Introduction

In recent years, nanotechnology has become an important part of people’s daily lives and environments. Nanoparticles (NPs) with sizes between 1 and 100 nm are widely used in household products, building materials, foods and cosmetics, sunscreens, water purification, toys, sports equipment and medicines1. For example, coatings with silica dioxide nanoparticles (SiO2-NPs) and titanium dioxide nanoparticles (TiO2-NPs) are used to create self-cleaning, water-repelling, and heat-resistant surfaces; graphene NPs and carbon nanotubes are widely used as composite materials to add strength with minimal weight to sporting equipment, such as tennis rackets, golf balls and clubs, and bicycles; silver nanoparticles (Ag-NPs) and copper nanoparticles (Cu-NPs), due to their strong antimicrobial properties, are widely used in clothing, linens, rugs, and towels; and platinum, palladium, rhodium, and cerium oxide NPs are used in automobile catalytic converters to make vehicle exhaust less harmful1.

Because of the widespread use of NPs, their biological effects on organisms and their toxicities have been studied extensively in recent years2,3,4,5,6. Among their effects, posttranslational modifications (PTMs) of histones such as acetylation, methylation, and phosphorylation have recently attracted research attention because they have been linked to a variety of biological processes and disease states7,8. PTMs of histones not only can alter the biophysical properties of the interactions between histones and DNA, but also can affect the binding of a variety of effector proteins that mediate further biological events. As a result, they can either overcome the inherently repressive nature of chromatin to establish a transcriptionally permissive state or participate in the silencing of a given gene8. One of the most consistent alterations induced by exposure to a broad range of NPs is increased phosphorylation of histone H2AX at serine 139 (γ-H2AX). It is well documented that γ-H2AX is generated as a response to various types of DNA lesions and is one of the earliest DNA damage responses9,10. In addition to the induction of γ-H2AX, exposure to NPs resulted in other types of histone modifications. For example, SiO2-NPs increased the acetylation of histones H3 and H4 in mouse Bhas 42 cells11. Exposure to gold nanoparticles reduced the extent of histone H3 lysine 27 (H3K27) trimethylation (Tri-Me-H3K27)12 and increased H3K9/H3K14 acetylation (Ac-H3K9/Ac-H3K14) and histone H3 serine 10 phosphorylation (p-H3S10)13. Recently, zinc oxide nanoparticles (ZnO-NPs) were also shown to enhance demethylation of H3K914 and decrease the level of histone H3K27 trimethylation15. Blanco et al.16 reported that treatment of Human lung adenocarcinoma (A549) cells with Ag-NPs induced dramatic deacetylation of histone H3. We also demonstrated that treatment of A549 cells, human breast adenocarcinoma (MCF-7) cells, and Human skin keratinocytes (HaCaT) cells with Ag-NPs resulted in increased p-H3S10, a modification associated with mitotic chromatin condensation17,18. However, the effect of NPs on the epigenome remains a developing area in the field of nanotoxicology research, and there are limited and inconclusive data as well as many unanswered questions12,19.

In this study, we investigated the histone H3 modifications induced by ten kinds of NPs under the same conditions. We observed that Al2O3-NP, CuO-NP or ZnO-NP induced the formation of γ-H2AX. In addition, among processes observed for tested nanoparticles, Ac-H3K9, Ac-H3K14 and Ac-H3 (global) production by CuO-NPs and ZnO-NPs were the most significant. We found that upregulation of Tri-Me-H3K27 is highly correlated with CuO-NP- and ZnO-NP-induced histone acetylation. Furthermore, we proved that CuO-NP- or ZnO-NP-induced histone H3 modification may occur via release of ions from nanoparticles inside cells.

Materials and methods

Materials

Al2O3 (cat. no. 544833; Size: <50 nm), Co3O4 (cat. no. 637025; Size: <50 nm), Cr2O3 (cat. no. 634239; Size: <100 nm), CuO (cat. no. 544868; Size: <50 nm), Fe3O4 (cat. no. 637106; Size: <50 nm), NiO (cat. no. 637130; Size: <50 nm), Sb2O3 (cat. no. 637173; Size: <250 nm), SiO2 (cat. no. S5130; Size: <7 nm), TiO2 (cat. no. 637254; Size: <25 nm), and ZnO (cat. no. 544906; Size: <100 nm) were purchased from Sigma-Aldrich, LLC. (St. Louis, MO, USA). The size distributions of NPs in suspension are shown in Figure S1. The average diameters for Al2O3, Co3O4, Cr2O3, CuO, Fe3O4, NiO, Sb2O3, SiO2, TiO2 and ZnO were 549.50, 184.88, 788.07, 233.00, 193.06, 295.88, 673.84, 1389.63, 402.46 and 402.34 nm, respectively.

Pretreatment of NPs

These NPs were pretreated as described previously18. In brief, NPs were suspended in Dulbecco’s modified Eagle’s medium (DMEM; Thermo Fisher Scientific Inc., Waltham, MA, USA) containing 0.5% (v/v) foetal bovine serum (FBS; Life Technologies Inc., Grand Island, NY, USA) at a final concentration of 1 mg/mL and were immediately sonicated in a bath-type sonicator (Bioruptor; Cosmo Bio Co., Ltd., Tokyo, Japan) for 1 min before being applied to cells.

Cell culture

Human lung adenocarcinoma cells (A549; provided by Shanghai Huiying Biological Technology Co., Ltd., Shanghai, China) were cultured in DMEM supplemented with 10% FBS and 100 U/mL penicillin-streptomycin at 37 °C in a humidified atmosphere containing 5% CO2. Adherent cell cultures were used in experiments during the logarithmic growth phase.

Treatment of cells with NPs

When the cells reached 70 ~ 80% confluence, the medium was changed to DMEM supplemented with 0.5% FBS. After being cultured for 24 h, the cells (approximately 1 × 105 cells / 35 mm dish) were treated with NPs (~ 1 mg/mL) for ~ 36 h. To remove the released metal ions from the medium, ethylene glycol-bis-(β-aminoethyl ether)-N, N,N′,N′-tetraacetic acid (EGTA) (5 mM) was added 0.5 h before the treatment with NPs.

Cell survival after treatment with NPs

After treatment with NPs, the cells were washed with PBS, trypsinized, and suspended in DMEM. Cells were mixed with Trypan blue solution (0.3%) (1:1), and at least 400 cells were counted under a microscope. Dead cells showed a distinctive blue colour.

Western blotting analysis

Cells treated with NPs were lysed in lysis buffer, and western blotting was performed as previously described18. In brief, A549 cells treated with NPs were harvested, and protein was extracted. The amount of protein was estimated using a DC protein assay kit (Bio-Rad, Hercules, CA). Samples containing ~ 20 µg protein were separated on 12.5% polyacrylamide gels [sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE)] and blotted onto polyvinylidene fluoride transfer membranes. Membranes were stained with 0.25% Coomassie Brilliant Blue R250, and the images of histones were scanned for proof of equal loading. Primary antibodies against γ-H2AX (cat. no. 9718), Ac-H3K9 (cat. no. 9649), Ac-H3K14 (cat. no. 7627), Tri-Me-H3K4 (cat. no. 9751), Di-Me-H3K9 (cat. no. 4658), and Tri-Me-H3K27 (cat. no. 9733) (Cell Signaling Technology, Inc., MA, USA) (1:1000), and Ac-H3 (global) (cat. no. 06-599) (Millipore Co., Billerica, MA, USA) (1:1000) were used, followed by secondary antibodies conjugated with horseradish peroxidase (Jackson ImmunoResearch Laboratories, West Grove, PA) (1:1000). Histone bands were visualized with an enhanced chemiluminescence detection kit (Pierce Biotechnology, Rockford, IL, USA), and the intensity of each band corresponding to histones was determined using ImageJ (version 1.38).

Evaluating the uptake of NPs

Cells treated with NPs were washed three times with PBS to remove free NPs. The cells were resuspended in DMEM, and the number of particles taken up was analysed with FCM (CytoFLEX; Beckman Coulter, Inc., Indianapolis, IN, USA). A profile of the sample was obtained by examining both forward-scattered (FS) and side-scattered (SS) light. As each cell intercepts the path of the laser beam, the light that passes around the cell is measured as FS light, indicating the cell size. The light scattered at a 90° angle to the axis of the laser beam is measured as SS light and is related to intracellular density. Changes in cellular side scatter after treatment with NPs can indicate the uptake potential of the particles20.

Measurement of the concentration of ions released by NPs

The concentrations of ions released from the surfaces of the NPs were analysed using inductively coupled plasma atomic-emission spectrometry (ICP-AES; Varian 730ES) as described previously21. Water suspensions of NPs (500 µL, 1 mg/mL) were added to a dialysis tube (Mini Dialysis kit, 8 kDa cut off; GE Healthcare, Piscataway, NJ, USA) and dialyzed against 10 mL of water for 24 h at room temperature. Ions of NPs were used to make a standard curve, and the ion concentration in the initial NP suspension was back-calculated.

Statistics

All experiments were repeated three times. Data are presented as the mean ± standard deviation (S.D.). The statistical analysis of all the experimental data was conducted using Origin 8.0 software. The significant differences (P < 0.05) among the measured values were evaluated by an ANOVA procedure.

Results

Cytotoxicity of treatment with NPs in A549 cells

The cytotoxic effect of treatment with NPs on A549 cells was examined. Cells were treated with ten kinds of NPs (~ 1 mg/mL), and the effect of culture time on cell survival was monitored (Fig. 1). Treatment with 1 mg/mL Al2O3-NPs, Co3O4-NPs, Fe3O4-NPs or Sb2O3-NPs caused a slight decrease in survival rates, whereas treatment with CuO-NPs or ZnO-NPs induced a marked decrease in A549 cell survival. Other NPs caused different degrees of reduction in the survival rate of A549 cells, which decreased in the order of SiO2-NPs > Cr2O3-NPs > NiO-NPs and TiO2-NPs. Next, we selected nonlethal concentrations (0.01 mg/mL) for follow-up experiments.

Fig. 1
figure 1

Cytotoxicity after treatment with NPs. A549 cells were treated with NPs (various concentrations) and cultured for 24 h. Cell survival was determined by a trypan blue exclusion assay.

Full size image

Changes in the γ-H2AX of A549 cells treated with NPs

Treatment with Al2O3-NPs, CuO-NPs or ZnO-NPs generated γ-H2AX in A549 cells in a time-dependent manner (Fig. 2 and Figure S2). The γ-H2AX changes caused by CuO-NPs and ZnO-NPs were consistent with the survival rate results. A slight γ-H2AX production was observed in almost all untreated cells after 24–36 h of culture. This may be because the apoptosis-related γ-H2AX caused by long-term culture. These results showed that the degree of DNA damage caused by Al2O3-NPs, CuO-NPs and ZnO-NPs was much higher than those caused by the other tested NPs, such as TiO2-NPs.

Fig. 2
figure 2

γ-H2AX after treatment with NPs. (A) γ-H2AX after treatment with NPs. A549 cells were treated with NPs (0.01 mg/mL) for various time points. H3 (CBB staining) was used as a standard for equal loading of proteins for SDS-PAGE; (B) γ-H2AX after treatment with Al2O3-NPs, CuO-NPs, ZnO-NPs or TiO2-NPs. γ-H2AX was determined using western blotting, in which the intensity of each band was extracted using ImageJ version 1.38. The extent of γ-H2AX in cells treated with Al2O3-NPs, CuO-NPs, ZnO-NPs or TiO2-NPs versus untreated control cells was calculated. All the above blots had already been cropped in order to improve the clarity and conciseness of the presentation. The original blots are presented in Supplementary Fig. 7.

Full size image

Changes in the histone H3 modification of A549 cells treated with NPs

Next, we investigated the effects of NP treatment on histone acetylation in A549 cells. Among them, Al2O3-NPs caused a reduction in Ac-H3K9 and Ac-H3K14, and SiO2-NPs caused an increase in Ac-H3 (global) (Fig. 3A and Figure S3). In contrast, CuO-NPs and ZnO-NPs caused an increase in histone acetylation (Ac-H3K9, Ac-H3K14 and Ac-H3 (global)) at all test time points (Fig. 3B, C). As with the results for γ-H2AX, the histone H3 acetylation changes caused by CuO-NPs and ZnO-NPs were also the most significant. Furthermore, the other six kinds of NPs did not cause significant changes in histone acetylation (Fig. 3A and Figure S3).

Fig. 3
figure 3

Histone H3 acetylation after treatment with NPs. (A) Ac-H3K9, Ac-H3K14 and Ac-H3 (global) after treatment with NPs. A549 cells were treated with NPs (0.01 mg/mL) for various time points. H3 (CBB staining) was used as a standard for the equal loading of proteins for SDS-PAGE; (B) Ac-H3K9, Ac-H3K14 and Ac-H3 (global) after treatment with CuO-NPs; (C) Ac-H3K9, Ac-H3K14 and Ac-H3 (global) after treatment with ZnO-NPs. Ac-H3K9, Ac-H3K14 and Ac-H3 (global) were detected using western blotting, in which the intensity of each band was extracted using ImageJ version 1.38. The extent of Ac-H3K9, Ac-H3K14 or Ac-H3 (global) in cells treated with CuO-NPs, ZnO-NPs or TiO2-NPs versus untreated control cells was calculated. All the above blots had already been cropped in order to improve the clarity and conciseness of the presentation. The original blots are presented in Supplementary Fig. 7.

Full size image

In addition, we also observed the changes in histone methylation (Tri-Me-H3K4, Di-Me-H3K9 and Tri-Me-H3K27) after treatment with NPs (Fig. 4). The Al2O3-NPs induced increases in Tri-Me-H3K27. After treatment with Co3O4-NPs, a decrease in Tri-Me-H3K4 and an increase in Di-Me-H3K9 were observed (Fig. 4A and Figure S4). Consistent with the results of histone acetylation, the upregulation of Tri-Me-H3K27 caused by CuO-NPs and ZnO-NPs was the most significant. The difference is that CuO-NPs induced a slight increase in Tri-Me-H3K4 production, while ZnO-NPs induced the opposite effect. In addition, ZnO-NPs caused significant upregulation of Di-Me-H3K9, while CuO-NPs did not cause a significant change (Fig. 4B, C). Furthermore, the other six kinds of NPs also did not cause significant changes in histone methylation (Fig. 4A and Figure S4). These results showed that the histone H3 modifications caused by the CuO-NPs and ZnO-NPs were also much stronger than those caused by other tested NPs.

Fig. 4
figure 4

Histone H3 methylation after treatment with NPs. (A) Tri-Me-H3K4, Di-Me-H3K9 and Tri-Me-H3K27 after treatment with NPs. A549 cells were treated with NPs (0.01 mg/mL) for various time points. H3 (CBB staining) was used as a standard for the equal loading of proteins for SDS-PAGE; (B) Tri-Me-H3K4, Di-Me-H3K9 and Tri-Me-H3K27 after treatment with CuO-NPs; (C) Tri-Me-H3K4, Di-Me-H3K9 and Tri-Me-H3K27 after treatment with ZnO-NPs. Tri-Me-H3K4, Di-Me-H3K9 and Tri-Me-H3K27 were determined with western blotting, in which the intensity of each band was extracted using ImageJ version 1.38. The extent of Tri-Me-H3K4, Di-Me-H3K9 or Tri-Me-H3K27 in cells treated with CuO-NPs, ZnO-NPs or TiO2-NPs versus untreated control cells was calculated. All the above blots had already been cropped in order to improve the clarity and conciseness of the presentation. The original blots are presented in Supplementary Fig. 7.

Full size image

Uptake of NPs and histone H3 modification

To confirm the relationship between NPs uptakes and histone H3 modification, the levels of intracellular uptake for ten kinds of NPs were measured using FCM. We have previously reported that SS light in an FCM analysis was useful for determining the intercellular uptake of metal NPs20. SS light indicates the intercellular density, which increases with the uptake of NPs. Figure 5A shows the change in SS intensity after treatment with 0.1 mg/mL of ten kinds of NPs for 1 h. The accumulation of NPs in cells was obviously different and most significant for TiO2-NPs. The SS intensity increased in a dose-dependent manner after treatment with CuO-NPs, ZnO-NPs or TiO2-NPs (Fig. 5B). Furthermore, in FS histograms, the FS reflecting cell size did not change (Figure S5). When a large amount of NPs was incorporated into the cells, FS decreased due to reflection of light away from the forward scatter detector22. These results indirectly indicated that NPs actually accumulated intracellularly, but not on the surface.

Fig. 5
figure 5

Uptake of NPs and histone H3 modification. (A) A549 cells were treated with NPs (0.1 mg/mL) for 1 h. SS was analysed using FCM; (B) Dose-dependent changes in SS intensity after treatment with CuO-NPs, ZnO-NPs or TiO2-NPs; (C) Correlation between uptake of the NPs and histone H3 modifications. γ-H2AX, Ac-H13K9, Ac-H3K14 and Tri-Me-H3K27 were detected using western blotting after 24 h of treatment with CuO-NPs, ZnO-NPs or TiO2-NPs (0.01 mg/mL), and the intensities of the bands were extracted using ImageJ version 1.38. The extent of γ-H2AX, Ac-H13K9, Ac-H3K14 and Tri-Me-H3K27 in cells treated with CuO-NPs, ZnO-NPs or TiO2-NPs versus untreated control cells was calculated. The mean SS (the mean for replicate treatments of cells for 1 h) (ratio: treated/untreated) was analysed with FlowJo version 10.7.2.

Full size image

The changes in γ-H2AX, Ac-H3K9, Ac-H3K14 and Tri-Me-H3K27 were greater for ZnO-NPs and CuO-NPs than for TiO2-NPs, although the cellular uptake of ZnO-NPs and CuO-NPs was lower than that of TiO2-NPs (Fig. 5C). Fe3O4-NPs, Sb2O3-NPs, Cr2O3-NPs and TiO2-NPs induced increases in SS intensity, indicating incorporation into the cells, however, the change in histone H3 modification was small and less than those of ZnO-NPs and CuO-NPs (Figs. 3, 4 and 5). These results showed that accumulation of NPs in cells is not directly related to changes in histone H3 modification.

Release of ions from NPs and histone H3 modification

We previously reported that intracellular Ag-NPs release Ag ions that alter actin filament dynamics, leading to activation of Aurora kinases and formation of p-H3S10 through a mechanism clearly different from that occurring during mitosis17,18. We conducted experiments to clarify the mechanism by which NPs induce histone H3 modifications and to determine whether histone H3 modifications arise from ion release from intracellular NPs.

To clarify the role of ions released from NPs in histone H3 modifications, the concentration of ions released was determined using ICP-AES. The amounts of ions released from CuO-NPs, ZnO-NPs, Cr2O3-NPs, Sb2O3-NPs, Fe3O4-NPs and TiO2-NPs were 4.37, 15.75, 1.28, 1.27, 0.44 and 0.87 ppm, respectively (Fig. 6A). Furthermore, to clarify the relationship between the ions released from NPs and histone H3 modification, the doses of ions and the intensities of γ-H2AX, Ac-H13K9, Ac-H3K14 and Tri-Me-H3K27, which were quantified from the western blot images, were plotted (Figs. 2, 3 and 4). The extent to which ions were released from CuO-NPs and ZnO-NPs corresponded to the extent of histone H3 modification. Cr2O3-NPs, Fe3O4-NPs, Sb2O3-NPs and TiO2-NPs, which released small amounts of ions, appeared close to the origin (Fig. 6B). These results show that the amount of ions released from NPs is highly correlated with histone H3 modification.

Fig. 6
figure 6

Potential mechanisms for histone modification after treatment with NPs. (A) Ions released from NPs were analysed using ICP-AES. Water suspensions of NPs (500 µL, 1 mg/mL) were dialyzed against 10 mL of water for 24 h at room temperature; (B) Correlation between ions released from NPs and histone H3 modifications. γ-H2AX, Ac-H13K9, Ac-H3K14 and Tri-Me-H3K27 were detected using western blotting after 24 h of treatment with CuO-NPs, ZnO-NPs or TiO2-NPs (0.01 mg/mL), and the intensities of the bands were extracted using ImageJ version 1.38. The extent of γ-H2AX, Ac-H13K9, Ac-H3K14 and Tri-Me-H3K27 in cells treated with CuO-NPs, ZnO-NPs or TiO2-NPs versus untreated control cells was calculated; (C) Histone H3 modification after treatment with CuO-NPs or ZnO-NPs in the presence of EGTA (5 mM). A549 cells were treated with EGTA for 0.5 h and treated with CuO-NPs or ZnO-NPs (0.01 mg/mL) for various time points. H3 (CBB staining) was used as a standard for the equal loading of proteins for SDS-PAGE.

Full size image

Furthermore, A549 cells were pretreated with EGTA (5 mM) for 0.5 h, which can chelate metal ions. Consistent with our speculation, EGTA completely inhibited the histone H3 modification changes induced by CuO-NPs or ZnO-NPs (Fig. 6C). Furthermore, EGTA did not affect the uptake of CuO-NPs or ZnO-NPs (Figure S6). These results suggest that the release of ions from NPs inside cells may be an important factor for changes in histone H3 modification caused by NPs.

Cellular uptake of NPs may differ from cell to cell and may involve diffusion, phagocytosis, and endocytosis23. We previously proved that Ag ions released continuously from AgNPs and incorporated into cells may be important for p-H3S10 formation24. In this paper, intracellular accumulations of TiO2-NPs, Cr2O3-NPs and Sb2O3-NPs that did not cause histone H3 modification changes were significantly higher than those of CuO-NPs and ZnO-NPs (Fig. 5A, B). However, the extent to which ions were released from CuO-NPs and ZnO-NPs corresponded to the extent of histone H3 modification (Fig. 6A, B). In addition, we have preliminary data suggesting that Ag-NPs-induce changes in other H3 modifications in a time-dependent manner, and the changes in H3 modification are highly related to the amount of silver ions released. These results indicate that the concentrations of ions, but not the accumulation of NPs in cells, were key to the CuO-NP- and ZnO-NP-induced changes in histone H3 modification.

Discussion

We previously reported that treatment of A549 cells and MCF-7 cells with Ag-NPs for 4 h induced the formation of γ-H2AX25. Several reports have shown that exposure of mammalian cells to NPs results in γ-H2AX induction. Treatment with TiO2-NPs induced the formation of γ-H2AX in A549 cells, human dermal fibroblasts and human skin fibroblasts26,27,28,29. Increased γ-H2AX has also been shown in human cells after exposure to Au-NPs13 and other NPs, including CuO-NPs, CeO2-NPs, As2O3-NPs, ZnO-NPs, SiO2-NPs, and Fe3O4-NPs30,31,32,33,34,35. However, in this study, we observed that only Al2O3-NPs, CuO-NPs and ZnO-NPs caused generation of γ-H2AX. The other measured NPs did not show a clear induction trend (Fig. 2). Many studies have shown that DNA damage caused by NPs is highly correlated with the generation of reactive oxygen species (ROS). Although there are some exceptions, it is accepted that when DNA double-strand breaks (DSBs) are induced, γ-H2AX foci are quickly formed, and the γ-H2AX signal reflects the number of DSBs in the cell9,10. As DSBs can also be formed by indirect repair of DNA damage and collision of replication forks at other sites of DNA damage, including oxidative bases, DNA adducts, SSBs, and DNA crosslinks, γ-H2AX is now considered as a DNA damage marker10. DSBs are among the most deleterious types of DNA damage because they can potentially lead to a loss or alteration of genetic material36. Al2O3-NPs, CuO-NPs or ZnO-NPs caused significant γ-H2AX production, and phosphorylation was induced immediately after the treatment and lasted up to 36 h. Our data showed that Al2O3-NPs, CuO-NPs and ZnO-NPs may have genetic toxicity when their DNA damage is not effectively repaired. Furthermore, after treatment with CuO-NP and ZnO-NP for 36 h, a decrease in γ-H2AX was clearly observed. This is most likely due to cell death rather than γ-H2AX repair.

In previous reports, Ac-H3K9 and Tri-Me-H3K4 were commonly enriched at transcription initiation to regulate gene activation, and dynamic changes in the Ac-H3K9 signal showed a very high correlation with that of the Tri-Me-H3K4 signal37,38. In this study, we observed that CuO-NPs or ZnO-NPs induced upregulation of the Ac-H3K9 signal. However, the Tri-Me-H3K4 signal did not increase but was significantly suppressed in the ZnO-NP exposure group. Interestingly, we discovered that dynamic changes in the Tri-Me-H3K27 signal were highly correlated with the γ-H2AX and Ac-H3K9 signals (Figs. 3 and 4). Tri-Me-H3K27 is usually associated with repressive chromatin but has also been observed to accumulate within 5 min at damage sites generated by lasers, as well as at restriction enzyme-induced DSBs and H2O2-induced damage foci39,40,41. These reports are consistent with our findings. We believe that CuO-NP- or ZnO-NP-induced Tri-Me-H3K27 may play an important role in DNA damage repair (DDR).

On the other hand, γ-H2AX also plays an important role in enhancing the repair of DNA lesions by facilitating opening of the chromatin structure and forming a platform for accumulation of DDR factors10. γ-H2AX was shown to precede and initiate the accumulation of repair factors and checkpoint proteins, including Mre11-Rad50-Nbs1 (MRN), mediator of DNA damage checkpoint protein 1 (MDC1), breast cancer 1 (BRCA1), p53-binding protein 1 (53BP1), and E3 ubiquitin-protein ligase 68 and 168 (RNF68 and RNF168)42,43,44,45,46. Moreover, the lysine methyltransferase EZH2, a Di- or Tri-Me-H3K27-specific lysine methyltransferase found in the PRC2 complex, is also recruited to damage sites and catalyses H3K2747. Upon homing endonuclease I-SceI-induced DSBs or oxidative damage induced by H2O2, EZH2 accumulates at promoters of actively transcribed genes together with other silencing factors, including SIRT1, DNMT1 and DNMT3B, which suggests that EZH2 may function in repressing specific genes in response to DNA damage40,41,48. EZH2-mediated Tri-Me-H3K27 may function as an important mechanism for coordinating transcription both at damage sites and in genes. This evidence suggests that different histone H3 modifications, such as phosphorylation, methylation, and acetylation, play important roles in different aspects of DDRs. Despite their distinct roles, different histone H3 modifications also work cooperatively to generate proper and efficient cellular responses to DNA damage.

Histone modification is a dynamic process that is tightly controlled by the balance between “writers” and “erasers”49. “Writers”, including histone phosphorylases, acetyltransferases, and methyltransferases, introduce a particular chemical histone modification, whereas “erasers”, including histone phosphatases, deacetylases, and demethylases, are responsible for removal of chemical modifications. Similar exposure-related global histone hypoacetylation was reported after exposure of human breast cancer cells to cadmium telluride quantum dots50. Recently, Zhang et al.15 reported that treatment of human bladder cancer T24 cells with 0.01 mg/mL ZnO-NPs for 48 h decreased the level of histone H3K27 trimethylation globally and at the RUNX3 gene promoter. Mechanistically, these changes were attributed to downregulation of EZH2 expression. Additionally, increased expression of the G9a and GLP histone methyltransferase genes and downregulation of the GCN5, P300, and CBP histone acetyltransferase genes were observed in HaCaT cells treated with ZnO-NPs14. Elucidating the enzymes related to CuO-NP- and ZnO-NP-induced histone H3 modification changes will be our main research direction in the future. In addition, other parameters such as exposure concentration, particle size, surface charge and shape may affect NP induced histone modifications. However, these parameters of different NPs were not compared in the study. Therefore, further researches will focus on investigating the impact of the physicochemical properties on the toxicological effects of nanoparticles.

In this study, we found that Al2O3-NPs-, CuO-NPs- or ZnO-NPs-induced the formation of γ-H2AX. In addition, compared with other tested nanoparticles, Ac-H3K9, Ac-H3K14 and Ac-H3 (global) caused by CuO-NPs and ZnO-NPs were most significant. We found that the upregulation of Tri-Me-H3K27 is highly correlated with CuO-NP- and ZnO-NP-induced histone acetylation. Furthermore, we proved that CuO-NP- or ZnO-NP-induced histone H3 modification may occur via release of ions from nanoparticles inside cells. In previous reports, we found that p-H3S10 can be used to evaluate the toxicity of Ag-NPs and Ag ion release in combination with the detection of SS from FCM. Accumulated evidence also demonstrates that epigenetic alterations may be used to detect toxicity caused by NPs and, more importantly, to predict the toxicity of NPs in preclinical assessments. We would like to clarify the molecular mechanism in future work to further understand the preclinical significance.