Nanotoxicity of different sizes of graphene (G) and graphene oxide (GO) in vitro and in vivo

Highlights

• The toxicity of G and GO was differed based on sizes, oxidation state, exposure concentrations, and sensitivity of the experimental model used.

• GFNs reduced cell viability in the following order of size; S > L > M with stronger impact of G than GO.

• Toxic potential of GO > G in terms of DNA damages, ROS generation and varying gene expression in vitro and in vivo.

• G induced acute toxicity on Tox2 bacteria with size-dependent manner: L>M>S, and M-GO showed obvious genetic toxicity on RecA bacteria.

Abstract

Graphene family nanomaterials (GFNs) have attracted significant attention due to their unique characteristics and applications in the fields of biomedicine and nanotechnology. However, previous studies highlighted the in vitro and in vivo toxicity of GFNs with size and oxidation state differences are still elusive. Therefore, we prepared graphene (G) and graphene oxide (GO) of three different sizes (S-small, M-medium, and L-large), and characterized them using multiple surface-sensitive analytical techniques. In vitro assays using HEK 293T cells revealed that the small and large sizes of G and GO significantly reduced the cell viability and increased DNA damage, accompanying with activated reactive oxygen species (ROS) generation and induced various expressions of associated critical genetic markers. Moreover, the bacterial assays highlighted that G and GO caused strong acute toxicity on Tox2 bacteria. Effects of G were higher than GO and showed size dependent effect: L > M > S, while the medium size of GO induced mild genetic toxicity on RecA bacteria. In vivo assays revealed that exposure to G and GO caused the developmental toxicity, induced ROS generation, and activated related pathways (specifically GO) in zebrafish. Taken together, G showed stronger ability to decrease the survival rate and induce the acute toxicity, while GO showed obvious toxicity in terms of DNA damages, ROS generation, and abnormal gene expressions. Our findings highlighted that G and GO differentially induced toxicity based on their varying physical characteristics, especially sizes and oxidation state, and exposure concentrations and sensitivity of the employed in vitro and in vivo models. In short, this study provided deep insights on the negative effects of GFNs exposure.

Introduction

Graphene family nanomaterials (GFNs) have enormous applications in fields of nanotechnology and biomedicine for their biocompatibility and excellent physicochemical properties (Feng and Liu, 2011; Geim and Novoselov, 2007; Yang et al., 2013a). Among GFNs, graphene (G) and graphene oxide (GO) are well-known ideal nanomaterials, frequently used in diagnostic, therapeutic, and preventive medical products, especially as novel carriers for drug delivery (Mainardes and Silva, 2004; Shen et al., 2012). The wide range application of G and GO may lead to their inevitable release into environments, thus pose risks to human, animals, fungi, and plants (Gurunathan and Kim, 2016; Kunzmann et al., 2011; Nogueira et al., 2015). The high doses of non- or poorly biodegradable materials would probably preclude such materials from being developed any further (Leroux, 2017). Both G and GO are the single layer of carbon atoms arranged in a two-dimensional crystal lattice, while GO has special chemical modification with highly reactive oxygen functional groups that effectively works as a stabilizing agent in water and can be covalently attached to small molecules (Dreyer et al., 2010; Geim and Novoselov, 2007; Rourke et al., 2011). Sizes and surface oxidation of G and GO nanomaterials potentially affect their properties and associated toxicity on various models (Linares et al., 2014; Wibroe et al., 2016; Yang et al., 2012). Therefore, to unveil the possible nexus among varying sizes of G and GO and associated toxic impacts on different biological systems (human cells, zebrafish, and bacteria) is of critical importance to ensure their safe utility.

Previously, the in vivo and in vitro toxicity of GFNs were explored ranging from their biodistribution in biological systems and response of various molecular pathways (Hillaireau and Couvreur, 2009; Wang et al., 2015; Wu et al., 2016; Zhang et al., 2017). Our previous study and others also highlighted that G and GO nanosheets induced DNA damage, generated reactive oxygen species (ROS) and activated the base excision repair (BER) signaling pathway both in cells and zebrafish larvae (Akhavan et al., 2012; Lu et al., 2017). Especially, the ROS generation and associated oxidative stress in cells have been reported as a critical indicator of GFNs mediated toxicity, which consequently induced DNA damages and reduced cell survival rate (Chang et al., 2011; Yang et al., 2013b; Zhang et al., 2010). Previous in vivo studies revealed that GO entered into the chorion of embryos via endocytosis, and induced developmental abnormalities in terms of altered heart rate, tail flexure, cardiac/yolk sac edema, and also induced Parkinson's disease-like symptoms in zebrafish (Chen et al., 2016; Liu et al., 2014; Ren et al., 2016). However, the toxicological relationship of G and GO with their different sizes in reliable in vivo system needs more investigations. In toxicological assessments, zebrafish was previously used as a credible in vivo model to study the development and growth, the molecular biomarker of oxidative stress (ROS levels), and various molecular pathways related to DNA damage responses such as the BER and PI3K pathways (Dai et al., 2014; Ersahin et al., 2015; Lu et al., 2017; Valavanidis et al., 2006).

Previous in vitro studies indicated that nano-sized GO entered into the cells by interacting with the plasma membrane and its surface receptors, implying the importance of size for internalization and intracellular fate of GO (Lammel et al., 2013; Yue et al., 2012). The sharp edges of G and GO nanosheets revealed direct interactions with the cell membrane and its disruption in cells and bacterial models (Hu et al., 2011; Yang et al., 2013b). In cells, a membrane-integrated protein, epidermal growth factor receptor (EGFR) usually activated in response to the physical stress and mechanical injury (Balestreire and Apodaca, 2007; Iwasaki et al., 2000). In addition, EGFR regulated cell growth and differentiation, and activated NFκB signaling pathway to repair DNA damage (Szumiel, 2006; Toulany and Rodemann, 2010). Therefore, those potential mechanisms involved in cellular damages and inner response pathways induced by different sized G and GO needed deep investigations.

To elucidate the toxicity of chemicals/nanomaterials, the luminous bacteria test (LBT) was increasingly applied because it showed advantages of being sensitive to the low concentration and a wide range of pollutants, and the bioluminescence was observed directly related to the level of chemical toxicity (El-Alawi et al., 2002; Sorensen et al., 2006). For better understanding the environmental impacts of different sizes of G and GO materials on bacterial strains, the modified lux gene bacteria of Tox2 strain (Acinetobacter sp. Tox2) was applied to detect the acute toxicity and RecA strain (Acinetobacter sp. RecA) to detect the genetic toxicity. RecA bacteria with luxCDABE contained all five genes of the lux cassette to detect xenobiotics without adding other substrates (Winson et al., 1998). Thus, the changes in the luminous intensity of RecA bacteria are positively correlated with gene expression, and this bacterial model can be used to probe genetic toxicity (Quillardet and Hofnung, 1993; Song et al., 2009). Tox2 bacteria promptly responded to the acute chemical exposure, which can disrupt the respiratory system of Tox2 by weakening their luminous intensity (Lopez-Roldan et al., 2012). The strong antibacterial functions of G and GO on various strains closely related to membrane damages, however, the bacterial membrane is the place where the respiratory action and ROS generation take place (Akhavan and Ghaderi, 2010; Ji et al., 2016). Thus, except relative luminous units (RLU) of bacteria, the ROS generation was also detected simultaneously as the key indicator of toxicity induced by different sizes of G and GO materials.

In this study, three sizes of G and GO were prepared and then characterized by the surface sensitive analytical techniques such as atomic force microscopy (AFM), transmission electron microscopy (TEM), dynamic light scattering (DLS), and X-ray photoelectron spectroscopy (XPS). In vitro toxicity was screened in terms of cell viability, DNA damages, ROS generation and responses of related genetic markers in HEK 293T cells after exposure to different sizes of G and GO. Additionally, the luminous Tox2 & RecA bacterial strains were employed to evaluate the acute and genetic toxicity as well as the potential relationship between ROS generation after G and GO treatment. Finally, the in vivo toxicity of G and GO was evaluated in terms of developmental toxicity, ROS generation and responses of BER and PI3K pathways in zebrafish larvae.