DNA-damage and cell cycle arrest initiated anti-cancer potency of super tiny carbon dots on MCF7 cell line

While carbon-based materials have spearheaded numerous breakthroughs in biomedicine, they also have procreated many logical concerns on their overall toxicity. Carbon dots (CDs) as a respectively new member have been extensively explored in nucleus directed delivery and bioimaging due to their intrinsic fluorescence properties coupled with their small size and surface properties. Although various in vitro/in vivo studies have shown that CDs are mostly biocompatible, sufficient information is lacking regarding genotoxicity of them and underlying mechanisms. This study aims to analyze the real-time cytotoxicity of super tiny CDs (2.05 ± 0.22 nm) on human breast cancer cells (MCF7) and human primary dermal fibroblast cell cultures (HDFa) by xCELLigence analysis system for further evaluating their genotoxicity and clastogenicity to evaluate the anti-tumor potential of CDs on breast adenocarcinoma. As combined with flow cytometry studies, comet assay and cytokinesis-block micronucleus assay suggest that the CDs can penetrate to the cell nuclei, interact with the genetic material, and explode DNA damage and G0/G1 phase arrest in cancer cells even at very low concentrations (0.025 ppm) which provide a strong foundation for the design of potentially promising CD-based functional nanomaterials for DNA-damage induced treatment in cancer therapy.

www.nature.com/scientificreports/ or biological carbon sources like food wastes, fruits, seeds and shells, plant extracts have been moving toward their application to biological sciences [36][37][38] . Toxicity studies by various research groups have shown that CDs exhibit very low toxicity compared to heavy metal-based quantum dots when internalized by the cells 26,[39][40][41][42][43] . In almost all of the cytotoxicity studies performed to date, CDs have been demonstrated to cause a negligible effect on cell viability at concentrations sufficient for drug delivery and bio-imaging. In the last 2 years, we have been focused on new types of CDs by tailoring the surface properties of them with the target of evaluating their potential in biological applications 16,44,45 . Very recently, utilizing different polymers as surface passivating agent and surface doping agent, we were able to show the immunostimulant and adjuvant-like effect of such CDs where no cytotoxic effect was observed 44,45 . CDs have been widely researched for nucleus-targeted delivery, nucleus labeling, photodynamic therapy, and optical monitoring of anticancer drugs [46][47][48][49][50] , however, so far there are few scientific reports in existence for addressing the genotoxic activities of CDs when they administrated on their own [51][52][53][54][55][56] . In a recent study, researchers showed the genotoxic responses of rat alveolar macrophages (NR8383) to amine-grafted graphene QDs which resulted in significant alterations in the expression of 2,898 genes after exposure for 24 h in which most of the down-regulated genes were reported the as they were responsive to "cell cycle" 57 . In a study on the use of graphene quantum dots (GQDs) as nucleus labeling of several cell lines (L929, HT-1080, MIA PaCa-2, HeLa, and MG-63 cells), researchers showed that internalization mechanism of the GQDs by healthy cells differ from tumor cells while GQDs showed to be entering into the nucleus regardless of the cell type. Moreover, the authors showed an altered number of L929 cells in the S phase as an indication of promoted cell proliferation in the presence of GQDs 50 . Yue et al. 25 reported that ruthenium incorporated CDs with no apparent cytotoxicity have the concentration-dependent DNA photocleavage ability on HeLa cells upon the light irradiation (6.5 mW/cm 2 ) showing the potential of CDs in imaging and photodynamic therapy (seeTable 2).
Nerium oleander (Oleander) is one of the most poisonous dwarf evergreen shrubs in the world 58 . Extracts from various parts of the plant show also anti-cancer 59,60 , anti-microbial 61 , anti-inflammatory 62 , anti-diabetic 63 , and neuroprotective activities 62 . Common ingredients of Oleander extract include polysaccharides containing rhamnose, galactose, arabinose, mannose, glucose, and galacturonic acid [64][65][66][67] . Other components and their concentrations may vary depending on the extraction method 68 . For instance, extracts of N. oleander leaves contain steroids, flavonoids, and terpenoids, etc. 67 . Our group lately presented the synthesis routes for Oleander based CDs using both thermal and microwave-based synthesis methodologies 69,70 while we have shown that extract type (water or ethanol extraction) is one of the important parameters where the highest fluorescence and the lowest size was observed using water-based Oleander extract as a carbon source for CD synthesis.
The present study covers the effects of super tiny CDs on the cell viability of MCF7 tumor cells and normal HDFa cells together with the CD-induced differentiation in cell-cycle progression, genotoxicity, and clastogenicity on MCF7 cells. Our results suggest that CDs, alone or in combination with chemotherapeutics, may be exploited for the development of potentially promising functional nanomaterials for DNA-damage induced treatment in cancer therapy. They have the potential that could be extended to be used as new generation biolabeling and imaging agents as well. However, the possible influence of the cell cycle on cellular uptake of CDs and the mechanism of its effect on MCF7 cells needs further investigation. equipment. ELMA TI-H 5 model ultrasonic bath was used during the extraction process. The thermal synthesis was conducted using a Neuve muffle furnace. Characterization studies of CDs were performed on a Shimadzu UV-1800 UV-Vis spectrophotometer, Agilent Cary Eclipse fluorescence spectrophotometer, Malvern Zeta Sizer Nano ZS, and Perkin Elmer frontier FT-IR. X-ray photoelectron spectroscopy (XPS) screening was performed using the Specs-Flex XPS spectrometer (Al Kα 1,486.7 eV). Morphology of CDs was monitored by a JEOL JEM-1400 series 120 kV Transmission Electron Microscope (TEM) and the FEI Tecnai G2 F30 HR-TEM at 300 kV. Particle core radius was calculated by measuring at least 100 individual particles using Image J program. Cell-seeding calculations were carried out with the Cedex XS analyzer (Innovatis Inc.). xCELLigence system (ACEA Biosciences Inc.) was used as a real-time cell sorter. BD FACSAria III flow cytometer (BD Biosciences, US) and BD CELLQuest Pro software (BD Biosciences, US) were used for cell-cycle analysis. Fluorescence imaging was performed by a fluorescence microscope (Olympus BX51) equipped with a CCD-4230 video camera. www.nature.com/scientificreports/ preparing plant extracts. The fresh leaves of N. oleander were washed twice with distilled water and dried at incubator at 70 °C for 2 days. Dried leaves were then grinded to powders. The water extraction procedure was performed using ultra-pure water with the final concentration 12.5 g dry leaf/100 mL in ddH 2 O in an ultrasonic bath for 5 h at room temperature (RT). After extraction, clear extracts were obtained by centrifugation at 5,000 rpm for 15 min and stored at + 4 °C.

Materials and
Synthesis of cD using N. oleander leaf extract. 1 mL aqueous extract was dispersed 1 g PEG solution pre-prepared in 2 mL of ddH 2 O/ethanol solution (1:1 v/v). The mixture then placed in a muffle furnace and left for the caramelization process for 45 min at 300 °C. Resulting brown solid was cooled down to RT and dissolved in 6 mL ddH 2 O. CDs were separated by three cycles of centrifugation at 13,500 rpm for 20 min. The supernatants were collected and dried in a vacuum oven at 60 °C for 1-2 days.
Quantum yield of cDs. The quantum yield of the CD (diluted samples to obtain an absorbance value of less than 0.10) was determined using quinine sulfate in 0.1 M H 2 SO 4 (quantum yield: 54%) as the standard sample following the procedure published previously 71 .
In vitro cytotoxicity of CDs on MCF7 and the HDFa cell lines by xCELLigence.

In vitro genotoxicity of CDs on MCF7 cells by comet assay (SCGE). Using the information obtained
from the cytotoxicity studies, CD concentrations of 0.25, 2.5, and 50 ppm were selected for performing the alkaline comet assays. MCF7 cells in log-phase were plated in 96-well plates and incubated for 2 days. The negative control (Untreated cells that were grown in growth media, NC), and the positive control (Cells treated with 20 mM hydrogen peroxide) were inoculated in series. The comet assay was performed in alkaline conditions (pH > 13) as described previously 72 . Briefly, after cells were exposed to the CDs for 48 h and 72 h, the cells were collected and trypsinized. Centrifuged cells at 1,100 rpm for 5 min were counted and 1-3 × 10 4 cells were resuspended in 75 μL molten 0.5% low-melting-point agarose at 37 °C. The resuspended cells in agarose were put onto dry microscope slides pre-coated with 1% normal-melting agarose, and the agar solidified by keeping at RT for 10 min. The slides were thereafter immersed in cold lysing solution (2.5 M NaCl, 100 mM EDTA, 10 mM Tris, 1% Triton X-100) for at least 1 h at 4 °C. Afterward, they were transferred to an electrophoresis tank containing freshly-made electrophoresis buffer (1 mM EDTA, 300 mM NaOH; pH > 13), where they were kept for 20 min at room temperature to allow DNA unwinding. Electrophoresis was performed in the same buffer at RT for 15 min at 24 V and 300 mA (0.8 V/cm). The slides were then neutralized thrice with 0.4 M Tris buffer (pH 7.5), air-dried, and fixed in ethanol. All preparative steps were conducted under dark conditions to prevent additional DNA damages. Slides stained with ethidium bromide (0.1 mg/mL, 1:4) were analyzed under a fluorescence microscope equipped with a CCD-4230 video camera.
Slide scoring in comet assay. Comet images were analyzed following the method reported by Collins et al. 73 . The percentage of DNA in the comet tail from 100 cells per sample (duplicate, each with 50 cells/slide) was used as a measure of the amount of DNA damage. An intensity score from class 0 (undamaged) to class 4 (ultra-high damage) was assigned to each cell 74  www.nature.com/scientificreports/ mixtures of acetic acid and methanol), and several drops of formaldehyde were added to preserve the cytoplasm. Immediately after centrifugation at 1,500 rpm (1,000g) for 10 min, cells were fixed again in Carnoy's solution.
Finally, the cells were dropped onto clean microscopic slides, air-dried, and stained with Giemsa. A total of 1,000 binucleated cells for each sample were examined microscopically for micronuclei as previously described 76 . The levels of chromosomal damage were reported as the fold induction of micronuclei compared with the untreated control. CBMN results were accepted only when (1) they were separated from the main nuclei, but included within the corresponding cytoplasm, (2) they had a chromatin material similar to that of the main nuclei, (3) they were coplanar to the main nuclei, (iv)they were 1/16th to 1/3rd of the mean diameter of the main nuclei. In the CBMN study, toxicity was evaluated by classifying cells according to the number of nuclei 77 . Statistical analysis. At least three independent experiments were carried out in triplicate for each evaluation. Data were expressed as the mean ± error (SE) and analyzed by repeated-measures ANOVA followed by the least significant difference post hoc test Bonferroni for the comet assay. An independent t test was used for the CBMN test. In all tests for comet assay and CBMN test differences were considered significant at p < 0.001 and p < 0.05 respectively. All the data analysis was carried out using software STATISTICA for comet assay and STATA MP/11 for the CBMN test.

Results and discussion
Characterization of physicochemical properties of CDs. Many studies have attempted to elucidate the mechanisms of nanoparticle toxicity and distinguish between their bulk counterparts [78][79][80] . It is well-known that the toxicity of nanoparticles (NPs) was highly dependent on their physicochemical characteristics 81 . Table 1 shows some physical properties of as-synthesized and purified CDs (Fig. 1). PEG was used as a surface passivation agent which mainly acts as surface functionalization precursors for developing highly tunable photoluminescence properties by stabilizing the dangling bonds and controlling the surface functional groups and surface states 15,82,83 . Use of N. Oleander as carbon source and PEG as passivating agent revealed super tiny CDs with a  (Fig. 2a) and core radius of 1.79 ± 0.33 nm measured by DLS and TEM imaging, respectively. The HR-TEM image (inset in Fig. 1) shows that the CDs displayed a highly crystalline structure with a 0.21 nm lattice spacing that is attributed to the graphitic (sp 2 ) carbon 84 . Surface ζ-potential of − 23.5 ± 6.21 mV which could be the sign of a surface with the higher density of oxygen-rich groups 74,75 .
CDs are exceptional carbon-based materials with superior optical properties that can be used in biomedical applications as they are trackable by fluorescence imaging. UV-Vis spectrum of CDs was presented in Fig. 2b. CDs showed a broad absorption peak located at around 245 nm which corresponds to a typical absorption of sp 2 carbon network which was also supported by the HR-TEM imaging ( Fig. 1) 16,85,86 . As can be seen in Fig. 2c, CDs showed the highest fluorescent emission at 466 nm upon excitation at 365 nm with the highest quantum yield of 7.12%. The fluorescence emission spectra of samples demonstrated an excitation-dependent feature originates from a combination of quantum confinement effects and the distribution of different emissive surface traps presented on the surface of CDs 82,87 .
FT-IR spectroscopy was carried out to characterize the surface properties of CDs. As depicted in Fig. 2d, CD showed a sharp peak around 3,600 cm −1 corresponds to free O-H groups present on the particle surface. The as-prepared CDs showed peaks belong to C-H stretches in methyl and methylene groups (2,800-3,000 cm −1 ) 88 . Peaks corresponding to C-O stretching located at 1,075-1,250 cm −1 might be associated with the partial oxidation of CD surfaces. Sharp peaks at 1,380 and 1,460 cm −1 were attributed to CH 2 vibrations (Fig. 2d). Many oxygen-rich functional groups including C-O-C (1,150 cm −1 ), C-OH (1,250 cm −1 ) and C-OH stretching peak at 1,380 cm −1 could be indicative of a C-O-C asymmetric stretch or C-H bending arising from a methyl functional groups presenting on the CDs 35,49,89 . Figure 3a depicts the XPS wide scan spectrum of the synthesized CDs. Two bands of the XPS survey spectrum at around 284.5 eV and 531.5 represented O1s and C1s, respectively, which indicates the atomic ratio of O/C is 32.8/67.2 as calculated from the survey spectrum. The high-resolution C1s XPS spectrum (Fig. 3b) was deconvoluted into three contributions at 283.4, 284.7, and 285.4 eV, which are associated with carbon in the states of sp 2 C (C=C, C-C) and C-OR, respectively [90][91][92] . The deconvoluted O1s spectrum (Fig. 3c) had three components peaking at 529.4, 530.9, and 531.8 eV, which are due to the C=O, C-OH, and C-O-C groups, respectively 91,92 . XPS results support the UV-Vis spectrophotometry and FT-IR results that CDs are purely composed of C=C core and have a highly oxygenated and reactive surface.

In vitro cytotoxicity of CDs on MCF7 and HDFa cells.
Although MTT assay is one of the most used methods for the assessment of in vitro cytotoxicity of materials, there are reports on the interference of carbonbased nanomaterials with the colorimetric assays [93][94][95] . The xCELLigence is a technology that gives the possibility www.nature.com/scientificreports/ to measure the cellular growth in real-time by measuring the net adhesion of cells on a custom-designed gold electrode following the changes in electrical impedance. Thus, it gives more pre-sized information on the longterm screening of cell viability and prevents the material-dye interaction-based false responses 96 The cytotoxicity of the CDs were conducted on MCF7 and HDFa cells. The normalized cell index (CI) is indicative of the level of adhesion and therefore associated with the viability of the cells. Figure 4a,b illustrates the normalized cell index of the MCF7 treated with varying concentrations of the CDs (0.0025-50 ppm) as compared to the PBS as the negative control. In all cases, cells were treated after 24 h following the seeding and the cell index was normalized at the point of treatment. The Oleander extract as control of the starting material resulted in cell death within a few hours after the treatment resulting in a sharp decrease in CI. After 24 h of treatment, CD at the highest concentration (50 ppm) revealed a significant decrease in cell growth as compared to the untreated controls. At lower concentrations, CDs did not show a similar growth trend as the untreated cells while no dose-dependent response was observed even after 72 h of treatment. These results agree well with experimental data taken from the literature that increased negative net charge and small size of CD increased the risk of cytotoxicity of them to tumor cells at high doses 49,97 . On the other hand, the story in the case of HDFa cells was different. As depicted in Fig. 4c, CDs did not show any cytotoxic effect at neither of the concentrations but also they significantly induced cell proliferation at CD concentrations of 0.25 ppm and below, while at 72 h of treatment with CDs even at the highest concentration resulted in increased cell proliferation which reveals that presence of CDs protected the cells from natural death. The CI value increased sixfold for the cells exposed to CDs at a concentration as low as 0.25 ppb as compared to the untreated cells. Oleandrin extract (50 ppm) was, as expected, showed to be cytotoxic for both of the cells regardless of the time of the exposure. Although the main reason is not clear yet, higher tolerability of non-cancerous cells, decreased or retarded cellular accumulation of CDs in healthy cells as compared to the cancerous cells have been highlighted in many recent reports to explain the enhancement of the cell proliferation of healthy cells after the CD exposure 50,81,98-100 . Yao et al. 101 showed that CDs specifically interact with some cellular proteins of tumor cells and downregulate the level of some proteins and the activity of enzymes that are related to tumor cell invasion ability. evaluation of genotoxicity and clastogenicity of cDs. The genotoxicity and clastogenicity of CDs on MCF7 cells were investigated by comet assay and CBMN, respectively. Figure 5 represents AU values indicating CD-induced DNA damage at various levels. AU values of negative control were calculated to be 18.50 ± 4.20 and 33.75 ± 6.24 after 48 and 72 h respectively. In contrast, the level of DNA damage jumped to 337.75 ± 9.81 (48 h)   (20 mM). Increased levels of DNA damages induced by carbon dots occurred with respect to the negative control (Fig. 5). Even at the lowest CD concentration (0.25 ppm), AU value was elevated more than twofold as compared to the untreated cells. The difference between selected concentration groups was significant (p < 0.05). CD-based oxidative DNA damage can be correlated with several factors that strongly define the extent of CD-induced DNA-damage, such as the small size of these nanoparticles together with the surface charge and surface functional groups of them which alters the interaction of CDs with the cell 90,102-106 . Comet tail formation attributed to the DNA breaks or failed DNA repair mechanisms induced by oxidative stress. As supported by FT-IR, XPS, and ʐ-potential results, CD surface bearing a high amount of oxygenated functional groups could lead to the production of ROSs, and as a consequence, the oxidative stress may lead the genomic instability 97,107,108 . A recent study by Zhou et al. 90 also indicated that modulating the oxygenated groups, most effectively the number of ketonic carboxyl groups, ROS production by carbon dots can also be controlled, the highest the oxygenated groups on the surface the highest the ROS production capability. Clastogenic damage was also evaluated using CBMN for MCF7 cell lines (Fig. 6a,b) CDs contributed to a significant increase in DNA damage as compared to the negative control for all doses administered (0.25, 2.5 and 50 ppm) after their exposure for 48 h and 72 h (p < 0.01 and p < 0.05, respectively). A concentration-dependent enhancement of the level of clastogenic damage was observed at 48 h (Fig. 6a). After 72 h of exposure, the clastogenicity of CDs at 0.25 ppm was higher than CDs administrated at higher concentrations (Fig. 6b), and the micronucleus frequencies (MN%) decreased as compared to the lower concentrations after 48 h treatment. The difference between the MN value obtained for 0.25 ppm at two incubation times was significant (p < 0.05). The reduction overall in cellular damage at the 72nd hour can be explained by the induction of cellular damage repair mechanisms, which alters the protection of the tumor cell from further chemical attacks and reveal a subsequent recovery from existing ones at increased exposure time. The slight decrease in cell damage for higher CD concentrations can also be explained with the same phenomena. Decreased nanoparticle accumulation in tumor cells at increased exposure times have been also reported by many other researchers which might decrease the effect of CDs on tumor cells 100 . Another explanation of the occurring plateau might be the DNA damage which   www.nature.com/scientificreports/ performed. Cell proliferation is dependent on the cell-cycle progression, in which cells pass through the G0/G1 phase to the S phase and the G2/M phase. As depicted in Fig. 7a, after 48 h of treatment, 57.5%, 34.5% and 8.0% of untreated control cells were in the G0/G1, S and G2/M phases, respectively. Cells treated with CDs at even 0.25 ppm resulted in a decrease in the number of cells in the S phase (p < 0.05), and consequently a slight increase in the number of cells in the G0/G1 phases as compared to the control group. At increased CD concentrations, it has also been shown that induced cell cycle arrest in the G0/G1 phase is happening in a concentration-dependent manner, and the data were statistically significant as compared to the control group (p < 0.001). The effect of CDs on the cell cycle was shown to be time-dependent. As depicted in Fig. 6a,b, at 72 h, the changes in the cell cycle phases differentiated dramatically; 89.5%, 90.5% and 94.2% of MCF7 cells were in the G0/G1 phase respectively for cells treated with 0.25, 2.5 and 50 ppm CD (p < 0.005 for different concentrations of CDs), while only 76.6% of the control population was in the G0/G1 phase (p < 0.001). The significant differences (p < 0.001) in cell cycles for changing CD concentrations showing that prolonged delay in the G0/G1 phase is induced by CDs even at the lowest concentrations (0.25 ppm) which resulted in slower cell growth and delayed the entrance to S phase (Fig. 7a). As compared to untreated cells, CD (2.5 ppm) treated cells in S phase decreased by around 35% and 82% after incubation for 48 h and 72 h, respectively (Fig. 7b). Together with the genotoxicity and clastogenic analysis, the significant arrest of the cell population in G0/G1 phase, and the dramatic decrease in the population of cells in the S and G2/M phases support that CDs could trigger MCF7 cell apoptosis by inducing anomalies at G0/G1 phase by DNA damage and mutagenic stimulate (Fig. 7c). A comparison of the present study and previous literature is provided in Table 2. Most of the listed studies showed that negatively charged CDs with a size lower than 10 nm which can freely enter or interact with the nucleus while there are a couple of examples reporting nucleus targeting by larger sized CDs 47,92,100 , relatively few of them addressing the CD-induced anomalies in cell-cycle 49,108,110,111 . Our results suggest that CDs could have the potential as both drug carriers interacting with the cell-nucleus and therapeutic agents against tumor cells (Scheme 1). However, more studies should be done to handle this enormous therapeutic potential taking account the safety manners.

conclusions
Although benign biocompatibility is attributed to carbon, systematic, and reliable biosafety assessment of carbonbased nanomaterials still needs to be conducted. Carbon dots serve as a useful and usable platform for a wide range of biological and biomedical applications including bio-imaging and nucleus targeted delivery/imaging. Herein, a systematic toxicity analysis of CDs produced from the extract of N. oleander via thermal synthesis method was conducted. In contradiction to many of the previous studies, we concluded that CDs have concentration and time-dependent cytotoxic potentials (at 50 ppm) over MCF7 cells while they comforted the proliferation of healthy HDFa cells even at highest CD concentrations. CDs caused severe DNA damage evident by the formation of COMET tail, micronuclei in MCF7 cells even at concentrations as low as 0.25 ppm. The interference of CDs with cell-cycle progression resulted in cell arrest in G0/G1 phases which showed that they can interact with genetic material and could trigger MCF7 cell apoptosis. Inducing Oxidative Stress Responses and interference with the cell cycle machinery could be due to the structure and surface properties of CDs while other mechanisms could also be involved. Although further studies are warranted to investigate the mutagenicity or carcinogenicity potency of CD in mammalian cells, this work shows evidence that CDs with super tiny size and high amount of oxygen on the surface can specifically affect the cellular function of tumor cells, and thus they have the potential to be used alone as an anti-cancer therapeutic material that can selectively target cancer cells by inducing series of DNA damage.