Biomaterials 35 (2014) 5041e5048
Contents lists available at ScienceDirect
Biomaterials journal homepage: www.elsevier.com/locate/biomaterials
The in vitro and in vivo toxicity of graphene quantum dots Yu Chong a, b,1, Yufei Ma a,1, He Shen a, Xiaolong Tu a, Xuan Zhou a, Jiaying Xu b, Jianwu Dai a, Saijun Fan b, c, **, Zhijun Zhang a, * a Suzhou Key Laboratory of Nanobiomedicine, Division of Nanobiomedicine & Collaborative Innovation Center of Suzhou Nano Science and Technology, Suzhou Institute of Nano-tech and Nano-bionics, Chinese Academy of Sciences, 398 Ruoshui Road, Suzhou 215123, China b School for Radiological & Interdisciplinary Sciences, Medicine College of Soochow University, Suzhou 215123, China c Institute of Radiation Medicine, Chinese Academy of Medical Sciences, Tianjin 300192, China
a r t i c l e i n f o
a b s t r a c t
Article history: Received 21 January 2014 Accepted 11 March 2014 Available online 28 March 2014
Graphene quantum dots (GQD) generate intrinsic fluorescence, and improves aqueous stability of graphene oxide (GO) while maintaining wide chemical adaptability and high adsorption capacity. Despite GO’s remarkable advantages in bio-imaging, bio-sensing and other biomedical applications, its biosafety issues are still unclear. Here we report a detailed and systematic study on the in vitro and in vivo toxicity of GQD. The GQD sample was prepared through a facile oxidation approach and fully characterized by means of AFM, TEM, FTIR, XPS and elemental analysis. In vitro experiments showed that GQD exhibits very low cytotoxicity owing to its ultra-small size and high oxygen content. Then, the in vivo biodistribution experiment of GQD revealed no material accumulation in main organs of mice and fast clearance of GQD through kidney. In order to mimic clinic drug administration, mice were injected with GQD and GO (as comparison) multiple times for in vivo toxicity tests. We found that GQD showed no obvious influence on mice owing to its small size, while GO appeared toxic, even caused death to mice due to GO aggregation inside mice. In brief, GQD possesses no obvious in vitro and in vivo toxicity, even under multi-dosing situation. Ó 2014 Published by Elsevier Ltd.
Keywords: Graphene quantum dots Cytotoxicity In vivo test In vitro test Biocompatibility
1. Introduction Carbon materials have excited scientists and engineers for decades [1,2], not only due to their various allotropes and abundant availability, but also because of the exciting electronic, mechanical and physicochemical properties of versatile carbon structures. Among various applications of carbon nanomaterials [3e5], the biomedical applications [6e8] especially drug delivery [9,10] and diagnostic potential [11] of the carbon structures, graphene and carbon nanotube in particular, have been attracting increasing interests in recent years. Furthermore, carbon materials have been exploited with advanced therapeutic techniques such as photothermal [12,13] and photodynamic [14] therapy. However, pristine carbon-based nanomaterials are hydrophobic related to benzene
* Corresponding author. Tel.: þ86 512 62872556; fax: þ86 512 62603079. ** Corresponding author. Institute of Radiation Medicine, Chinese Academy of Medical Sciences, Tianjin 300192, China. Tel.: þ86 22 85685301; fax: þ86 22 85683033. E-mail addresses: [email protected] (S. Fan), [email protected] (Z. Zhang). 1 These authors contributed equally to the manuscript. http://dx.doi.org/10.1016/j.biomaterials.2014.03.021 0142-9612/Ó 2014 Published by Elsevier Ltd.
structures, which leads to aqueous instability and thus restricts direct use in biological applications. A popular way to stabilize these carbon nanomaterials is to functionalize them with hydrophilic polymers (mostly polyethylene glycol, PEG), and it has been proved to be successful with no obvious toxicities both in vitro and in vivo [15,16]. In addition, animal studies suggested that PEGylated carbon nanotubes or graphene accumulated in the main organs and stayed for over one month at an administration dose of 20 mg/ kg [17,18], raising much concern about toxicity tolerance despite of no toxic evidence of carbon materials to mice [19]. Another strategy to improve the aqueous solubility and stability of carbon materials is to make ultra-small carbon nanoparticles, so that Brownian motion of the nanoparticles can provide sufficient energy to prevent the aggregation, and that nanoparticles become less hydrophobic due to the Oxygen-rich functional groups at the edges. Thus, nanometer-sized carbon materials, carbon quantum dots (Cdots) [20] and especially graphene quantum dots (GQD) [21], are rising stars in the family of carbon materials because of their excellent solubility and stability in water, strong fluorescence, and the retained advantages of graphene. Rapidly increasing number of reports about GQD have explored the synthesis methods (from oxidation, hydrolysis to electrolysis) [22], the origin of
5042
Y. Chong et al. / Biomaterials 35 (2014) 5041e5048
fluorescence [23], and biological applications such as imaging and sensing [22,24e26]. Although GQD have shown great potential in biomedical applications [14,22,27], the toxicity of GQD at in vitro and in vivo level is still unclear. Only quite few reports studied the in vitro and in vivo toxicity of GQD [28,29], while several examples investigated the in vivo distribution and toxicity of carbon dots [30,31]. In addition, there is no report on whether size of graphene-based nanomaterials relates to the toxicity in the region of sub-50-nm. Thus, more profound and systematical study on the toxicity of GQD and GO at in vitro and in vivo level is highly desired for its practical applications. Our group has made many efforts to develop graphene oxide (GO) for versatile theranostic agents [7,32e34], and we also found that GQD appeared to be a better drug delivery platform than GO due to its high water solubility and stability [27]. Before applying GQD to biomedical applications, here we report the in vitro and in vivo toxicity of a GQD sample. The GQD sample was generated by a facile synthetic method with high yield, and their chemical structure was fully characterized to reveal the structure-toxicity relationship. We utilized several cellular assays (WST-1, cell apoptosis, LDH, ROS) to analyze the effect of GQD on the cells, and then systematically test the in vivo biodistribution and toxicity of GQD. For in vivo toxicity evaluation, multiple doses of GQD were injected into mice in order to mimic clinic drug administration and to practically study the effect of GQD on mice [29,30]. GO of 10e 30 nm was utilized for comparison on size effect in the in vivo experiments because the toxicity of GO is relatively clearer than that of GQD [13,16,18]. Through a close look into the toxicity of GQD at the cellular and animal levels, these findings will reveal, in the region of sub-50-nm, the size effect of graphene materials on their toxicity, and inspire people to develop biologically compatible nanomaterials. 2. Materials and methods 2.1. Synthesis and characterization 2.1.1. Synthesis of GQD GQD was obtained through modified oxidative cutting method reported in literature [21]. Generally, 0.2 g of graphite powder (Alfa Aesar) was added into a mixture of 15 mL of concentrated H2SO4 and 5 mL of fuming HNO3. The reaction mixture was sonicated for 10 min, and then stirred for 30 min at 120 C. The mixture was cooled, diluted with 100 mL of deionized (DI) water, and then neutralized to pH 7e8 with Na2CO3. In order to remove salts, the resulted solution was filtered and further dialyzed in a dialysis bag (50 kDa cutoff) for 2 days. After concentrating the solution on a rotary evaporator and filtration through 0.22 mm filters, a black solution of GQD (4 mg/mL, measured through absorption at 230 nm) was obtained. Production yield was calculated to be about 60%, which was much higher than that of previous reports [21,22,29,35]. 2.1.2. PEGylation of GQD 20 mg of GQD was diluted with 50 mM phosphate buffer (pH 5.5) to the concentration of ca. 0.2 mg/mL, and activated with 20 mg of 1-ethyl-3-(3dimethylaminopropyl)-carbodiimide (EDC, Sigma). After 15 min, 200 mg of PEG (2000 Da, amine terminus, Jenkem Technology) with amine terminus was added into the solution and the solution was further stirred for 2 h for complete reaction. At last, the solution was concentrated and rinsed with water through spin-dialysis (3000 Da cutoff, 5000 rpm) to generate GQD-PEG. 2.1.3. Cy7 labeling of GQD 20 mg of EDC was mixed with 100 mL of 50 mM phosphate buffer (pH 5.5) containing 2 mg of Cy7 (carboxyl terminus, Fanbo Biochemicals Co. Ltd., Beijing) for 15 min of COOH activation, and then 20 mg of GQD-PEG was added into the solution. The solution was further stirred for 2 h, before it was concentrated and rinsed with water through spin-dialysis (3000 Da cutoff, 5000 rpm) to generate GQD-PEG-Cy7. 2.1.4. Synthesis of GO-PEG GO was synthesized using Hummer’s method and then converted to carboxylated GO according to previous protocol [11]. 20 mL of GO solution at a concentration of 0.2 mg/mL was mixed with 24 mg of six-armed branched PEG (10 kDa, amine terminus, Jenkem Technology) and sonicated for 5 min, following by EDC additions (4 mg 3) under continuous sonication for additional 1 h. Large aggregates were removed through centrifugation at 13,000 rpm for 1 h, and then the supernatant
was collected and washed five times through spin-dialysis (100 kDa cutoff, 4000 rpm) to generate GO-PEG. 2.2. Material characterization UVeVis spectra were collected with a PerkinElmer Lambda 25 spectrophotometer, and fluorescence spectra were obtained on a Hitachi F-4600 fluorescence spectrophotometer. Atomic force microscope (AFM) images were conducted on a Vecco Dimension 3100 Atomic force microscope, and transmission electron microscopy (TEM) was a Tecnai G2 F20 S-Twin from FEI. CHN elemental analyses were performed on an Elementar Vario EL analyzer. Fourier transform infrared (FTIR) spectra were conducted within the 4000-400 cm1 wavenumber range using a Thermo Nicolet 6700 FTIR spectrometer with the KBr pellet technique. X-ray photoelectron spectroscopy (XPS) spectra were acquired with a Kratos Axis UltraDLD spectrometer using a monochromatic Al Ka source at 1486.6 eV. Thermogravimetric analysis (TGA) was carried out with a Seiko TGA 6200 at a heating rate of 10 C/min under nitrogen. 2.3. Biological tests 2.3.1. Cell viability The cell viability was evaluated with WST-1 (Beyotime, China). Briefly, HeLa (or A549) cells were seeded on each well of 96-well plate at a density of 8 103 cells per well, and cultured in a humidified 5% CO2 incubator at 37 C for 24 h. Then GQD or GQD-PEG was introduced into cells with various test concentrations (10, 20, 40, 80 and 160, 320, and 640 mg/mL) in culture medium, and cells cultured without nanomaterials were taken as the control. The cells were washed with PBS after 24 h of incubation, following by addition of 100 mL of fresh culture medium and 10 mL of WST-1 into each well and additional 1 h of incubation at 37 C. At last, the optical density (OD) of each well at 450 nm was recorded on a PerkinElmer Victor 4 microplate reader. 2.3.2. Cell apoptosis and necrosis Cell apoptosis and necrosis were examined using the annexin V and PI assay kit (Beyotime, China). In brief, GQD-PEG at concentrations ranging from 10 to 160 mg/ mL in culture medium was added into HeLa cells precultured for 24 h and exposed for another 24 h. After that, the cells were washed three times with PBS, then trypsinized and subsequently washed with PBS once, resuspended in 300 mL PBS, and incubated with 5 mL Annexin V-FITC and 1 mL PI for 10 min at room temperature. At last, the samples were diluted with 200 mL of binding buffer and analyzed with a Beckman Coulter FC500 flow cytometer (>1 104 cells for each sample). Excitation wave was set at 488 nm, and the fluorescence intensities of Annexin V (green) and PI (red) was collected at 530 and 575 nm, respectively. 2.3.3. LDH Release and oxidative stress For LDH (lactate dehydrogenase) assay, HeLa cells seeded in a 96-well plate were incubated with different concentrations of GQD-PEG at 37 C for 24 h, and then 20 mL of LDH standard solution (Beyotime, China) was added into each well for additional 1 h incubation at 37 C. After that, 100 mL of cell supernatant was aspirated and transferred into another plate according to the instruction, and the absorbance of supernatant was examined at 490 nm to analyze the level of LDH release. ROS (Reactive Oxygen Species) was measured with a procedure similar to the LDH assay. After incubating with GQD-PEG at various concentrations, the cells were washed three times and mixed with 1 mL of 10 mM DCFH-DA (Beyotime, China) for additional incubation for 20 min. Next, cells were washed three times with DMEM without FBS to eliminate the DCFH-DA outside cell membrane, and then collected in suspension and measured with a flow cytometer (excitation at 488 nm and emission at 525 nm). 2.3.4. Biodistribution All animal experiments were conducted under protocols approved by the Soochow University Laboratory Animal Center. Healthy female Balb/c mice (Suzhou Industrial Park Animal Technology Co., Ltd.) at an average age of 6e7 weeks were housed under a standard condition of a 12 h light/dark cycle with free access to food. After injected with 4T1 cells (108 cells/mL 100 mL) and housed for 3 days, Balb/c mice bearing 4T1 murine breast cancer tumors were intravenously (i.v.) or intraperitoneally (i.p.) injected with GQD-PEG-Cy7 (200 mL of 1.5 mg/mL solution for each mouse; a dose of 15 mg/kg), and then were sacrificed at time points of 1, 4, 12, 24 or 48 h after injection. Various organs and tissues were collected and spectrally imaged by the IVISÒ Lumina imaging system. The averaged fluorescence intensity of Cy7 of each imaged organ was calculated for a semiquantitative biodistribution analysis. 2.3.5. In vivo toxicity assessment To simulate practical drug administration, multiple-dosing toxicity test was performed to further investigate the biosafety of GQD-PEG. 48 healthy female Balb/c mice (6e7 weeks) were randomly divided into four different experimental groups, including groups of control, GQD-PEG by i.p., GQD-PEG by i.v. and GO-PEG by i.p. injection. The mouse was injected with the materials (20 mg/kg) or saline every other day for 14 days. Then, half of them were killed 1 day post injection for acute toxicity study, and the others were further monitored for over a month. Mice were
Y. Chong et al. / Biomaterials 35 (2014) 5041e5048 weighted every 3 days during 2 months of the experiment, and finally sacrificed for further tests. About 0.8 mL of blood from each mouse was collected for blood chemistry tests and complete blood panel analysis before the mouse was euthanatized. Major organs from those mice were harvested, fixed in 4% neutral buffered formalin, processed routinely into paraffin, sectioned at 8 mm, stained with hematoxylin and eosin (H&E) and examined under a digital microscope. Statistics were based on standard deviations of 6 mice per group.
5043
a large number of OH groups (3% of H content), which is further agreed by IR peaks at w3400 and w3300 cm1. This high O and OH content can be possible only if the graphene sheet is mono-layered and the lateral size is smaller than 5 nm, as have been proved by TEM and AFM. The small size and high O content of GQD in turn help them to be solubilized and highly stabilized in water or serum, implying that GQD would present good biocompatibility.
3. Results and discussion 3.2. Cellular toxicity 3.1. Synthesis and characterization of GQD The GQD sample was obtained through modified oxidative cutting method (Fig. 1a) [21]. Briefly, graphite powder in H2SO4/ HNO3 was heated at 120 C for 30 min, then the black solution was neutralized with Na2CO3 and dialyzed to obtain final GQD product with a yield as high as 60%. Compared with the previous studies [14,21,22,27,35], which focused on improving the fluorescence of GQD, this method offers a facile preparation with high production yield, and thus enables further exploration of GQD for widespread applications. Fig. 1b shows the UVevis and fluorescence spectra of the GQD sample. There are two distinguished peaks of GQD in the UV region, 230 nm (pep* transition from benzene structures) and 290 nm (se p* transitions from functional groups) [36]. Due to the small size of GQD and thus small sp2 domains, GQD absorbs very little in the visible light region and appears to be yellowish brown solution. The GQD solution displays yellow fluorescence under UV irradiation at 365 nm, and presents peaks at 445 nm and 503 nm in the fluorescence spectroscopy upon 350 nm of excitation. As had been reported in the literature [21], the peak position of GQD fluorescence depends on the excitation wavelength with variation of 100 nm. Using quinine sulfate as reference, we measured the quantum yield of the GQD samples of 3%, which is fair compared to the previous reports [14,21,29,35,37] and is reasonable for a highly yielded production. TEM image (Fig. 1c) shows that GQD sizes vary from 3 to 5 nm, and AFM (Fig. 1d) indicates GQD height of 0.5e1 nm, which corresponds to 1-2 graphene layers. These results suggest the planar structure of GQD. Elemental analysis, XPS, FTIR and TGA (Fig. S1) were utilized to reveal the chemical component of the GQD sample. Elemental analysis and XPS data of GQD indicate that GQD contains high content of O (36%, which means about 16:7 of C/O molar ratio) and
For cytotoxicity assessment of the GQD sample, cell viability, cell apoptosis and necrosis, LDH and ROS level were examined to systematically evaluate the effect of GQD on cell growth. WST-1 assay, which reveals the mitochondrial function of cells from the level of succinate dehydrogenase firstly demonstrated no toxicity of GQDPEG to HeLa cells and A549 cells. Fig. 2a indicates that over 95% of HeLa cells remained alive after 24 h of incubation with GQD-PEG, even when GQD concentration went up to 160 mg/mL. Similar phenomena were found in the case of A549 cells (Fig. S2), in which GQD-PEG shows about 85% of cell viability at 640 mg/mL of GQD concentration. Moreover, immunochemistry experiment for apoptosis and necrosis further demonstrated negligible effects of GQD at the cellular level. As shown in Fig. 2b, GQD did not induce any apoptosis or necrosis of HeLa cells even at the concentration of 160 mg/mL of GQD, and the apoptosis levels are irrelevant to the dose of the GQD. These results are different from that of Yang group [35] and Lee group [29] who reported little toxicity of GQD (80 w 90% of cell viability at low dose) to various cell lines, respectively, but are consistent with low toxicity of hyaluronic acid derivatives of carbon dots at low concentrations [30]. Therefore, low cytotoxicity may come from the PEGylation or the inherent properties of the GQD sample. Considering low toxicity of bare GQD sample (Fig. S3), we ascribed the good biocompatibility of GQD to their high oxygen content (w36%, as elemental analysis and XPS data have showed). Other techniques including LDH assay and ROS production were utilized to further study the cytotoxicity of GQD-PEG. While graphene and graphene oxide generally impact cell membrane integrity [38e40] and oxidative stress [41], GQD presented comparable LDH release and ROS level to control group, according to LDH and ROS assay (Fig. 2c, d). These results also suggest that GQD shows no noticeable cytotoxicity, possibly due to the high oxygen content of GQD.
Fig. 1. (a), GQD samples were obtained through oxidation from graphite; (b), UVevis spectroscopy (dotted line) shows strong absorption in UV region, and emission spectra (solid lines) irradiated at 350 (black), 400 (blue) and 488 nm (red) respectively indicate excitation-dependent fluorescence of GQD; (c), TEM image suggests 3e5 nm of lateral sizes of GQD; (d) AFM image and the height diagram (inset) of a GQD sample present 0.5e1 nm of GQD height. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
5044
Y. Chong et al. / Biomaterials 35 (2014) 5041e5048
Fig. 2. Results from (a) WST-1 assay, (b) cell apoptosis and necrosis, (c) LDH assay, and (d) ROS generation of GQD all suggest no obvious toxicity of GQD-PEG to HeLa cells at GQD concentration as high as 160 mg/mL. Values represent means SE, n ¼ 3.
3.3. In vivo biodistribution and clearance of GQD In vivo biodistribution of GQD can provide essential information of GQD’s behavior post administration, such as accumulation sites and clearance routes. In order to achieve this goal with high sensitivity, Cy7, a commonly used NIR fluorescent dye, was anchored on GQD-PEG to fully understand GQD’s behaviors in vivo. Balb/c mice were intravenously (i.v.) or intraperitoneally (i.p.) injected with Cy7-labeled GQD-PEG (200 mL of 1.5 mg/mL solution for each mouse; at a dose of 15 mg/kg), and then sacrificed post 1, 4, 12, 24 and 48 h, respectively. The organs and tissues of mice were harvested and imaged under irradiation at 710 nm. The calculated results, by averaging Cy7 fluorescence intensity of the organs for a semiquantitative biodistribution analysis, suggested that the biodistribution and clearance of GQD-PEG were independent of administration routes. As shown in Fig. 3 and Fig. S4, GQD-PEG mainly accumulates in kidneys and tumor sites compared with the control group, and leaked from the kidneys very quickly. It should be noted that the observation of very high content of GQD in kidneys and low intensity in the other organs in the present work is different from that of the carbon-based nanoparticles in other studies. For example, Lee et al. [29,42] concluded that GQD would disperse in most organs including liver, kidneys, lung and spleen; Hahn et al. [30], Liu et al. [43] and Chen et al. [31] separately found that polymer-coated carbon dots were typically metabolized through liver (into bile) and kidneys (into urine), and that the particles were cleared out in 24 h, especially in the first hour regardless of injection routes; Liu group [13,16] reported that PEGylated GO accumulated and stayed mainly in liver and spleen for more than one month. Based on the facts of strong fluorescence in urine (data not shown) and the absence of GQD in reticuloendothelial system (RES), we think that the differences between our
GQD-PEG and other similar carbon nanomaterials originate from the smaller size and higher O content of GQD. In addition, sustained Cy7 fluorescence was observed in tumor, suggesting PEG-induced accumulation of GQD-PEG due to the Enhanced Permeability and Retention (EPR) effect [44]. Overall, the fast clearance in kidneys and low dispersion in the other organs demonstrate good biocompatibility of GQD-PEG. 3.4. In vivo toxicity In order to further investigate the biosafety of GQD-PEG, multiple-dosing which simulates clinical drug administration was applied to the study of in vivo toxicity, since single-dosing experiment mostly presented low toxicity of nanomaterials without obvious accumulation [29e31,43]. In our experiment, the mice were intravenously (i.v.) or intraperitoneally (i.p.) injected with 20 mg/kg of GQD-PEG every other day for 14 days, and then monitored for up to 40 days. GO-PEG (10e30 mm wide and 0.5e 2 nm in height, Fig. S5) was also examined for comparison purpose, and only administration of i.p. injections was tested since the toxicity of GO-PEG upon i.v. injection had been studied in details [16]. Fig. 4 shows that no obvious difference upon various administration routes of GQD-PEG was found compared to the control group, and in each case the mice survived; however, 3/12 of mice injected with GO-PEG died, each after the 4th, 5th and 6th administration, despite that the live animals displayed no significant loss of body weight compared to the GQD groups and the control group. These mice died without any sign, and presented no damage to main organs but dark livers and spleens. This is the first report of mice dying of GO-PEG although people have doubted the biosafety of GO-PEG (based on the aqueous instability and aggregation in organs) for a long time. [16,19] The design of multiple-
Y. Chong et al. / Biomaterials 35 (2014) 5041e5048
5045
Fig. 5. The weight indexes of main organs (kidneys, liver, and spleen) show the abnormal livers and spleens (p < 0.05 compared to control group) collected at day 40 from mice injected with GO-PEG, and demonstrate no obvious toxicity of GQD-PEG. Values represent means SE, n ¼ 6.
Fig. 3. (a), PL intensities of livers (black), spleens (red), kidneys (cyan), lungs (blue), hearts (green) and tumors (purple) indicate the GQD concentrations in these organs collected at various time points from mice i.v. injected with GQD-PEG-Cy7; (b), the photographs of kidneys and tumors under NIR illumination clearly demonstrate the changes of GQD concentration in these organs. Values represent means SE, n ¼ 3. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
dosing experiments is the key to these unusual and surprising results, offering a deep insight into the toxicity of nanomaterials. By doing seven injections of samples, we demonstrated that GQD-PEG possessed superior biocompatibility to GO-PEG, possibly due to the fast clearance of ultra-small GQD-PEG.
The main organs of mice were collected post 1 or 40 days of seven injections with the nanomaterials. The organs from mice treated with GQD-PEG display no clear difference with the control group from their appearance (Fig. S6); on the contrary, the livers and spleens of mice with GO-PEG look darker than the others, in consistent with the aggregation of GO in animal organs reported previously by Liu et al. [18]. In addition, the weight indexes (Fig. 5) of livers and spleens from mice with GO-PEG are larger than the others, indicating chronic damage of GO-PEG to the livers and spleens of mice due to slow clearance of GO-PEG [16]. To further investigate the origin of the damages and expose whether GQDPEG caused structural damage, the organs were sliced and stained with hematoxylin and eosin (H&E). As Fig. 6 and Fig. S8 show, a lot of dark spots with tens of micrometers in diameter (much larger than the size of GO sheets) appear in the livers and spleens from mice with GO-PEG, while nothing abnormal can be observed in the organs from mice with GQD-PEG. These dark spots could diminish over time and may be the aggregation of GO-PEG, similar to what Liu and his colleagues observed [18]. These evidences prove that GO-PEG aggregation may induce significant damage to the organs of mice, even cause the death of animals. In
Fig. 4. (a), mice were injected seven times with GQD-PEG and GO-PEG, respectively, and three of them died from multiple injections of GO-PEG. Half of survived mice were sacrificed post 1 day of injections for the study of short-term toxicity; (b), the survived mice were weighted every three days and no difference from the control group was observed. Values represent means SE, n ¼ 6.
5046
Y. Chong et al. / Biomaterials 35 (2014) 5041e5048
Fig. 6. H&E stained tissue slices (liver, kidney, and spleen) of mice injected with GO-PEG (i.p.) and GQD-PEG (i.p. or i.v.) at the dose of 20 mg/kg per time. Mice were sacrificed at day 1 post seven injections in 14 days. While aggregated GO was found in the main organs (indicated by golden circle) of mice injected with GO-PEG, no black dots were found in the groups of GQD-PEG. Bars indicated by white stripes in the image are 50 mm. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
addition, GQD-PEG shows excellent biocompatibility, due to their small size and thus fast clearance from kidneys. To make a comparative study on the potential toxic effect of GOPEG and GQD-PEG on the treated mice, we carried out blood biochemistry and hematology analysis. Fig. 7 and Fig. S9 show the levels of standard hematology markers, such as white blood cell (WBC), red blood cell (RBC), hematocrit (HCT), mean corpuscular volume (MCV), hemoglobin (HGB), platelet (PLT), mean corpuscular hemoglobin (MCH), and mean corpuscular hemoglobin concentration (MCHC). The WBC is sensitive to the physiological response, but the values are still in the normal range despite of significant differences between groups. All the other parameters in the
nanomaterial-treated groups appear to be normal compared with the control groups, and be within the normal ranges [45]. In addition, we present the biochemistry results of the mice in Figs. S8 and S10, including total protein (TP), alanine transaminase (ALT), aspartate transaminase (AST), albumin (ALB), globulin (GLOB), the ratio of albumin and globulin (A/G), blood urea nitrogen (BUN), and creatinine (CREA). The values of ALT, AST, BUN, and CREA fluctuate a little compared to the control groups but they still remained within the normal ranges. The results of blood biochemistry and hematology analysis suggest no obvious toxicity of GQD-PEG and GOPEG. It is interesting that the survived mice with GO-PEG showed no damage from the materials, which agrees well with previous
Fig. 7. Hematology results of the GQD-PEG and GO-PEG treated mice post 1 day of injections. These results show mean and standard deviation of white blood cells (WBC), red blood cells (RBC), hematocrit (HCT), mean corpuscular volume (MCV), hemoglobin (HGB), platelets (PLT), mean corpuscular hemoglobin (MCH), and mean corpuscular hemoglobin concentration (MCHC). Bars represent mean standard deviation. Values represent means SE, n ¼ 3.
Y. Chong et al. / Biomaterials 35 (2014) 5041e5048
5047
Fig. 8. Blood biochemistry results of the GQD-PEG and GO-PEG treated mice post 1 day of injections. These results show mean and standard deviation of total protein (TP), alanine transaminase (ALT), aspartate transaminase (AST), albumin (ALB), globulin (GLOB), the ratio of albumin and globulin (A/G), blood urea nitrogen (BUN), and creatinine (CREA). Bars represent mean standard deviation. Values represent means SE, n ¼ 3.
study by other scientists [16,18], and suggests the need of more investigation on the toxicity of GO-PEG (see Fig. 8). 4. Conclusions We have studied the synthesis, characterization, in vitro toxicity, in vivo biodistribution and toxicity of a GQD sample. Analysis of WST-1 assay, cell apoptosis, LDH production and ROS level clearly demonstrated good biocompatibility of GQD and GQD-PEG at cellular level. NIR imaging of the mice administrated with GQD-PEG suggested that GQD could be metabolized quickly through kidneys and be retained in the tumor, regardless of injection routes. With multiple-dosing experiments of the nanomaterials, we found for the first time that GO-PEG showed some in vivo toxicity, probably due to GO aggregation and thus thrombus in mice. Our results, from either died or survived mice, also exposed GO-PEG accumulation in reticuloendothelial system. In contrast, mice injected with GQDPEG presented excellent biocompatibility, based on the mice growth, organ appearance and slices, as well as hematology and blood chemistry analyses. Taking together, GQD exhibits no apparent in vitro and in vivo toxicity of GQD, resulting from its small size and high O content compared with that of the widely used GOPEG. These findings may facilitate the development of safe and efficient GQD-based bio-imaging and sensing platform, and drug delivery systems. Acknowledgment The authors are grateful to the financial support from National Natural Science Foundation of China (No. 51361130033) and the Ministry of Science and Technology of China (Nos. 2010CB933504, 2014CB965003), and the Natural Science Foundation of Jiangsu Province (No. BK20130358). Y. Ma appreciates the Jiangsu Planned Projects for Postdoctoral Research Funds. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.biomaterials.2014.03.021
References [1] Baughman RH, Zakhidov AA, de Heer WA. Carbon nanotubesethe route toward applications. Science 2002;297:787e92. [2] Geim AK, Novoselov KS. The rise of graphene. Nat Mater 2007;6:183e91. [3] Chung DDL. Electrical applications of carbon materials. J Biomater Sci 2004;39: 2645e61. [4] Zhu Y, Murali S, Cai W, Li X, Suk JW, Potts JR, et al. Graphene and graphene oxide: synthesis, properties, and applications. Adv Mater 2010;22:3906e24. [5] Compton OC, Nguyen ST. Graphene oxide, highly reduced graphene oxide, and graphene: versatile building blocks for carbon-based materials. Small 2010;6: 711e23. [6] Kostarelos K, Bianco A, Prato M. Promises, facts and challenges for carbon nanotubes in imaging and therapeutics. Nat Nanotechnol 2009;4:627e33. [7] Shen H, Zhang L, Liu M, Zhang Z. Biomedical applications of graphene. Theranostics 2012;2:283e94. [8] Chung C, Kim Y-K, Shin D, Ryoo S-R, Hong BH, Min D-H. Biomedical applications of graphene and graphene oxide. Acc Chem Res 2013;46:2211e24. [9] Liu Z, Tabakman SM, Chen Z, Dai H. Preparation of carbon nanotube bioconjugates for biomedical applications. Nat Protoc 2009;4:1372e81. [10] Mendes RG, Bachmatiuk A, Buchner B, Cuniberti G, Rummeli MH. Carbon nanostructures as multi-functional drug delivery platforms. J Mater Chem B 2013;1:401e28. [11] Yang W, Ratinac KR, Ringer SP, Thordarson P, Gooding JJ, Braet F. Carbon nanomaterials in biosensors: should you use nanotubes or graphene? Angew Chem Int Ed 2010;49:2114e38. [12] Robinson J, Welsher K, Tabakman S, Sherlock S, Wang H, Luong R, et al. High performance in vivo near-IR (>1 mm) imaging and photothermal cancer therapy with carbon nanotubes. Nano Res 2010;3:779e93. [13] Yang K, Zhang S, Zhang G, Sun X, Lee S-T, Liu Z. Graphene in mice: ultrahigh in vivo tumor uptake and efficient photothermal therapy. Nano Lett 2010;10: 3318e23. [14] Markovic ZM, Ristic BZ, Arsikin KM, Klisic DG, Harhaji-Trajkovic LM, Todorovic-Markovic BM, et al. Graphene quantum dots as autophagy-inducing photodynamic agents. Biomaterials 2012;33:7084e92. [15] Liu Z, Cai W, He L, Nakayama N, Chen K, Sun X, et al. In vivo biodistribution and highly efficient tumour targeting of carbon nanotubes in mice. Nat Nanotechnol 2007;2:47e52. [16] Yang K, Wan J, Zhang S, Zhang Y, Lee S-T, Liu Z. In vivo pharmacokinetics, long-term biodistribution, and toxicology of PEGylated graphene in mice. ACS Nano 2010;5:516e22. [17] Liu Z, Davis C, Cai W, He L, Chen X, Dai H. Circulation and long-term fate of functionalized, biocompatible single-walled carbon nanotubes in mice probed by Raman spectroscopy. Proc Natl Acad Sci 2008;105:1410e5. [18] Yang K, Gong H, Shi X, Wan J, Zhang Y, Liu Z. In vivo biodistribution and toxicology of functionalized nano-graphene oxide in mice after oral and intraperitoneal administration. Biomaterials 2013;34:2787e95. [19] Bianco A. Graphene: safe or toxic? The two faces of the medal. Angew Chem Int Ed 2013;52:4986e97. [20] Luo PG, Sahu S, Yang S-T, Sonkar SK, Wang J, Wang H, et al. Carbon “quantum” dots for optical bioimaging. J Mater Chem B 2013;1:2116e27. [21] Peng J, Gao W, Gupta BK, Liu Z, Romero-Aburto R, Ge L, et al. Graphene quantum dots derived from carbon fibers. Nano Lett 2012;12:844e9.
5048
Y. Chong et al. / Biomaterials 35 (2014) 5041e5048
[22] Shen J, Zhu Y, Yang X, Li C. Graphene quantum dots: emergent nanolights for bioimaging, sensors, catalysis and photovoltaic devices. Chem Commun 2012;48:3686e99. [23] Shang J, Ma L, Li J, Ai W, Yu T, Gurzadyan GG. The origin of fluorescence from graphene oxide. Sci Rep 2012;2:792. [24] Song Y, Shi W, Chen W, Li X, Ma H. Fluorescent carbon nanodots conjugated with folic acid for distinguishing folate-receptor-positive cancer cells from normal cells. J Mater Chem 2012;22:12568e73. [25] Zhu S, Tang S, Zhang J, Yang B. Control the size and surface chemistry of graphene for the rising fluorescent materials. Chem Commun 2012;48: 4527e39. [26] Zheng XT, Than A, Ananthanaraya A, Kim D-H, Chen P. Graphene quantum dots as universal fluorophores and their use in revealing regulated trafficking of insulin receptors in adipocytes. ACS Nano 2013;7:6278e86. [27] Zhang L, Xing Y, He N, Zhang Y, Lu Z, Zhang J, et al. Preparation of graphene quantum dots for bioimaging application. J Nanosci Nanotechnol 2012;12: 2924e8. [28] Wu C, Wang C, Han T, Zhou X, Guo S, Zhang J. Insight into the cellular internalization and cytotoxicity of graphene quantum dots 2013;2:1613e9. [29] Nurunnabi M, Khatun Z, Huh KM, Park SY, Lee DY, Cho KJ, et al. In vivo biodistribution and toxicology of carboxylated graphene quantum dots. ACS Nano 2013;7:6858e67. [30] Goh EJ, Kim KS, Kim YR, Jung HS, Beack S, Kong WH, et al. Bioimaging of hyaluronic acid derivatives using nanosized carbon dots 2012;13:2554e61. [31] Huang X, Zhang F, Zhu L, Choi KY, Guo N, Guo J, et al. Effect of injection routes on the biodistribution, clearance, and tumor uptake of carbon dots. ACS Nano 2013;7:5684e93. [32] Zhang L, Xia J, Zhao Q, Liu L, Zhang Z. Functional graphene oxide as a nanocarrier for controlled loading and targeted delivery of mixed anticancer drugs. Small 2010;6:537e44. [33] Zhang L, Lu Z, Zhao Q, Huang J, Shen H, Zhang Z. Enhanced chemotherapy efficacy by sequential delivery of siRNA and anticancer drugs using PEI grafted graphene oxide. Small 2011;7:460e4.
[34] Zhang Y, Chen B, Zhang L, Huang J, Chen F, Yang Z, et al. Controlled assembly of Fe3O4 magnetic nanoparticles on graphene oxide. Nanoscale 2011;3:1446e50. [35] Zhu S, Zhang J, Qiao C, Tang S, Li Y, Yuan W, et al. Strongly greenphotoluminescent graphene quantum dots for bioimaging applications. Chem Commun 2011;47:6858e60. [36] Saxena S, Tyson TA, Shukla S, Negusse E, Chen H, Bai J. Investigation of structural and electronic properties of graphene oxide. Appl Phys Lett 2011;99:013104-1e013104-3. [37] Feng L, Liu Z. Graphene in biomedicine: opportunities and challenges. Nanomedicine 2011;6:317e24. [38] Li Y, Yuan H, von dem Bussche A, Creighton M, Hurt RH, Kane AB, et al. Graphene microsheets enter cells through spontaneous membrane penetration at edge asperities and corner sites. Proc Natl Acad Sci 2013;110:12295e300. [39] Tu Y, Lv M, Xiu P, Huynh T, Zhang M, Castelli M, et al. Destructive extraction of phospholipids from Escherichia coli membranes by graphene nanosheets. Nat Nanotechnol 2013;8:594e601. [40] Wang A, Pu K, Dong B, Liu Y, Zhang L, Zhang Z, et al. Role of surface charge and oxidative stress in cytotoxicity and genotoxicity of graphene oxide towards human lung fibroblast cells. J Appl Toxicol 2013;33:1156e64. [41] Chang Y, Yang S-T, Liu J-H, Dong E, Wang Y, Cao A, et al. In vitro toxicity evaluation of graphene oxide on A549 cells. Toxicol Lett 2011;200:201e10. [42] Nurunnabi M, Khatun Z, Nafiujjaman M, Lee D-G, Lee Y-K. Surface coating of graphene quantum dots using mussel-inspired polydopamine for biomedical optical imaging. ACS Appl Mater Interfaces 2013;5:8246e53. [43] Tao H, Yang K, Ma Z, Wan J, Zhang Y, Kang Z, et al. In vivo NIR fluorescence imaging, biodistribution, and toxicology of photoluminescent carbon dots produced from carbon nanotubes and graphite. Small 2012;8:281e90. [44] Seymour L. Passive tumor targeting of soluble macromolecules and drug conjugates. Crit Rev Ther Drug Carrier Syst 1991;9:135e87. [45] Reference ranges of hematology data of healthy female Balb/c mice were obtained from Charles River Laboratories (updated in 2011): http://www. criver.com/files/pdfs/rms/balbc/rm_rm_r_balb-c_mouse_clinical_pathology_ data.aspx.
Contents lists available at ScienceDirect
Biomaterials journal homepage: www.elsevier.com/locate/biomaterials
The in vitro and in vivo toxicity of graphene quantum dots Yu Chong a, b,1, Yufei Ma a,1, He Shen a, Xiaolong Tu a, Xuan Zhou a, Jiaying Xu b, Jianwu Dai a, Saijun Fan b, c, **, Zhijun Zhang a, * a Suzhou Key Laboratory of Nanobiomedicine, Division of Nanobiomedicine & Collaborative Innovation Center of Suzhou Nano Science and Technology, Suzhou Institute of Nano-tech and Nano-bionics, Chinese Academy of Sciences, 398 Ruoshui Road, Suzhou 215123, China b School for Radiological & Interdisciplinary Sciences, Medicine College of Soochow University, Suzhou 215123, China c Institute of Radiation Medicine, Chinese Academy of Medical Sciences, Tianjin 300192, China
a r t i c l e i n f o
a b s t r a c t
Article history: Received 21 January 2014 Accepted 11 March 2014 Available online 28 March 2014
Graphene quantum dots (GQD) generate intrinsic fluorescence, and improves aqueous stability of graphene oxide (GO) while maintaining wide chemical adaptability and high adsorption capacity. Despite GO’s remarkable advantages in bio-imaging, bio-sensing and other biomedical applications, its biosafety issues are still unclear. Here we report a detailed and systematic study on the in vitro and in vivo toxicity of GQD. The GQD sample was prepared through a facile oxidation approach and fully characterized by means of AFM, TEM, FTIR, XPS and elemental analysis. In vitro experiments showed that GQD exhibits very low cytotoxicity owing to its ultra-small size and high oxygen content. Then, the in vivo biodistribution experiment of GQD revealed no material accumulation in main organs of mice and fast clearance of GQD through kidney. In order to mimic clinic drug administration, mice were injected with GQD and GO (as comparison) multiple times for in vivo toxicity tests. We found that GQD showed no obvious influence on mice owing to its small size, while GO appeared toxic, even caused death to mice due to GO aggregation inside mice. In brief, GQD possesses no obvious in vitro and in vivo toxicity, even under multi-dosing situation. Ó 2014 Published by Elsevier Ltd.
Keywords: Graphene quantum dots Cytotoxicity In vivo test In vitro test Biocompatibility
1. Introduction Carbon materials have excited scientists and engineers for decades [1,2], not only due to their various allotropes and abundant availability, but also because of the exciting electronic, mechanical and physicochemical properties of versatile carbon structures. Among various applications of carbon nanomaterials [3e5], the biomedical applications [6e8] especially drug delivery [9,10] and diagnostic potential [11] of the carbon structures, graphene and carbon nanotube in particular, have been attracting increasing interests in recent years. Furthermore, carbon materials have been exploited with advanced therapeutic techniques such as photothermal [12,13] and photodynamic [14] therapy. However, pristine carbon-based nanomaterials are hydrophobic related to benzene
* Corresponding author. Tel.: þ86 512 62872556; fax: þ86 512 62603079. ** Corresponding author. Institute of Radiation Medicine, Chinese Academy of Medical Sciences, Tianjin 300192, China. Tel.: þ86 22 85685301; fax: þ86 22 85683033. E-mail addresses: [email protected] (S. Fan), [email protected] (Z. Zhang). 1 These authors contributed equally to the manuscript. http://dx.doi.org/10.1016/j.biomaterials.2014.03.021 0142-9612/Ó 2014 Published by Elsevier Ltd.
structures, which leads to aqueous instability and thus restricts direct use in biological applications. A popular way to stabilize these carbon nanomaterials is to functionalize them with hydrophilic polymers (mostly polyethylene glycol, PEG), and it has been proved to be successful with no obvious toxicities both in vitro and in vivo [15,16]. In addition, animal studies suggested that PEGylated carbon nanotubes or graphene accumulated in the main organs and stayed for over one month at an administration dose of 20 mg/ kg [17,18], raising much concern about toxicity tolerance despite of no toxic evidence of carbon materials to mice [19]. Another strategy to improve the aqueous solubility and stability of carbon materials is to make ultra-small carbon nanoparticles, so that Brownian motion of the nanoparticles can provide sufficient energy to prevent the aggregation, and that nanoparticles become less hydrophobic due to the Oxygen-rich functional groups at the edges. Thus, nanometer-sized carbon materials, carbon quantum dots (Cdots) [20] and especially graphene quantum dots (GQD) [21], are rising stars in the family of carbon materials because of their excellent solubility and stability in water, strong fluorescence, and the retained advantages of graphene. Rapidly increasing number of reports about GQD have explored the synthesis methods (from oxidation, hydrolysis to electrolysis) [22], the origin of
5042
Y. Chong et al. / Biomaterials 35 (2014) 5041e5048
fluorescence [23], and biological applications such as imaging and sensing [22,24e26]. Although GQD have shown great potential in biomedical applications [14,22,27], the toxicity of GQD at in vitro and in vivo level is still unclear. Only quite few reports studied the in vitro and in vivo toxicity of GQD [28,29], while several examples investigated the in vivo distribution and toxicity of carbon dots [30,31]. In addition, there is no report on whether size of graphene-based nanomaterials relates to the toxicity in the region of sub-50-nm. Thus, more profound and systematical study on the toxicity of GQD and GO at in vitro and in vivo level is highly desired for its practical applications. Our group has made many efforts to develop graphene oxide (GO) for versatile theranostic agents [7,32e34], and we also found that GQD appeared to be a better drug delivery platform than GO due to its high water solubility and stability [27]. Before applying GQD to biomedical applications, here we report the in vitro and in vivo toxicity of a GQD sample. The GQD sample was generated by a facile synthetic method with high yield, and their chemical structure was fully characterized to reveal the structure-toxicity relationship. We utilized several cellular assays (WST-1, cell apoptosis, LDH, ROS) to analyze the effect of GQD on the cells, and then systematically test the in vivo biodistribution and toxicity of GQD. For in vivo toxicity evaluation, multiple doses of GQD were injected into mice in order to mimic clinic drug administration and to practically study the effect of GQD on mice [29,30]. GO of 10e 30 nm was utilized for comparison on size effect in the in vivo experiments because the toxicity of GO is relatively clearer than that of GQD [13,16,18]. Through a close look into the toxicity of GQD at the cellular and animal levels, these findings will reveal, in the region of sub-50-nm, the size effect of graphene materials on their toxicity, and inspire people to develop biologically compatible nanomaterials. 2. Materials and methods 2.1. Synthesis and characterization 2.1.1. Synthesis of GQD GQD was obtained through modified oxidative cutting method reported in literature [21]. Generally, 0.2 g of graphite powder (Alfa Aesar) was added into a mixture of 15 mL of concentrated H2SO4 and 5 mL of fuming HNO3. The reaction mixture was sonicated for 10 min, and then stirred for 30 min at 120 C. The mixture was cooled, diluted with 100 mL of deionized (DI) water, and then neutralized to pH 7e8 with Na2CO3. In order to remove salts, the resulted solution was filtered and further dialyzed in a dialysis bag (50 kDa cutoff) for 2 days. After concentrating the solution on a rotary evaporator and filtration through 0.22 mm filters, a black solution of GQD (4 mg/mL, measured through absorption at 230 nm) was obtained. Production yield was calculated to be about 60%, which was much higher than that of previous reports [21,22,29,35]. 2.1.2. PEGylation of GQD 20 mg of GQD was diluted with 50 mM phosphate buffer (pH 5.5) to the concentration of ca. 0.2 mg/mL, and activated with 20 mg of 1-ethyl-3-(3dimethylaminopropyl)-carbodiimide (EDC, Sigma). After 15 min, 200 mg of PEG (2000 Da, amine terminus, Jenkem Technology) with amine terminus was added into the solution and the solution was further stirred for 2 h for complete reaction. At last, the solution was concentrated and rinsed with water through spin-dialysis (3000 Da cutoff, 5000 rpm) to generate GQD-PEG. 2.1.3. Cy7 labeling of GQD 20 mg of EDC was mixed with 100 mL of 50 mM phosphate buffer (pH 5.5) containing 2 mg of Cy7 (carboxyl terminus, Fanbo Biochemicals Co. Ltd., Beijing) for 15 min of COOH activation, and then 20 mg of GQD-PEG was added into the solution. The solution was further stirred for 2 h, before it was concentrated and rinsed with water through spin-dialysis (3000 Da cutoff, 5000 rpm) to generate GQD-PEG-Cy7. 2.1.4. Synthesis of GO-PEG GO was synthesized using Hummer’s method and then converted to carboxylated GO according to previous protocol [11]. 20 mL of GO solution at a concentration of 0.2 mg/mL was mixed with 24 mg of six-armed branched PEG (10 kDa, amine terminus, Jenkem Technology) and sonicated for 5 min, following by EDC additions (4 mg 3) under continuous sonication for additional 1 h. Large aggregates were removed through centrifugation at 13,000 rpm for 1 h, and then the supernatant
was collected and washed five times through spin-dialysis (100 kDa cutoff, 4000 rpm) to generate GO-PEG. 2.2. Material characterization UVeVis spectra were collected with a PerkinElmer Lambda 25 spectrophotometer, and fluorescence spectra were obtained on a Hitachi F-4600 fluorescence spectrophotometer. Atomic force microscope (AFM) images were conducted on a Vecco Dimension 3100 Atomic force microscope, and transmission electron microscopy (TEM) was a Tecnai G2 F20 S-Twin from FEI. CHN elemental analyses were performed on an Elementar Vario EL analyzer. Fourier transform infrared (FTIR) spectra were conducted within the 4000-400 cm1 wavenumber range using a Thermo Nicolet 6700 FTIR spectrometer with the KBr pellet technique. X-ray photoelectron spectroscopy (XPS) spectra were acquired with a Kratos Axis UltraDLD spectrometer using a monochromatic Al Ka source at 1486.6 eV. Thermogravimetric analysis (TGA) was carried out with a Seiko TGA 6200 at a heating rate of 10 C/min under nitrogen. 2.3. Biological tests 2.3.1. Cell viability The cell viability was evaluated with WST-1 (Beyotime, China). Briefly, HeLa (or A549) cells were seeded on each well of 96-well plate at a density of 8 103 cells per well, and cultured in a humidified 5% CO2 incubator at 37 C for 24 h. Then GQD or GQD-PEG was introduced into cells with various test concentrations (10, 20, 40, 80 and 160, 320, and 640 mg/mL) in culture medium, and cells cultured without nanomaterials were taken as the control. The cells were washed with PBS after 24 h of incubation, following by addition of 100 mL of fresh culture medium and 10 mL of WST-1 into each well and additional 1 h of incubation at 37 C. At last, the optical density (OD) of each well at 450 nm was recorded on a PerkinElmer Victor 4 microplate reader. 2.3.2. Cell apoptosis and necrosis Cell apoptosis and necrosis were examined using the annexin V and PI assay kit (Beyotime, China). In brief, GQD-PEG at concentrations ranging from 10 to 160 mg/ mL in culture medium was added into HeLa cells precultured for 24 h and exposed for another 24 h. After that, the cells were washed three times with PBS, then trypsinized and subsequently washed with PBS once, resuspended in 300 mL PBS, and incubated with 5 mL Annexin V-FITC and 1 mL PI for 10 min at room temperature. At last, the samples were diluted with 200 mL of binding buffer and analyzed with a Beckman Coulter FC500 flow cytometer (>1 104 cells for each sample). Excitation wave was set at 488 nm, and the fluorescence intensities of Annexin V (green) and PI (red) was collected at 530 and 575 nm, respectively. 2.3.3. LDH Release and oxidative stress For LDH (lactate dehydrogenase) assay, HeLa cells seeded in a 96-well plate were incubated with different concentrations of GQD-PEG at 37 C for 24 h, and then 20 mL of LDH standard solution (Beyotime, China) was added into each well for additional 1 h incubation at 37 C. After that, 100 mL of cell supernatant was aspirated and transferred into another plate according to the instruction, and the absorbance of supernatant was examined at 490 nm to analyze the level of LDH release. ROS (Reactive Oxygen Species) was measured with a procedure similar to the LDH assay. After incubating with GQD-PEG at various concentrations, the cells were washed three times and mixed with 1 mL of 10 mM DCFH-DA (Beyotime, China) for additional incubation for 20 min. Next, cells were washed three times with DMEM without FBS to eliminate the DCFH-DA outside cell membrane, and then collected in suspension and measured with a flow cytometer (excitation at 488 nm and emission at 525 nm). 2.3.4. Biodistribution All animal experiments were conducted under protocols approved by the Soochow University Laboratory Animal Center. Healthy female Balb/c mice (Suzhou Industrial Park Animal Technology Co., Ltd.) at an average age of 6e7 weeks were housed under a standard condition of a 12 h light/dark cycle with free access to food. After injected with 4T1 cells (108 cells/mL 100 mL) and housed for 3 days, Balb/c mice bearing 4T1 murine breast cancer tumors were intravenously (i.v.) or intraperitoneally (i.p.) injected with GQD-PEG-Cy7 (200 mL of 1.5 mg/mL solution for each mouse; a dose of 15 mg/kg), and then were sacrificed at time points of 1, 4, 12, 24 or 48 h after injection. Various organs and tissues were collected and spectrally imaged by the IVISÒ Lumina imaging system. The averaged fluorescence intensity of Cy7 of each imaged organ was calculated for a semiquantitative biodistribution analysis. 2.3.5. In vivo toxicity assessment To simulate practical drug administration, multiple-dosing toxicity test was performed to further investigate the biosafety of GQD-PEG. 48 healthy female Balb/c mice (6e7 weeks) were randomly divided into four different experimental groups, including groups of control, GQD-PEG by i.p., GQD-PEG by i.v. and GO-PEG by i.p. injection. The mouse was injected with the materials (20 mg/kg) or saline every other day for 14 days. Then, half of them were killed 1 day post injection for acute toxicity study, and the others were further monitored for over a month. Mice were
Y. Chong et al. / Biomaterials 35 (2014) 5041e5048 weighted every 3 days during 2 months of the experiment, and finally sacrificed for further tests. About 0.8 mL of blood from each mouse was collected for blood chemistry tests and complete blood panel analysis before the mouse was euthanatized. Major organs from those mice were harvested, fixed in 4% neutral buffered formalin, processed routinely into paraffin, sectioned at 8 mm, stained with hematoxylin and eosin (H&E) and examined under a digital microscope. Statistics were based on standard deviations of 6 mice per group.
5043
a large number of OH groups (3% of H content), which is further agreed by IR peaks at w3400 and w3300 cm1. This high O and OH content can be possible only if the graphene sheet is mono-layered and the lateral size is smaller than 5 nm, as have been proved by TEM and AFM. The small size and high O content of GQD in turn help them to be solubilized and highly stabilized in water or serum, implying that GQD would present good biocompatibility.
3. Results and discussion 3.2. Cellular toxicity 3.1. Synthesis and characterization of GQD The GQD sample was obtained through modified oxidative cutting method (Fig. 1a) [21]. Briefly, graphite powder in H2SO4/ HNO3 was heated at 120 C for 30 min, then the black solution was neutralized with Na2CO3 and dialyzed to obtain final GQD product with a yield as high as 60%. Compared with the previous studies [14,21,22,27,35], which focused on improving the fluorescence of GQD, this method offers a facile preparation with high production yield, and thus enables further exploration of GQD for widespread applications. Fig. 1b shows the UVevis and fluorescence spectra of the GQD sample. There are two distinguished peaks of GQD in the UV region, 230 nm (pep* transition from benzene structures) and 290 nm (se p* transitions from functional groups) [36]. Due to the small size of GQD and thus small sp2 domains, GQD absorbs very little in the visible light region and appears to be yellowish brown solution. The GQD solution displays yellow fluorescence under UV irradiation at 365 nm, and presents peaks at 445 nm and 503 nm in the fluorescence spectroscopy upon 350 nm of excitation. As had been reported in the literature [21], the peak position of GQD fluorescence depends on the excitation wavelength with variation of 100 nm. Using quinine sulfate as reference, we measured the quantum yield of the GQD samples of 3%, which is fair compared to the previous reports [14,21,29,35,37] and is reasonable for a highly yielded production. TEM image (Fig. 1c) shows that GQD sizes vary from 3 to 5 nm, and AFM (Fig. 1d) indicates GQD height of 0.5e1 nm, which corresponds to 1-2 graphene layers. These results suggest the planar structure of GQD. Elemental analysis, XPS, FTIR and TGA (Fig. S1) were utilized to reveal the chemical component of the GQD sample. Elemental analysis and XPS data of GQD indicate that GQD contains high content of O (36%, which means about 16:7 of C/O molar ratio) and
For cytotoxicity assessment of the GQD sample, cell viability, cell apoptosis and necrosis, LDH and ROS level were examined to systematically evaluate the effect of GQD on cell growth. WST-1 assay, which reveals the mitochondrial function of cells from the level of succinate dehydrogenase firstly demonstrated no toxicity of GQDPEG to HeLa cells and A549 cells. Fig. 2a indicates that over 95% of HeLa cells remained alive after 24 h of incubation with GQD-PEG, even when GQD concentration went up to 160 mg/mL. Similar phenomena were found in the case of A549 cells (Fig. S2), in which GQD-PEG shows about 85% of cell viability at 640 mg/mL of GQD concentration. Moreover, immunochemistry experiment for apoptosis and necrosis further demonstrated negligible effects of GQD at the cellular level. As shown in Fig. 2b, GQD did not induce any apoptosis or necrosis of HeLa cells even at the concentration of 160 mg/mL of GQD, and the apoptosis levels are irrelevant to the dose of the GQD. These results are different from that of Yang group [35] and Lee group [29] who reported little toxicity of GQD (80 w 90% of cell viability at low dose) to various cell lines, respectively, but are consistent with low toxicity of hyaluronic acid derivatives of carbon dots at low concentrations [30]. Therefore, low cytotoxicity may come from the PEGylation or the inherent properties of the GQD sample. Considering low toxicity of bare GQD sample (Fig. S3), we ascribed the good biocompatibility of GQD to their high oxygen content (w36%, as elemental analysis and XPS data have showed). Other techniques including LDH assay and ROS production were utilized to further study the cytotoxicity of GQD-PEG. While graphene and graphene oxide generally impact cell membrane integrity [38e40] and oxidative stress [41], GQD presented comparable LDH release and ROS level to control group, according to LDH and ROS assay (Fig. 2c, d). These results also suggest that GQD shows no noticeable cytotoxicity, possibly due to the high oxygen content of GQD.
Fig. 1. (a), GQD samples were obtained through oxidation from graphite; (b), UVevis spectroscopy (dotted line) shows strong absorption in UV region, and emission spectra (solid lines) irradiated at 350 (black), 400 (blue) and 488 nm (red) respectively indicate excitation-dependent fluorescence of GQD; (c), TEM image suggests 3e5 nm of lateral sizes of GQD; (d) AFM image and the height diagram (inset) of a GQD sample present 0.5e1 nm of GQD height. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
5044
Y. Chong et al. / Biomaterials 35 (2014) 5041e5048
Fig. 2. Results from (a) WST-1 assay, (b) cell apoptosis and necrosis, (c) LDH assay, and (d) ROS generation of GQD all suggest no obvious toxicity of GQD-PEG to HeLa cells at GQD concentration as high as 160 mg/mL. Values represent means SE, n ¼ 3.
3.3. In vivo biodistribution and clearance of GQD In vivo biodistribution of GQD can provide essential information of GQD’s behavior post administration, such as accumulation sites and clearance routes. In order to achieve this goal with high sensitivity, Cy7, a commonly used NIR fluorescent dye, was anchored on GQD-PEG to fully understand GQD’s behaviors in vivo. Balb/c mice were intravenously (i.v.) or intraperitoneally (i.p.) injected with Cy7-labeled GQD-PEG (200 mL of 1.5 mg/mL solution for each mouse; at a dose of 15 mg/kg), and then sacrificed post 1, 4, 12, 24 and 48 h, respectively. The organs and tissues of mice were harvested and imaged under irradiation at 710 nm. The calculated results, by averaging Cy7 fluorescence intensity of the organs for a semiquantitative biodistribution analysis, suggested that the biodistribution and clearance of GQD-PEG were independent of administration routes. As shown in Fig. 3 and Fig. S4, GQD-PEG mainly accumulates in kidneys and tumor sites compared with the control group, and leaked from the kidneys very quickly. It should be noted that the observation of very high content of GQD in kidneys and low intensity in the other organs in the present work is different from that of the carbon-based nanoparticles in other studies. For example, Lee et al. [29,42] concluded that GQD would disperse in most organs including liver, kidneys, lung and spleen; Hahn et al. [30], Liu et al. [43] and Chen et al. [31] separately found that polymer-coated carbon dots were typically metabolized through liver (into bile) and kidneys (into urine), and that the particles were cleared out in 24 h, especially in the first hour regardless of injection routes; Liu group [13,16] reported that PEGylated GO accumulated and stayed mainly in liver and spleen for more than one month. Based on the facts of strong fluorescence in urine (data not shown) and the absence of GQD in reticuloendothelial system (RES), we think that the differences between our
GQD-PEG and other similar carbon nanomaterials originate from the smaller size and higher O content of GQD. In addition, sustained Cy7 fluorescence was observed in tumor, suggesting PEG-induced accumulation of GQD-PEG due to the Enhanced Permeability and Retention (EPR) effect [44]. Overall, the fast clearance in kidneys and low dispersion in the other organs demonstrate good biocompatibility of GQD-PEG. 3.4. In vivo toxicity In order to further investigate the biosafety of GQD-PEG, multiple-dosing which simulates clinical drug administration was applied to the study of in vivo toxicity, since single-dosing experiment mostly presented low toxicity of nanomaterials without obvious accumulation [29e31,43]. In our experiment, the mice were intravenously (i.v.) or intraperitoneally (i.p.) injected with 20 mg/kg of GQD-PEG every other day for 14 days, and then monitored for up to 40 days. GO-PEG (10e30 mm wide and 0.5e 2 nm in height, Fig. S5) was also examined for comparison purpose, and only administration of i.p. injections was tested since the toxicity of GO-PEG upon i.v. injection had been studied in details [16]. Fig. 4 shows that no obvious difference upon various administration routes of GQD-PEG was found compared to the control group, and in each case the mice survived; however, 3/12 of mice injected with GO-PEG died, each after the 4th, 5th and 6th administration, despite that the live animals displayed no significant loss of body weight compared to the GQD groups and the control group. These mice died without any sign, and presented no damage to main organs but dark livers and spleens. This is the first report of mice dying of GO-PEG although people have doubted the biosafety of GO-PEG (based on the aqueous instability and aggregation in organs) for a long time. [16,19] The design of multiple-
Y. Chong et al. / Biomaterials 35 (2014) 5041e5048
5045
Fig. 5. The weight indexes of main organs (kidneys, liver, and spleen) show the abnormal livers and spleens (p < 0.05 compared to control group) collected at day 40 from mice injected with GO-PEG, and demonstrate no obvious toxicity of GQD-PEG. Values represent means SE, n ¼ 6.
Fig. 3. (a), PL intensities of livers (black), spleens (red), kidneys (cyan), lungs (blue), hearts (green) and tumors (purple) indicate the GQD concentrations in these organs collected at various time points from mice i.v. injected with GQD-PEG-Cy7; (b), the photographs of kidneys and tumors under NIR illumination clearly demonstrate the changes of GQD concentration in these organs. Values represent means SE, n ¼ 3. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
dosing experiments is the key to these unusual and surprising results, offering a deep insight into the toxicity of nanomaterials. By doing seven injections of samples, we demonstrated that GQD-PEG possessed superior biocompatibility to GO-PEG, possibly due to the fast clearance of ultra-small GQD-PEG.
The main organs of mice were collected post 1 or 40 days of seven injections with the nanomaterials. The organs from mice treated with GQD-PEG display no clear difference with the control group from their appearance (Fig. S6); on the contrary, the livers and spleens of mice with GO-PEG look darker than the others, in consistent with the aggregation of GO in animal organs reported previously by Liu et al. [18]. In addition, the weight indexes (Fig. 5) of livers and spleens from mice with GO-PEG are larger than the others, indicating chronic damage of GO-PEG to the livers and spleens of mice due to slow clearance of GO-PEG [16]. To further investigate the origin of the damages and expose whether GQDPEG caused structural damage, the organs were sliced and stained with hematoxylin and eosin (H&E). As Fig. 6 and Fig. S8 show, a lot of dark spots with tens of micrometers in diameter (much larger than the size of GO sheets) appear in the livers and spleens from mice with GO-PEG, while nothing abnormal can be observed in the organs from mice with GQD-PEG. These dark spots could diminish over time and may be the aggregation of GO-PEG, similar to what Liu and his colleagues observed [18]. These evidences prove that GO-PEG aggregation may induce significant damage to the organs of mice, even cause the death of animals. In
Fig. 4. (a), mice were injected seven times with GQD-PEG and GO-PEG, respectively, and three of them died from multiple injections of GO-PEG. Half of survived mice were sacrificed post 1 day of injections for the study of short-term toxicity; (b), the survived mice were weighted every three days and no difference from the control group was observed. Values represent means SE, n ¼ 6.
5046
Y. Chong et al. / Biomaterials 35 (2014) 5041e5048
Fig. 6. H&E stained tissue slices (liver, kidney, and spleen) of mice injected with GO-PEG (i.p.) and GQD-PEG (i.p. or i.v.) at the dose of 20 mg/kg per time. Mice were sacrificed at day 1 post seven injections in 14 days. While aggregated GO was found in the main organs (indicated by golden circle) of mice injected with GO-PEG, no black dots were found in the groups of GQD-PEG. Bars indicated by white stripes in the image are 50 mm. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
addition, GQD-PEG shows excellent biocompatibility, due to their small size and thus fast clearance from kidneys. To make a comparative study on the potential toxic effect of GOPEG and GQD-PEG on the treated mice, we carried out blood biochemistry and hematology analysis. Fig. 7 and Fig. S9 show the levels of standard hematology markers, such as white blood cell (WBC), red blood cell (RBC), hematocrit (HCT), mean corpuscular volume (MCV), hemoglobin (HGB), platelet (PLT), mean corpuscular hemoglobin (MCH), and mean corpuscular hemoglobin concentration (MCHC). The WBC is sensitive to the physiological response, but the values are still in the normal range despite of significant differences between groups. All the other parameters in the
nanomaterial-treated groups appear to be normal compared with the control groups, and be within the normal ranges [45]. In addition, we present the biochemistry results of the mice in Figs. S8 and S10, including total protein (TP), alanine transaminase (ALT), aspartate transaminase (AST), albumin (ALB), globulin (GLOB), the ratio of albumin and globulin (A/G), blood urea nitrogen (BUN), and creatinine (CREA). The values of ALT, AST, BUN, and CREA fluctuate a little compared to the control groups but they still remained within the normal ranges. The results of blood biochemistry and hematology analysis suggest no obvious toxicity of GQD-PEG and GOPEG. It is interesting that the survived mice with GO-PEG showed no damage from the materials, which agrees well with previous
Fig. 7. Hematology results of the GQD-PEG and GO-PEG treated mice post 1 day of injections. These results show mean and standard deviation of white blood cells (WBC), red blood cells (RBC), hematocrit (HCT), mean corpuscular volume (MCV), hemoglobin (HGB), platelets (PLT), mean corpuscular hemoglobin (MCH), and mean corpuscular hemoglobin concentration (MCHC). Bars represent mean standard deviation. Values represent means SE, n ¼ 3.
Y. Chong et al. / Biomaterials 35 (2014) 5041e5048
5047
Fig. 8. Blood biochemistry results of the GQD-PEG and GO-PEG treated mice post 1 day of injections. These results show mean and standard deviation of total protein (TP), alanine transaminase (ALT), aspartate transaminase (AST), albumin (ALB), globulin (GLOB), the ratio of albumin and globulin (A/G), blood urea nitrogen (BUN), and creatinine (CREA). Bars represent mean standard deviation. Values represent means SE, n ¼ 3.
study by other scientists [16,18], and suggests the need of more investigation on the toxicity of GO-PEG (see Fig. 8). 4. Conclusions We have studied the synthesis, characterization, in vitro toxicity, in vivo biodistribution and toxicity of a GQD sample. Analysis of WST-1 assay, cell apoptosis, LDH production and ROS level clearly demonstrated good biocompatibility of GQD and GQD-PEG at cellular level. NIR imaging of the mice administrated with GQD-PEG suggested that GQD could be metabolized quickly through kidneys and be retained in the tumor, regardless of injection routes. With multiple-dosing experiments of the nanomaterials, we found for the first time that GO-PEG showed some in vivo toxicity, probably due to GO aggregation and thus thrombus in mice. Our results, from either died or survived mice, also exposed GO-PEG accumulation in reticuloendothelial system. In contrast, mice injected with GQDPEG presented excellent biocompatibility, based on the mice growth, organ appearance and slices, as well as hematology and blood chemistry analyses. Taking together, GQD exhibits no apparent in vitro and in vivo toxicity of GQD, resulting from its small size and high O content compared with that of the widely used GOPEG. These findings may facilitate the development of safe and efficient GQD-based bio-imaging and sensing platform, and drug delivery systems. Acknowledgment The authors are grateful to the financial support from National Natural Science Foundation of China (No. 51361130033) and the Ministry of Science and Technology of China (Nos. 2010CB933504, 2014CB965003), and the Natural Science Foundation of Jiangsu Province (No. BK20130358). Y. Ma appreciates the Jiangsu Planned Projects for Postdoctoral Research Funds. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.biomaterials.2014.03.021
References [1] Baughman RH, Zakhidov AA, de Heer WA. Carbon nanotubesethe route toward applications. Science 2002;297:787e92. [2] Geim AK, Novoselov KS. The rise of graphene. Nat Mater 2007;6:183e91. [3] Chung DDL. Electrical applications of carbon materials. J Biomater Sci 2004;39: 2645e61. [4] Zhu Y, Murali S, Cai W, Li X, Suk JW, Potts JR, et al. Graphene and graphene oxide: synthesis, properties, and applications. Adv Mater 2010;22:3906e24. [5] Compton OC, Nguyen ST. Graphene oxide, highly reduced graphene oxide, and graphene: versatile building blocks for carbon-based materials. Small 2010;6: 711e23. [6] Kostarelos K, Bianco A, Prato M. Promises, facts and challenges for carbon nanotubes in imaging and therapeutics. Nat Nanotechnol 2009;4:627e33. [7] Shen H, Zhang L, Liu M, Zhang Z. Biomedical applications of graphene. Theranostics 2012;2:283e94. [8] Chung C, Kim Y-K, Shin D, Ryoo S-R, Hong BH, Min D-H. Biomedical applications of graphene and graphene oxide. Acc Chem Res 2013;46:2211e24. [9] Liu Z, Tabakman SM, Chen Z, Dai H. Preparation of carbon nanotube bioconjugates for biomedical applications. Nat Protoc 2009;4:1372e81. [10] Mendes RG, Bachmatiuk A, Buchner B, Cuniberti G, Rummeli MH. Carbon nanostructures as multi-functional drug delivery platforms. J Mater Chem B 2013;1:401e28. [11] Yang W, Ratinac KR, Ringer SP, Thordarson P, Gooding JJ, Braet F. Carbon nanomaterials in biosensors: should you use nanotubes or graphene? Angew Chem Int Ed 2010;49:2114e38. [12] Robinson J, Welsher K, Tabakman S, Sherlock S, Wang H, Luong R, et al. High performance in vivo near-IR (>1 mm) imaging and photothermal cancer therapy with carbon nanotubes. Nano Res 2010;3:779e93. [13] Yang K, Zhang S, Zhang G, Sun X, Lee S-T, Liu Z. Graphene in mice: ultrahigh in vivo tumor uptake and efficient photothermal therapy. Nano Lett 2010;10: 3318e23. [14] Markovic ZM, Ristic BZ, Arsikin KM, Klisic DG, Harhaji-Trajkovic LM, Todorovic-Markovic BM, et al. Graphene quantum dots as autophagy-inducing photodynamic agents. Biomaterials 2012;33:7084e92. [15] Liu Z, Cai W, He L, Nakayama N, Chen K, Sun X, et al. In vivo biodistribution and highly efficient tumour targeting of carbon nanotubes in mice. Nat Nanotechnol 2007;2:47e52. [16] Yang K, Wan J, Zhang S, Zhang Y, Lee S-T, Liu Z. In vivo pharmacokinetics, long-term biodistribution, and toxicology of PEGylated graphene in mice. ACS Nano 2010;5:516e22. [17] Liu Z, Davis C, Cai W, He L, Chen X, Dai H. Circulation and long-term fate of functionalized, biocompatible single-walled carbon nanotubes in mice probed by Raman spectroscopy. Proc Natl Acad Sci 2008;105:1410e5. [18] Yang K, Gong H, Shi X, Wan J, Zhang Y, Liu Z. In vivo biodistribution and toxicology of functionalized nano-graphene oxide in mice after oral and intraperitoneal administration. Biomaterials 2013;34:2787e95. [19] Bianco A. Graphene: safe or toxic? The two faces of the medal. Angew Chem Int Ed 2013;52:4986e97. [20] Luo PG, Sahu S, Yang S-T, Sonkar SK, Wang J, Wang H, et al. Carbon “quantum” dots for optical bioimaging. J Mater Chem B 2013;1:2116e27. [21] Peng J, Gao W, Gupta BK, Liu Z, Romero-Aburto R, Ge L, et al. Graphene quantum dots derived from carbon fibers. Nano Lett 2012;12:844e9.
5048
Y. Chong et al. / Biomaterials 35 (2014) 5041e5048
[22] Shen J, Zhu Y, Yang X, Li C. Graphene quantum dots: emergent nanolights for bioimaging, sensors, catalysis and photovoltaic devices. Chem Commun 2012;48:3686e99. [23] Shang J, Ma L, Li J, Ai W, Yu T, Gurzadyan GG. The origin of fluorescence from graphene oxide. Sci Rep 2012;2:792. [24] Song Y, Shi W, Chen W, Li X, Ma H. Fluorescent carbon nanodots conjugated with folic acid for distinguishing folate-receptor-positive cancer cells from normal cells. J Mater Chem 2012;22:12568e73. [25] Zhu S, Tang S, Zhang J, Yang B. Control the size and surface chemistry of graphene for the rising fluorescent materials. Chem Commun 2012;48: 4527e39. [26] Zheng XT, Than A, Ananthanaraya A, Kim D-H, Chen P. Graphene quantum dots as universal fluorophores and their use in revealing regulated trafficking of insulin receptors in adipocytes. ACS Nano 2013;7:6278e86. [27] Zhang L, Xing Y, He N, Zhang Y, Lu Z, Zhang J, et al. Preparation of graphene quantum dots for bioimaging application. J Nanosci Nanotechnol 2012;12: 2924e8. [28] Wu C, Wang C, Han T, Zhou X, Guo S, Zhang J. Insight into the cellular internalization and cytotoxicity of graphene quantum dots 2013;2:1613e9. [29] Nurunnabi M, Khatun Z, Huh KM, Park SY, Lee DY, Cho KJ, et al. In vivo biodistribution and toxicology of carboxylated graphene quantum dots. ACS Nano 2013;7:6858e67. [30] Goh EJ, Kim KS, Kim YR, Jung HS, Beack S, Kong WH, et al. Bioimaging of hyaluronic acid derivatives using nanosized carbon dots 2012;13:2554e61. [31] Huang X, Zhang F, Zhu L, Choi KY, Guo N, Guo J, et al. Effect of injection routes on the biodistribution, clearance, and tumor uptake of carbon dots. ACS Nano 2013;7:5684e93. [32] Zhang L, Xia J, Zhao Q, Liu L, Zhang Z. Functional graphene oxide as a nanocarrier for controlled loading and targeted delivery of mixed anticancer drugs. Small 2010;6:537e44. [33] Zhang L, Lu Z, Zhao Q, Huang J, Shen H, Zhang Z. Enhanced chemotherapy efficacy by sequential delivery of siRNA and anticancer drugs using PEI grafted graphene oxide. Small 2011;7:460e4.
[34] Zhang Y, Chen B, Zhang L, Huang J, Chen F, Yang Z, et al. Controlled assembly of Fe3O4 magnetic nanoparticles on graphene oxide. Nanoscale 2011;3:1446e50. [35] Zhu S, Zhang J, Qiao C, Tang S, Li Y, Yuan W, et al. Strongly greenphotoluminescent graphene quantum dots for bioimaging applications. Chem Commun 2011;47:6858e60. [36] Saxena S, Tyson TA, Shukla S, Negusse E, Chen H, Bai J. Investigation of structural and electronic properties of graphene oxide. Appl Phys Lett 2011;99:013104-1e013104-3. [37] Feng L, Liu Z. Graphene in biomedicine: opportunities and challenges. Nanomedicine 2011;6:317e24. [38] Li Y, Yuan H, von dem Bussche A, Creighton M, Hurt RH, Kane AB, et al. Graphene microsheets enter cells through spontaneous membrane penetration at edge asperities and corner sites. Proc Natl Acad Sci 2013;110:12295e300. [39] Tu Y, Lv M, Xiu P, Huynh T, Zhang M, Castelli M, et al. Destructive extraction of phospholipids from Escherichia coli membranes by graphene nanosheets. Nat Nanotechnol 2013;8:594e601. [40] Wang A, Pu K, Dong B, Liu Y, Zhang L, Zhang Z, et al. Role of surface charge and oxidative stress in cytotoxicity and genotoxicity of graphene oxide towards human lung fibroblast cells. J Appl Toxicol 2013;33:1156e64. [41] Chang Y, Yang S-T, Liu J-H, Dong E, Wang Y, Cao A, et al. In vitro toxicity evaluation of graphene oxide on A549 cells. Toxicol Lett 2011;200:201e10. [42] Nurunnabi M, Khatun Z, Nafiujjaman M, Lee D-G, Lee Y-K. Surface coating of graphene quantum dots using mussel-inspired polydopamine for biomedical optical imaging. ACS Appl Mater Interfaces 2013;5:8246e53. [43] Tao H, Yang K, Ma Z, Wan J, Zhang Y, Kang Z, et al. In vivo NIR fluorescence imaging, biodistribution, and toxicology of photoluminescent carbon dots produced from carbon nanotubes and graphite. Small 2012;8:281e90. [44] Seymour L. Passive tumor targeting of soluble macromolecules and drug conjugates. Crit Rev Ther Drug Carrier Syst 1991;9:135e87. [45] Reference ranges of hematology data of healthy female Balb/c mice were obtained from Charles River Laboratories (updated in 2011): http://www. criver.com/files/pdfs/rms/balbc/rm_rm_r_balb-c_mouse_clinical_pathology_ data.aspx.
