[No public or meaningful name is available]

Endpoint summary

Key value for chemical safety assessment

Genetic toxicity in vitro

Description of key information

micronucleus test in mammalian cells (comparable to OECD 487):

- Ivask et al 2015: significant increase in the formation of micronuclei in human primary lymphocytes after exposure to concentrations reducing cell viability (by approx. 30%), non-significant increases in the T-lymphocyte cell line JURKAT and in the B-lymphocyte cell line (WIL2-NS) and also less cytotoxic in these cell lines.

DNA strand breaks in mammalian cells (Comet assay, non-guideline):

- Bengtson et al. 2016: no DNA damage observed in FE1 cells (murine pulmonary epithelial cell line) after exposure to 5 -200 µg/mL for 3 h and 24 h, no decrease in cell viability as determined via automated counting of cells stained with fluorescent DNA-reactive dyes (acridine orange and DAPI, using a Nucleocounter and Via1 -Cassettes)

- Chatterjee et al. 2016: slight but significant increase in the olive tail moment detected in BEAS-2B cells (human bronchial epithelial cell line) after exposure to 50 µg/mL, reduction of cell viability by approx. 20%

- Wang et al. 2013: significant and concentration-dependent increases in tail length and the percentage of DNA in the tail were found in HLF cells (human lung fibroblasts) after exposure to 1 to 100 µg/mL for 24 h, no decrease in cell viability (MTT assay) after exposure to 1 µg/mL for 24 h, concentration-dependent increase in cytotoxicity

- De Marzi et al. 2014: increase in tail moment (tail length x fraction of DNA in the tail) observed in A549 (human lung carcinoma cell line), Caco-2 (human colon carcinoma cell line), and Vero (monkey kidney cell line) cells after exposure to 10 µg/mL as lowest tested concentration for 24 h, no decrease of cell viability as determined by MTT assay in independent cell cultures

Link to relevant study records

Reference

Reference 1

Endpoint:

in vitro DNA damage and/or repair study

Type of information:

experimental study

Adequacy of study:

weight of evidence

Study period:

no details given

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

test procedure in accordance with national standard methods with acceptable restrictions

Qualifier:

no guideline available

Principles of method if other than guideline:

DNA strand break levels were determined using the comet assay. PBS was used as negative control, 7.5, 15, 30 µm hydrogen peroxide as positive control. DNA strand breaks were quantified as comet tail length (TL) and % DNA in the tail (%DNA).

GLP compliance:

not specified

Remarks:

no information on GLP compliance available in this publication

Type of assay:

comet assay

Specific details on test material used for the study:

TABLE I. Characterization of GO:

Number of layers 2 - 3
Lateral size T EM (µm) 2 - 3
(Transmission electron microscopy)
Lateral size STEM (µm) ~1
(Scanning transmission electron microscopy)
Surface area B E T (m2/g) -
(Brunauer-Emmet-Teller)
Z-average DLS (nm)b 157
(dynamic light scattering)
PDIc 0.354
(Polydispersity Index)
Zeta potential (mV)d -39.3 ± 1.5
pH 7.02

b Mean hydrodynamic size (6 repeated measurements) in cell culture medium determined with dynamic light scattering.
c Polydispersity Index.
d Mean 6 SEM across 3 repeated measurements.

Target gene:

not applicable

Species / strain / cell type:

mammalian cell line, other: FE1 cells

Details on mammalian cell type (if applicable):

CELLS USED
The spontaneously immortalized murine pulmonary epithelial cell line (FE1), derived from the transgenic mouse strain 40.6 MUTA-Mouse [White et al., 2003] was used.

MEDIA USED
FE1 cells were cultured in an incubator (37°C, 5% CO2) in cell culture medium (DMEM/ F.-121 GlutaMAX, Life Technologies, 31331-028) supplied by 2% heat-inactivated Fetal Bovine Serum (Gibco, 10106-169), 1% Penicillin (10,000 IU/ml) Streptomycin (10,000 pg/ml, Biological Industries, 949¬208), and 0.001% Epidermal growth factor (Sigma E4127).

Additional strain / cell type characteristics:

not applicable

Metabolic activation:

not applicable

Test concentrations with justification for top dose:

0, 5, 10, 25, 50, 100, 200 µg/ml

Untreated negative controls:

not specified

Negative solvent / vehicle controls:

yes

True negative controls:

no

Positive controls:

other:

Remarks:

As positive control, FE1 cells exposed to 7.5, 15, and 30 µM hydrogen peroxide

Details on test system and experimental conditions:

DNA strand break levels were determined using the comet assay.
As positive control, FE1 cells exposed to 7.5, 15, and 30 µM hydrogen peroxide for 30 min were included and showed statistically signiicantly increased DNA strand break levels compared to PBS exposed cells. Levels of DNA strand breaks were determined in FE1-cells after exposure to GO (5-200 µg/ml) following 3 hr and 24 hr of exposure.
In brief, 10 µl cell suspensions of FE1 cells, preserved in 17% DMSO + 83% fetal bovine serum, were embedded in agarose gel on 20-well Trevigen comet slides (Gaithersburg, MD). Slides were placed in lysis buffer overnight at 4°C. The next day, the slides were placed in alkaline solution for 30 min prior to alkaline electrophoresis (25 min, 1.15 V/cm, and 294 mA) with circulation (70 ml/min). After electrophoresis, slides were neutralized for 10 min. The slides were stained with SYBR Green for 30 min. FE1 cells exposed for 30 min at 4°C to PBS or 7.5, 15, 30 µm hydrogen peroxide were included as negative and positive control, respectively. Analysis and scoring of DNA strand breaks was performed with IMSTAR path-inder system (IMSTAR, Paris, France). DNA strand breaks were quantified as comet tail length (TL) and %DNA in the tail (%DNA).

Evaluation criteria:

For statistical analysis comet tail length (TL) and % DNA in the tail (% DNA) were normalized to the mean control level (0 µg/ml) for the respective experiment.

Statistics:

All statistical analyses were performed in R (v3.10) and Rstudio (v 0.98.1091). The statistical analyses were performed with One Way Analysis of Variance (ANOVA) and presented as mean +/- standard error of the means (SEM). Data were separated by individual particle and dose was set as categorical variable. In case of signiicant main effect of dose (signiicance level; 0.05), a pairwise comparison across doses was performed with Tukey's honest significant difference (HSD) test with adjusted correction (significant level; 0.05, confidence interval; 0.95).

Key result

Species / strain:

mammalian cell line, other: FE1 cells

Metabolic activation:

not applicable

Genotoxicity:

negative

Cytotoxicity / choice of top concentrations:

not specified

Vehicle controls validity:

valid

Untreated negative controls validity:

not valid

Positive controls validity:

valid

Additional information on results:

The levels of DNA strand breaks were presented as TL and % DNA. As positive control, FE1 cells exposed to 7.5, 15, and 30 µM hydrogen peroxide for 30 min were included and showed statistically significantly increased DNA strand break levels compared to PBS exposed cells. Levels of DNA strand breaks were determined in FE1-cells after exposure to GO (5-200 µg/ml) following 3 hr and 24 hr of exposure. Exposure had no effect on DNA strand break levels at doses up to 200 µg/ml at either time point.

Conclusions:

The few layered GO was genotoxic to FE1 murine lung epithelial cells at concentrations up to 200 µg/ml.

Executive summary:

This study analyzed the genotoxic effects of graphene oxide on FE1 cells ( murine pulmonary epithelial cell line).

The levels of DNA strand breaks were assessed with the comet assay and presented as TL and % DNA. As positive control, FE1 cells exposed to 7.5, 15, and 30 µM hydrogen peroxide for 30 min were included and showed statistically significantly increased DNA strand break levels compared to PBS exposed cells. Levels of DNA strand breaks were determined in FE1-cells after exposure to GO (5-200 µg/ml) following 3 hr and 24 hr of exposure to reflect both transient and prolonged genotoxicity. Exposure had no effect on DNA strand break levels at doses up to 200 µg/ml at either time point.

The result of this study is supported by an in vivo cytogenicity test of Bengtson et al. (2017) where no dose-dependent increase in DNA breaks was observed for graphene oxide after pulmonary exposure of Graphene oxide in mice, therefore the test substance graphene oxide is not considered as genotoxic.

Reference 2

Endpoint:

in vitro DNA damage and/or repair study

Type of information:

experimental study

Adequacy of study:

weight of evidence

Study period:

no details given

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

test procedure in accordance with national standard methods with acceptable restrictions

Qualifier:

no guideline available

Principles of method if other than guideline:

- Principle of test: An in vitro comet assay was performed on BEAS-2B cell line.The study was carried out by treatment of the cells (2.5 E4 cells/mL) with single layer graphene oxide (SLGO) and few layer graphene oxide (FLGO) at 10 and 50 mg/L for 24h.
- Parameters analysed / observed: DNA damage was expressed as the tail extent moment using an image analysis by the Komet 5.5 software (Kinetic Imaging Ltd.).

GLP compliance:

not specified

Remarks:

no information on GLP compliance available in this publication

Type of assay:

comet assay

Specific details on test material used for the study:

The commercially available SLGO, FLGO (4-8 layers) were purchased from Cheap Tubes Inc. (Brattleboro, VT, USA) in powdered and prepared stock in distilled water at 1000 mg/L with sonication. In addition the information supplied by the manufacturer, the GFNs (graphene family nanomaterials) were characterized by using atomic force microscopy (AFM) and Raman spectroscopy. Surface topography, height profile and lateral size distribution of the GFNs were examined by AFM (Park Systems XE-BiO) in non-contact mode. Raman spectroscopy was performed at room temperature with a Micro Raman system (UniRAM3500, UniNanoTech Co., Ltd., Korea) with a 532 nm laser. The calibration was initially made using an internal silicon reference at 500 cm-1 and gave a peak-position resolution of less than 1 cm-1. The spectra were measured from 0 to 5000 cm-1. The size distribution and ζ-potential of the MWCNTs [30 mg/L in Dulbecco modified eagle medium (DMEM/F12) culture media] were evaluated by using a photal dynamic light scattering spectrometer (DLS) (ELSZ-1000, Otsuka Electronics Co., Inc.).

Target gene:

not applicable

Species / strain / cell type:

mammalian cell line, other: BEAS-2B cells

Details on mammalian cell type (if applicable):

Human bronchial epithelial cells; cultured in DMEM/F12 (GIBCO), supplemented with 10% (v/v) fetal bovine serum and 1% (v/v) antibiotics, at 37 °C in a 5% CO2 atmosphere.

Additional strain / cell type characteristics:

not applicable

Metabolic activation:

not applicable

Test concentrations with justification for top dose:

10 and 50 mg/L

Vehicle / solvent:

The GFNs were freshly prepared in cell culture medium (DMEM/F12) at the desired concentrations with appropriate amount from the stock (1000 mg/L in distilled water) and was sonicated for 10 min before biological exposure.

Untreated negative controls:

no

Negative solvent / vehicle controls:

yes

True negative controls:

no

Positive controls:

no

Details on test system and experimental conditions:

BEAS-2B cells (human bronchial epithelial cells) were cultured in DMEM/F12 (GIBCO), supplemented with 10% (v/v) fetal bovine serum and 1% (v/v) antibiotics, at 37 °C in a 5% CO2 atmosphere.
- Cell density at seeding: 2.5 E4 cells/mL were seeded in 6 well plates before 24 h of treatment and treated with GFNs at 10 and 50 mg/L for next 24h.

DURATION
- Preincubation period: 24h
- Exposure duration: 24h
STAIN (for cytogenetic assays): 50 µL ethidiumbromide (5 µg/ mL)

METHODS OF SLIDE PREPARATION AND STAINING TECHNIQUE USED: after the exposure the cells were trypsinised and centrifuged at 1500 rpm for 3 min. 1-3 E4 cells were resuspended in 0.5% low-melting-point agarose (LMPA, Bio-Rad Laboratories, Hercules) at 1:2 ratio. The resuspended cells in agarose were put onto slides pre-coated with 1% normal-melting agarose. After solidification, the slides were immersed in cold lysing solution (2.5 M NaCl, 100 mM EDTA, 10 mM Tris-base, 1% N-laurolsarcosinate, 1% Triton X-100) for 1.5 h at 4°C, after which they were transferred to an electrophoresis tank containing freshly made electrophoresis buffer (1 mM EDTA, 300 mM NaOH; pH>13) and were kept for 20 min at room temperature to allow DNA unwinding. Electrophoresis was performed in the same buffer at room temperature for 20 min at 25 V and 300 mA. The slides were neutralized three times with 0.4 M Tris-chloride buffer (pH 7.5), air-dried, and fixed in 70% ethanol. The slides were analyzed using a luorescence microscope (Leica DM IL) at 40x magnification.

NUMBER OF CELLS EVALUATED: 50 cells per slide (3 slides per treatment) were examined.

Rationale for test conditions:

In a previous study it has shown that GNPs exhibited higher toxicity than GOs (SLGO & FLGO) in BEAS-2B cells [N. Chatterjee, J.S. Yang, K. Park, S.M. Oh,J. Park,J. Choi, Screening of toxic potential of graphene family nanomaterials using in vitro and alternative in vivo toxicity testing systems, Environ. Health Toxicol. 30 (2015)]. Based on effective concentrations (ECs) of the GFNs 10 mg/L (less than EC10 of each compound) and 50 mg/L (around EC20 of the GOs and less the EC50 of GNPs) doses were chosen for the evaluation of geno-epigenotoxic potentialities of the GFNs at low to moderate cytotoxicity levels.

Evaluation criteria:

DNA damage was expressed as the tail extent moment using an image analysis by the Komet 5.5 software (Kinetic Imaging Ltd.).

Statistics:

The statistical significance of differences among/between treatments was determined using one way analysis of variance (ANOVA). This was followed by a post-hoc test (Tukey, P< 0.05). All statistical analyses were carried out using SPSS 12.0KO (SPSS Inc.) and graphs were prepared in SigmaPlot (Version 12.0).

Key result

Species / strain:

mammalian cell line, other: BEAS-2B cells

Metabolic activation:

not applicable

Genotoxicity:

positive

Cytotoxicity / choice of top concentrations:

not specified

Vehicle controls validity:

valid

Untreated negative controls validity:

not valid

Positive controls validity:

not specified

Conclusions:

The results clearly indicated a genotoxic effect of single layer graphene oxide (SLGO) and few layer graphene oxide (FLGO).

Executive summary:

This study analyzed the genotoxic effects of single layer graphene oxide (SLGO) and few layer graphene oxide (FLGO) using a 24-h comet assay on BEAS-2B cells (human bronchial epithelial cells). The genotoxicity study, carried out with a Comet assay, showed an increase in DNA damages in both treated doses (10 mg/L and 50 mg/L for 24 h) and significant dose dependency was found in SLGO (p< 0.034).

Reference 3

Endpoint:

in vitro DNA damage and/or repair study

Type of information:

experimental study

Adequacy of study:

weight of evidence

Study period:

no details given

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

test procedure in accordance with national standard methods with acceptable restrictions

Qualifier:

no guideline available

Principles of method if other than guideline:

DNA strand break levels were determined using the comet assay. DNA strand breaks were quantified as comet tail length (TL) and % DNA in the tail (%DNA).

GLP compliance:

not specified

Remarks:

GLP compliance is not specified in this publication

Type of assay:

comet assay

Specific details on test material used for the study:

Atomic force microscopy images showed that the lateral width of the GO was 200 ~ 500nm and the thickness is about 1 nm, which indicated that GO was single- or two-layered sheets. The thickness of GO was about 1 nm, the Zeta Potential Distribution of GO was determined to be -65.1mV.

Species / strain / cell type:

other: human lung fibroblasts

Additional strain / cell type characteristics:

not applicable

Metabolic activation:

without

Vehicle / solvent:

cell culture medium: Dulbecco’s modified Eagle’s high glucose medium (DMEM/high; HyClone, Logan, Utah, USA) supplemented with 10% (v/v) fetal bovine serum (FBS; HyClone, Logan, Utah, USA)

Untreated negative controls:

other: cell culture medium was also used as solvent

Negative solvent / vehicle controls:

yes

True negative controls:

no

Positive controls:

not specified

Details on test system and experimental conditions:

METHOD OF APPLICATION: in medium; in agar (plate incorporation); preincubation; in suspension; as impregnation on paper disk
- Cell density at seeding (if applicable): HLF cells were plated in 12-well plates (Corning) at a concentration of 2x10^5 in 1.5 ml of culture medium.

DURATION
- Preincubation period: 12 h
- Exposure duration: 24 h

STAIN (for cytogenetic assays): SYBR Green I

METHODS OF SLIDE PREPARATION AND STAINING TECHNIQUE USED: The treated cells were harvested and washed three times with PBS, then embedded in 0.1% low melting agarose and were immediately spread on a
comet slide (self-made). The slide was allowed to solidify for 40 min at 4 °C. Cells were then subjected to lyse in alkaline buffer (2.5 M NaCl, 100 mM EDTA, 10 nM Tris and 1% Triton X-100) for 1.5 h in the dark at 4 °C. After that, the slide was transferred to an electrophoresis tank filled with fresh electrophoresis buffer and kept for 20 min at room temperature for DNA unwinding. Electrophoresis was operated in this buffer for 20 min at 18 V. Afterwards, the slide was immersed in neutralization buffer (0.5M Tris-HCl, pH 7.5) for 15 min then washed with PBS. The DNA was stained with SYBR Green I (self-made) and viewed with a Nikon fluorescence microscope (Japan).

NUMBER OF CELLS EVALUATED: more than 300 comets were analyzed for each concentration

DETERMINATION OF CYTOTOXICITY
- Method: mitotic index; cloning efficiency; relative total growth; other: MTT assay in independent samples (96 well plates)

Rationale for test conditions:

not specified

Evaluation criteria:

not specified

Statistics:

All these experiments were replicated three or more independent times and the data were presented as the mean +/- standard deviation (SD). Comparisons with untreated controls were carried out using Student’s t-test (Microsoft Excel; Microsoft Corporation, Redmond, WA, USA). Comparisons between different materials at the same concentrations were determined through oneway ANOVA performed in SPSS 16.0 (SPSS Inc., Chicago, IL, USA). A P-value < 0.05 was considered to be statistical significant.

Species / strain:

other: human lung fibroblasts

Metabolic activation:

without

Genotoxicity:

positive

Cytotoxicity / choice of top concentrations:

cytotoxicity

Vehicle controls validity:

valid

Untreated negative controls validity:

not applicable

Positive controls validity:

not examined

Additional information on results:

When electrophoreses were done with an intact nucleus, a comet-like tail implied the presence of a damaged DNA strand that would lag behind. The length of tail increased with the extent of the DNA damage. In this assay, tail length and the percentage of DNA in tail were analyzed. Data analysis of the comet assay showed that there were statistically significant increases in tail length and the percentage of DNA in tail after 24-h treatment with the tested concentrations of GO (1, 50 and 100 µg/mL).

Conclusions:

Exposure of human lung fibroblasts with 1, 50 and 100 µg/mL GO for 24 h caused a concentration-dependent, significant increase in tail length and the percentage of DNA in tail in an in vitro Comet assay.

Executive summary:

This study analyzed the genotoxicity of graphene oxide (GO) using a Comet assay on human lung fibroblasts. Cells were exposed to 1, 50 and 100 µg/mL GO for 24 h. Treatment with 50 and 100 µg/mL GO for 24 h significantly reduced cell viability (MTT assay) to approx. 70% (derived from graph), whereas 1 µg/mL GO did not impact on cell viability. In the Comet assay, a concentration-dependent increase in tail length and the percentage of DNA in tail was found. The increases were statistically significant for all tested concentrations of GO.

Reference 4

Endpoint:

in vitro DNA damage and/or repair study

Type of information:

experimental study

Adequacy of study:

weight of evidence

Study period:

no details given

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

test procedure in accordance with national standard methods with acceptable restrictions

Qualifier:

no guideline available

Principles of method if other than guideline:

- Principle of test: An in vitro comet assay was performed on A549, CaCo2 and Vero cell lines. The study was carried out by varying both graphene oxide (GO) concentration (10 µg/mL, 50 µg/mL, 100 mg/mL) and flakes sizes of 1320 nm and 130 nm.
- Parameters analysed / observed: The samples (containing single DNA fragments reflecting the amount of DNA breakage) were analysed with a fluorescence microscope and scored using Comet IV image analysis software. The analysis was performed measuring the tail moment (tail length x fraction of DNA in the tail) and comparing the treated samples versus the not-treated ones (water, solvent control).

GLP compliance:

not specified

Remarks:

no information on GLP compliance available in this publication

Type of assay:

comet assay

Specific details on test material used for the study:

Pristine GO flakes (micro-GO) with a maximum lateral size of 140 µm were prepared in water solution using a modified Hummers method. Part of the solution was then submitted to 26-h sonication to induce controlled fragmentation of the GO flakes (nano-GO). The sonication was performed with an Elmasonic SI OH bath (operating frequency 37 kHz - effective power 30 W). The GO characterization was performed by scanning electron microscopy (SEM) and Raman spectroscopy (RS). The SEM images were obtained using a Zeiss-Gemini LEO 1530 system without sputter coating the samples. RS was performed with a Horiba-JobinYvon LABRAM system (ʎ=633 nm, 1 µm spatial resolution, and ~2 cm-1 spectral resolution). In all cases, samples were prepared by spin coating (at room temperature and in dry air) 10 µL of a 0.5 mg/mL water dispersed GO (either sonicated or not) at 3,000 rpm for 1 min on 100 nm SiO2/Si. The RS data were acquired in air. The statistical analysis of the GO flakes size was carried out on the SEM images with ImageJ software. The size and the area of the GO flakes was determined with a pixel counting procedure. The total number of flakes counted for the analysis ranged between 1000 and 2000.

Species / strain / cell type:

mammalian cell line, other: A549

Details on mammalian cell type (if applicable):

A human alveolar adenocarcinoma cell line, CCL-185CELLS USED
- Number of passages if applicable: three in vitro culture passages were performed before starting each experiment.

MEDIA USED
- Type and identity of media including CO2 concentration if applicable: cells were cultured in a Dulbecco's modified eagle medium (DMEM) completed with 10% of heat-inactivated fetal bovine serum (FBS-HyClone), 2% L-glutamine (Sigma-Aldrich) and 0.5% of penicillin/streptomycin (Sigma-Aldrich) and maintained in controlled atmosphere (37°C, 5% C02)

Additional strain / cell type characteristics:

not specified

Species / strain / cell type:

mammalian cell line, other: CaCo2

Details on mammalian cell type (if applicable):

Human colorectal adenocarcinoma cell lines, HTB-37
- Number of passages if applicable: three in vitro culture passages were performed before starting each experiment.

MEDIA USED
- Type and identity of media including CO2 concentration if applicable:
cells were cultured in a Dulbecco's modified eagle medium (DMEM) completed with 10% of heat-inactivated fetal bovine serum (FBS-HyClone), 2% L-glutamine (Sigma-Aldrich) and 0.5% of penicillin/streptomycin (Sigma-Aldrich) and maintained in controlled atmosphere (37°C, 5% C02)

Additional strain / cell type characteristics:

not specified

Species / strain / cell type:

mammalian cell line, other: Vero

Details on mammalian cell type (if applicable):

Kidney epithelial cell line from an African green monkey, CCL-81)
- Number of passages if applicable: three in vitro culture passages were performed before starting each experiment.

MEDIA USED
- Type and identity of media including CO2 concentration if applicable:
cells were cultured in a Dulbecco's modified eagle medium (DMEM) completed with 10% of heat-inactivated fetal bovine serum (FBS-HyClone), 2% L-glutamine (Sigma-Aldrich) and 0.5% of penicillin/streptomycin (Sigma-Aldrich) and maintained in controlled atmosphere (37°C, 5% C02)

Additional strain / cell type characteristics:

not specified

Test concentrations with justification for top dose:

10 µg/mL, 50 µg/mL and 100 µg/mL, plus negative control

Vehicle / solvent:

none

Untreated negative controls:

no

Negative solvent / vehicle controls:

yes

True negative controls:

no

Positive controls:

yes

Positive control substance:

methylmethanesulfonate

Details on test system and experimental conditions:

The comet assay was performed on A549, CaCo2 and Vero cell lines. The cells were seeded into a 96-well plate with complete medium (104 cells in 100 µL/each well). The day after seeding, micro GO and nano GO solutions were added to the culture medium in order to obtain the final GO concentrations of 10 µg/mL, 50 µg/mL and 100 µg/mL, plus negative control. A 24-h exposure treatment was carried out.
After the lysis eyposure electrophoresis was performed for 40 min. The samples were stained with 4',6-diamidino-2-phenylindole (DAPI) and analysed with a fluorescence microscope (Nikon Eclipse E600 with super high pressure mercury lamp). Sixty randomly selected cells from each concentration were scored using Comet IV image analysis software.

Rationale for test conditions:

no details given

Evaluation criteria:

Sixty randomly selected cells from each concentration were scored using Comet IV image analysis software. The fluorescence microscope used was a Nikon Eclipse E600. The analysis was performed measuring the tail moment (tail length x fraction of DNA in the tail) and comparing the treated samples versus the not-treated ones (water, solvent control).

Statistics:

The means and standard deviations (SD) were calculated for descriptive statistical documentation. The data are the means ± SD, calculated for the three replicates performed for each experimental point. SPSS software was used for analytical statistics.

Species / strain:

mammalian cell line, other: A549 cell line

Metabolic activation:

not applicable

Genotoxicity:

positive

Cytotoxicity / choice of top concentrations:

not specified

Vehicle controls validity:

valid

Untreated negative controls validity:

not valid

Positive controls validity:

valid

Remarks:

methyl methanesulfonate (MMS) 400µM

Species / strain:

mammalian cell line, other: Vero cell line

Metabolic activation:

not applicable

Genotoxicity:

positive

Cytotoxicity / choice of top concentrations:

not specified

Vehicle controls validity:

valid

Untreated negative controls validity:

not valid

Positive controls validity:

valid

Remarks:

methyl methanesulfonate (MMS) 400µM

Species / strain:

mammalian cell line, other: CaCo2 cell line

Metabolic activation:

not applicable

Genotoxicity:

positive

Cytotoxicity / choice of top concentrations:

not specified

Vehicle controls validity:

valid

Untreated negative controls validity:

not valid

Positive controls validity:

valid

Remarks:

methyl methanesulfonate (MMS) 400µM

Additional information on results:

The results of the genotoxic response to micro GO and nano GO at 10, 50, and 100 µg/mL on the three studied cell lines were compared to the solvent control cells. Moreover, methyl methanesulfonate (MMS) as positive control was used at the concentration of 400 µM. The % of tail obtained with such treatment was 66 ± 20 for the three used cell lines.
In the case of treatment with pristine micro-GO flakes the genotoxic response was moderate (overall less than 20% of the tail) at 10 µg/mL, while it definitely increased at higher GO concentrations (50 and 100 µg/mL), reaching % tail values, at 100 mg/mL pristine micro-GO exposure, exceeding 80% and 60% for A549 and, respectively, Vero and CaCo2.
The results obtained upon treatment with sonicated nano-GO demonstrate a remarkable genotoxic response with % tail values ranging from 60 to 80% for Vero and CaCo2, and A549, respectively. Such tail values did not vary with the concentration of nano-GO.

Remarks on result:

other: result for nano- and micro-GO

Conclusions:

The results clearly indicated a size-dependent genotoxic effect of GO flakes.

Executive summary:

This study analyzed the genotoxic effects of chemically pristine GO using a 24-h comet assay on different cell lines: A549- human lung carcinoma, CaCo2- human colon carcinoma and Vero-monkey kidney cells. The study was designed to clearly assess the effect of pristine GO by varying the GO flake sizes without altering its surface chemistry. The genotoxicity study, carried out with a Comet assay, showed a size-dependent effect on the three cell lines. Micron-sized flakes of GO induced a genotoxic effect which increased with concentration (10 µg/mL < 50 µg/mL <100 µg/mL) while nano-sized GO flakes at the same concentration values showed a saturated high genotoxic response.

Reference 5

Endpoint:

in vitro cytogenicity / micronucleus study

Type of information:

experimental study

Adequacy of study:

weight of evidence

Study period:

no details given

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

comparable to guideline study with acceptable restrictions

Qualifier:

equivalent or similar to guideline

Guideline:

OECD Guideline 487 (In vitro Mammalian Cell Micronucleus Test)

Deviations:

yes

Remarks:

only one experimental condition (extended exposure), no metabolic activation system

GLP compliance:

not specified

Remarks:

no information on GLP compliance available in this publication

Type of assay:

in vitro mammalian cell micronucleus test

Specific details on test material used for the study:

For the genotoxicity study, Graphene oxide NP were diluted in cell culture medium (RPMI 1640 medium to 0.05 mg/mL (Sigma-Aldrich) containing 4.5 g/L glucose and supplemented with 50 µM glutamine, 10% FBS, 1 mM Na-pyruvate, and 10 mM HEPES buffer) and vortexed. The diluted NPs appeared precipitation free at room temperature.

Target gene:

not applicable

Species / strain / cell type:

primary culture, other: human primary lymphocytes

Additional strain / cell type characteristics:

not applicable

Cytokinesis block (if used):

4.5 µg/mL of cytochalasin B

Metabolic activation:

without

Test concentrations with justification for top dose:

6.25 µg/mL, 12.5 µg/mL, 25 µg/mL, 50 µg/mL, 100 µg/mL, 200 µg/mL, 400 µg/mL

Vehicle / solvent:

cell culture medium (RPMI 1640 containing 4.5 g/L glucose and supplemented with 50 µM glutamine, 10% FBS, 1 mM Na-pyruvate, and 10 mM HEPES buffer)

Untreated negative controls:

no

Negative solvent / vehicle controls:

yes

True negative controls:

no

Positive controls:

yes

Positive control substance:

other: hydrogen peroxide

Details on test system and experimental conditions:

METHOD OF APPLICATION: in medium
- Cell density at seeding (if applicable): 1 X 10^6 cells/mL

DURATION
- Exposure duration: exposed to test chemicals for 44 h, after which cytB was added. Lymphocytes were incubated for another 24-28 h.

STAIN (for cytogenetic assays): Hemacolor stains (Merck)

METHODS OF SLIDE PREPARATION AND STAINING TECHNIQUE USED: 100 µL of cells was deposited onto a microscope glass slide using the Cytospin system (Thermo Scientiic). The glass slide was then dried for 10 min at room temperature, dipped for 10 min into the Hemacolor fixing solution (Merck), quickly dried, then dipped 10 times in Hemacolor red reagent following by 6 dips in the Hemacolor blue reagent. The slides were then rinsed with deionized water and dried overnight. A coverglass was mounted using the DEPEX Mounting Medium.

NUMBER OF CELLS EVALUATED: From each treatment, 500 cells were counted (in the case of toxic concentrations, the cell count was lower, down to 100 cells per treatment).

DETERMINATION OF CYTOTOXICITY
- Method: conversion of resazurin (determined with independently exposed samples)

OTHER EXAMINATIONS:
- Determination of polyploidy:
- Determination of endoreplication:
- Methods, such as kinetochore antibody binding, to characterize whether micronuclei contain whole or fragmented chromosomes (if applicable):

- OTHER:

Evaluation criteria:

Scoring criteria were similar to those described by Fenech (2007). The nuclear division index (NDI) was based on mono-, bi-, and multinuclear cells, while no apoptotic or necrotic cells were taken into account. NDI = (mononuclear cells + 2 X binuclear cells + 3 X multinuclear cells) / (total number of cells). Also, the number of micronuclei per 500 binuclear cells was calculated.

Statistics:

A one-tailed t test was carried out to find the significant differences (p < 0.05) between the number of micronuclei in the nonexposed control and NP exposed cells.

Key result

Species / strain:

lymphocytes: human primary lymphocytes

Metabolic activation:

not applicable

Genotoxicity:

positive

Cytotoxicity / choice of top concentrations:

cytotoxicity

Vehicle controls validity:

valid

Untreated negative controls validity:

not valid

Positive controls validity:

valid

Additional information on results:

TEST-SPECIFIC CONFOUNDING FACTORS
- Effects of pH: not specified
- Effects of osmolality: not specified
- Evaporation from medium: not applicable
- Water solubility: not specified
- Precipitation: no indication in publication
- Definition of acceptable cells for analysis: not specified


CYTOKINESIS BLOCK (if used)
- Distribution of mono-, bi- and multi-nucleated cells:

NUMBER OF CELLS WITH MICRONUCLEI
- Number of cells for each treated and control culture:
- Indication whether binucleate or mononucleate where appropriate:

HISTORICAL CONTROL DATA (with ranges, means and standard deviation and confidence interval (e.g. 95%)
- Positive historical control data:
- Negative (solvent/vehicle) historical control data:

ADDITIONAL INFORMATION ON CYTOTOXICITY:
- Measurement of cytotoxicity used: [complete, e.g. CBPI or RI in the case of the cytokinesis-block method; RICC, RPD or PI when cytokinesis block is not used]
- Other observations when applicable: [complete, e.g. confluency, apoptosis, necrosis, metaphase counting, frequency of binucleated cells]

Positive Control: When 500 cells exposed to 100 µM H₂O₂were counted, the number of micronuclei per 500 binucleated cells was 15.0 in the case of T-lymphocyte cell line, 13.2 in the case of B-lymphocyte cell line and 10.3 in the case of blood lymphocytes.

Similar to H₂O₂, 500 binucleated cells of GO-exposed T- and B-lymphocytes and primary lymphocyte cells were counted, and their formed micronuclei were recorded. The results show that the number of micronuclei in response to GO increased at concentrations where significant decrease in cellular viability was observed (as shown by the resazurin assay of cellular metabolic state).

Half-Maximal Inhibitory Concentration (IC50) Values with 95% Confidence Intervals, of the Tested GO toward Human Lymphocytes in the Resazurin Assay:

Primary lymphocytes: 341.7 µg/mL

T-lymphoblastoma cells JURKAT: 401.9 µg/mL

B-lymphoblastoma cells WIL-2NS: 176.7 µg/mL

GO nanosheets cause half-maximal inhibitory concentrations ranging between 176 and 400 µg GO/mL

Conclusions:

GO sheets induced a statistically significant increase in the formation of micronuclei in primary human lymphocytes after exposure to 50 and 100 µg/mL, while for these concentrations the remaining cell viability was below 70%.

Executive summary:

The genotoxic effects of GO was tested in the Cytokinesis-Block Micronucleus Assay (CBMN) according to OECD TG 487. The induction of micronuclei by GO sheets was evaluated using human primary lymphocytes and H₂O₂ as a positive control chemical. The number of micronuclei (MN) per 500 binucleated cells was calculated, and cytotoxicity was evaluated via resazurin conversion.

Exposure of primary human lymphocytes to up to 400 µg GO/mL for 24 h significantly reduced cellular viability to approx. 30%, the half-maximal inhibitory concentration (IC50) of GO towards primary lymphocytes was estimated as 176.7 µg/mL. In the primary lymphocytes, a statistically significant formation of micronuclei was observed after exposure to 50 and 100 µg/mL, while for these concentrations the remaining cell viability was below 70%.

In this study, similar in vitro tests were conducted in parallel with human B-lymphocyte cell line WIL2-NS and human T-lymphocyte cell line JURKAT. In the T-lymphocyte cell line, an IC50 of 341.7 µg/mL was found, and treatment of the human B-lymphocyte cell line WIL2-NS with up to 400 µg GO/mL resulted in even lower cytotoxicity (estimated IC50 of 401.9 µg/mL). In both cell lines, the formation of micronuclei was increased but did not reach statistical significance after exposure to up to 400 µg/mL for 24 h.

Taken together, these results show that the number of micronuclei in response to GO increased at concentrations where significant decrease in cellular viability was observed.

Genetic toxicity in vivo

Description of key information

micronucleus test in mice (comparable to OECD 474):

- Liu et al. 2013: significant and dose-dependent increase in polychromatic erythrocytes with micronuclei after repeated intravenous administration of 1, 2, and 4 mg GO/kg bw for five consecutive days (corresponding to total doses of 5, 10, and 20 mg/kg bw, respectively), no modulation of PCE/NCE ratio indicate no direct toxicity of GO to erythrocytes under these conditions.

DNA strand breaks in mice (Comet assay, comparable to OECD 489):

- Bengtson et al. 2017: no consistent pattern of DNA damage induced in BAL, lung or liver cells after single intratracheal instillation of 18, 54 or 162 µg GO/mouse (corresponding to 0.91, 2.74, and 8.22 mg/kg bw, respectively) observed after 1, 3, 28 or 90 days post exposure. Intratracheal administration of GO also caused acute toxicity and a marked inflammatory response in the lung which peaked after 1 and 3 days.

Link to relevant study records

Reference

Reference 1

Endpoint:

in vivo mammalian cell study: DNA damage and/or repair

Type of information:

experimental study

Adequacy of study:

weight of evidence

Study period:

no details given

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

study well documented, meets generally accepted scientific principles, acceptable for assessment

Reason / purpose for cross-reference:

reference to same study

Qualifier:

equivalent or similar to guideline

Guideline:

OECD Guideline 489 (In vivo Mammalian Alkaline Comet Assay)

Deviations:

yes

Remarks:

only one exposure, no early sampling at 2-6 h after treatment but sampling at 1, 3, 28 or 90 days post exposure

GLP compliance:

not specified

Type of assay:

mammalian comet assay

Specific details on test material used for the study:

GO was prepared in 0.2 µm filtered, y-irradiated Nanopure Diamond UV water (Pyrogens: < 0,001 EU/mL, total organic carbon: < 3.0 ppb) added 0.1% Tween80® (TW80) to a final concentration of particles of 3.24 mg/mL. To achieve a homogenous dispersion, the final solution was then prepared by probe sonication on ice for 16 min with 10% amplitude (Branson Sonifier S-450D, Branson Ultrasonics Corp., Danbury, CT, USA) equipped with disruptor horn (model number 101-147-037). Following sonication, solution was further diluted to 1.08 mg/mL and sonicated for 2 minutes. Dilution was further diluted to 0.36 mg/mL and sonicated for 2 minutes. As vehicle control (VC), Nanopure water added 0.1% TW80 was prepared by procedure as described above. Suspensions were instilled in mice within 20 minutes after sonication.

Species:

mouse

Strain:

C57BL

Details on species / strain selection:

C57BL/6J mice

Details on test animals or test system and environmental conditions:

TEST ANIMALS
- Source: Taconic Europe (Ejby, Denmark)
- Age at study initiation: 7-weeks old
- Weight at study initiation: 19.7 ± 1 g
- Assigned to test groups randomly: grouped based on exposure and dose level (VC groups n = 8, exposed groups n = 7)
- Fasting period before study: not specified
- Housing: polypropylene cages with sawdust bedding and enrichment
- Diet (e.g. ad libitum): food (Altromin 1324) ad libitum
- Water (e.g. ad libitum): tap water ad libitum
- Acclimation period: not specified

ENVIRONMENTAL CONDITIONS
- Temperature (°C): 21 ± 1°C
- Humidity (%): 50 ± 10%
- Air changes (per hr): not specified
- Photoperiod (hrs dark / hrs light): 12-h light and 12-h dark cycle

Route of administration:

intratracheal

Vehicle:

0.2 pm filtered, y-irradiated Nanopure Diamond UV water (Pyrogens: < 0,001 EU/ml, total organic carbon: < 3.0 ppb) added 0.1% Tween80

Details on exposure:

single intratracheal instillation (50 pl/mouse) under isoflurane sedation

Duration of treatment / exposure:

not specified

Frequency of treatment:

single treatment

Post exposure period:

1, 3, 28 or 90 days

Dose / conc.:

18 other: µg/mouse

Remarks:

corresponds to 0.91 mg/kg bw

Dose / conc.:

54 other: µg/mouse

Remarks:

corresponds to 2.74 mg/kg bw

Dose / conc.:

162 other: µg/mouse

Remarks:

corresponds to 8.22 mg/kg bw

No. of animals per sex per dose:

only female mice, vehicle control groups n = 8, exposed groups n = 7

Control animals:

yes, concurrent vehicle

Positive control(s):

Carbon Black Printex90 (P90), provided by Degussa (Frankfurt, Germany) was included as reference material (162 µg/mouse) based on findings from previous studies showing inflammatory and genotoxic response in mice following single instillation (Bourdon et al., 2012).
- Justification for choice of positive control(s): The authors routinely include P90 as a reference material (Kyjovska et al., 2015a, Kyovska et al., 2015b, Poulsen, 2015, Wallin et al., 2016, Saber et al., 2012a, Saber et al., 2012b, Saber et al., 2016, Saber et al., 2012c) and P90 has been described and characterized in detail (Saber, 2012a, Jacobsen et al., 2008).
- Route of administration: intratracheal
- Doses / concentrations: 162 µg/mouse, corresponds to 8.22 mg/kg bw

Tissues and cell types examined:

Bronchoalveolar lavage, lung, liver

Details of tissue and slide preparation:

All mice were anesthetized by i.p. injection of 0.1 mL ZRF solution (Zoletil 250 mg, Rompun 20 mg/mL, Fentanyl 50mg/mL in sterile isotone saline). Blood was withdrawn from the heart and stabilized using 36 µl K2EDTA and followed by collection of BAL where lungs were flushed twice with 0.8 ml 0.9% sterile saline. Total BAL recovery was about 1.4 mL. BAL samples were immediately stored on ice until further preparation. BAL cells were prepared on glass slides and stained with May-Grünewald-Giemsa staining. Details about preparation method have been described previously (Kyovska et al. 2015a). Images of BAL cells were acquired at 100x on an Olympus BX 43 microscope with a Qimaging Retiga4000R camera. Uneven illumination in brightfield images was corrected using ImageJ (Schneider et al., 2012) and the Calculator Plus plugin via the formula: Corrected image = (Image / background) * 255. The background image was a maximum projection of 3 background brightfield images without BAL cells.

Statistics:

Statistical analysis was performed in Minitab v.17.1.0 (Minitab Inc., State College, PA, USA), except for neutrophils where SAS®, version 9.3 for the Windows platform was used. All data are presented as mean ± standard error of the mean (Mean±SEM).
Statistical analysis of DNA damage (%DNA) was performed on data normalized to the mean %DNA of PBS-exposed A549 cells on slides included in each electrophoresis.

Sex:

female

Genotoxicity:

negative

Toxicity:

not specified

Vehicle controls validity:

valid

Negative controls validity:

not examined

Positive controls validity:

other: Significant increases in DNA strand breaks in BAL at day 3 and in liver at day 90.

Table 4. DNA strand breaks. Level of DNA damage (Mean ± SEM) in BAL, lung and liver assessed with the comet assay (% DNA) 1, 3, 28 and 90 days post exposure to VC, GO, rGO or P90 (n = 7-8).

		Day 1	Day 3	Day 28	Day 90
	Dose	BAL
VC	0	0.44 ± 0.06	0.60 ± 0.04	1.19 ± 0.17	1.00 ± 0.11
GO	18	0.61 ± 0.03	0.98 ± 0.04*	1.87 ± 0.15*	1.18 ± 0.13
	54	0.54 ± 0.03	0.68 ± 0.03	-	-
	162	0.55 ± 0.07	0.51 ±0.06	-	-
P90	162	0.39 ± 0.02	0.75 ± 0.02**	0.87 ± 0.06	0.81 ± 0.05
		Lung
VC	0	0.54 ± 0.02	0.97 ± 0.08	1.47 ± 0.12	2.05 ± 0.16
GO	18	0.46 ± 0.04	0.99 ± 0.07	1.67 ± 0.18	1.64 ± 0.13
	54	0.42 ± 0.03	1.09 ± 0.06	-	-
	162	0.43 ± 0.04	1.21 ± 0.14	-	-
P90	162	0.78 ± 0.12	1.04 ± 0.07	1.58 ± 0.08	2.15 ± 0.19
		Liver
VC	0	1.75 ± 0.12	1.56 ± 0.21	1.04 ± 0.10	0.66 ± 0.03
GO	18	1.71 ± 0.30	1.62 ± 0.13	0.87 ± 0.09	0.72 ± 0.03
	54	2.06 ± 0.20	1.92 ± 0.26	-	-
	162	1.56 ± 0.22	1.65 ± 0.15	-	-
P90	162	1.43 ± 0.19	1.62 ± 0.19	0.80 ± 0.10	0.91 ± 0.04**

Data were normalized to the mean level of %DNA of H2O2-exposed A549 cells included on each slide during each electrophoresis. All samples from BAL, lung and liver were divided onto different electrophoresis and further divided according to time point post exposure to minimize day-to-day variation.

*, **, ***: statistically significantly different from corresponding VC at level p < 0.05, p < 0.01, p < 0.001, respectively.

Conclusions:

The test item graphene oxide did not show a consistent pattern of a genotoxic potential in this in vivo Comet assay in mice.

Executive summary:

The genotoxic potential of the test item graphene oxide was examined in an in vivo Comet assay in mice. The study conduct was generally comparable to OECD TG 489, however sampling was at 1, 3, 28 and 90 days after a single intratracheal exposure, and Carbon Black Printex90 was used as positive control.

Mice were exposed to 18, 54 and 162 µg graphene oxide/animal. DNA strand breaks were examined in bronchoalveolar lavage cells, lung and liver cells. At 1 day post exposure, no significant increase was observed for any concentration of graphene oxide in BAL, lung or liver. At day 3 post exposure, no statistically significant increase was observed in lung and liver, while the result obtained for 18 µg/mouse reached statistical significance. However, this effect was not concentration-dependent and was accompagnied by an acute inflammatory response in lung cells at day 3. DNA strand breaks induced by 18 µg graphene oxide/mice in BAL cells were also statistically significant (p < 0.05) at day 28 post-exposure, however due to the acute toxicity observed at day 3, only animals exposed to the lowest concentration could be further observed after day 3. Therefore, a concentration-dependence cannot be determined from these data at the later time points, and in addition, cytotoxicity was not determined. Therefore, the biological significance and relevance of this effect cannot be assessed from these data. No statistically significant increase in DNA strand breaks was observed in BAL cells after 90 days, and in lung and liver at 28 days and 90 days post exposure to graphene oxide.

Therefore, these data do not provide a clear evidence of a genotoxic potential of the test item graphene oxide in mice after intratracheal instillation.

Reference 2

Endpoint:

in vivo mammalian somatic cell study: cytogenicity / erythrocyte micronucleus

Type of information:

experimental study

Adequacy of study:

weight of evidence

Study period:

no details given

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

comparable to guideline study with acceptable restrictions

Qualifier:

equivalent or similar to guideline

Guideline:

OECD Guideline 474 (Mammalian Erythrocyte Micronucleus Test)

Deviations:

yes

Remarks:

four animals/sex/dose group, sampling time not specified

GLP compliance:

not specified

Type of assay:

mammalian erythrocyte micronucleus test

Specific details on test material used for the study:

GO was dispersed in deionized water at the concentration of 1 mg/mL, sonicated at 30 W for 3 min by a probe type sonifier (Automatic Ultrasonic Processor UH-500A, China) using a pulse function (pulse on 2.0 s, pulse off 2.0 s), and the sonication was repeated three times. Next, GO dispersion was centrifuged at 5000 rpm for 5 min to obtain GO aqueous dispersion.

Species:

mouse

Strain:

other: Kunming mice

Sex:

male/female

Details on test animals or test system and environmental conditions:

TEST ANIMALS
- Source: Wei Tong Li Hua experimental animal (Beijing, China)
- Age at study initiation: not specified
- Weight at study initiation: 25–30 g

ENVIRONMENTAL CONDITIONS
- Temperature (°C): 25 +/- 4°C,
- Humidity (%): 55 +/- 5%

Route of administration:

intravenous

Vehicle:

deionized water

Details on exposure:

intravenous

Duration of treatment / exposure:

once a day for five consecutive days

Frequency of treatment:

once a day for five consecutive days

Post exposure period:

not specified

Dose / conc.:

1 mg/kg bw/day

Dose / conc.:

2 mg/kg bw/day

Dose / conc.:

4 mg/kg bw/day

No. of animals per sex per dose:

4

Control animals:

yes

Positive control(s):

cyclophosphamide
- Route of administration: intravenous
- Doses / concentrations: 50 mg/kg bw

Tissues and cell types examined:

bone marrow

Details of tissue and slide preparation:

After humanitarian executions, femur bones of mice were excised and the bone marrow flushed into test tubes, and then the micronucleus assay was carried out following standard protocols, as recommended by Schmid (1975) and Krishna (2000).

Evaluation criteria:

For analysis of the micronucleated cells, 2000 polychromatic erythrocytes per animal were scored to determine the mutagenic property of GO. To detect possible cytotoxic effects, the poychromatic erythrocyte/normochromatic erythrocyte ratio in 200 erythrocytes per mouse was calculated.

Statistics:

Statistical differences were evaluated using the t-test and considered significance at P < 0.05.

Sex:

male/female

Genotoxicity:

positive

Toxicity:

no effects

Vehicle controls validity:

not applicable

Negative controls validity:

valid

Positive controls validity:

valid

Additional information on results:

RESULTS OF DEFINITIVE STUDY
- Induction of micronuclei (for Micronucleus assay): Intravenous injection of GO at 1, 2, 4 mg/kg for 5 consecutive days, corresponding respectively to the total doses of 5, 10 and 20 mg/kg, clearly induced the formation of micronucleated polychromatic erythrocytes (MNPCEs) in mice (see table below).
- Ratio of PCE/NCE (for Micronucleus assay): The calculated ratio of PCE/NCE in bone marrow preparations showed no statistically significant differences when compared to the negative control group, indicating that GO at 20 mg/kg didn’t directly kill erythrocytessee table below (see table below).
- Statistical evaluation: There were significant differences between GO treatment groups and the negative control group (1.16 +/- 0.64 MNPCEs per 1000 erythrocytes), further confirming the mutagenic effect of GO.

Table. Polychromatic erythrocytes with micronuclei (MNPCE) observed in bone marrow cells of female (F) and male (M) mice treated with GO and respective controls

Groups	Dose (mg/kg)	Number of MNPCE per animal^a								MNPCE (Mean±SD)	PCE/NCE^b (Mean±SD)
		F1	F2	F3	F4	M1	M2	M3	M4
Negative	0	1	2	1	1	2	1	0	1	1.1±0.64	1.88±0.25
GO1	5	3	3	1	3	2	3	2	2	2.4±0.74*	1.78±0.27
GO2	10	3	4	5	3	3	2	3	5	3.5±1.07**	1.67±0.24
GO4	20	6	8	5	9	7	6	7	8	7.0±1.31***	1.69±0.34
Positive	50	12	10	13	9	10	8	9	11	10.3±1.67***	1.72±0.18

^a Two thousand cells were analyzed per animal, for a total of 16,000 cells per group.

^b PCE/NCE: the ratio of polychromatic erythrocytes to normochromatic erythrocytes, and determined on total of 1000 erythrocytes counted.

Signifcantly different from negative control (*P < 0.05, **P < 0.01, and ***P < 0.001).

Conclusions:

This study indicates that GO induced the formation of MNPCE in mice in vivo and its potence at the dose of 20 mg/kg was equivalent to that of cyclophosphamid (50 mg/kg), which is a classic mutagene.

Executive summary:

The mutagenic potential of the test item Graphene oxide (GO) in mice was studied in the micronucleus test, similar to OECD test guideline 474.

Three group of Kunming mice (4 female/ 4 male/group) were administered intravenously with GO solution at 1, 2 and 4 mg/kg once a day for five consecutive days, corresponding respectively to the total doses of 5, 10 and 20 mg/kg. The other two group mice received physiological saline and cyclophosphamide (50 mg/kg) as the negative and positive controls, respectively.

After humanitarian executions, femur bones of mice were excised and the bone marrow flushed into test tubes, and then the micronucleus assay was carried out following standard protocols, as recommended by Schmid (1975) and Krishna (2000). For analysis of the micronucleated cells, 2000 polychromatic erythrocytes per animal were scored to determine the mutagenic property of GO. To detect possible cytotoxic effects, the poychromatic erythrocyte/normochromatic erythrocyte ratio in 200 erythrocytes per mouse was calculated.

Cyclophosphamide, a potent mutagen, induced formation of MNPCE at 50 mg/kg. The micronuclei were round or oval in shape and had a diameter about 1/20 to 1/5 that of PCEs, consistent with previous report (Gollapudi et al., 1995). Intravenous injection of GO at 1, 2, 4 mg/kg for 5 consecutive days, corresponding respectively to the total doses of 5, 10 and 20 mg/kg, clearly induced the formation of MNPCE in mice. As shown in Table 3 (see above), the frequencies of MNPCEs in GO treated mice exhibited dose-dependent property, and increased from 2.4 +/- 0.74 to 7.0 +/- 1.31 per 1000 erythrocytes with GO dose increasing from 5 to 20 mg/kg. There were significant differences between GO treatment groups and the negative control group (1.16 +/- 0.64 MNPCEs per 1000 erythrocytes), further confirming the mutagenic effect of GO. Moreover, the calculated ratio of PCE/NCE in bone marrow preparations showed no statistically significant differences when compared to the negative control group, indicating that GO at 20 mg/kg didn’t directly kill erythrocytes (see Table). Together, the results suggested that GO may not be a dangerous cytotoxin, but is a potential inducer of mutagenesis.

Justification for classification or non-classification

The investigation of the genotoxic potential of graphene oxide has been inconclusive so far, however, some research studies on the biocompatibility and genotoxicity of graphene oxide materials indicate a potential to interfere with the integrity of DNA.

Among the identified in vitro genotoxicity studies, both positive and negative results were reported from Comet assays, and one study indicates the formation of micronuclei. However, recent studies demonstrate that the genotoxic potential of carbon-based nanomaterials (NM) crucially depends on the NMs' characteristics and how they behave during exposure. More specifically, the potential to induce DNA damage and consequently the formation of micronuclei was found to strongly correlate with the agglomeration size of NM (Luoro, 2018). The level of agglomeration/aggregation is influenced by tissue culture medium components and clearly differs between various experimental setups. However, this behaviour of the test materials is only rarely specified in the above-mentioned studies. Therefore, the relevance of individual in vitro studies, indicating either negative or positive results, for the registered test item is not clear and further studies are necessary to elucidate interactions of NM with the respective test systems.

In the two available in vivo studies, the chosen administration routes, intravenous injection and intratracheal instillation of high doses cannot be considered as relevant to human exposure. Interestingly, according to available data both administration routes may lead to a non-realistic overload in the lung and subsequently trigger local inflammatory responses.

The administration via intravenous injection may have led to an increased amount of test item reaching the systemic circulation and, in turn, the target tissue (i.e., the bone marrow) which would not reflect the potentially limited systemic uptake (if any) under real exposure conditions. While intravenous injection led to the formation of micronuclei in erythrocytes in mice (Liu et al., 2013), an unclear pattern of DNA damage was found in the lung under strongly inflammatory conditions (Bengtson et al. 2017). Furthermore, as the bioavailability of NM in the lung after inhalation exposure clearly depends on the particle size and the respiratory fraction, also the relevance of results obtained after intratracheal instillation for the registered substance is unclear.

In addition, beside the particle form (lateral size and number of graphene layers), the effects of graphene oxide also seem to depend on the physico-chemical properties influenced e.g. by the number and nature of oxygen groups and their functionalisation. Therefore, the specific characterisation of the utilised test materials is crucial for the assessment of its genotoxic properties, and the available data on the tested graphene oxide material is not necessarily relevant for the registered substance.

Therefore, the available studies do not yet allow a conclusion on classification with respect to genetic toxicity of graphene oxide, but specific tests on the genetic toxicity and the toxicokinetic behaviour of the registered graphene oxide are needed.