Effect of Long-Term Low-Dose Arsenic Exposure on DNA Methylation and Gene Expression in Human Liver Cells

Millions of people around the world are exposed to elevated levels of arsenic through food or drinking water. Epidemiological studies have linked chronic arsenic exposure to an increased risk of several cancers, cardiovascular disease, central nervous system neuropathies, and genotoxic as well as immunotoxic effects. In addition to the induction of oxidative stress and inhibition of DNA repair processes, epigenetic effects, including altered DNA methylation patterns resulting in aberrant gene expression, may contribute to carcinogenicity. However, the underlying mechanisms by which chronic micromolar concentrations of arsenite affect the methylation status of DNA are not fully understood. In this study, human HepG2 hepatocarcinoma cells were treated with 0.5–10 μM sodium arsenite for 24 h, 10, or 20 days. During these periods, the effects on global DNA methylation, cell cycle phase distribution, and gene expression were investigated. While no impact on DNA methylation was seen after short-term exposure, global hypomethylation was observed at both long-term exposure periods, with concomitant induction of the DNA methyltransferase genes DNMT1 and DNMT3B, while DNMT3A was slightly down-regulated. Pronounced time- and concentration-dependent effects were also seen in the case of genes involved in DNA damage response and repair, inflammation, oxidative stress response, and metal homeostasis. These results suggest that chronic low-dose arsenite exposure can lead to global hypomethylation. As an underlying mechanism, the consistent down-regulation of DNA methyltransferase genes could be excluded; alternatively, interactions at the protein level could play an important role.


Introduction
It is estimated that more than 100 million people worldwide are exposed to inorganic arsenic through drinking water or food, and there is high concern for adverse health effects [1,2].On the other hand, arsenic trioxide (ATO) has been established as a therapeutic agent for acute promyelocytic leukemia (APL) [3].From a toxicological point of view, epidemiological studies suggest that chronic low-dose exposure to inorganic arsenic in humans is significantly associated with an increased risk of lung, skin, and bladder cancer, central nervous system neuropathies, and immunotoxicity [1,2].
Arsenic and its inorganic compounds have been classified as carcinogenic to humans (Group 1) by both the German MAK Commission and the International Agency for Research on Cancer [4,5].Although inorganic arsenic is a known carcinogen, it is not a strong mutagen.However, it can act as a co-mutagen and interfere with the cellular response to DNA-damaging exposures, such as UV radiation and benzo[a]pyrene [4,[6][7][8].Current evidence suggests that the induction of oxidative stress and interactions with DNA repair processes play crucial roles [1].Furthermore, various studies have observed widespread disruption of transcriptional activity following arsenic exposure, indicating that the expression of genes, for example, those involved in oxidative stress and the DNA damage response, is affected by arsenic exposure [7,9,10].Arsenic also appears to be involved in the disturbance of DNA methylation reactions, which could explain the extensive changes in gene expression [11].
The methylation of cytosine at so-called CpG islands is a fundamental determinant of chromatin structure and particularly important for regulating transcriptional activity.Thus, long-term transcriptional silencing generally correlates with the extent of CpG island methylation near gene promoters.The global disruption of DNA methylation is lethal for mammals, while gene-specific DNA hypermethylation has been associated with the loss of expression of tumor suppressor genes.On the other hand, genome-wide hypomethylation has been found to accompany oncogenic transformation, leading to genetic instability and spontaneous mutations.In humans, DNA methylation is catalyzed by three methyltransferases (DNMTs), namely, DNMT1, DNMT3A, and DNMT3B.DNMT1 appears to be responsible for maintaining methylation patterns during replication, while DNMT3A and DNMT3B play crucial roles in early embryogenesis and X-chromosome inactivation [12,13].
The methylation of cytosine to 5-methylcytosine occurs via the transfer of a methyl group from S-adenosylmethionine (SAM) to the C5 position of cytosine in CpG dinucleotides.However, SAM is also required during arsenite metabolism following exposure.This suggests that in the presence of excess arsenite, SAM, which is normally used for cellular methylation reactions, is used by arsenite methyltransferase (AS3MT) for arsenic methylation.This may lead to the depletion of SAM in the cell and the accumulation of S-adenosylhomocysteine (SAH), which, in turn, may suppress SAM-dependent methyltransferase.Since arsenic metabolism mediated by AS3MT primarily occurs in the liver and possibly in other AS3MT-expressing cell types, this mechanism of DNMT inhibition would be expected to be limited to these cell types [7,11].Potential arsenite-induced changes in DNA methylation patterns have been found to result in the suppression of genes involved in cell cycle regulation, such as p21, subsequently leading to abnormal cell proliferation [14].In addition to altered DNA methylation patterns, ROS-induced cell signaling also appears to impact cell cycle progression and cell proliferation [15].
Since most cell culture studies have been performed after short-term exposure of up to 24 h, and epigenetic changes would rather be expected to be evident at later time points, the present study aimed to compare short-term and long-term exposure to arsenite in human liver cells.Special focus was placed on gene expression profiles, cell cycle regulation, and the effect of arsenite on DNA methylation in epigenetic regulation.We found that arsenite clearly activates the transcriptional response to oxidative stress, cell cycle control, metal homeostasis, and DNA damage response in short-and long-term experiments, with pronounced time-dependent effects in some gene clusters.Furthermore, significant changes in the expression of genes coding for DNA methyltransferases were observed after chronic arsenite exposure, accompanied by global hypomethylation.

Results
To investigate the effect of arsenite after short-and long-term treatment, in the first step, the cytotoxicity was assessed.HepG2 cells were incubated with NaAsO 2 in a concentration range of 0.5 µM to 50 µM for 24 h.The specific cytotoxicity profile was also used to determine the exposure range for the long-term experiments.To gain additional insights into the transcriptional toxicity profile of NaAsO 2 , HT RT-qPCR analysis was performed in HepG2 cells.To examine the impact of arsenic on epigenetic mechanisms, in addition to the investigation of genes relevant for epigenetic modifications, the global methylcytosine content was determined by HPLC.Furthermore, the effect of NaAsO 2 on cell cycle phase distribution was examined using flow cytometry.For short-term incubations, HepG2 cells were treated with corresponding concentrations of NaAsO 2 for 24 h.The long-term experiments were scheduled to span a period of 10 and 20 days.

Cytotoxicity
As the first step, the cytotoxic potential of arsenite was investigated, with the aim of establishing appropriate exposure concentrations for short-and long-term incubations within the non-cytotoxic concentration range.The cytotoxicity of NaAsO 2 after 24 h of treatment was investigated in HepG2 cells using the CellTiter-Glo ® luminescence cell viability assay (Promega).In this test system, beetle luciferin is converted to oxyluciferin and photons.The luminescence produced correlates with the ATP content after complete lysis and thus reflects the number of viable cells.The results are shown in Figure 1.

Cytotoxicity
As the first step, the cytotoxic potential of arsenite was investigated, with the aim o establishing appropriate exposure concentrations for short-and long-term incubation within the non-cytotoxic concentration range.The cytotoxicity of NaAsO2 after 24 h o treatment was investigated in HepG2 cells using the CellTiter-Glo ® luminescence cell vi bility assay (Promega).In this test system, beetle luciferin is converted to oxyluciferin an photons.The luminescence produced correlates with the ATP content after complete lys and thus reflects the number of viable cells.The results are shown in Figure 1.The incubation with low micromolar NaAsO2 concentrations did not show any e fects on the cell viability of HepG2 cells.However, significant dose-dependent cytotox effects were observed starting at 10 µM NaAsO2, reducing the cellular ATP content b about 15%.Severe cytotoxicity was observed when HepG2 cells were treated with 50 µM NaAsO2, resulting in a 45% reduction in ATP content compared to the untreated contro Based on these data, a concentration range of 0.5 µM to 2 µM was selected for most long term incubations, with an additional concentration range of up to 10 µM for the 20-da incubation period at some endpoints.The determination of cytotoxicity during the tw long-term exposure periods was carried out via determination of the relative cell coun (RCC) (Figure S1).A significant decrease in RCC was observed in both long-term exposu periods starting from a concentration of 5 µM NaAsO2.At the highest exposure concen tration of 10 µM NaAsO2, this resulted in a 17% reduction in RCC after both 10-day an 20-day incubation periods.

Gene Expression Analysis
To investigate the cellular response to NaAsO2 in HepG2 cells, gene expression pro files were established using high-throughput RT-qPCR.This method enabled the simu taneous assessment of the transcription of 95 genes in 96 samples.The gene set include markers associated with metal homeostasis, (oxidative) stress response, DNA damage an repair, cell cycle control, and apoptosis.A detailed description of this method has bee The incubation with low micromolar NaAsO 2 concentrations did not show any effects on the cell viability of HepG2 cells.However, significant dose-dependent cytotoxic effects were observed starting at 10 µM NaAsO 2 , reducing the cellular ATP content by about 15%.Severe cytotoxicity was observed when HepG2 cells were treated with 50 µM NaAsO 2 , resulting in a 45% reduction in ATP content compared to the untreated control.Based on these data, a concentration range of 0.5 µM to 2 µM was selected for most long-term incubations, with an additional concentration range of up to 10 µM for the 20-day incubation period at some endpoints.The determination of cytotoxicity during the two long-term exposure periods was carried out via determination of the relative cell count (RCC) (Figure S1).A significant decrease in RCC was observed in both long-term exposure periods starting from a concentration of 5 µM NaAsO 2 .At the highest exposure concentration of 10 µM NaAsO 2 , this resulted in a 17% reduction in RCC after both 10-day and 20-day incubation periods.

Gene Expression Analysis
To investigate the cellular response to NaAsO 2 in HepG2 cells, gene expression profiles were established using high-throughput RT-qPCR.This method enabled the simultaneous assessment of the transcription of 95 genes in 96 samples.The gene set included markers associated with metal homeostasis, (oxidative) stress response, DNA damage and repair, cell cycle control, and apoptosis.A detailed description of this method has been previously published by our group [16].Additionally, genes related to DNA methylation and histone modifications were included within the present study.Relative gene expression was calculated by normalizing the treated samples to the untreated control and expressed as log 2 values.Reductions of at least 50% (log 2 -fold change ≤ −1) and duplications (log 2 -fold change ≥ 1) compared to the untreated control were considered relevant; in addition, concentration-dependent trends were also taken into account.The cells were incubated with NaAsO 2 for 24 h, 10 days, or 20 days.A selection of significantly altered genes is depicted in Figures 2-6.The log 2 values for all genes are provided in Supplementary Table S1.The results for the entire gene panel, presented as heatmaps, are shown for the 24 h, 10-day, and 20-day incubations in Figures S2 and S3.
Int. J. Mol.Sci.2023, 24, 15238 4 of 18 previously published by our group [16].Additionally, genes related to DNA methylation and histone modifications were included within the present study.Relative gene expression was calculated by normalizing the treated samples to the untreated control and expressed as log2 values.Reductions of at least 50% (log2-fold change ≤ −1) and duplications (log2-fold change ≥ 1) compared to the untreated control were considered relevant; in addition, concentration-dependent trends were also taken into account.The cells were incubated with NaAsO2 for 24 h, 10 days, or 20 days.A selection of significantly altered genes is depicted in Figures 2-6.The log2 values for all genes are provided in Supplementary Table S1.The results for the entire gene panel, presented as heatmaps, are shown for the 24 h, 10-day, and 20-day incubations in Figures S2 and S3.previously published by our group [16].Additionally, genes related to DNA methylation and histone modifications were included within the present study.Relative gene expression was calculated by normalizing the treated samples to the untreated control and expressed as log2 values.Reductions of at least 50% (log2-fold change ≤ −1) and duplications (log2-fold change ≥ 1) compared to the untreated control were considered relevant; in addition, concentration-dependent trends were also taken into account.The cells were incubated with NaAsO2 for 24 h, 10 days, or 20 days.A selection of significantly altered genes is depicted in Figures 2-6.The log2 values for all genes are provided in Supplementary Table S1.The results for the entire gene panel, presented as heatmaps, are shown for the 24 h, 10-day, and 20-day incubations in Figures S2 and S3.concentrations of NaAsO2 for 24 h, 10 days, or 20 days.Results are presented as log2-fold changes in gene expression, compared to the respective untreated control.Shown are the means ± SD of three independent experiments performed in duplicate.Statistical analysis was performed to determine the differences between the exposed cells and the negative control using one-way ANOVA followed by Dunnett's post hoc test: * (p ≤ 0.05), ** (p ≤ 0.01).Comparing the gene expression profiles of HepG2 cells after 24 h, 10 days, and 20 days of exposure to NaAsO2, both dose-and time-dependent alterations in the transcriptional activity of a large number of genes were evident.As shown in Figure 2, the metallothionein-encoding genes MTX1 and MT2A were significantly induced at all exposure times.The largest change, a 20-fold change, was observed for the MT2A gene after 20 days of incubation with 10 µM NaAsO2.Other examples of up-regulated genes include HMOX1 and IL-8, which encode markers of inflammatory and oxidative stress responses (Figure 3).The most pronounced change was observed in the case of HMOX1, which already showed significant inductions in the sub-micromolar and low micromolar concentration range of arsenite throughout the incubation period.At the highest dose, the strongest concentration-dependent activation of HMOX1 was observed after 10 days of incubation, with a 64-fold induction.Low micromolar concentrations of arsenite also resulted in significant inductions of the cell cycle-and apoptosis-regulating genes EGFR and CDKN2B following acute exposure.Exposure for 10 days further enhanced these effects, whereas after 20 days of exposure, inductions were only observed in the concentration range of 5 µM NaAsO2 and above.Sub-micromolar concentrations, on the other hand, led to the significant repression of both genes during the longest exposure time (Figure 4).Significant differences between the short-term and long-term incubations were observed for the DNA damage response and repair genes.As shown in Figure 5, after 24 h of incubation with NaAsO2, significant reductions of DDB2 were already observed in the low micromolar concentration range, which increased with rising concentrations.At the highest dose, 4.2- Comparing the gene expression profiles of HepG2 cells after 24 h, 10 days, and 20 days of exposure to NaAsO 2 , both dose-and time-dependent alterations in the transcriptional activity of a large number of genes were evident.As shown in Figure 2, the metallothioneinencoding genes MTX1 and MT2A were significantly induced at all exposure times.The largest change, a 20-fold change, was observed for the MT2A gene after 20 days of incubation with 10 µM NaAsO 2 .Other examples of up-regulated genes include HMOX1 and IL-8, which encode markers of inflammatory and oxidative stress responses (Figure 3).The most pronounced change was observed in the case of HMOX1, which already showed significant inductions in the sub-micromolar and low micromolar concentration range of arsenite throughout the incubation period.At the highest dose, the strongest concentrationdependent activation of HMOX1 was observed after 10 days of incubation, with a 64-fold induction.Low micromolar concentrations of arsenite also resulted in significant inductions of the cell cycle-and apoptosis-regulating genes EGFR and CDKN2B following acute exposure.Exposure for 10 days further enhanced these effects, whereas after 20 days of exposure, inductions were only observed in the concentration range of 5 µM NaAsO 2 and above.Sub-micromolar concentrations, on the other hand, led to the significant repression of both genes during the longest exposure time (Figure 4).Significant differences between the short-term and long-term incubations were observed for the DNA damage response and repair genes.As shown in Figure 5, after 24 h of incubation with NaAsO 2 , significant reductions of DDB2 were already observed in the low micromolar concentration range, which increased with rising concentrations.At the highest dose, 4.2-fold suppression was observed.After longer incubation times, however, no impact was seen any more at 10 days incubation, and after 20 days incubation, even an induction was observed at 5 µM NaAsO 2 and above.In addition to DDB2, BRCA1 also showed a trend towards repression after 24 h incubation, but pronounced activation after long-term incubation of 10 or 20 days with NaAsO 2 , leading to significant inductions by sub-micromolar NaAsO 2 concentrations after 20 days.Regarding genes involved in epigenetic alterations, significant concentration-dependent inductions of DNMT1 and DNMT3B were detected at the two long-term exposure times (Figure 6).At the highest dose of 10 µM NaAsO 2 , 7-fold activation was detected after 20 days.On the contrary, DNMT1A showed significant repression after both the short-term incubation and the two long-term incubations.

Cell Cycle Regulation
Both short-term and long-term NaAsO 2 treatments affected the expression of genes related to cell cycle regulation and proliferation.To assess the consequences at the functional level, the cell cycle distribution in HepG2 cells was analyzed via flow cytometry.Figure 7 shows the cell cycle phase distribution of HepG2 cells after 24 h (A) and after 10 or 20 days (B) of incubation with NaAsO 2 .
fold suppression was observed.After longer incubation times, however, no impact was seen any more at 10 days incubation, and after 20 days incubation, even an induction was observed at 5 µM NaAsO2 and above.In addition to DDB2, BRCA1 also showed a trend towards repression after 24 h incubation, but pronounced activation after long-term incubation of 10 or 20 days with NaAsO2, leading to significant inductions by sub-micromolar NaAsO2 concentrations after 20 days.Regarding genes involved in epigenetic alterations, significant concentration-dependent inductions of DNMT1 and DNMT3B were detected at the two long-term exposure times (Figure 6).At the highest dose of 10 µM NaAsO2, 7fold activation was detected after 20 days.On the contrary, DNMT1A showed significant repression after both the short-term incubation and the two long-term incubations.

Cell Cycle Regulation
Both short-term and long-term NaAsO2 treatments affected the expression of genes related to cell cycle regulation and proliferation.To assess the consequences at the functional level, the cell cycle distribution in HepG2 cells was analyzed via flow cytometry.Figure 7 shows the cell cycle phase distribution of HepG2 cells after 24 h (A) and after 10 or 20 days (B) of incubation with NaAsO2.Considering the basal cell cycle distribution in control cells, the highest population was found in the G1 phase (~59%) and about 34% of the cells remained in the G2/M phase, while the S phase (~7%) represented the smallest cellular fraction.After 24 h of treatment with NaAsO 2 , a significant change in the cell cycle phase distribution compared to the basal distribution was observed, starting at a concentration of 15 µM.At the highest concentration of 20 µM NaAsO 2 , there was an 8.3% increase in cells in the G2/M phase with a simultaneous 7.5% decrease in cells in the G1 phase.Only the population in the S phase showed a similar level over the entire concentration range.After long-term incubations, treatment with NaAsO 2 had no effect on the distribution of cell cycle phases throughout the applied concentration range.

DNA Methylation
Long-term incubation with NaAsO 2 showed concentration-dependent induction of the genes DNMT1 and DNMT3B, which are particularly involved in DNA methylation.To follow this impact at the functional level, global DNA methylation was investigated via HPLC-UV analysis.The amount of methyl-CMP present in each sample was expressed relative to the total dCMP present (methyl-dCMP + dCMP) and relative to the untreated control levels.Figure 8A shows the results after 24 h of short-term incubation and Figure 8B after the corresponding long-term incubations.
and the negative control using one-way ANOVA followed by Dunnett's post hoc test: * (p ≤ 0.05 Considering the basal cell cycle distribution in control cells, the highest populat was found in the G1 phase (~59%) and about 34% of the cells remained in the G2/M pha while the S phase (~7%) represented the smallest cellular fraction.After 24 h of treatm with NaAsO2, a significant change in the cell cycle phase distribution compared to basal distribution was observed, starting at a concentration of 15 µM.At the highest c centration of 20 µM NaAsO2, there was an 8.3% increase in cells in the G2/M phase wit simultaneous 7.5% decrease in cells in the G1 phase.Only the population in the S ph showed a similar level over the entire concentration range.After long-term incubatio treatment with NaAsO2 had no effect on the distribution of cell cycle phases through the applied concentration range.

DNA Methylation
Long-term incubation with NaAsO2 showed concentration-dependent induction the genes DNMT1 and DNMT3B, which are particularly involved in DNA methylati To follow this impact at the functional level, global DNA methylation was investiga via HPLC-UV analysis.The amount of methyl-CMP present in each sample was expres relative to the total dCMP present (methyl-dCMP + dCMP) and relative to the untrea control levels.Figure 8A   Short-term incubation with NaAsO 2 did not significantly alter the global methylation level in HepG2 cells.However, regarding the two long-term incubation periods of 10 and 20 days, exposure to NaAsO 2 caused significant and dose-dependent global hypomethylation.Already, in the low micromolar concentration range, starting at 1 µM, a significant decrease in 5-methyl-dCMP levels was observed after 10 days of exposure compared to the control.At the highest concentration of 2 µM, 10-day exposure to NaAsO 2 resulted in a 6.3% decrease in 5-methyl-dCMP levels.In the 20-day long-term incubation, significant concentration-dependent decreases in global methylation were observed even at 0.5 µM.The highest incubation concentration of 10 µM NaAsO 2 resulted in an 8% decrease in detectable 5-methyl-dCMP levels.

Discussion
For arsenic, it is well established that the induced carcinogenic effects occur upon sustained exposure over extended periods.As a result, long-term exposures in in vitro studies also appear to be of particular interest.Complementary to already existing longterm in vitro studies, it was the aim of the present work to investigate the toxicological and epigenetic effects of sub-micromolar to low micromolar, and thus, non-cytotoxic concentrations of arsenite following short-and long-term exposure in human liver cells.The applied concentrations are considered relevant for human exposure levels when compared to urine levels of inorganic arsenite.Even though the reference values, and thus, background levels of the general population are usually lower (for example, 15 µg/L urine (200 nM) in Germany), in occupational settings, levels up to 340 µg/L (4.5 µM) have been reported [17].In Bangladesh, due to the highly elevated levels of inorganic arsenic in drinking water, even levels of up to 2166 µg/L urine (29 µM) were reported [18].
In the present study, it was observed that chronic exposure to sub-micromolar concentrations of arsenite resulted in global DNA hypomethylation and significant alterations in various gene expression clusters.
Altered epigenetic modifications, particularly DNA methylation, are believed to represent one underlying mechanism of arsenic-induced carcinogenesis.One potential cause of global hypomethylation could be the depletion of S-adenosylmethionine (SAM), although this relationship has not been fully elucidated.The enzymatic metabolism of inorganic arsenic in the liver to the methylated intermediates, monomethylarsonic acid (MMA(III)) and dimethylarsinic acid (DMA(III)), is mediated by S-adenosylmethionine-dependent arsenic methyltransferase (As3MT).However, SAM is also a co-factor that is generally required for cellular methylation reactions, including DNA methylation [11,[19][20][21].To further investigate the potential effect of arsenite metabolism on DNA methylation following long-term incubation, we employed the human liver cell line HepG2, which is proficient in As3MT [22][23][24].
The acute cytotoxic potential of NaAsO 2 was assessed by determining the intracellular ATP levels after 24 h of exposure.At 10 µM NaAsO 2 , a significant decrease in basal ATP levels was evident, which further declined with increasing concentrations, reaching 45% after incubation with 50 µM arsenite.In the long-term experiments, a significant decrease in cell count was observed, starting at 5 µM NaAsO 2 in both experimental periods.The cytotoxic effect of arsenite has been previously demonstrated in various cell lines, although the precise mechanisms remain incompletely understood [25,26].One significant specific mechanism appears to be arsenite-induced cellular oxidative stress, leading to the inhibition of oxidative phosphorylation and a decline in cellular ATP levels [27].
To assess the effects of arsenite at the transcriptional level and to identify potential differences in gene expression profiles between short-and long-term exposed cells, gene expression analyses were performed using a high-throughput RT-qPCR method [16].The most pronounced changes were observed in the cluster associated with the oxidative stress response in both short-and long-term arsenite treatments, particularly involving alterations in target genes of the KEAP1-NRF2 signaling pathway.The strongest induction was observed in the HMOX1 gene, encoding heme oxygenase 1, which serves as a sensitive indicator for oxidative stress.Remarkably, a significant effect was already evident after 24 h even at the lowest sub-micromolar arsenite concentration, which was sustained in the long-term experiment.Other sensor genes for oxidative stress, such as HSPA1A, were strongly induced after short-term exposure, although to a lesser extent.However, this effect was not maintained in the long-term exposure, suggesting an early stress-induced response.Additionally, there were minor but dose-dependent inductions of GCLC and TXNRD1, which were significantly amplified after 10 days of exposure.Conversely, a declining trend of induction was observed at the longest incubation time of 20 days, indicating a timedependent arsenite-induced response.GCLC and TXNRD1 are also target genes of the NRF2 transcription factor, which regulates cellular levels of glutathione (GSH) and thioredoxin (TXN) as part of the cellular antioxidant response.The genes NFKB1 and NFKBIA, whose expression is mediated by the redox-sensitive transcription factor NF-κB, exhibited mild repression throughout the entire exposure period.The arsenite-induced induction of genes involved in the oxidative stress response, as previously demonstrated in short-term experiments in other cell lines [28], appears to persist even in the long-term studies.
Remarkably, a significant concentration-dependent induction of interleukin-8 (IL-8) was observed only during the two long-term exposure periods, and not after short-term exposure.IL-8 is a marker of inflammatory response, but plays also a crucial role in cell transformation, tumor growth, and metastasis [29,30].For example, long-term exposure to arsenite has been shown to be involved in the transformation of human bronchial epithelial (HSE) cells, involving the HIF-2α-mediated up-regulation of IL-6 and IL-8 [30].Therefore, IL-8 appears to play an important role in later steps of arsenite-induced carcinogenicity.
In addition to the induction of genes involved in the oxidative and inflammatory stress response, significantly increased expression levels of MT1X and MT2A were observed upon NaAsO 2 exposure.These genes encode metallothioneins, indicative of elevated intracellular metal ion concentrations; their expression was significantly induced even at sub-micromolar concentrations in the long-term experiment.These results suggest the involvement of the zinc-binding, metal-regulatory transcription factor MTF-1, which is activated by the release of zinc from metallothionein (MT) upon metal overload or oxidative stress [31].Interestingly, the up-regulation of SLC30A1, which encodes a proton-coupled zinc antiporter involved in the efflux of zinc from cells to prevent toxicity, was also observed with increasing exposure duration [32].
Concentration-dependent changes in gene expression profiles were also obtained for cell cycle-and apoptosis-related genes in the short-term incubations.Significantly increased inductions were observed for CDKN2B, which encodes the cell cycle regulator p15, as well as EGFR and CDKN1A.These changes in the gene expression profiles of cell cycle-regulating genes were maintained in the 10-day long-term exposure; interestingly, at this time point, the up-regulation of CDKN2B occurred at lower concentrations, while EGFR resulted in similar up-regulation as compared to 24 h incubation.The picture changed considerably after 20-day incubation: here, both CDKN2B and EGFR were downregulated at low concentrations of arsenite, but up-regulated at higher concentrations, even though less pronounced when compared to the shorter incubation times.Time-dependent differences in EGFR-related effects were also reported in the literature.Thus, in human BEAS-2B cells, increased EGFR protein levels were observed after acute (24 h) and chronic (24 week) arsenite exposure, albeit involving different mechanisms.While the elevated expression of TGFα was observed after long-term treatment, accompanied by increased EGFR phosphorylation and elevated cell surface EGFR levels, no concomitant increase in TGFα expression was evident after 24 h, suggesting time-dependent differences in the molecular mechanisms of EGFR activation [33].
To investigate these different effects of arsenite on the mRNA expression of cell cycle-regulating genes at a functional level, flow cytometric analyses were conducted, also focusing on a comparison between short-term and long-term exposed cells.Acute treatment with NaAsO 2 at a concentration of 15 µM or higher led to G 2 /M arrest, indicative of DNA damage after short-term exposure, in agreement with the results on the transcriptional level.This supports previous studies where pronounced G2/M arrests after acute arsenite exposure have been observed in various cell lines [15,34,35].However, in the long-term experiments with 10-and 20-day sub-micromolar and micromolar NaAsO 2 incubations, no effect on cell cycle progression was observed within the present study.In contrast, in a study by Ganapathy et al., a delayed exit from mitosis was demonstrated in BEAS-2B cells and keratinocytes following one month of chronic exposure to 0.5 µM sodium arsenite [36].Time-and cell line-dependent effects may play a role in this discrepancy.Additionally, it should be noted that the flow cytometric analyses conducted in this study did not distinguish between G2 and M arrests.
Considerable differences in gene expression profiles between short-and long-term exposed cells were observed in genes involved in the DNA damage response and repair.As previously reported [37,38] and also seen within the present study, short-term incubation with arsenite resulted in the concentration-dependent repression of DNA repair factors involved in basically all major DNA repair pathways.However, opposite effects were observed in the long-term experiments, where significant inductions were observed, including the genes BRCA1, DDB2, and LIG1 involved in double-strand break and nucleotide excision repair.These results suggest that the down-regulation of DNA repair may only be a transient reaction of the cell, compensated by the up-regulation of DNA repair genes upon chronic arsenite exposure.
Significant differences were also observed in the cluster of genes associated with epigenetics between cells exposed to short-and long-term arsenite treatment.Since arsenite may interfere with DNA methylation, special focus was placed on the genes encoding DNA methyltransferases.Interestingly, in addition to time-dependent effects, pronounced differences were also seen between the expression patterns of the different DNA methyltransferases.In the case of DNMT3A, significant down-regulation was observed at all time points; this effect was observed even at very low concentrations after long-term exposure.In contrast, DNMT1 showed moderate induction at all time points.The expression levels of DNMT3B, however, were not altered after 24 h of incubation, but were dose-dependently elevated after both 10 and 20 days of incubation, starting at sub-micromolar NaAsO 2 concentrations.This supports our observation that changes in epigenetic modifications require long-term exposure conditions.The genes DNMT1, DNMT3A, and DNMT3B encode DNA methyltransferases, which perform different functions in the methylation process.The de novo DNA methylation during embryogenesis has been proposed to be carried out by the DNA methyltransferases DNMT3A and DNMT3B [39,40], while DNMT1 was thought to maintain specific methylation patterns in a replication-dependent manner [41,42].However, an increasing number of studies suggest that the classification into de novo and maintenance DNA methyltransferases may be an oversimplification.Thus, DNMT3A and DNMT3B appear to also play important roles in the maintenance of methylation patterns [43,44], and DNMT1 is involved in establishing DNA methylation patterns, as well [45,46].
The pronounced induction of two DNMT-encoding genes, namely, DNMT1 and DNMT3B, at the transcriptional level may suggest DNA hypermethylation.Therefore, the extent of global DNA methylation was determined via HPLC-UV.Acute arsenite exposure did not show any impact on global 5-mC content.Interestingly, however, 10-day long-term incubation with 1 µM already led to significant hypomethylation, which was further amplified with increasing concentration.This hypomethylation was also observed after 20-day exposure, even at the sub-micromolar concentration of 0.5 µM, with pronounced dose-dependency at higher arsenite concentrations.Altogether, long-term exposure to low concentrations of NaAsO 2 resulted in global hypomethylation in human liver cells, which was not detected in short-term incubations.
Global hypomethylation can contribute to malignant transformation by promoting chromosomal defects that can lead to genetic instability and by promoting spontaneous mutations [11].The induction of global hypomethylation due to long-term arsenite exposure confirms several observations reported in the literature [47][48][49].However, controversial results were obtained with respect to the role of DNMT expression levels as one potential underlying mechanism of hypomethylation.Within the present study, down-regulation was only observed in the case of DNMT3A.In contrast, Rea et al. observed a notable reduction in the transcript levels of all DNMTs in BEAS-2B cells following 36-day exposure to 0.5 µM sodium arsenite, which also resulted in decreased protein levels of DNMT1, DNMT3A, and DNMT3B.Nevertheless, Zhao et [19,[47][48][49].Therefore, data concerning the role of gene expression changes in global hypomethylation remain inconsistent at present.One other potential consequence of arsenite exposure consists in the consumption of S-adenosylmethionine (SAM) during arsenic metabolism, which, in turn, could lead to the accumulation of Sadenosylhomocysteine (SAH).Both the depletion of SAM and the accumulation of SAH can inhibit DNMT activity at both the mRNA and protein levels through negative feedback.
It is important to note, however, that the inhibition of DNMT activity primarily occurs in the liver, and possibly in other cell types expressing As3MT, due to arsenic metabolism mediated by AS3MT.Therefore, this mechanism does not appear to be the general reason for arsenite-induced hypomethylation.In cells unable to metabolize arsenic, activation of the trans-sulfuration pathway as a consequence of oxidative stress appears to be a possible cause of global hypomethylation [11].
In addition to the indirect inhibition of methyltransferases through the limitation of the co-factor SAM, there is a possibility that direct inhibition at the protein level could be a cause of global hypomethylation.Potentially very sensitive targets are zinc-binding structures, so-called zinc fingers, where zinc complexes four amino acids, including two to four cysteines and up to two histidines.Interactions with these structures have been shown to be particularly relevant in arsenite-induced DNA repair inhibition [50].Thus, the most sensitive target related to DNA repair affected by trivalent arsenicals is poly(ADP-ribosyl)ation, inhibited by low nanomolar concentrations of arsenite, MMA(III), and DMA(III).Furthermore, all three trivalent arsenicals inhibited isolated PARP1, indicating a direct interaction with this enzyme [25,51,52].Also, trivalent but not pentavalent arsenicals have been shown to release zinc from a 37-amino-acid peptide resembling the zinc finger domain of the human XPA protein (XPAzf) [53], albeit via different mechanisms.Whereas equimolar concentrations of MMA(III) mediated zinc release, forming mono-and diarsenical derivatives of XPAzf, a 10-fold excess of arsenite was required to partially oxidize XPAzf, yielding one or two disulfide bonds [54].Interestingly, zinc finger structures of the CXXC and ADD classes have also been identified in DNA methyltransferases [55,56].The CXXC zinc finger structures are present in various DNA methyltransferases, including DNMT1, and play a crucial role in the recognition and binding of specific DNA sequences.They consist of conserved cysteine residues that coordinate zinc ions, forming a stable protein structure.It has been shown for DNMT1 that the binding of the CXXC zinc finger domain to CpG-rich DNA regions is essential for its catalytic activity [55,57].The impact of arsenite on other DNA methyltransferases such as DNMT3A has also been investigated.DNMT3A contains an ADD domain with a zinc finger motif in its DNA binding domain, facilitating the establishment of new DNA methylation patterns during development and DNA repair.It has been demonstrated that arsenite directly binds to the cysteine-rich zinc-binding ADD domain of DNMT3A [58].Finally, due to the complexation of zinc to cysteines within zinc-binding structures, they may also be particularly sensitive to oxidative stress induced by arsenite [55].
In summary, the existing evidence points to different cell-specific mechanisms of action of arsenic in DNMT activity and global DNA methylation patterns.According to the results obtained within the present study, hypomethylation appears not to be due to the inhibition of DNMT mRNA expression.Therefore, the ability of arsenite to influence methyltransferase activity in human liver cells may occur through the depletion of cofactors or direct inactivation of the respective proteins, potentially via interaction with zinc-binding structures present in DNMTs.This aspect needs to be further investigated.

Cell Culture and Drug Treatment
The human liver cancer cell line HepG2 was cultured as a monolayer in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% heat-inactivated fetal bovine serum (FBS), 100 U/mL penicillin, and 100 µg/mL streptomycin.The cells were cultured at 37 • C with 5% CO 2 in the air and 100% humidity.Logarithmically growing cells were incubated with NaAsO 2 in the short-term experiments for 24 h, and in the long-term experiments for 10 or 20 days.For the long-term experiments of 10 or 20 days, the cells were passaged every 5 days and re-incubated with the arsenite-containing medium.The non-exposed control cells were concurrently cultivated and sub-cultured over the same period to discern the effects of continuous passaging from those of sustained exposure to arsenic.

Cytotoxicity Assays
For the analysis of ATP content, the CellTiter-Glo ® Luminescent Cell Viability Assay Kit (Promega GmbH, Walldorf, Germany) was used.Cell cultivation and incubation were performed in 96-well plates.Specifically, 3 × 10 4 cells per well were seeded.Logarithmically growing cells were treated with NaAsO 2 according to the instructions provided in the section on drug treatment.After removing the incubation medium, the cells were washed twice with PBS, and 100 µL of fresh medium was added to the wells.The CellTiter-Glo ® Luminescent Cell Viability Assay was performed according to the manufacturer's instructions.The plate was equilibrated in the dark at room temperature for 30 min, and 100 µL of the CellTiter-Glo ® reagent was added.Chemiluminescence was measured after brief shaking and an additional 10 min stabilization period of the signal using the Infinite ® 200 Pro microplate reader (Tecan Group Ltd., Männedorf, Switzerland).The ATP content and, consequently, the decrease in cell viability were expressed as percentages of the untreated control cells.

High-Throughput RT-qPCR
For the gene expression analysis, 5 × 10 5 cells were seeded.Logarithmically growing cells were treated with NaAsO 2 for 24 h, 10 days, or 20 days.Subsequently, they were washed with PBS, trypsinized, resuspended in ice-cold PBS containing 10% FBS, and collected via centrifugation.RNA was then isolated using the MN NucleoSpin ® RNA Plus Kit (Macherey-Nagel, Düren, Germany) according to the manufacturer's instructions.The subsequent high-throughput RT-qPCR was performed via Fluidigm Dynamic Arrays using the BioMark™ system, as previously described [16].The original gene set was partly replaced with 16 genes of epigenetic regulation (DNMT1, DNMT3A, DNMT3B, HDAC1, HDAC2, HDAC3, HDAC10, KDM3A, MeCP2, MBD4, SETD2, TET1, TET2, TET3, EHMT2, EP300) and 4 genes of cell cycle control and senescence (CDKN2A, CDKN2D, FOXO1, FOXO3).All genes are presented in Table S1 and all data of the newly established primers are listed in Table S2.The data were analyzed using Fluidigm Real-Time PCR Analysis and GenEx software, version 5.3.6.170.Normalization was performed using up to five reference genes (ACTB, B2M, GAPDH, GUSB, and HPRT1).Changes in the transcriptional values of the target genes were presented as log 2 -fold change compared to the respective untreated controls, with the relative quantities calculated using the ∆∆Cq method [59].

Conclusions
The present study investigated the impact of acute and chronic arsenite exposure on global DNA methylation, gene expression profiles, and cell cycle regulation in human liver cells.The results demonstrate pronounced differences between the short-and long-term treatment of HepG2 cells, providing insight on time-dependent changes in the cellular response.Thus, clear time-and concentration-dependent effects were observed at the transcriptional level for genes involved in metal homeostasis, the oxidative stress response, cell cycle regulation, and DNA repair.Regarding genes related to epigenetic regulation, both long-term exposure periods of 10 and 20 days resulted in significant inductions of DNMT1 and DNMT3B, with time-dependent amplifying effects.Interestingly, under the same conditions, significant concentration-dependent hypomethylation was observed during both long-term exposures, with a pronounced effect already occurring at a concentration of 0.5 µM NaAsO 2 after 20 days of exposure.In contrast, no impact on global methylation levels was observed during the short-term exposure of 24 h.Global hypomethylation can contribute to malignant transformation by promoting chromosomal defects that lead to genetic instability and spontaneous mutations.Since the even up-regulation of two DNMT-coding genes was observed, our data rather point towards the inhibition of DNMTs at the protein level or a potential co-factor deficiency caused by arsenic metabolism.For example, the zinc-binding structures of DNA methyltransferases may represent potentially sensitive targets; this aspect needs to be further investigated.

Figure 1 .
Figure 1.Cellular ATP content of HepG2 cells after 24 h of treatment with NaAsO2.Mean values standard deviations (SD) from three independent experiments are shown.Statistical analysis w performed to determine the differences between the exposed cells and the negative control usin one-way ANOVA followed by Dunnett's post hoc test: * (p ≤ 0.05), ** (p ≤ 0.01).

Figure 1 .
Figure 1.Cellular ATP content of HepG2 cells after 24 h of treatment with NaAsO 2 .Mean values ± standard deviations (SD) from three independent experiments are shown.Statistical analysis was performed to determine the differences between the exposed cells and the negative control using one-way ANOVA followed by Dunnett's post hoc test: * (p ≤ 0.05), ** (p ≤ 0.01).

Figure 2 .
Figure 2. Impact of NaAsO2 exposure on the mRNA expression of metallothionein genes MT1X and MT2A, which are significantly altered in their expression in HepG2 cells.Cells were incubated with different concentrations of NaAsO2 for 24 h, 10 days, or 20 days.Results are presented as log2-fold changes in gene expression, compared to the respective untreated control.Shown are the means ± SD of three independent experiments each performed in duplicate.Statistical analysis was performed to determine the differences between the exposed cells and the negative control using oneway ANOVA followed by Dunnett's post hoc test: ** (p ≤ 0.01).

Figure 3 .Figure 2 .
Figure 3. Impact of NaAsO2 exposure on the mRNA expression of IL-8 and HMOX1, which are significantly altered in their expression in HepG2 cells.Cells were incubated with different

Figure 2 .
Figure 2. Impact of NaAsO2 exposure on the mRNA expression of metallothionein genes MT1X and MT2A, which are significantly altered in their expression in HepG2 cells.Cells were incubated with different concentrations of NaAsO2 for 24 h, 10 days, or 20 days.Results are presented as log2-fold changes in gene expression, compared to the respective untreated control.Shown are the means ± SD of three independent experiments each performed in duplicate.Statistical analysis was performed to determine the differences between the exposed cells and the negative control using oneway ANOVA followed by Dunnett's post hoc test: ** (p ≤ 0.01).

Figure 3 .Figure 3 .
Figure 3. Impact of NaAsO2 exposure on the mRNA expression of IL-8 and HMOX1, which are significantly altered in their expression in HepG2 cells.Cells were incubated with differentFigure 3. Impact of NaAsO 2 exposure on the mRNA expression of IL-8 and HMOX1, which are significantly altered in their expression in HepG2 cells.Cells were incubated with different concentrations of NaAsO 2 for 24 h, 10 days, or 20 days.Results are presented as log 2 -fold changes in gene expression, compared to the respective untreated control.Shown are the means ± SD of three independent experiments performed in duplicate.Statistical analysis was performed to determine the differences between the exposed cells and the negative control using one-way ANOVA followed by Dunnett's post hoc test: * (p ≤ 0.05), ** (p ≤ 0.01).

Figure 4 .
Figure 4. Impact of NaAsO2 exposure on the mRNA expression of the cell cycle-regulating genes CDKN2B and EGFR, which are significantly altered in their expression in HepG2 cells.Cells were incubated with different concentrations of NaAsO2 for 24 h, 10 days, or 20 days.Results are presented as log2-fold changes in gene expression, compared to the respective untreated control.Shown are the means ± SD of three independent experiments performed in duplicate.Statistical analysis was performed to determine the differences between the exposed cells and the negative control using one-way ANOVA followed by Dunnett's post hoc test: ** (p ≤ 0.01).

Figure 5 .
Figure 5. Impact of NaAsO2 exposure on the mRNA expression of DNA repair genes BRCA1 and DDB2, which are significantly altered in their expression in HepG2 cells.Cells were incubated with

Figure 4 .
Figure 4. Impact of NaAsO 2 exposure on the mRNA expression of the cell cycle-regulating genes CDKN2B and EGFR, which are significantly altered in their expression in HepG2 cells.Cells were incubated with different concentrations of NaAsO 2 for 24 h, 10 days, or 20 days.Results are presented as log 2 -fold changes in gene expression, compared to the respective untreated control.Shown are the means ± SD of three independent experiments performed in duplicate.Statistical analysis was performed to determine the differences between the exposed cells and the negative control using one-way ANOVA followed by Dunnett's post hoc test: ** (p ≤ 0.01).

Figure 4 .
Figure 4. Impact of NaAsO2 exposure on the mRNA expression of the cell cycle-regulating genes CDKN2B and EGFR, which are significantly altered in their expression in HepG2 cells.Cells were incubated with different concentrations of NaAsO2 for 24 h, 10 days, or 20 days.Results are presented as log2-fold changes in gene expression, compared to the respective untreated control.Shown are the means ± SD of three independent experiments performed in duplicate.Statistical analysis was performed to determine the differences between the exposed cells and the negative control using one-way ANOVA followed by Dunnett's post hoc test: ** (p ≤ 0.01).

Figure 5 .
Figure 5. Impact of NaAsO2 exposure on the mRNA expression of DNA repair genes BRCA1 and DDB2, which are significantly altered in their expression in HepG2 cells.Cells were incubated with

Figure 5 .
Figure 5. Impact of NaAsO 2 exposure on the mRNA expression of DNA repair genes BRCA1 and DDB2, which are significantly altered in their expression in HepG2 cells.Cells were incubated with different concentrations of NaAsO 2 for 24 h, 10 days, or 20 days.Results are presented as log 2 -fold changes in gene expression, compared to the respective untreated control.Shown are the means ± SD of three independent experiments each performed in duplicate.Statistical analysis was performed to determine the differences between the exposed cells and the negative control using one-way ANOVA followed by Dunnett's post hoc test: * (p ≤ 0.05), ** (p ≤ 0.01).

Figure 6 .
Figure 6.Impact of NaAsO2 exposure on the mRNA expression of genes involved in epigenetic regulation, DNMT1, DNMT3A, and DNMT3B, which are significantly altered in their expression in HepG2 cells.Cells were incubated with different concentrations of NaAsO2 for 24 h, 10 days, or 20 days.Results are presented as log2-fold changes in gene expression, compared to the respective untreated control.Shown are the means ± SD of three independent experiments each performed in duplicate.Statistical analysis was performed to determine the differences between the exposed cells and the negative control using one-way ANOVA followed by Dunnett's post hoc test: * (p ≤ 0.05), ** (p ≤ 0.01).

Figure 6 .
Figure 6.Impact of NaAsO 2 exposure on the mRNA expression of genes involved in epigenetic regulation, DNMT1, DNMT3A, and DNMT3B, which are significantly altered in their expression in HepG2 cells.Cells were incubated with different concentrations of NaAsO 2 for 24 h, 10 days, or 20 days.Results are presented as log 2 -fold changes in gene expression, compared to the respective untreated control.Shown are the means ± SD of three independent experiments each performed in duplicate.Statistical analysis was performed to determine the differences between the exposed cells and the negative control using one-way ANOVA followed by Dunnett's post hoc test: * (p ≤ 0.05), ** (p ≤ 0.01).

Figure 7 .
Figure 7. Analysis of cell cycle distribution of HepG2 cells after treatment with NaAsO2 for 24 h (A) and long-term incubation for 10 and 20 days (B).The cell cycle distribution was analyzed via DAPI staining using flow cytometry.Means ± standard deviations of three independent experiments are

Figure 7 .
Figure 7. Analysis of cell cycle distribution of HepG2 cells after treatment with NaAsO 2 for 24 h (A) and long-term incubation for 10 and 20 days (B).The cell cycle distribution was analyzed via DAPI staining using flow cytometry.Means ± standard deviations of three independent experiments are shown.Statistical analysis was performed to determine the differences between the exposed cells and the negative control using one-way ANOVA followed by Dunnett's post hoc test: * (p ≤ 0.05).
shows the results after 24 h of short-term incubation and Fig 8B after the corresponding long-term incubations.

Figure 8 .Figure 8 .
Figure 8. Effects of 24 h (A) and 10-or 20-day (B) NaAsO2 exposure on global DNA methylation HepG2 cells, presented as global DNA methylation compared to the untreated control.Genomi methyl-dCMP content was calculated as the proportion of total dCMP (5me-dCMP)/(5me-dCM Figure 8. Effects of 24 h (A) and 10-or 20-day (B) NaAsO 2 exposure on global DNA methylation in HepG2 cells, presented as global DNA methylation compared to the untreated control.Genomic 5-methyl-dCMP content was calculated as the proportion of total dCMP (5me-dCMP)/(5me-dCMP + dCMP) expressed in relation to the untreated control.Shown are the mean values ± SD of three independent experiments performed in duplicate.Statistical analysis was performed to determine the differences between the exposed cells and the negative control using one-way ANOVA followed by Dunnett's post hoc test: * (p ≤ 0.05), ** (p ≤ 0.01).
al. demonstrated increased DNMT ex-pression levels, Benbrahim-Tallaa et al. reported unchanged DNMT expression levels, and Cui et al. observed a reduction in DNMT1 mRNA expression after 48 h of exposure of HepG2 cells to arsenite