© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 5391 www.advmat.de www.MaterialsViews.com wileyonlinelibrary.com COMMUNICATION Wendi Zhang, Chi Wang, Zhongjun Li, Zhenzhen Lu, Yiye Li, Jun-Jie Yin, Yu-Ting Zhou, Xingfa Gao, Ying Fang,* Guangjun Nie,* and Yuliang Zhao* Unraveling Stress-Induced Toxicity Properties of Graphene Oxide and the Underlying Mechanism W. Zhang, Dr. C. Wang, Z. Li, Dr. Z. Lu, Dr. Y. Li, Prof. Y. Fang, Prof. G. Nie, Prof. Y. Zhao CAS Key Laboratory for Biomedical Effects of Nanomaterials & Nanosafety National Center for Nanoscience and Technology Beijing 100190, China E-mail: fangy@nanoctr.cn; niegj@nanoctr.cn zhaoyuliang@ihep.ac.cn Dr. J.-J. Yin, Y.-T. Zhou Center for Food Safety and Applied Nutrition Food and Drug Administration College Park, MD 20740, USA Prof. X. Gao, Prof. Y. Zhao Institute of High Energy Physics CAS, Beijing 1000049, China DOI: 10.1002/adma.201202678 The unique physiochemical properties of graphene nanosheets, single atomic layers of carbon arranged in a hexagonal lattice, endows graphene and its derivatives (such as graphene oxide, GO) with a great range of potential applications, especially in biomedical fields, such as cellular imaging, [1] drug delivery, [2] biosensors, [3,4] and photothermal therapy. [5] However, studies of in vivo toxicity and the underlying chemical mechanism(s) are very limited, [6,7] which largely impedes the development of graphene materials for in vivo applications. To date, investiga- tions regarding the toxicity of graphene and its derivatives are mostly conducted on in vitro models under normal conditions. Graphene and GO nanosheets have been reported to possess cytotoxic effects on both bacteria and mammalian cells, [8–13] where disruption of the bacterial cell membrane and oxidative stress have been proposed to be involved in the toxicity. [9–10] However, there are also contradictory results demonstrating high biocompatibility of GO and functionalized GO toward mammalian cells. [2,12,14–16] Although different physiochemical properties, such as the density of functional groups, size and conductivity, have been proposed to be related to the differ- ential antibacterial activity of GO-based materials, there is no consensus on the issue of systemic toxicity of these nanoma- terials in eukaryotic organisms, which raises the need for fur- ther studies of the biological effects of GO and its derivatives in vivo. Early investigations on in vivo administration of GO nanosheets showed a dose-dependent toxicity which led to lung injury in mice. [6,7] However, whether the GO and its derivatives exert certain systemic toxic effects in multicellular organisms and the underlying chemical mechanism(s) is still unknown. To date, the toxicity of GO derivatives was investigated under normal physiological conditions, however, no studies have yet been carried out under pathophysiological conditions. The concept of susceptible population to nanotoxicity proposed by previous studies [17] suggested the necessity to further unveil the toxicity of GO derivatives under pathophysiological condi- tions. Here, we used a simple animal model, Caenorhabditis elegans ( C. elegans) to investigate in vivo the toxic profile of GO and functionalized GO modified with PEGylated poly-L-lysine (PLL-PEG) (GO/PP) under both normal and stress conditions. C. elegans is a well established model for toxicological testing in biomedical and environmental fields. [18] With this simple model, the potential in vivo toxicity of nanosheets under stress was studied. GO was prepared by a modified Hummer’s method. [19] The plain GO was soluble in water but aggregated in buffer solu- tion, a phenomenon also observed by Liu et al. [16] For enhanced stability in a nematode culture medium, GO nanosheets were coated with PLL-PEG through electrostatic adsorption between the negative charges on GO and the positive charges of PLL. The resulting GO/PP was soluble in both the buffer and nema- tode culture medium. The average sizes of the GO and GO/PP nanosheets were 80 and 106 nm, respectively, as measured by dynamic light scattering (DLS) (Figure S1). The surface mor- phologies of GO and GO/PP were determined by atomic force microscopy (AFM), and the height of GO was about 1 nm, indi- cating the presence of single-layer nanosheets (Figure S2). The characteristic peaks at about 1600 cm -1 in Raman spectroscopy revealed that the differences in the carbon structure of GO and GO/PP was small (Figure S3). In addition, the survey and C 1s spectra of the nanosheets were determined by X-ray photoelec- tron spectroscopy (XPS). Compared with GO (Figure S4A), the increased area proportion of the peak (286.4 eV), which is due to the contributions from C–O and C =O to the C–C peak (284.9 eV) (Figure S4B), and the enhanced N proportion by 2.0% in GO/PP can be assigned to the functionalized PLL-PEG (Figure S4). Different surface modifications or structures have been shown to significantly influence the physicochemical properties of graphene and graphene-based materials, [20–22] which further alteres their toxicity properties. [9] In our work, the in vivo toxicity profiles of GO, and especially GO/PP, have been studied. Initially, the uptake of GO/PP by C. elegans was determined. We used FITC-PLL (Sigma) instead of PLL to synthesize GO/ FITC-PLL-PEG (GO/F-PP) to visualize the nanosheets in the worm’s body. The uptake of FITC-labeled nanosheets in the worms was investigated by laser scanning confocal microscopy. After being incubated with GO/F-PP for 4 h, green fluorescence was observable through the length of worm’s body and in the anterior part of intestine, suggesting that GO/F-PP (and GO/ PP) could be taken up by the nematodes (Figure S5). Adv. Mater. 2012, 24, 5391–5397