Graphene oxide-silver nanocomposites induce oxidative stress and apoptosis in caprine fetal broblast cells

Background: Graphene oxide (GO) has drawn much attention as excellent platform to which silver nanoparticles (AgNPs) can be anchored for the production of biomedical nanocomposites. Yet, the potential toxicity of reduced graphene oxide- silver nanoparticles (GO-AgNPs) nanocomposites to animal and human is complex to evaluate and remains largely unknown. Results: Our data indicated that GO-AgNPs caused cytotoxicity in dose-dependent manner. GO-AgNPs induced signicant cytotoxicity by the loss of cell viability, over-production of reactive oxygen species (ROS), increased leakage of lactate dehydrogenase (LDH) and level of malondialdehyde (MDA), increased expression of pro-apoptotic genes and decreased expression of anti-apoptotic genes. Conclusions: This study demonstrated that GO-AgNPs potentially induced oxidative stress, which resulted in toxicity and cell apoptosis in caprine fetal broblast cell due to an increased generation of ROS. For antibacterial applications of GO-AgNPs nanocomposite in animal, the toxical effect must be further evaluated.

spleen [24][25][26]. In addition to the dependence of toxicity on surface functionalization, graphene, GO and reduced GO differentially induced toxicity is especially dependent on their sizes and oxidation state as well as the exposure concentrations and sensitivity of the employed in vitro and in vivo models [27,28]. Therefore, the higher concentration of graphene family nanomaterials seems unsafe to human and environmental health for long-term exposure [27].
In recent decades, graphene and graphene related nanocomposite have attracted much attention both in the industry and scienti c community due to their unique physiochemical and biological properties [11].
Silver nanoparticles (AgNPs) are one of the most frequently used nanoparticles in a variety of biomedical applications [29]. Recent advances in nanotechnology have widened the potential combination of AgNPs with graphene-based nanocomposite, which offers a novel graphene-silver hybrid nanomaterials with unique functions in biomedical nanotechnology, and nanomedicine [30]. For example, AgNPs attaching on the surface of GO sheets can prevent AgNPs from aggregating, allow a more controlled release of AgNPs + ions and lead to the increase antibacterial and anticancer activity [31,32]. Thus, humans, animal and the environment may be constantly exposed to a wide range of commercial products containing residues of AgNPs. The potential mechanisms involved in cellular damages and inner response pathways induced by grapheme related nanomaterials needed to be deeply investigated. AgNPs have been broadly used as antibacterial agents and for biomedical application in the healthcare industry. Our previous study demonstrated that AgNPs (1 µg/mL) exhibited substantial activity against multidrug resistant (MDR) bacteria in dairy goats [33]. Since, AgNPs can be translocated to the blood stream and distributed throughout vital organs such as the liver, kidney, lung and brain. Therefore, for safe applications of GO-AgNPs as antibacterial agents for treatment of mastitis in the bovine and caprine, investigation of its toxicity and their potential risks is needed. The present study was conducted to investigate the effect of GO-AgNPs nanocomposites on caprine fetal broblast cells in vitro and to determine the underlying molecular mechanisms of their cytotoxicity. To our knowledge, this is the rst report in caprine fetal broblast cells to demonstrate cellular responses and functional aspects of GO-AgNPs.

Results
Characterization of GO-AgNPs TEM analysis was conducted to con rm the structural and surface morphology of the GO-AgNPs composites. The size distribution of the GO-AgNPs was about 20 nm as shown in the image of TEM (Fig.  1). GO-AgNPs images clearly showed transparent, single-layer sheets containing ake-like wrinkles in which AgNPs were homogeneously arranged on the micron scale of the GO sheets, which presented no evidence of agglomeration.

Effect of GO-AgNPs on caprine fetal broblast cells viability
For assessment of the potential cytotoxic effect of GO-AgNPs on caprine fetal broblast cells, the cell viability following GO-AgNPs treatment was determined by CCK-8 assay. As shown in Fig. 2, there were no signi cant differences in cell viability between control cells and those exposed to 1 µg/mL GO-AgNPs for 24 h, however, the viability of cells was signi cantly reduced when the concentration increased (4,8,12 and 16 µg/mL), suggesting that GO-AgNPs induced toxicity in caprine fetal broblast cells in a dosedependent manner.

Effect of GO-AgNPs on cell morphology
The morphologies of caprine fetal broblast cells after exposure to GO-AgNPs for 24 h were shown in Fig.   3. Caprine fetal broblast cells that had been exposed to 4 µg/mL and 8 µg/mL GO-AgNPs exhibited marked morphological changes and showed cell membrane breakage, with obvious reduction in number of cells in the group of 8 µg/mL GO-AgNPs. Scale bar = 20 μm.

Effect of GO-AgNPs on reactive oxygen species (ROS) production
To study whether GO-AgNPs induced oxidative impact involving in the apoptosis, the intracellular ROS level in caprine fetal broblast cells was analyzed. As shown in Fig. 4, the level of intracellular ROS in caprine fetal broblast cells was signi cantly increased (P < 0.05) when the cells were treated with 4 and 8 µg/mL of GO-AgNPs for 24 h compared to the control group.

Effects of GO-AgNPs on apoptosis
We went further to test the effect of GO-AgNPs on cell apoptosis. An Annexin V/PI apoptosis kit was ued to quantify the percentage of caprine fetal broblast cells undergoing apoptosis and dying by ow cytometry. The results suggested that the GO-AgNPs induced signi cant apoptosis and dead in caprine fetal broblast cells (Fig. 5).

Effects of GO-AgNPs on superoxide dismutase (SOD) production
Effects of GO-AgNPs on the production of the antioxidant indicator SOD in caprine fetal broblast cells were determined with SOD assay kit. As shown in Fig. 6, SOD activity decreased signi cantly (P < 0.05) in caprine fetal broblast cells treated with 4 µg/mL GO-AgNPs for 24 h when compared to the control group. Furthermore, treatment of caprine fetal broblast cells with 8 µg/mLGO-AgNPs decreased signi cantly (P < 0.01) in SOD activity compared to non-treated caprine fetal broblast cells.
Effects of GO-AgNPs on malondialdehyde (MDA) production The production of MDA in caprine fetal broblast cells was determined with the MDA assay kit after treated with different concentrations of GO-AgNPs (0, 4 and 8 µg/mL) for 24 h. The results showed that levels of the oxidative damage indicator MDA increased signi cantly (P < 0.05) in the 4 and 8 µg/mL treated groups, when compared to the control group (Fig. 7).
Effects of GO-AgNPs on lactate dehydrogenase (LDH) Caprine fetal broblast cells were treated with different concentrations of GO-AgNPs (0, 4 and 8 µg/mL) for 24 h, and the level of leakage of LDH was measured. The results indicated that GO-AgNPs signi cantly increased the leakage level of LDH in caprine fetal broblast cells compared to the control group ( Fig. 8; P < 0.05).

Effects of GO-AgNPs on caspase-3 activity
To con rm whether caspase-3 is involved in apoptosis of caprine fetal broblast cells induced by GO-AgNPs (4 and 8 µg/mL), caspase-3 activity was measured by the caspase-3 kit. The activity of caspase-3 in the 4 and 8 µg/mL group was signi cantly (P < 0.05) higher after treatment than that in the control group ( Fig. 9).

Effects of GO-AgNPs on gene expression
To elucidate the possible molecular mechanisms underlying the negative effect of GO-AgNPs, the mRNA levels of pro-and anti-apoptotic, cell death and survival related genes caspase-3, Cyt-C, BAX, Smac,p53 and BCL2 were investigated in caprine fetal broblast cells treated with GO-AgNPs (4 and 8 µg/mL) for 24 h and compared with control. The results showed that the level of caspase-3, Cyt-C, BAX, Smac and p53 were signi cantly (P < 0.05) upregulated in the GO-AgNPs treated groups, compared to the control group (Fig. 10). However, the level of anti-apoptosis gene BCL2 in the GO-AgNPs treated group was signi cantly (P < 0.05) downregulated, compared with the control group (Fig. 10).

Discussion
Nanotechnology has become an indispensable eld tool to develop various kinds of nanoparticles with unique properties [34]. As an e cient support material, graphene sheets can disperse and stabilize silver nanoparticles preventing their agglomeration, which open up a way for the nanomaterial development.
Therefore, the combination of graphene and AgNPs based nanocomposites has been widely produced to enhanced antibacterial and anticancer activity [35]. The AgNPs can easily enter cells, thus affect the physiology of organisms, which may have the potential toxicity both on human and animal health or ecosystems [26]. Therefore, the adverse effects of GO-AgNPs composites have been considered as a major limitation for its broad applications. Numerous studies have proved the toxicological effects of GO-AgNPs nanocomposite on animal and human normal cells [12,35,36]. In the present study, a GO-AgNPs nanocomposite was synthesized using quercetin and was con rmed its surface and structural morphology, and the uniform distribution of AgNPs on the GO sheets with TEM, and investigated its toxic effect on caprine fetal broblast cells.
Humans could be directly exposed to AgNPs of many consumer products via the route of inhalation, dermal contact and oral ingestion [26]. Therefore, toxical effect of these products is required to ensure that they are safe to use. As cell lines are considered as an experimental models for testing nanomaterial toxic, several studies have demonstrated that AgNPs induced toxicity via oxidative stress and apoptosis in mouse and rat cell lines [37,38]. Cell viability is an important index of cytotoxicity, which is de ned as the potential of tested AgNPs to induce cell death. AgNPs with a smaller size have the ability to enter into cells more easily by uptake and are distributed throughout the cytosol [39], and contacted with cell organelles. While the smaller AgNPs (6 nm) was reported to be non-toxic to mouse broblast line and human keratinocyte cell line [40]. Similar results were reported that 5 µg/mL rGO-Ag nanocomposite did not induce cytotoxicity in human normal cells (CHANG cell) but could slightly induce toxic effect on HepG2 cells [38], which ascribe the differences in toxicity mechanisms to cell type [41]. The present data showed that 20 nm GO-AgNPs reduced cell growth, viability and induced morphological changes in a concentration-dependent manner. In our previous study, GO-AgNPs signi cantly decreased the human ovarian cancer cell viability with an IC 50 of 5 µg/mL [42], which is lower than that in the present study, suggesting that caprine fetal broblast cells are less sensitive to GO-AgNPs than human cancer or mouse cells. AgNPs with different sizes and surface coatings or without coatings are likely to contribute to these different results. Lopes et al [43] reported that coating-AgNPs had a better dispersion than AgNPs and was more fully exposed to the cells. When compared to L02 cells, HepG2 cells are more sensitive to AgNPs at the exposure level of 20-160 μg/mL [44].
One of the main mechanisms of toxicity induced by nanomaterial is that it causes oxidative stress through the generation of ROS and causes damage to cellular components including DNA damage, abnormal activation of transcription factors, depletion of antioxidant molecules, binding and disabling of proteins, and damage to the cell membrane [26]. Oxidative stress inducing ROS is one of the proposed toxicological mechanisms of various nanomaterials such as Ag or Ag-graphene nanocomposites, can cause mitochondrial damage, and initiation of lipid peroxidation [42,45,46]. Cytotoxicity of AgNPs is associated with increased production of ROS, which play an important role in apoptosis induced by AgNPs [47]. Compared to pristine AgNPs, GO-AgNPs signi cantly induced generation of ROS in the macrophage in a dose-dependent manner [48]. In current study, the level of ROS increased signi cantly at a 1.4-and 1.8-fold in caprine fetal broblast cells after treatment with 4 and 8 µg/mL GO-AgNPs for 24 h. Our results are consistent with previous reports on various cancer cell lines with graphene and graphenerelated materials [30,42]. The up-regulated ROS level in caprine fetal broblast cells alters the mitochondrial functions and plays a key role in apoptosis induction, which proved by the data of Annexin V/PI double labeling assay and increasing level of caspase-3. The present data suggests that the possible mechanisms of GO-AgNPs-induced toxicity in caprine fetal broblast cells include the stimulation of oxidative stress, which results in apoptosis and is responsible for the upregulation of proapoptotic and down regulation of anti-apoptotic genes in caprine fetal broblast cells [30].
The increased level of MDA and LDH is generally considered to imply the cell injury. One of the adverse effects of oxidative stress is the lipid peroxidation of cell membranes. Many type of cells treated with AgNPs and GO showed signi cantly increased levels of MDA, which is one of the nal products of polyunsaturated fatty acids peroxidation in the cells [17,30,[49][50][51]. Assessing the release of intracellular LDH in cell, which resulted from the breakdown and alteration in the permeability of the plasma membrane, is one of marker for estimating cytotoxicity [30,46]. rGO-Ag increases LDH leakage in human cancer cells, thus resulting in cell death [30,42,52]. The present data indicated that the mechanism of increased level of MDA and LDH in GO-AgNPs-treated caprine fetal broblast cells may be due to the strong hydrophobic interactions with the cell membrane and ROS formation, which in uenced the viability and proliferation of cells, suggesting the possible cytotoxicity of GO-AgNPs on caprine fetal broblast cells.
The apoptosis in cell is a highly conserved mechanism, and ROS is an important factor involved in the apoptotic process [53]. ROS induced by nanomaterial could result in nuclear DNA damages and leakage of lipids, proteins and carbohydrates in the cell [33,51]. ROS production and lipid peroxidation induced by GO-AgNPs affected cellular redox homeostasis and decrease antioxidant levels [51]. It is well known that SOD plays an important role in antioxidant defense against oxidative stress in cells that can combat the accumulation of ROS and reduce the oxidative injury. A decrease of SOD activity is an indicator of impairment of the protective mechanisms and signi cantly contributes to cell damage [54]. It has been reported that, AgNPs directly interacted with SOD and CAT, altered the expression and activity of antioxidant enzymes (CAT, SOD and GPX) [54]. The present data showed that the level of SOD signi cantly decreased in GO-AgNPs-treated caprine fetal broblast cells. It may suggest that GO-AgNPs decrease levels of antioxidant molecules in the cells, which may be the reason for the cause of cytotoxicity.
Apoptotic and anti-apoptotic genes play an important role in cell survival and death. Several studies have reported that the mechanisms of GO-and graphene oxide silver nanocomposite-induced toxicity in numerous human cancer cell lines due to include oxidative stress, DNA damage, and apoptosis [12]. A similar study reported that GO-AgNPs can cause oxidative damage, leakage of LDH, and enhance expression of apoptotic genes of p53, caspase-3, caspase-9, Bax, and c-myc, thus lead to mitochondrial dysfunction and trigger apoptosis [55]. All apoptotic pathways appear to terminate in the activation of the caspase family of proteases [56]. Moreover, oxidative stress induced by GO-AgNPs is reported to increase the total expression of Bax in a dose dependent manner and also down-regulate the expression of the anti-apoptotic gene of BCL-2 [52]. The present data showed that GO-AgNPs up-regulated the expression of pro-apoptotic genes such as caspase-3, Bax, Smac and c-myc, and down-regulated anti-apoptotic genes such as Bcl-2. Similarly, rGO-Ag reported to cause dynamic balance troubles in the level of Bcl-xl and Bcl-2, and downregulation of c-myc triggers apoptosis along with p53 [55], which may induced apoptosis in caprine fetal broblast cells.

Conclusion
Collectively, the synthesized GO-AgNPs nanocomposite displayed cytotoxicity to caprine fetal broblast cells at the concentration of 4 and 8 µg/mL with apoptotic induction by loss of proliferation and cell viability, enhancing production of ROS, decrease levels of anti-apoptotic genes and increase expression of pro-apoptotic genes. In addition to the dependence of toxicity on GO, AgNPs may play an important role in cellular toxicity, suggesting that the combination of GO and AgNPs served either as competence factors or acted synergistically to enhance the cytotoxicity towards caprine fetal broblast cells. These data provide toxicological information in assessing the toxical effects of GO-AgNPs in cell of large animals.

Chemicals
All chemicals and reagents were purchased from Sigma-Aldrich (St. Louis, MO, USA) unless otherwise stated.

Synthesis and characterization of GO-AgNPs
GO-AgNPs nanocomposites were synthesized using the biomolecule quercetin as described previously [42], then lyophilized and keep in lab at 4°C till usage. Brie y, 50 mg GO was dispersed in 30 mL water and sonicated for 60 min. One mM AgNO 3 were dissolved in 15 mL water in a 500 mL round-bottom flask. 30 mL of the GO dispersion was added, followed by addition of 5 mL of aqueous 1 mM quercetin, and then stirred at 60°C for 12 h. The resultant mixture was washed and centrifuged three times with water. Physicochemical characterization of GO-AgNPs were checked by fourier-transform infrared and Xray diffraction. The size and shape were observed under a transmission electron microscope (TEM; HT7800, Hitachi High-Technologies Corporation, Tokyo, Japan).

Cell culture
Caprine fetal broblast cells were isolated from 70-day old fetuses that were recovered surgically from a Boer goat obtained from Yangzhou University farm as previously described [57]. Brie y, the pregnant goats were anaesthetized using intramuscular injected with serazine hydrochloride (0.02 mL/kg body weight) and fetuses were collected. After removal of the head and internal organs, the remaining tissues of fetus were dissociated into small pieces using scissors and digested with 0.25% trypsin (Thermo Cell viability assay The cell viability was assessed by using an in vitro cell-counting assay kit (CCK-8; Rockville, MD, USA) as described previously [45]. Caprine fetal broblast cells were seeded on 96-well or 6-well plate and cultured for 24 h to allow adherence and stabilization. GO-AgNPs were sonicated for 20 min before use. Then the GO-AgNPs suspension was dispersed in DMEM/F12 for different concentrations (1, 4, 8, 12 and 16 μg/mL) for 24 h at 37℃. After culture, 10 μL CCK-8 was added into each well and were incubated for 30 min at 37℃ in the dark. The absorbance at 450 nm was measured using a microplate reader (BioTek Synergy 2, USA). According to the toxic data of LC 50 , the 4 and 8 μg/mL were selected for further experiments. This study was designed and blinded throughout all stages of the methodological process.

Cell morphology
Caprine fetal broblast cells were seeded into a 24-well plate for 24 h, and then treated with the different concentrations of GO-AgNPs (0, 4 and 8 μg/mL) for 24 h. The cell morphology was observed using an Olympus BX-UCB microscope (Tokyo, Japan).
Annexin V-FITC/PI staining assay Caprine fetal broblast cells were seeded in a 75mm culture plate and treated with different concentrations of GO-AgNPs (0, 4 and 8 μg/mL) for 24 h. Cell apoptosis of caprine fetal broblast cells was detected by Annexin V-FITC and Propidium iodide (PI) staining assay according to manufacturer's instructions (Bipec Biopharma Corporation, USA). The cells were harvested, centrifuged for 5 min, rinsed with phosphate buffered saline (PBS) twice and resuspended in 500 μL binding buffer containing 5 μL PI and 5 μL V-FITC, and then incubated for 15 min at the room temperature in dark. The cell suspension was determined by ow cytometry to analyze the apoptotic rate.

Measurement of ROS production
Dichlorodihydro uorescein diacetate (DCFH-DA) was used to detect intracellular ROS induced by different concentrations (0, 4 and 8 μg/mL) of GO-AgNPs [45]. Caprine fetal broblast cells were incubated in 10 μM DCFH-DA for 30 min at 37°C. The cells were rinsed with PBS twice, and then the intracellular accumulation of ROS was measured by ow cytometry (Beckman-Coulter, USA).

Measurement of total SOD enzyme activity
The SOD activity in caprine fetal broblast cells was detected following the the SOD assay kit (Beijing Solarbio Science & Technology, Beijin, China) [45]. Cells were treated with 0, 4 and 8 μg/mL of GO-AgNPs for 24 h. Cells were washed with PBS twice, and lysed with lysis buffer on the ice. The lysates were then centrifuged for 15 min. Then, the supernatant was analsysed with a U-vis spectrophotometer (Nanodrop, Thermo, USA) at 550 nm.
Measurement of MDA production MDA, a convenient index for detecting the extent of lipid peroxidation reactions, was performed using the MDA assay kit (Beijing Solarbio Science & Technology, Beijin, China) according to the manufacturer's instructions [58]. Cells were plated into 6-well plates at a density of 1.0×10 5 cells per well and cultured for 24 h to allow adherence before exposure to different concentrations (0, 4 and 8 μg/mL) of GO-AgNPs for 24 h. Then the cells were washed with PBS twice and MDA activities were quantitated by reading optical densities using Synergy 2 multi-mode microplate reader (BioTek, USA) at 532 nm.

Measurement of LDH production
Caprine fetal broblast cells were seeded in a 24-well culture plate, and treated with 0, 4 and 8 μg/mL of

Statistical analysis
The assessors were blinded to any stages of the methodological process. All results were expressed as mean ± S.D. and analyzed by Origin 8.0 and SPSS 18.0 (IBM Corp., Armonk, NY, USA). The statistical signi cance of the changes between tested groups and control group were analyzed by one-way ANOVA followed Dunnett's multiple comparison. The level of statistical signi cance was set at P < 0.05. All experiments were performed at least three times.