Effects of H19/SAHH/DNMT1 on the oxidative DNA damage related to benzo[a]pyrene exposure

The mechanisms that long noncoding RNA (lncRNA) H19 binding to S-adenosylhomocysteine hydrolase (SAHH) interacted with DNA methyltransferase 1 (DNMT1) and then regulated DNA damage caused by polycyclic aromatic hydrocarbons (PAHs) remain unclear. A total of 146 occupational workers in a Chinese coke-oven plant in 2014 were included in the final analyses. We used high-performance liquid chromatography mass spectrometry (HPLC–MS) equipped to detect urine biomarkers of PAHs exposure, including 2-hydroxynaphthalene (2-NAP), 2-hydroxyfluorene (2-FLU), 9-hydroxyphenanthrene (9-PHE) and 1-hydroxypyrene (1-OHP). The levels of SAM and SAH in plasma were detected by HPLC-ultraviolet. By constructing various BEAS-2B cell models exposed to 16 μM benzo[a]pyrene (BaP) for 24 h, toxicological parameters reflecting distinct mechanisms were evaluated. We documented that urinary 1-hydroxypyrene (1-OHP) levels were positively associated with blood H19 RNA expression (OR: 1.51, 95% CI: 1.03–2.19), but opposite to plasma SAHH activity (OR: 0.63, 95% CI: 0.41–0.98) in coke oven workers. Moreover, by constructing various BEAS-2B cell models exposed to benzo[a]pyrene (BaP), we investigated that H19 binding to SAHH exaggerated DNMT1 expressions and activity. Suppression of H19 enhanced the interaction of SAHH and DNMT1 in BaP-treated cells, decreased eight-oxoguanine DNA glycosylase 1 (OGG1) methylation, reduced oxidative DNA damage and lessened S phase arrest. However, SAHH or DNMT1 single knockdown and SAHH/DNMT1 double knockdown showed the opposite trend. A H19/SAHH/DNMT1 axis was involved in OGG1 methylation, oxidative DNA damage and cell cycle arrest by carcinogen BaP.


Background
Long noncoding RNAs (lncRNAs), a group of transcription of RNA from non-protein coding regions of the genome, are defined as endogenous cellular RNAs of longer than 200 nucleotides. They are implicated in a number of molecular functions, such as regulators of protein activity, molecular scaffolds, regulation of cell cycle and controllers of chromatin remodelling (Chen 2016). H19, a multifunctional lncRNA, interacts with other proteins to exert functions as a ribonucleoprotein in tumorigenesis (Wang et al. 2019). Similar to Zhou et al. (2015), a previous research by our team has revealed that H19 binds to S-adenosylhomocysteine hydrolase (SAHH) and inhibits SAHH function, thus blocking LINE-1 methylation in vitro cells (Fu et al. 2018). SAHH has been reported as a highly conserved enzyme capable of reversibly hydrolysing S-adenosylhomocysteine (SAH) and participates in S-adenosylmethionine (SAM)-dependent transmethylation reactions (Oksana et al. 2013). Gene methylation (5-methyl-cytosine), accomplished by the joint action of three SAM-dependent DNA methyltransferases (DNMT1, DNMT3A and DNMT3B), plays a critical role in mammalian development and physiology (Oksana et al. 2004). However, very little is known which type of DNA methyltransferase (DNMT1, DNMT3A or DNMT3B) would play a prominent role in the mechanism of H19 binding to SAHH involved in gene methylation. The genotoxic potency of polycyclic aromatic hydrocarbons (PAHs) is driven by excessive reactive oxygen species (ROS) production generated in the redox-cycling processes of their metabolic activation which overwhelms their elimination, and thus causing oxidative DNA lesion (Idowu et al. 2019). Base excision repair (BER) is responsible for repairing oxidative lesion (Whitaker et al. 2017). Eight-oxoguanine DNA glycosylase 1 (OGG1) acts as the key enzyme participated in the first step of the BER pathway (Gonçalves et al. 2017). Gene methylation, the important form of epigenetic modification in eukaryote genome, is known to regulate DNA damages caused by PAHs (Kanduri, 2015). Increased methylation of OGG1 leads to the accumulation of DNA damage and further mutations (Alvarado-Cruz et al. 2017). Although the aspects of genetic alterations in population exposed to PAHs are well-documented (Moorthy et al. 2015), abnormal epigenetic regulation is still poorly understood.
Therefore, in order to explore effects of PAHs exposure on peripheral blood H19 RNA expressions and plasma SAHH activity, we performed a cross-sectional observational study among occupational workers. Moreover, benzo[a]pyrene (BaP), a potent carcinogen of PAHs, can cause a wide range of cell cycle perturbations, including S-phase accumulation, diminished capacity for DNA replication and inhibition of cell proliferation. We hypothesized that H19/SAHH/DNMTs involved in OGG1 methylation, oxidative DNA damage and cell cycle arrest in human after BaP exposure. To address our objectives, we used human lung epithelial cell lines treated with BaP as a model and expected to illuminate the molecular mechanism of H19/SAHH/DNMTs regulating OGG1 methylation, oxidative DNA damage and cell cycle arrest by carcinogen BaP. Firstly, the RNA immunoprecipitation (RIP), co-immunoprecipitation (Co-IP), lncRNAs fluorescence in situ hybridization (FISH) and immunofluorescence analysis were used to detect the interaction between H19, SAHH and DNMTs; and then, the activity and expression of SAHH and DNMTs were detected by western blot and enzyme-linked immunosorbent assay (ELISA); finally, the level of OGG1 methylation, oxidative DNA damage and cell cycle arrest was detected to reveal that H19 plays an important role in BaP-induced alterations by regulating the activity and expression of SAHH and DNMTs.

Participants
A total of 146 occupational workers, who had worked for at least 1 year in a coke-oven plant in Taiyuan (Shanxi, China), were enrolled in the final analyses. And the workers who had worked on top-, side-and bottom-oven in the cokeoven plant were mainly young and middle-aged (youth age defined by WHO-2017: 15-44 years old). Exclusion criteria included pulmonary diseases, syphilis, cancer, kidney diseases and other factors which may interfere with research. And the participants were not exposed to known occupational carcinogens or dust during the past 3 months. Our research was approved by the Ethical Review Board of the School of Public Health, Shanxi Medical University, Shanxi, People's Republic of China (No. 2019LL116). Peripheral blood (5 mL) and morning first urine (20 mL) samples from all study population were collected in the morning of the same day. Plasma, which would be used to test the levels of SAM and SAH, was separated within 4 h after collection and then stored at − 80 °C.

Plasma SAHH activity
After pre-treated by 240 µL trichloroacetic acid (400 g/L) with an ice bath for 30 min, 1.2 mL plasma was centrifuged at 14,000 rpm at 4 °C for 20 min, and then supernatant was separated and filtered. HPLC-ultraviolet detection was used to measure plasma SAM and SAH concentrations according to a method presented elsewhere (Illamola et al. 2019). We quantified plasma SAM and SAH concentrations in accordance with the peak area with standard solutions of 0, 0.05078, 0.10156, 0.20312, 0.40625, 0.81250, 1.62500 and 3.25000 mg/L for SAM and 0, 0.03125, 0.06250, 0.12500, 0.25000, 0.50000 and 1.00000 mg/L for SAH to establish the retention time. The quality control data were detailed in Table S1.

Cell treatments
Eight types of cells (WT, si-H19, si-SAHH, si-DNMT1, si-H19 + si-SAHH, si-H19 + si-DNMT1, si-SAHH + si-DNMT1 and si-NC) were treated with BaP (Sigma-Aldrich, USA) at a single dose of 16 μM for 24 h to reflect distinct mechanisms. The dose and time in this study were selected based on our preliminary experiments, which suggested that cell viability decreases less than 15% (Yang et al. 2009).

RNA extraction and real-time quantitative reverse transcription PCR (RT-PCR)
TRIzon Reagent (CWbiotech, Beijing, China) was used to isolate total RNA from participants' peripheral blood or cell lines and then cDNA was synthesized by the use of Prime Script RT Master Mix (Takara, Dalian, China) according to the method described in previous researches (Fu et al. 2018). Quantitative RT-PCR was also performed as described previously. Table S2 lists the RT-PCR primers. The primers for β-actin, H19, SAHH, GAPDH and DNMT1 were used as previously reported (Fu et al. 2018). The primer for OGG1 was obtained from QuantiTect Primer Assays.

RNA immunoprecipitation analysis
RNA immunoprecipitation (RIP) experiments were performed as described previously (Fu et al. 2018). Reverse transcription was in a 30 mL reaction, followed by RT-PCR (see Table S2 for primer sequences).

DNMT1 activity assay
The DNMT1 activity of cells was quantitated following the manufacturer's protocol (EpiQuik™ DNA Methyltransferase Activity/Inhibition Assay Kit, Epigentek, Germany) and was determined by monitoring the fluorescence product at 450 nm using SpectraMax M2.

Co-immunoprecipitation analysis
The extractions of proteins were obtained from cells using the same method as the previous research from cells (Fu et al. 2018). For interactional assay of SAHH and DNMT1, identical amounts of SAHH in extractions of cells and equal volume of mouse anti-DNMT1 antibody (GeneTex) were added in 1.5-mL tubes and rotated overnight at 4 °C. Then 40-μL protein A/G beads (CWbiotech) were added and incubated with rocking for 4 h at 4 °C. Protein A/G beads and protein bound to the beads were washed three times with 1 mL of ice-cold PBS to reduce the non-specific binding to the beads. Finally, the proteins bound to beads were analysed by western blot and probed for SAHH using rabbit anti-SAHH antibody.

DNA extraction and gene methylation analysis
Genomic DNA extraction from cultured cells was performed using a Universal Genomic DNA Kit (CWbiotech) following the manufacturer's instructions. Genomic DNA (500 ng) was subsequently treated with sodium bisulphite by EZ DNA Methylation-Gold Kit (Zymo Research, USA) and amplified by PCR following the manufacturer's instructions. These PCR amplifications were isolated using the Pyrosequencing Work Station and sequenced by Pyromark Q96 MD pyrosequencing instrument (Qiagen, Germany) per the manufacturer's instructions. The dispensation order is 5′-ATC GAG ATG GCT GAT CGA TGT CGG TTG ATG TAT CGT -3′, and the analytic sequence of OGG1 is 5′-YGG AGA ATT GGG GTA YGA AGY GGG GTT TTG ATG ATT YGTA-3′. We detected DNA methylation levels at four CpG loci of OGG1 in the first exon (Fig. 4a, left top).

Oxidative DNA damage assessment
Total genomic DNA was extracted from cells as described previously. DNA was hydrolysed with nuclease P1 in a water bath at 37 °C for 20 min, and then regulated pH to 7.5-8.5 by 1 M Tris. Hydrolytic DNA was incubated with alkaline phosphatase (1 unit/100 μg DNA) at 37 °C for 30 min, and then boiled for 10 min. The content of 8-OHdG was measured following the manufacturer's protocol (8-OHdG ELISA Kit, Abcam), and then the absorbance of samples was detected at 450 nm using SpectraMax M2.

Cell cycle analysis
Cells were collected into 1.5-mL centrifuge tubes for flow cytometry by propidium iodide (PI) staining using DNA Content Quantitation Assay (Cell Cycle) (Solarbio) per the manufacturer's instructions. Flow cytometry was carried out by the same method as the previous research (Andres et al. 1998).

Statistical analysis
Statistical representation of data was done by Statistical Package for Social Sciences version 22.0 (IBM SPSS, Chicago, IL, USA) software and SAS software version 9.4 (SAS Institute Inc., Cary, NC, USA).
Continuous data in vitro cell experiments were summarized as mean ± SD of at least three replicates and compared among means using ANOVA. Post hoc Dunnett analyses were used to analyse the differences between groups when the P value of the ANOVA was < 0.05. Covariance analysis was used to clarify whether there were differences among different types of cells independent of exposure. P values < 0.05 were deemed to be statistically significant.
We substituted the concentrations below the LOD with 50% LOD, including four urinary PAHs metabolites, SAM and SAH. Continuous data in the cross-sectional study were expressed as medians (Med) and interquartile ranges (IQR), and compared with Kruskal-Wallis H test. Categorical data in the cross-sectional study were expressed as N (%) and compared with chi-square test. To investigate the correlations and estimate the potential dose-repose relationship, we used logistic regression models adjusted confounders and restricted cubic spline model, respectively. By assigning the median value in quartiles as a continuous variable, the linear trend across 1-OHP concentration increasing with the quartile can be estimated. The cut points for peripheral blood H19 RNA expressions and plasma SAHH activity were defined as 14.42 and 0.83 in the present study, which were equivalent to the median.

Results
Effects of PAHs exposure on peripheral blood H19 RNA expressions and plasma SAHH activity among 146 occupational workers Table 1 summarizes the basic characteristics of study population based on the quartiles of urinary 1-OHP concentrations. There were no significant differences in age, smoking status, drinking status, heating mode and education (P > 0.05); even so, the marginal significances were showed in employment time (P = 0.053). The results indicated that with the rise of urinary 1-OHP levels, the concentrations of urinary 2-NAP, 2-FLU, 9-PHE and ΣOH-PAHs metabolites were also significantly elevated (P < 0.001). In addition, the detailed distributions of peripheral blood H19 RNA expressions and plasma SAHH activity were given in Fig. 1a with violin plots. An elevated median of peripheral blood H19 RNA expressions was showed coupled with the increased quartile of urinary 1-OHP levels (P = 0.479), whereas the median of plasma SAHH activity was slightly reduced (P = 0.844).
The results of collinearity diagnostics showed that there was no collinearity between the four urinary PAHs metabolites, and also indicated that age was not collinear with employment time (Table S3). Figure 1c represents that peripheral blood H19 RNA expressions were significantly increased when 1-OHP levels were lower than 0.09 μg/mM creatinine, and then maintained on the plateau phase while there is a marginal positive linear association between urinary 1-OHP levels and peripheral blood H19 RNA expressions (4th vs. 1st quartile = 3.44, 95% CI: 1.04-11.44, P trend = 0.058) (Fig. 1b). Nevertheless, the plasma SAHH activity was significantly reduced when 1-OHP levels were lower than 0.06 μg/mM creatinine (Fig. 1c), showing a non-linear association between plasma SAHH activity and urinary 1-OHP concentrations (P overall = 0.089, P non-linearity = 0.030) (Fig. 1c). However, the results did not reveal that 2-NAP, 9-PHE, 2-FLU and ΣOH-PAHs metabolites have significant effects on the expression of H19 RNA in peripheral blood and plasma SAHH activity (P > 0.05, Table S4).

H19 binding to SAHH interacts with DNMT1 in BEAS-2B cells exposed to BaP
As H19 interacts with SAHH and shows characteristics of a ribonucleoprotein particle (Fu et al. 2018), we reasoned that H19 binding to SAHH interacts with DNMT1 in BEAS-2B cells treated with BaP. In support of the assumption, we accessed Co-IP experiments to examine whether anti-DNMT1 antibody would be able to immunoprecipitate SAHH from complex. As shown in Fig. 3a, Co-IP analysis showed that a nearly 0.5-fold enrichment of SAHH in the DNMT1 antibody complex compared with WT cells. In addition, compared with the WT and si-NC cells, SAHH protein expression in si-H19 cells was significantly increased after BaP exposure (Fig. 3a, right column 3). However, the enrichment of SAHH in DNMT1-containing complex between WT and si-NC cells was similar (Fig. 3a, left column 2 and right column 2). Additionally, co-localization between endogenous H19, SAHH and DNMT1 was evaluated by confocal microscopy, predominantly in nucleus and perinuclear cytoplasm (Fig. 3b, white arrow). We found that the fluorescence intensity of DNMT1 was increased after BaP exposure for 24 h (Fig. 3b, column 4). The intracellular co-localization of H19 and SAHH was consistent with the previous results (Fig. 3b, column 5). The increased fluorescence of purple intensity, which displayed the intracellular co-localization of H19 and DNMT1, indicated a stronger steric interaction after BaP exposure for 24 h, while fluorescence intensity was decreased in si-H19 cells (Fig. 3b, column 6). In addition, the increased fluorescence intensity of SAHH and DNMT1 indicated a stronger steric interaction after BaP exposure for 24 h and transfected with si-H19 (Fig. 3b, column 7). Moreover, we also observed the increased fluorescence of blue violet, which displayed the intracellular co-localization of H19, SAHH and DNMT1, indicated a stronger steric interaction after BaP exposure for 24 h (Fig. 3b, column 8).

H19/SAHH/DNMT1 affects oxidative DNA damage and cell cycle arrest
Given the previous study conducted by our team indicating that OGG1 methylation mediated oxidative DNA damage and cell cycle arrest associated with urinary 1-OHP concentrations in coke oven workers (Fu et al. 2019), we took an in vitro cell approach to examine whether H19/SAHH/ DNMT1 would regulate oxidative DNA damage and cell cycle arrest in accordance with the alteration of OGG1 methylation for BaP exposure. Figure 4c and d exhibits the effects H19/SAHH/DNMT1 on the oxidative DNA damage and cell cycle arrest. After transfection, there was no significant change in the levels of 8-OHdG compared with WT cells (Fig. 4c, left column). However, the 8-OHdG levels were significantly increased in cells treated with BaP (Fig. 4c, right column). After BaP exposure, H19 single knockdown slightly reduced 8-OHdG levels (Fig. 4c, right column 3), whereas SAHH or DNMT1 single knockdown significantly elevated 8-OHdG levels (Fig. 4c, right columns 4 and 5). After BaP treatment, SAHH/DNMT1 double knockdown further dramatically increased 8-OHdG levels (Fig. 4c, right   Fig. 1 Associations between urinary 1-OHP and peripheral blood H19 RNA expression, plasma SAHH activity (n = 146). a A violin plot showed the distributions of peripheral blood H19 RNA expression and plasma SAHH activity. P values of the numerical variable were calculated by Mann-Whitney U test. The red line represents mean, the blue line represents median, the blue indicates the range from 25 th percentile to 75 th percentile, and the wathet blue indicates the range from 5th percentile to 95th percentile. b Effects of urinary 1-OHP on peripheral blood H19 RNA expression and plasma SAHH activity. Q, quartile. Data were presented as odds ratio (OR) and 95% confidence interval (95% CI). Model 1: adjusted for employment time, education, age, drinking status, smoking status and heating mode. Model 2: additional adjusted for urinary 2-FLU, 2-NAP, 9-PHE. The median of blood H19 RNA expression as cut point: < 14.42 (0), ≥ 14.42 (1). The median of plasma SAHH activity as cut point: < 0.83 (0), ≥ 0.83 (1). c Restricted cubic spline model for peripheral blood H19 RNA expression and plasma SAHH activity across urinary 1-OHP levels. Adjusted for employment time, education, age, drinking status, smoking status, heating mode, 2-FLU, 2-NAP and 9-PHE. The median of peripheral blood H19 RNA expression as cut point: < 14.42 (0), ≥ 14.42 (1). The median of plasma SAHH activity as cut point: < 0.83 (0), ≥ 0.83 (1) ◂ Fig. 2 H19 interacts with SAHH at specific sites and then regulates DNMT1 in BEAS-2B cells exposed to BaP. a RIP analysis was performed to test the interaction between SAHH protein and H19 mRNA. The relative fold enrichment of H19 in SAHH-containing RNPs of WT cells without BaP treatment was set as 1. The results were representative of three independent experiments with similar results (n = 6). *P < 0.05: compared to the relative fold enrichment of H19 in SAHH-containing RNPs of WT cells without BaP treatment. NS: P > 0.05. b After transfection, RT-PCR was performed to detect H19 RNA levels. The results were representative of three independent experiments with similar results (n = 9). *P < 0.05: compared to WT cells without BaP treatment. NS: P > 0.05. c Western blot was performed to detect DNMT1 protein expression. The numbers at the bottom of the blots show DNMT1 protein expressions by using β-tubulin as an invariant internal control. The results were representative of three independent experiments with similar results (n = 9). *P < 0.05: compared to WT cells without BaP treatment. d DNMT1 activ-ity levels in five types of BEAS-2B (WT, si-NC, si-H19, si-SAHH, si-H19 + si-SAHH) without or with 16 μM BaP treatment for 24 h. *P < 0.05 vs. WT cells treated without BaP. #P < 0.05 vs. WT cells treated with 16 μM BaP for 24 h. NS, not significant. e DNMT3A protein expression exposed to no or 16 μM BaP for 24 h. Numbers underneath the blots indicate DNMTs protein expressions after normalization against β-tubulin loading controls with that in WT cells without BaP treatment arbitrarily set as 1. The results were representative of three independent experiments with similar results (n = 9). *P < 0.05: compared to WT cells without BaP treatment. f DNMT3B protein expression without or with 16 μM BaP treatment for 24 h. Numbers underneath the blots indicate DNMTs protein expressions after normalization against β-tubulin loading controls with that in WT cells without BaP treatment arbitrarily set as 1. The results were representative of three independent experiments with similar results (n = 9). *P < 0.05: compared to WT cells without BaP treatment Fig. 3 H19 binding to SAHH interacts with DNMT1 in BEAS-2B cells exposed to BaP. a Co-immunoprecipitation was performed to measure the binding of endogenous SAHH and DNMT1. Levels of Co-IP-purified SAHH were determined with western blot and presented as fold enrichment in anti-DNMT1 relative to input SAHH. Relative fold enrichment of SAHH in DNMT1-containing proteins of WT cells without BaP treatment was set as 1. The results were representative of three independent experiments with similar results (n = 5). IgG was used as a control. *P < 0.05 vs. WT cells treated without BaP. b Co-localization studies between endogenous H19, SAHH and DNMT1. Representative images showed the localization of nucleus (DAPI, sky blue), H19 (red), SAHH (green) and DNMT1 (blue). Orange spots represent the colocalization between H19 and SAHH. Purple spots represent the colocalization between H19 and DNMT1. Aquamarine blue spots represent the colocalization between SAHH and DNMT1. Blue violet spots represent the colocalization between endogenous H19, SAHH and DNMT1 column 8), whereas H19/SAHH and H19/DNMT1 double knockdown abrogated this effect (Fig. 4c, right columns 6  and 7). Furthermore, the results of S phase arrest were parallel to 8-OHdG levels (Fig. 4d).

Discussion
In this cross-sectional study, we found that urinary 1-OHP levels were positively associated with peripheral blood H19 RNA expression in occupational workers, while urinary 1-OHP levels were negatively associated with plasma SAHH activity. Using human lung epithelial cell lines (BEAS-2B) with BaP treatment as a model, we observed that H19 binding to SAHH exaggerate DNMT1 expressions and activity with BaP treatment. Suppression of H19 enhanced the interaction of SAHH and DNMT1 induced by BaP. And H19/SAHH/DNMT1 may regulate OGG1 methylation, oxidative DNA damage and S phase arrest in BEAS-2B cells exposed to BaP. As far as we know, this represents the first case that a H19/SAHH/DNMT1 axis involving in OGG1 methylation, oxidative DNA damage and cell cycle arrest by carcinogen PAHs both in human and cells.
LncRNA H19 has been closely related with human several diseases in recent studies (Chen et al. 2020;Wang et al. 2020), and its upregulated expression profile has been observed in a variety of human malignancies (Wang et al. 2019). Numerous studies found that PAHs are known to trigger lung cancer in animals and humans (Petit et al. 2019). 1-OHP has been widely used as a biomarker of PAHs exposure and a urinary metabolite of pyrene in occupationally exposed population, which accounts for 1.2-13.4% of PAHs The schematic diagram shows that there are four CpG loci in the first exon of OGG1 which we selected. Heat map represents the coefficient of r of the Spearman correlation between the mean of four CpG loci and each locus in OGG1. b H19/SAHH/DNMT1 regulates OGG1 methylation in cells treated with BaP. Relative OGG1 methylation was presented with that in WT cells without BaP treatment arbitrarily set as 1. The results were representative of three independent experiments with similar results (n = 6). *P < 0.05: compared to WT cells without BaP treatment. c H19/SAHH/DNMT1 affects oxidative DNA dam-age in cells treated with BaP. Relative 8-OHdG levels were presented with that in WT cells without BaP treatment arbitrarily set as 1. The results were representative of three independent experiments with similar results (n = 9). *P < 0.05: compared to WT cells without BaP treatment. d H19/SAHH/DNMT1 affects cell cycle in cells treated with BaP. Relative cell cycle ratio (S phase) was presented with that in WT cells without BaP treatment arbitrarily set as 1. The results were representative of three independent experiments with similar results (n = 10). *P < 0.05: compared to WT cells without BaP treatment exposure in coke-oven plant (Olujimi et al. 2018). Additionally, H19 binds to and inhibits SAHH function in population epidemiological study and cell experiment (Zhou et al. 2015;Oksana et al. 2013). In this cross-sectional study, a positive correlation was also observed between urinary 1-OHP levels and peripheral blood H19 RNA expression in occupational workers, while a negative correlation between urinary 1-OHP levels and plasma SAHH activity was observed. Our previous vitro cell experiments exposed to BaP also yielded the consistent results (Fu et al. 2018).
One-carbon metabolism comprises complex biological networks in which input nutrients are processed through a series of chemical reactions to cycle carbon units. The produced metabolites are then made available for important processes including cellular biosynthesis, methylation, regulation of redox status and aa homeostasis (Fe Dra et al. 2017). Essentially, one-carbon metabolism involves three pathways: the folate and methionine cycles, and the transsulfuration pathway. The folate and methionine cycles overlap upon the synthesis of methyltetrahydrofolate (MTHF) necessary for the generation of methionine through methylation of homocysteine (Tibbetts and Appling 2010). Methionine is then converted into the fundamental metabolite SAM, the universal cellular methyl donor required for DNA, RNA, protein and lipid methylation. SAHH is an enzyme, which catalyses the hydrolysis of SAH which is formed after the donation of the methyl group of SAM to a methyl acceptor in methylation reaction (Li et al. 2013;Oksana et al. 2013). Human SAHH forms a homotetramer consisting of 432 aa and is composed of four identical chemically identical and functionally equivalent subunits (Porcelli et al. 2000), each with three domains: a small C-terminal domain (386-432 aa), a cofactor-binding domain (197-351 aa) and a substratebinding domain (1-181 aa and 355-385 aa) (Beluzić et al. 2008). Nonetheless, we know little about the aa residues involved in the catalytic mechanism of SAHH. And then, WT-SAHH (1-432 aa), M1-SAHH (1-150 aa), M2-SAHH (151-300 aa) and M3-SAHH (300-432 aa) mutant versions of SAHH were constructed to explore association between aa sites of SAHH catalytic activity and the function of lncRNA H19. Whether exposed BaP or not, we observed an accessorial enrichment of H19 in SAHH-containing RNPs in WT-SAHH, M1-SAHH and M2-SAHH cells, whereas M3-SAHH cells did not detectably elevate. Furthermore, H19 RNA expressions were inhibited in SAHH-overexpressing cells compared to WT cells, particularly in M3-SAHH cells. Collectively, H19 may bind to the 1-300 aa chain of SAHH to regulate oxidative DNA damage and cell cycle arrest exposed to BaP.
Otherwise, SAHH offers a single way in which catalyses the reversible SAH hydrolysis to relief from SAM-dependent methylation inhibition in eukaryotes (Oksana et al. 2013). Gene methylation, maintained strictly by the action of three DNMTs (DNMT1, DNMT3A and DNMT3B) in normal cells, is a key controller in a vast array of biological processes, such as DNA replication and repair (Oksana et al. 2004). Thus, the increase of SAHH activity would affect SAM-dependent DNMTs, which leads to the change of gene methylation. However, not all DNMTs are sensitive to SAHH. In one report, H19/SAHH/DNMT3B circuit regulates genome-wide methylation in human tumour cell models (Zhong et al. 2017), whereas Ponnaluri et al. demonstrated that SAHH could enhance DNMT1 activity and interact with DNMT1 during S-phase in vitro cell experiment (Ponnaluri et al. 2018). In this experiment, we found that whether treated cells with BaP or not, H19 or SAHH single knockdown and H19/SAHH double knockdown attenuated DNMT3A and DNMT3B protein expressions. However, H19 or SAHH single knockdown and H19/SAHH double knockdown attenuated DNMT1 protein expressions and activity after BaP exposure. Moreover, the interaction of SAHH with DNMT1 would be strengthened by BaP and inhibition of H19 enhanced the interaction of SAHH and DNMT1 induced by BaP. Furthermore, co-localization between endogenous H19, SAHH and DNMT1 was observed in nucleus and perinuclear cytoplasm. Our investigations revealed that H19/SAHH might regulate BaP-induced abnormal gene methylation via binding to DNMT1 and exaggerate its expression and activity.
PAHs have attracted much attention due to their carcinogenicity and its metabolism mechanisms occurred by cytochrome P450-mediated oxidase system may lead to oxidative DNA damage (Campo et al. 2020) and abnormal cell cycle distribution, especially S phase arrest (Andrysik et al. 2007). BaP, a potent carcinogenic PAH, can cause cell cycle block, including elevated S-phase cells ratio, weakened DNA replication capacity and inhibited cell proliferation (Hruba et al. 2010). Accordingly, we conducted the vitro cell experiment with BEAS-2B cells treated with BaP as a simulated condition and hope to provide insights into the epigenetic regulation mechanism of H19/SAHH/DNMT1. Moreover, 8-OHdG, a ROSinduced DNA base modification, is a sensitive and stable biomarker in the evaluation of DNA damage by oxidation factors (Davalli et al. 2018). Previous study also examined that a significantly elevated 8-OHdG levels (OR = 2.63, 95% CI = 1.04-6.66) and S phase arrest (OR = 2.76, 95% CI = 1.18-6.45) was associated with high levels of urinary 1-OHP in 385 study population (Fu et al. 2019). In this experiment, we have observed that suppression of H19 reduced oxidative DNA damage and recovered S phase arrest, while suppression of SAHH and DNMT1 shows the opposite trend. It was validated that H19 binding to SAHH interacts with DNMT1 to regulate oxidative DNA damage and cell cycle arrest in human lung epithelial cell lines after BaP exposure.
Further, BER is thought to be the vital guardian pathway participated in the removal of the common oxidative lesion 8-OHdG, which is initiated by OGG1 involving in the first step of this repair process (Castillejos et al. 2000). OGG1 is bifunctional enzyme: it is able to remove 8-OHdG paired with C and therefore distinguish between 8-OHdG and the vast majority of normal bases (Klungland and Bjelland 2007). Previous studies have observed associations between PAHs exposure with global or gene-specific DNA methylation alterations (Herbstman et al. 2012). Both PAHs exposure and its related damages (e.g. oxidative stress) have been associated with DNA methylation alterations (Herbstman et al. 2012). Moreover, exposure to some PAHs in PM 10 has reported to show a higher methylation level in specific DNA repair genes, such as OGG1 and APEX (Alvarado-Cruz et al. 2017). Findings from the previous reports were consistent, which workers highly exposed to PAHs had an OGG1 Pos.4 hypermethylation in comparison with the low-exposed employees (Hernandez-Cortes et al. 2018). Based on the aforementioned information, it led us to be curious about the role of OGG1 methylation regulated by H19/SAHH/ DMNT1 plays in oxidative DNA damage and cell cycle arrest related to BaP exposure. As expected, the significant rise occurred in OGG1 methylation with BaP treatment, whereas the decrease was observed when H19 was downregulated. Conversely, SAHH or DNMT1 single knockdown exacerbated OGG1 methylation. Interestingly, although H19 single knockdown obviously reduced OGG1 methylation, H19/SAHH and H19/DNMT1 double knockdown abrogated this effect while SAHH/DNMT1 double knockdown aggravated the alteration. It is further confirmed that H19/SAHH/ DNMT1 can enhance the suppression of SAM-dependent biological methylation in BaP-treated cells, thereby reducing OGG1 methylation. As stated above, oxidative DNA damage and S phase arrest were consistent to OGG1 methylation. It also prompted that the alterations OGG1 methylation may be related with oxidative DNA damage and S phase arrest in BaP-treated cells. These parallel results indicated that H19/ SAHH/DNMT1 plays a critical role in oxidative DNA damage and cell cycle arrest by PAHs.
In summary, our results support the hypothesis that H19/ SAHH/DNMT1 axis may be involved in OGG1 methylation, oxidative DNA damage and cell cycle arrest by carcinogen BaP. The results in this study suggest that the possible oxidative DNA damage and cell cycle arrest might reciprocally influence each other and form a vicious cycle, leading to greater severity DNA damage and even cancer. Further research should be aimed at a more detailed study of the specific mechanisms by which methylation occurs and plays a role in DNA damage.
Further, we have only studied oxidative DNA damage and cell cycle arrest in human lung epithelial cell lines, but have not shown any results on reactive oxygen species and χH2AX, which are the main markers of oxidative DNA damage. So, we will continue to have studied the effects of the parameters on ROS levels and expression of χH2AX in BEAS-2B cell line and lung cancer cell line, and published the research results in succession.

Conclusions
Urinary 1-OHP, the vital urinary metabolite for PAHs exposure, was positively associated with peripheral blood H19 RNA expression, while was negatively associated with plasma SAHH activity in occupational workers. BaP, the human carcinogen PAH, contributed to the binding among H19, SAHH and DNMT1, which synergistically involved in regulating OGG1 methylation, oxidative DNA damage and cell cycle arrest in human lung epithelial cell lines. Hence, we concluded from the study that H19/SAHH/DNMT1 plays an essential role in the OGG1 methylation, oxidative DNA damage and cell cycle arrest.