Introduction
Graphene is a nanomaterial that was developed in 2004 using a simple technique (
Sasidharan et al., 2013). It is an allotrope (form) of carbon consisting of a single layer of carbon atoms arranged in a hexagonal lattice (
Geim, 2009). Nowadays, graphene nanoparticles (GNPs) are used in the health-care sector as a gene transducer and biosensor for the diagnosis of diseases and tissue engineering (
Singh, 2018). GNPs have also attracted much attention because of their great potential for food safety and packaging (
Becaro et al., 2015), agriculture (
Hill et al., 2015), and industrial applications. This means workers worldwide are potentially exposed to GNPs. Consequently, concerns have been raised about the toxicity of GNPs (
Goede et al., 2018). Several studies on the cytotoxicity of GNPs have indicated that the reactive oxygen species (ROS) increases in cells. ROS can damage proteins, DNA, and lipids, and may lead to many diseases (
Seabra et al., 2014). Apoptosis, the programmed cell death (
Sanchez et al., 2011), and necrosis, a type of cell injury that results in the premature death of cells in living tissue by autolysis, are important processes that accelerate cell death by GNPs (
Li et al., 2012). The genotoxicity evaluation of GNPs has confirmed its carcinogenic potential in animal models (
Zhang et al., 2015). Wang et al. (
2011) have reported that a dose of 0.4 mg graphene oxide (GO) could cause granuloma formation in the kidneys, lungs, liver, and spleen and could not be filtered by the kidneys.
Recent studies have shown that when respiratory exposure to nanoparticles occurs, macrophages and neutrophils in the pulmonary system can be reactivated. In that case, it could produce an inflammatory reaction (
Kennedy et al., 2009).
Since in vitro systems can assess the primary mechanisms of toxicity, they are known to be useful methods for the evaluation of toxicity of nanomaterials. Some of the advantages of in vitro systems are efficiency, rapidity, cost-effectiveness, and the fact that they provide base studies for animal research (
Huang et al., 2010). However, the results of in vitro methods vary because of the differences in size, surface structure, functionalization, charge, impurities, aggregations, and the corona effect, and the sample preparation of nanomaterials. Therefore, cytotoxicity studies on nanomaterials need to be extended (
Jastrzebska et al., 2012).
There are still uncertainties about the toxicity of GNPs. Research on the toxicological effects of GNPs is still limited, and even occupational exposure limits (OELs) have not yet been defined for GNPs (
Mihalache et al., 2017). A better understanding of the relationship between the cytotoxicity of nanomaterials and the OEL would significantly extend our knowledge for establishing safe workplaces. A recent review showed that most of the studies on the cytotoxicity of GNPs are limited to the evaluation of the total lethal concentration (TLC) and inhibitory concentration of 50% (IC
50) (
Jastrzebska et al., 2012). The current study evaluated the “no observed adverse effect concentration” (NOAEC) as the highest experimental point without any adverse effects. Therefore, this study aimed to evaluate the cytotoxicity of GNPs on A549 epithelial cells of the human lung and provide basic data for identifying toxicological aspects of graphene clearly. This may be helpful in future risk assessment of exposed working groups in industries.
Method and materials
Synthesis of GNPs
GNPs were purchased with a purity of 99.98% from the Research Institute of Petroleum Industry (RIPI, Iran) (
Sereshta et al., 2013). GNPs are synthesized from graphite by the modified Hummer method through RIPI. In this method, some graphite powder is added to a mixture of concentrated H
2SO
4 and fuming HNO
3. The mixture is sonicated for 10 min and stirred for 30 min at 120°C. Then, it is cooled, diluted with deionized water, and neutralized with Na
2CO
3. The solution is filtered and dialyzed in a dialysis bag for 2 days. After the solution is concentrated on a rotary evaporator and filtered through 0.22-mm filters, a black solution is obtained. To obtain the graphene, the oxygen functional groups have to be eliminated. Hydrazine hydrate is used as the reducer. Finally, the solution is sonicated using the ultrasound method, heated in an oil bath, and refluxed under magnetic stirring at 70°C. The product is filtered and washed with additional ethanol to remove the residual hydrazine hydrate. Then, it is dried in an oven at 100°C.
Characterization of GNPs
The particle size and distribution of GNPs were measured by a transmission electron microscope (TEM; CM30-Philips, Japan) and a scanning electron microscope (SEM; S4160-Hitachi, Japan). The specific surface area and pore volume were calculated using the Brunauer–Emmett–Teller method (
Romero et al., 2012). Because nanoparticle size and size distribution are important factors that affect in vitro toxicity, it was needed to characterize these parameters via dynamic light scattering (DLS). DLS is used as a simple method for analyzing suspension stability and measurement of particle size in solution to evaluate any nanoparticles such as metal and metal oxide and carbon-based nanomaterials (
Murdock et al., 2007). The hydrodynamic sizes of GNPs in the dispersion media were examined using the DLS technique (Malvern Instruments Ltd, Zetasizer version 6.01) in a laboratory service company (Daypetronic, Iran).
Preparation of stock solution
The GNPs were dispersed using the generic nano-genotoxic dispersion protocol standard operation procedure (
Jensen et al., 2011). In total, 15.36 g GNP was used to prepare a 2.56 mg/mL stock dispersion in 6 mL of ethanol and bovine serum albumin water (BSA-water). Then, 0.5% velum ethanol (96% or higher) and 99.5% velum sterile-filtered BSA-water (0.05% w/v) were used to disperse the GNPs.
Cell culture and exposure of GNPs
Human epithelial cell line, A549, was bought from the Cell Bank (Pasteur Institute of Iran, Tehran, Iran). The cells were cultured in Dulbecco Modified Eagle Medium (DMEM; BIO-IDEA, Iran) supplemented with 10% fetal bovine serum (BIO-IDEA, Iran), 100 U/mL penicillin, and 100 l µg/mL streptomycin (pen stripe). Then, the cells were incubated at 37°C in a humidified atmosphere (5% CO2 and 95% air).
After 24 h of incubation, the cells were seeded in a 96-well culture plate (1 × 103 cells/mL) and allowed to be attached to the plate for 24 h. The suspensions of the GNPs were dispersed by sonication (160 W, 20 kHz, 5 min) and diluted in various concentrations. Then, these were suspended in DMEM of 10 different concentrations (0.1, 1, 10, 50, 100, 200, 300, 500, 600, 1000 µg/mL) for 24, 48, and 72 h. The cells maintained in DMEM without GNPs were used as the control group. Two control groups, negative control and blank, were used in this study. In the negative control group, the wells contained only culture medium and cells. While, the wells containing culture medium and GNPs was specified as the blank control.
Cell morphology
The A549 epithelial cells were plated in the 96-well plates (103 cells per well) and incubated for 24 h. The GNPs were introduced to the cells in a predetermined concentration in the culture medium. The cells cultured in the medium without the addition of the GNPs were the control. The cell morphology was observed under an optical microscope (Olympus 1x71, equipped with Olympus DP72 Camera 12.8 megapixel, Japan) after 24 h.
Cell viability
The effect of GNPs on cell viability was determined by the MTT (3-[4, 5-dimethylthiazolyl-2]-2, 5-diphenyltetrazolium bromide) assay. The MTT assay is a colorimetric assay based on the ability of viable cells to reduce a soluble yellow tetrazolium salt to a purple Formosan crystal by the mitochondrial succinate dehydrogenase activity of viable cells.
According to this assay, the A549 epithelial cells were washed twice with phosphate-buffered saline (PBS). Ten microliters of MTT (5 mg/mL) supplemented with 150 µL complete culture medium were added to the plates and incubated for approximately 4 h. The surface medium was vacated and 150 µL dimethyl sulfoxide was added to dissolve the insoluble purple Formosan product into a colored solution. The cell plates were put into a shaker for 20 min. Finally, the optical density was read using a microplate reader (ELX800, BioTek model, Winooski, USA) at 570 nm.
In MTT method, some nanomaterials such as carbon nanotubes may interact with or adsorb the dye/dye products leading to invalid results. Therefore, the cells were washed twice with PBS before reading the adsorption of GNPs at 570 nm.
Statistical analysis and determination of toxicological indicators
The data were evaluated using SPSS-V16 software. To determine the relationship between concentration/time variables and GNP-induced cell death, the statistical analysis of variance (ANOVA) was used. The critical value for statistical significance was p < 0.05 with 95% confidence.
The collected data were also entered into the Minitab 18.1 software and the probit regression model was used to achieve the toxicological indicators. In addition, the descriptive data from three independent experiments were reported as the mean ± standard deviation.
Discussion
The results of this study show that the GNPs reduced cell viability at high concentrations/doses. In other words, GNPs can be characterized as a toxin for the A549 epithelial cells of the human lung. There was a relatively good correlation between cytotoxicity and the exposure time. In addition, toxicological indicators illustrated the relationship between dose/time and response for GNPs. Several studies have confirmed the adverse effects of GNPs on organs and cells, too (
Shang et al., 2014;
Su et al., 2015). A study indicated that GNPs with 100–110 nm diameters had an increasing cytotoxic potential for pheochromocytoma and PC12 cells in rats. In addition, lactate dehydrogenase (LDH), ROS, and apoptosis were also released at 10–100 µg/mL for 48 h (
Zhang et al., 2010).
Another study reported that GNPs with a size of 500–1000 nm increased the ROS levels and triggered apoptosis in murine RAW 264.7 macrophages in 5–100 mg/mL concentrations after a 48-h exposure (
Li et al., 2012). Ma et al. found that the cytotoxicity of GO depends on the size of nanoparticles. Small nanosheets of graphene enter the cells mainly by endocytosis, whereas large graphene sheets are adsorbed onto the plasma membrane. Then, they enter the cells by phagocytosis, which may lead to the enhanced production of inflammatory cytokines and recruitment of immune cells (
Ma et al., 2015). Therefore, the cytotoxicity of GNPs depends on the size of the nanoparticles. In the current study, GNPs at concentrations lower than 0.19 µg/mL exhibited remarkable cytotoxicity in the lung cells during the 24-h exposure period, while the diameters of GNPs were 2–50 nm. The results of the current study are in agreement with previously mentioned studies.
The mode of dispersion of the nanomaterials in the exposure tests is decisive for the outcome of in vitro studies. The agglomeration of GNPs, due to their differential dispersion ability, critically affects their ability to interact with the cells and, in turn, affects their internalization within the cells (
Stone et al., 2009). In this study, DMEM containing 5% serum was chosen for dispersion, whereas the agglomeration of GNPs was moderate, and the size of the GNPs was 323.3 nm. In another study, the graphene agglomeration occurred following the dispersion in double-distilled water cells and on the cellular membrane, while the mean size of the nanoparticles was 349 ± 24 nm (
Majeeda et al., 2016). It can be concluded that the agglomeration of GNPs might be related to their size.
The functionalized surface of GNPs may also affect cytotoxicity. Majeeda et al. found that solubility and toxicity could be influenced by changes in the surface properties. With respect to graphene surface chemistry, the pristine graphene exhibited the highest toxicological behavior (
Majeeda et al., 2016). GNP-COOH and GNP-NH2 caused DNA damage, genotoxicity, and hypomethylation in human bronchial epithelial cells (BEAS-2B cells) at 10 and 50 mg/L after 24 h of exposure. Chatterjee et al. reported GNP toxicity higher than GNP-COOH and GNP-NH2 in a severe form of DNA damage. So, the concentration of 50 mg/mL was reported as the lowest effective concentration (
Chatterjee et al., 2016). In this study, GNPs were pure and without a functionalized surface, but the TLC (or 99% death) of GNPs was obtained at 40.65 mg/mL. The obtained TLC results confirm that pristine GNPs are more toxic than functionalized surface GNPs.
According to the study by Chang et al., when GNPs were oxidized to GOs, cell viability decreased at a high concentration. They found that the GO did not enter the A549 cell and thus had no obvious cytotoxicity. Although the GO created ROS, IC
20 was reported for 200 µg/mL in 24 h (
Chang et al., 2011). In another study, cytotoxicity was observed at the highest GO concentration of 100 µg/mL (
Schinwald et al., 2013;
Zhang et al., 2012). When even graphene is changed to reduced graphene oxide (rGO), toxicity occurred in the highest concentration (100 µg/mL) (
Akhavan et al., 2012;
Jaworski et al., 2015). Chong synthesized the graphene quantum dots (GQD) from graphite and GO and evaluated the cytotoxicity. This compound of GO, which could improve the aqueous stability of GO, resulted in approximately 85% A459 cell viability at a concentration of 640 µg/mL (
Chong, 2014). In the present study, IC
50 was achieved at around 134.77 µg/mL in 24 h. The results of the current study justified the high toxicity of pristine GNPs, which is consistent with the results of similar studies.
This study indicated that cytotoxicity depends on the exposure period. GNPs were more cytotoxic after 72 h of exposure compared to a 24-h and 48-h exposure period. Chang et al. (
2011) have reported similar results after 24, 48, and 72 h of exposure to GO and confirmed that the cytotoxicity of GNPs depends on the exposure period. Hence, cytotoxicity was dose- and time-dependent.
The obtained NOAEC results showed that GNPs are toxic for A459 cells in a 0.19 µg/mL concentration after an exposure for 24 h. In other words, IC
10/24 for pristine GNPs was 0.19 µg/mL. Chang et al. (
2011) reported IC
20/24 of 200 µg/mL for GO. Since pristine GNPs are more toxic than other forms of GNPs (such as GO, rGO, and GQD), it can be proposed that the structure of nanomaterials can influence the NOAEC. Zhang et al. (
2010) observed that a low dose of GNPs (0.01 µg/mL) could decrease PC12 cell viability, while it could not affect metabolic activity, LDH release, and ROS. It can be concluded that 0.01 µg/mL was the NOAEC indicator for GNPs. In other studies,
Li et al. (2012) have reported that less than 20% of the RAW 264 cells remained in a concentration of 20 mg/mL. This indicates that the type of cell may influence the cytotoxicity.
Although the MTT measurement is a reliable toxicity marker, the assay has some inherent limitations. This measurement, which is a colorimetric assay, was used to determine the effect of cytotoxicity; it can evaluate cell viability without considering the toxicity mechanisms. It is recommended that toxicity mechanisms, such as oxidative stress, apoptosis, LDH, superoxide dismutase, and autophagy, which can influence cytotoxicity, be considered to evaluate the cytotoxicity of GNPs. A better understanding of toxicity mechanisms will justify the cytotoxicity and the toxicological indicators of GNPs.