization cocktails were made as described in the GeneChip Expression Analysis Technical Manual (701021, Rev. 5, section 2, chapt. 2, “Eukaryotic Target Hybridization”) and hybridized to human genome U133 plus2.0 chips at 60 rpm and 45°C for 16 h using the Hybridization Oven 640 110 V (no. 800138; Affymetrix, Santa Clara, CA). Human genome U133 plus2.0 chips (Affymetrix) comprise 54,000 probe sets and provide comprehensive coverage of the transcribed human genome on a single array to analyze expression levels of >47,000 transcripts and variants, including 38,500 well-characterized human genes plus ∼6,500 new genes. GeneChips were stained with streptavidin-phycoerythrin using a Fluidics Station 450 (00-0079; Affymetrix). After being washed extensively, GeneChips were scanned using a GeneChip Scanner 3000 (00-0074; Affymetrix). Data were analyzed using GeneChip Operating Software version 1.1 (no. 690036; Affymetrix) according to GeneChip Expression Analysis Data Analysis Fundamentals (chapt. 4, “First-Order Data Analysis and Data Quality Assessment”; and chapt. 5, “Statistical Algorithms Reference”; http://www.affymetrix.com/support/technical/manuals.affx). To allow comparison, all chips were scaled to a target intensity of 500 on the basis of all probe sets on each chip. Comparison of GeneChip array data was obtained using custom analysis services (Kurabo Industries, Osaka, Japan). Kurabo Industries is the authorized service provider for Affymetrix Japan (Tokyo, Japan). Genes that were significantly upregulated (top 100 genes; see Supplemental Table 1; http://ajpcell.physiology.org/cgi/content/full/00481.2005/DC1) or downregulated (top 100 genes; Supplemental Table 2) in two independent experiments are summarized. Microarray data were deposited in the National Center for Biotechnology Information's Gene Expression Omnibus (GenBank accession no. GSE3364).
Data are means ± SE. Statistical evaluations were performed using an unpaired Student's t-test. Values of P < 0.05 were considered statistically significant.
Fullerene induces cytotoxic morphological changes in HUVECs.
To examine the direct effects on vascular ECs, cultured HUVECs were treated with C60(OH)24 for 24 h. Fullerenes (1–100 μg/ml) induced cytotoxic morphological changes in HUVECs such as vacuole formation in the cytosol and decreased cell density in a dose-dependent manner (Fig. 1, A–D). Figure 1, E (Control) and F [100 μg/ml C60(OH)24], represent low-magnification pictures, and cell density was clearly decreased after treatment with fullerene.
Fullerene increases release of LDH from HUVECs.
To assess EC injury by fullerene quantitatively, we examined the effects of fullerenes on endothelial LDH release, a marker of cell death and injury of the plasma membrane. Although 10 μg/ml C60(OH)24 showed slight cytotoxic morphological changes (Fig. 1C), only the maximal concentration of C60(OH)24 (100 μg/ml, 24 h, n = 8) significantly increased LDH release into culture medium (Fig. 2A) (LDH increased 2.4 ± 0.2-fold vs. controls, n = 11; P < 0.01). To further explore the degree of cell injury after fullerene treatment, we calculated the living cell number using WST-8. A maximal dose of 100 μg/ml C60(OH)24 killed 58.0 ± 1.7% of cells (n = 3; P < 0.01 vs. controls) (Fig. 2B).
Fullerene has antiproliferative effects on HUVECs.
To examine the effects of fullerene on cell growth, HUVECs at ∼30% confluence were treated with C60(OH)24 (1–100 μg/ml) for 24 h and then the total number of living cells was measured using WST-8. HUVEC growth was inhibited by C60(OH)24 in a dose-dependent manner (Fig. 3A). Quantitative analysis (Fig. 3B) revealed that only the maximal concentration of 100 μg/ml C60(OH)24 significantly inhibited cell growth (64.2 ± 4.7%, n = 6; P < 0.01 vs. controls).
Fullerene does not induce apoptosis in HUVECs.
We next examined whether fullerene induces apoptosis in vascular ECs. Serum starvation but not a maximal dose of C60(OH)24 (100 μg/ml, 24 h) induced cleavage of caspase-3 (17 kDa) and PARP (Fig. 4), which are markers for the activation of the apoptotic cascade (5, 20). This suggests that fullerene does not induce apoptosis in HUVECs. Notably, protein bands for both caspase-3 and PARP were weak in C60(OH)24-treated samples. This finding is consistent with fullerene-treated samples in other immunoblot analysis experiments. We speculate that this phenomenon is due to protein degeneration by fullerene.
Fullerene induces accumulation of polyubiquitinated proteins in HUVECs.
Because activation of the ubiquitin-proteasome system represents another death pathway, protein polyubiquitination by fullerene was examined. C60(OH)24 (1–100 μg/ml, 24 h, n = 8) induced protein polyubiquitination in a dose-dependent manner (Fig. 5A). Proteasome activity assay showed that 100 μg/ml C60(OH)24 (24 h) did not directly modify it (Fig. 5B) (1.09 ± 0.27-fold increase vs. controls; n = 3). Proteasome activity in cell lysates was suppressed almost completely by a proteasome inhibitor, MG132 (1 μM).
Ultrastructural features of HUVECs: fullerene facilitates autophagic cell death.
We next performed ultrastructural analysis using transmission electron microscopy. The cytoplasm of untreated HUVECs (control) (Fig. 6A) contained small vesicles. In vascular smooth muscle cells, the formation of small vesicles reportedly occurs under normal physiological conditions to remove abnormal proteins and cytoplasmic macromolecules (15). Treatment of HUVECs with the maximal dose of C60(OH)24 (100 μg/ml, 24 h) caused extensive vacuolization and internalization of fullerene (Fig. 6B). Fullerene aggregates were observed primarily within autophagosome-like vesicles (Fig. 6, C and D). To show that vesicles represented autophagosomes, Western blot analysis was performed to detect LC3 II because conversion of LC3 I (cytosolic isoform) to LC3 II (membrane-bound form) are frequently used markers for autophagosomes (9, 32). Figure 7 clearly shows that 100 μg/ml C60(OH)24 (24 h) increased levels of the LC3 II isoform (n = 4).
Chronic effects of low-dose fullerene on EC toxicity.
We examined the chronic effects of low concentrations of C60(OH)24 (1–10 μg/ml, up to 10 days) on EC toxicity. During this time, cells were subcultured three times (passages 3–6). Media containing fullerene were changed every 2 days. Chronic treatment with 1 μg/ml C60(OH)24 for 10 days had no significant effects on EC toxicity (data not shown). Figure 8 shows the morphological features of HUVECs treated without (control) or with 10 μg/ml C60(OH)24. Fullerene was treated soon after splitting cells from passages 3 to 4. On day 2, both control and fullerene-treated cells reached subconfluence, but fullerene-treated cells showed clear morphological changes such as cytosolic vacuole formation and spindlelike cell shape. After the second passage, control cells reached confluence within 4 days. In contrast, the attachment of cells (day 3) was bad in fullerene-treated groups and cell growth speed was slow. On day 8, fullerene-treated cells reached confluence, but the shapes of the cells were bad (spindlelike) and vacuole formation was commonly observed, suggesting the possibility that fullerene-resistant types of cells survived and increased.
Finally, microarray analysis was performed using total RNA from HUVECs treated with the maximal dose of C60(OH)24 (100 μg/ml, 24 h). Results from 2 independent samples are summarized in Supplemental Tables 1 and 2. Of note, although these were not top 100 genes, several genes related to the ubiquitin-proteasome system were significantly upregulated by fullerene [HECT (a COOH-terminal catalytic homologous to E6-AP-COOH terminus domain), C2, and WW domain containing E3 ubiquitin protein ligase 2 (ratio, 2.3 to 1), ubiquitin-specific protease 31 (ratio, 1.7 to 1), ubiquitin-specific protease 32 (ratio, 1.7 to 1), and ubiquitin-conjugating enzyme E2 (ratio, 1.5 to 1)].
The major findings of the present study are that water-soluble fullerene directly affects vascular ECs to cause cytotoxic injury or cell death and inhibition of cell growth. To the best of our knowledge, this study provides the first demonstration of the direct effects of water-soluble fullerene on vascular endothelium. In other human cells, including dermal fibroblasts, liver carcinoma cells (HepG2), neuronal astrocytes, and T-lymphocytes (Jurkat cells), recent reports have noted that water-soluble fullerene shows cytotoxic effects, presumably via production of reactive oxygen species (22, 24). Notably, several types of water-soluble fullerene derivatives are available [e.g., hydroxyl fullerene used herein, dendritic C60 monoadduct (22), malonic acid C60 (22) and nano-C60 (24) (basically pristine C60)], and cytotoxicity to cells varies depending on the fullerene subtype used, presumably because of surfactant chemistry, including a balance between hydrophobicity and hydrophilicity (3).
Some novel mechanistic insights of this study are that fullerene causes EC injury or cell death by increasing the accumulation of polyubiquitinated proteins in the cytosol and facilitating excessive autophagic cell death. EC injury and death are closely related to the initiation of atherosclerosis (14, 23). Furthermore, through nitric oxide production, ECs offer important protective functions against ischemic heart disease, including myocardial infarction, by inhibiting platelet aggregation (16) and lowering blood pressure (27). We thus propose that exposure to nanomaterials is a potential risk for cardiovascular disease, including atherosclerosis and ischemic heart disease. However, because quantitatively significant EC toxicity from water-soluble fullerene was observed only at high dosage, further validations (particularly in vivo) are needed.
We recently found that carbon black (CB), a chemically inert carbon nanoparticle present in diesel exhaust particles (13), shows endothelial cytotoxicity by mechanisms different from those of fullerene (31). Specifically, microarray analysis revealed that CB stimulated the induction of several proinflammatory mediators, including E-selectin, ICAM-I, IL-8, heme oxygenase-1, and prostaglandin endoperoxide synthase 2. However, such effects were not observed in fullerene-stimulated ECs. Furthermore, CB did not cause accumulation of polyubiquitinated proteins in ECs. Although the underlying mechanisms remain unclear, our findings indicate that the results reported herein could be specific to fullerene. Another report (21) examined the effects of several nanomaterials, including metals (TiO2, SiO2, Co, Ni, polyvinyl chloride), on endothelial toxicity. The report showed that only Co2+ particles possessed cytotoxic effects on ECs, supporting the concept that cytotoxicity varies among different nanoparticles.
Several pathways lead to cell death (6). Two major types of programmed cell death have been distinguished: the caspase-mediated process of apoptosis and the caspase-independent process involving autophagy.
In the present study, we found that exposure to fullerene did not induce cleavage of caspase-3 or PARP, which are the hallmarks of apoptosis (5, 20). Furthermore, no clear chromatin condensation was apparent in the nuclei (Fig. 6B). The present results thus indicate that the induction of apoptosis does not seem to be responsible for fullerene-induced cell death and/or injury.
Conversely, fullerene markedly increased the accumulation of polyubiquitinated proteins (Fig. 5A). Normally, under physiological conditions, cells degrade ubiquitinated proteins using 20S proteasomes. Fullerene does not seem to modify proteasome activity directly (Fig. 5B). Ultrastructural analysis revealed the excess formation of phagosome-like vesicles in the cytosol (Fig. 6B). Vesicles have been determined to be autophagosomes on the basis of Western blot analysis using specific marker LC3 II (9, 32) (Fig. 7). Similarly, the formation of phagosomes was observed after treating alveolar macrophages with single-wall and multiwall carbon nanotubes (8). Autophagy (type II programmed death) is known to be an alternative pathway to exclude unnecessary proteins and cellular organelles. We thus propose that accumulation of ubiquitinated proteins and abnormal activation of an autophagic death pathway could be responsible at least for cell death or injury caused by water-soluble fullerene.
EC injury and/or death appears to represent a primary mechanism for the initiation of atherosclerosis (14, 23). Injury and/or denudation of ECs trigger the attachment of leukocytes to the subendothelial region and promote transendothelial migration of cells, allowing the initiation of atherosclerosis. EC injury also leads to the loss of beneficial functions, including antithrombogenic and blood pressure-lowering functions via nitric oxide, which leads to the progression of ischemic heart disease, including myocardial infarction. In addition to cell death and injury, fullerene also inhibited EC growth. Impairment of EC growth may be related to impairment of angiogenesis. Because angiogenesis is crucial to the maintenance of vascular integrity by forming collateral vessels in response to tissue ischemia (11), fullerene inhibition of EC growth may be related to the progression of ischemic heart disease. Collectively, the present findings support the concept that exposure to fullerene could be a risk for atherosclerosis and ischemic heart disease.
The present study used 1–100 μg/ml fullerene concentrations for in vitro experiments. The pathophysiological concentrations of fullerene are barely known. In addition to engineered nanomaterials, traffic-derived nanoparticles are known to represent a risk for cardiovascular disorders, including atherosclerosis and ischemic heart disease (2, 4). The maximal concentration of particulate matter <2.5 μm in Chongqing, one of the biggest cities in China, was ∼700 μg/m3 (daily average) (26), indicating that an individual could inhale ∼10,000 μg of particulate matter during the course of 24 h there. This value is equivalent to ∼1 μg/ml when the extracellular fluid volume is 12 L in a 60-kg person. The fullerene dosage used in this study was up to 100-fold higher than this level.
The kinetics of water-soluble fullerene in vivo have not yet been completely determined. Normally, inhaled microparticles are cleaned off by alveolar macrophages via phagocytosis. However, this is not applicable to nanoparticles (2, 19), which appear to translocate to extrapulmonary sites via blood and lymph and thus reach other tissues (19). Using radiolabeled water-soluble fullerene administered intravenously to rats, Yamago et al. (28) demonstrated that most fullerenes moved rapidly to the liver (within 1 h) and then were distributed to various other tissues, including spleen, lung, kidney, heart, and brain. Extraction seems slow, and >90% was retained in the body 1 wk later, raising concern about chronic toxic effects. Although we could not clearly observe EC cytotoxicity after acute treatment with low-dose fullerene (10 μg/ml, 24 h), treatment for 8 days seems to enhance toxicity (Fig. 8). The effects of chronic exposure to low-dose fullerene in vivo need to be examined, particularly with regard to the cardiovascular system.
In summary, in the present study, we examined the direct effects of water-soluble fullerene on vascular endothelial cells to explore the potential toxicity of fullerene in humans, especially regarding the cardiovascular systems. We found that fullerene causes cytotoxic injury or cell death in vascular ECs, indicating that exposure to fullerene could represent a risk for atherosclerosis and ischemic heart disease. Because cytotoxicity by water-soluble fullerene occurs only at high doses, further validation experiments using blood vessels and animal models are warranted.
This study was supported by Health and Labor Sciences Research Grant for Research on Risk of Chemical Substance Grant H17-Chemistry-008 (to N. Iwai) and Grant-in-Aid for Scientific Research Grant 17790176 (to H. Yamawaki) from the Ministry of Education, Culture, Sports, Science, and Technology, Japan.
We are grateful to Dr. T. Yoshimori (National Institute of Genetics, Mishima, Japan) for providing LC3 antibody.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2006 the American Physiological Society