The Novel Role of Arsenic (+3 oxidation state) Methyltransferase in Arsenic Genotoxicity

Background: Arsenic (+3 oxidation state) methyltransferase (AS3MT) is the key enzyme in methylation metabolism of arsenic. It is closely related to DNA methylation, but little is known about the novel molecular mechanisms. Methods: 79 workers and 41 individuals in the control group were recruited. Arsenic, relative indexes, 28 relative RNAs, and base modications of exon 5-8 of p53 were detected. Enzyme linked immunosorbent assay(ELISA) was performed to detect the expression of AS3MT protein in all subjects. A series of methods were used to analyze the relationships between them. The AS3MT protein was detected in A549 and 16HBE cells after treated using sodium arsenite, MMA and DMA for 48 hours. Small interfering RNA (siRNA) transfection was used to investigate the role of AS3MT in arsenite-induced tumorigenesis. The cell proliferation and apoptosis were assessed with MTT assay, EdU assay, HO/PI double staining and JC-1 assay. The real-time quantitative PCR (qRT-PCR) and Western Blot analyses were used to evaluate the expression of genes. The p53 luciferase reporter gene assay and Co-immunoprecipitation (Co-IP) were used to identify the interactions of target proteins. Results: AS3MT RNA is closely related to p53, a series of ncRNAs and mRNAs, and likely to have causal correlations. Base modications of p53, miR-548 and miR-190 have signicant distinctive effects, but arsenic may play limited roles. AS3MT is over expression in lung cancer patients who have not exposed to arsenic, human lung adenocarcinoma and bronchial epithelial cells with arsenic treatment for 48h. AS3MT protein is induced in arsenic exposed population. Down regulation of AS3MT inhibit proliferation and promotes apoptosis of cells. Mechanistically, AS3MT specically bind with c-Fos, and block the binding ability between c-Fos and c-Jun. Additionally, knockdown of AS3MT mediated by siRNA enhance the phosphorylation level of p53 Ser392 through activating p38 MAPK. These probably lead to activation of Discussion: Here showed that AS3MT RNA plays a great role in the genotoxicity and carcinogenesis which started by arsenic, but inuenced by other factors. Up regulation of AS3MT can directly act on cell, and affect cell proliferation and apoptosis through activation of p53 signaling and up regulation in downstream targets. 16HBE after knockdown of AS3MT. of late apoptosis signicantly in siAS3MT cells. levels of cleaved-caspase3 and cleaved-caspase-7 in siAS3MT cells. the deregulation of apoptosis and/or up-regulation of tumor cell a key role in tumorigenesis the HaCaT cell cycle induced circ100284/ miR-217/ Cyclin p53 activation blocked apoptosis p53 and its down-stream gene p21, Bax, Puma, and Fas were over in A549and 16HBE cells after AS3MT silencing. P21 is a negative regulator of the cell cycle, as p53 binds to the p21 promoter and activates its transcription, which inhibit DNA replication and impede entry into mitosis by binding to PCNA(31, However, previous research revealed that PCNA and CDKs were signicantly low expression in siAS3MT groups So, it been speculated that AS3MT is involved in regulation p53/p21 axis, and promote cell proliferation. Similarly, p53 positively regulates the expression of Bax, Puma and Fas, and promote cells apoptosis(33).These results imply that AS3MT, as a new proto-oncogene, can promote cell proliferation and inhibit apoptosis by p53 signaling pathway.


Background
Arsenic (+3 oxidation state) methyltransferase (AS3MT) is the key enzyme in the methylation metabolism of arsenic (1)(2)(3). It promotes methyl transfer to form methyl arsenate and dimethyl arsenate, which play an important role in the formation and development of toxicity (4). AS3MT is closely relative to DNA methylation. It probably disrupts DNA methylation in genome, especially in promoter regions of some genes which possess important functions (4,5). Scientists used to believe that polymorphisms of AS3MT were associated with special carcinoma by affecting arsenic methylation capacity (6). Previous results indicate that AS3MT is not only involved in arsenic metabolism, but also could regulate the development of tumours (7). However, it is still unknown whether AS3MT could directly act on cells and promotes the growth of arsenic-induced tumours by regulating cell apoptosis and proliferation.
Lung tissue is considered to be the most sensitive site for arsenic toxicity, and long-term exposure to arsenic is closely associated with pulmonary malignant tumor (8). The preliminary investigation indicated chronic exposure to arsenic can increases the risk of lung cancer (9). Reports revealed NaAsO 2 can induce the oncogenic transformation of human lung bronchial epithelial cells via STAT3/miR-21/PDCD4 signaling pathway (10). These results provide a basis for the study of arsenic carcinogenesis process.
The imbalance of cell proliferation and apoptosis is a crucial mechanism of tumorigenesis and progression. P53, as a typical tumor suppressor gene, play an important role in regulating this unbalance state. Studies have shown that p53 mutation or de ciency has been found in about 50% of human malignant tumors (11). P53 abnormal expression is in uenced not only by phosphorylation, but also its transcriptional level. It is regulated by activator protein-1 (AP-1) that mainly composed of Jun and Fos (12,13). Under normal circumstances, p53 protein remains at very low level in cells. When external stimuli, DNA damage or the activation of proto-oncogenes, the expression of p53 is increased and half-life is prolonged(14, 15). There is an increase tendency of p53 in human peripheral blood mononuclear with arsenic treatment and AS3MT knockdown (16). But, the molecular mechanism of how AS3MT regulate the expression of p53 is entirely unknown.
According to the preliminary bioinformatics software analysis and function veri cation experiment in cell culture, combined with the comprehensive analysis of literature reports, it is found that miRNAs, such as miR-548 and miR-190, play an important regulatory role in the process of AS3MT, and affect methylation metabolism of arsenic (17,18). It may lead to genetic toxicity with great differences. The initial role of these RNAs is probably not to target arsenic, but become the main regulatory factor with human exposure to arsenic ceaselessly, and possess the potential to gradually evolve into the major regulatory factor in toxicity and methylation metabolism of arsenic. Such as the remarkable feature of miR-548 is low conservation and high evolution (19). MiR-548c-3p may play an important role in the methylation metabolism and toxic effects of arsenic. It may through adaptive changes of key bases.
In this study, the expression pattern of AS3MT RNA and protein, and in uencing factors were explored in vivo and in vitro, and plan to understand the processes and metabolism involved in carcinogenesis.

Study subjects
It is recruited that total of 76 workers in arsenic plants which were producing As 2 O 3 and 23 farm laborers who resided in villages away from the arsenic plants more than 50 km and had similar living conditions to the arsenic exposure workers. Their demographic characteristics such as age, sex, education, years of service, chronic health problem, smoking and drinking and family medical history were collected by a questionnaire. In obtaining informed consent, 5.0mL venous blood were collected and placed in -80°C freezer for enzyme-linked immunosorbent assay, and used acid-washed tubes to collect 20.0 mL spot urine and immediately transferred to -20°C freezer for subsequent measurement. The project was permitted by the ethics committee of Yunnan center for disease control and prevention.

Reagents and standards
Arsenate (Na 3 AsO 4 ·12H 2 O), arsenite (NaAsO 2 ), HCI, NaOH, and NaBH 4 are purchased from Shanghai Chemical Co. (Shanghai, China). All reagents used in this study are analytical grade and As free (<0.01 mg/L). It was used a mixed As standard of 1000 mg/L methylarsonic acid (MMA) and dimethylarsinic acid (DMA) (Tri Chemical Laboratories Inc., Yamanashi, Japan). Inorganic arsenic (iAs) standard of 1000 mg/L were acquired from the National Center for Standard Reference Materials (Beijing, China). Standard reference material of freeze-dried urine (SRM 2670) for toxic metals was obtained from the U.S. National Institute of Standards and Technology (NIST, Gaithersburg, MD, USA).

Sample collection
Written instructions regarding the hygienic conditions for collection of samples and polyethylene containers which treated with hydrochloric acid and rinsed with deionized water were provided to all participants. Subjects were asked to provide the rst morning void urine. At the same time, blood samples were collected, and total DNA and RNA were extracted in 12 hours.

Determination of As metabolisms
1.0 mL of frozen urine sample was thawed at room temperature and digested with 2N-NaOH at 95˚C for 3 h, followed by dilution with Milli-Q water. Next, the atomic absorption spectrophotometer (AA-6800) with an As speciation pretreatment system (ASA-2SP, Shimadzu Co. Kyoto, Japan) was performed to analyze the concentrations of iAs, MMA and DMA in urine. Arsenic speciation analysis was based on the well-established hydride generation of volatile arsines, followed by cryogenic separation in liquid nitrogen. The detection limit of this method was 2ng. Mg/g creatinine was used to indicate urinary arsenic concentration.
DNA was used to detect base modi cations of exon 5-8 of p53. PCR primers were designed to amplify 4 exones of p53, andβactin sequence (control fragment). High △△Ct (△△Ct=Cti-Cto, Cti is one exon of p53, Cto isβ-actin sequence) means high base modi cation, which related to severe DNA damages.
The total RNA from A549 and 16HBE cells was extracted using the TRIzol® Reagent (Vazyme, Nanjing, China). The cDNA was synthesized using the HiFiScript cDNA Synthesis Kit(CoWin Biotech,Beijing, China), and then analyzed by qRT-PCR in a LightCycler® 96 instrument Real-Time PCR system (Roche Molecular Systems, California,USA) using the SYBR Green PCR Master Mix (CoWin Biotech, Beijing, China). The sequences of primers are listed in Supplementary Table 1. 2.6 Enzyme-linked immunosorbent assayThe level of AS3MT protein was quantitatively detected by ELISA method. The peripheral venous blood samples were collected from arsenic exposed population and control subjects, and tested by using commercial human AS3MT ELISA Kits (Qcheng, Shanghai, China)according to the manufacturer's instruction. The results were measured by the enzyme mark instrument at the absorbance of 450 nm.

Cell culture and treatment
A549 adenocarcinomic cells and 16HBE human bronchial epithelial cells were purchased from the Kunming Institute of Zoology (Kunming, China) in 2019. The 60T-16HBE cells, which were transformed by 16HBE cells exposed to 2.5 µM NaAsO2 for 60 passages, were provided by Dr. Che WJ. A549 cells were cultured in RPMI-1640/10% fetal bovine serum (FBS)/1% penicillin and streptomycin. 16HBE and 60T-16HBE cells were maintained MEM medium supplemented with10%FBS and 1% penicillin and streptomycin. 2.5 ×10 5 cells were resuspended and plated at 6-well plate. After 22h, A549 and 16HBE cells were exposed to different concentrations of NaAsO 2 , MMA and DMA for 48 hours.

siRNA transfection
A549 and 16HBE cells were plated at 1×10 5 per well in 6-well plate and cultured for another 19 hours, then transfect using RFect siRNA transfection reagent (Biogenerating Biotechnologies, Changzhou, China) according to the manufacturer's instruction. The transfection e ciency was measured by observing with uorescence microscope and detecting the AS3MT using qRT-PCR and western blot after transfected for 72 hours. Small-interfering RNA duplexes (siRNA) against AS3MT (siAS3MT#1 and siAS3MT#2) and siRNA control (siCtrl) were purchased from GenePharma (Shanghai,China). The sequences of siAS3MT and siCtrl were listed in Supplementary Table 2.

Cell viability
For NaAsO2 treatment, A549, 16HBE and 60T-16HBE cells were plated at 8000 per well in 96-well plate. After 22h, cells were treated with different concentrations of NaAsO2 for 48h. For transient transfection, A549 and 16HBE cells were plated at 3000 per well in 96-well plate and transfected with siAS3MT for 72h. After that, 10µL of CCK-8 (Beyotime, Beijing, China) was added directly to per well. After 1-2h incubation, the results were measured by the enzyme mark instrument at the absorbance of 450 nm.

Cell proliferation assay
BeyoClickTM EdU Cell Proliferation Kit with Alexa Fluor 555 (Beyotime, Beijing, China) was used to assess the effect of AS3MT silencing on cell proliferation according to the manufacturer's protocol. In brief, A549 and 16HBE cells were plated at 0.75 × 10 5 per well in 6-well plate. Cells transfected after 72 hours, the cells were gently rinsed twice with 1×PBS, and incubated with EdU working solution for 2 h at 37℃ in darkness. Then, the cells were gently rinsed by 1×PBS and xed with 4% paraformaldehyde for 30min at room temperature. After incubating with 0.3% TritonX-100 in PBS, and covered with Click reaction solution. The cell nuclear was stained with Hoechst 33342, and photographed by an inverted uorescence microscope with excitation wavelength at 565 nm and 460 nm.

Cell apoptosisassay
A549 and 16HBE cells were plated at 1 × 10 5 per well in 6-well plate or 3000 per well in 96-well plate and transfected for another 72 hours. Next, cells apoptosis were measured by HO/PI double staining (BestBio,Shanghai,China) and mitochondrial membrane potential(JC-1)(Beyotime, Beijing, China) according to those manufacturer's instructions. For HO/PI double staining, the cells in 6-well plates were gently rinsed twice with 1×PBS, and stained with Hoechst 33342 (10 ng/ml) and PI (10 ng/ml) for 20 min at 26℃ in darkness. Finally, the condensed or fragmented nuclei of apoptotic cells were observed by an inverted uorescence microscope with excitation wavelength at 350 nm and 488 nm. For JC-1 assay, cells in 96-well plates were gently rinsed twice with 1×PBS, and stained with JC-1 staining solution (5µg/ml) for 15min at 37 ℃ in the darkness. The mitochondrial membrane potential was detected by a multifunctional microplate reader with excitation wavelength at 485 nm and 550 nm. Mitochondrial depolarization is described by an increase in the green/red uorescence intensity ratio.

Luciferase reporter gene assay
The p53 transcription activity was measured by the Fire y Luciferase Reporter Gene Assay Kit(Beyotime, Beijing,China) according to the manufacturer's protocols. A549 and 16HBE cells were plated at 1×10 5 per well in 6-well plate. After 18h, the cells were co-transfected with the luciferase reporter pp53-TA-luc(Beyotime,Beijing,China) and siAS3MT using RFect siRNA transfection reagent and RFect Plasmid DNA Transfection Reagent (Biogenerating Biotechnologies,Changzhou,China) respectively. Cells transfected after 72 hours, the cells were harvested and lysed, and protein concentration was normalized by Enhanced BCA Protein Assay Kit (Beyotime, Shanghai, China). Finally, the luciferase activity was measured by multifunctional microplate reader and described as a fold change compared to siCtrl+ pp53-TA-luc group. A549 and 16HBE cells were separately divided into three groups: siCtrl + pp53-TA-luc group, siAS3MT#1+ pp53-TA-luc group and siAS3MT#2+ pp53-TA-luc group.

Protein binding ability assays
The Co-immunoprecipitation assay was used to evaluate binding ability of c-jun to c-Fos. In brief, cells transfected after 72 hours, A549 and 16HBE cells were lysed and incubated with anti-c-Jun antibody or anti-c-Fos antibody. Finally, the c-Fos and c-Jun protein levels were assessed by western blot,and c-Jun and c-Fos were used as an internal control respectively.

Western blot analysis
Total protein was extracted from A549, 16HBE and 60T-16HBE cells using RIPA lysis buffer containing protease inhibitors and phosphatase inhibitors. The protein concentration was normalized using Enhanced BCA Protein Assay Kit. 30µg of protein was subjected to 10% SDS-PAGE, and then electronically transferred to PVDF membranes (Roche, German) via semi-dry transfer method (BioRad, USA).The membranes were blocked with 5xProtein Free Rapid Blocking Buffer (EpiZyme, PS108, China) and incubated with primary antibodies overnight at 4°C. The membranes were washed four times using TBST and then incubated with the corresponding secondary antibodies at room temperature, but Mouse anti-Rabbit IgG light chain for immunoprecipitation was used. The bands were detected by Gel-Pro Analyzer software. Antibodies information is listed in Supplementary Table 3.

Statistical analysis
The concentrations of iAs, MMA and DMA are log-transformed to improve the normality of measures. After assessing the association among lncRNAs, miRNAs, mRNAs, base modi cations of 4 fragments of p53 and three arsenic species were performed for the analysis under different levels of arsenic trioxide by correlation analysis, covariance and independentsamples ttest. Then the association between arsenic species, 4 fragments of p53 and all lncRNAs, miRNAs and mRNAs were investigated. All statistical tests were two-side, with p < 0.05 is considered statistically signi cant. Statistical analyses were performed using SPSS software for above data (Version 23, USA).
Then data analyses were performed using ImageJsoftware, and GraphPad Prism7.0 software. The difference between two groups and multiple group comparisons were analyzed by Student's t-test and one-way ANOVA, respectively. Wilcoxon rank sum test was used for the non-normal distribution of data.
Pearson correlation coe cient was performed to analyze associations between concentrations of different arsenic species and AS3MT expression. The data were described as the means±standard error of the mean (SEM). Differences were considered statistically signi cant at p<0.05.

Results
There are no essential differences in sex, smoking and alcohol consumption, and other factors between the exposed workers and control group in molecular epidemiological study. It did not been found that the number of years a person had worked in the plant affect the results. The workers had always worked and moved around the plant. Based on the characteristic of chemical components of ore and production techniques ow adopted, there are few other occupational hazard factors except for arsenic pollution in the selected plants. 3.3 Correlation analysis between log(iAs), or log(MMA), or log(DMA), or PMI, or SMI and relative lncRNAs, or mRNAs, or miRNAs, or base modi cations of exon5-8 of p53(r p) As shown in Table 1,3., the base modi cations of exon 5 or 6 of p53 is positive correlation with log(iAs) or log(MMA),but base modi cation of exon 8 of p53 is negative correlation with log(iAs) or log(MMA) (p 0.05). The HAIR, MALAT1, mRNAs of trbp, dicer, ago2, base modi cations of exon 6 of p53 are positive correlation with log(DMA) (p 0.05). The mRNAs of lin28b,dicer, ago2, mdm2, bcl2, base modi cations of exon 7 and 8 of p53 are positive correlation with PMI, but MiR-190 is negative correlation with PMI (p 0.05). The MALAT1, mRNAs of dicer, ago2, sox4, mdm2, noxa, bcl2, base modi cations of exon 7 and 8 of p53 are positive correlation with PMI (p 0.05).

The expression of AS3MT is correlated with arsenic exposure
Enzyme-linked immunosorbent assay was performed to assess the expression of AS3MT protein in peripheral blood samples from all subjects. The results reveal long-term exposure to arsenic signi cantly increased AS3MT protein compared with the control group (Supplementary Figure 4A). Furthermore, the inter relationship were evaluated between AS3MT expression and the urinary arsenic species levels. Pearson correlation coe cient shows AS3MT protein is positively linked to tAs, iAs, MMA and DMA (Supplementary Figure 4B-E). All subjects were divided into 2 groups based on PMI or SMI. Results show that the AS3MT protein was higher in high PMI group compared to low PMI group (Supplementary Figure 4F-G). AS3MT in both cell lines treated with NaAsO2 was evidently higher than that of MMA and DMA at the same concentration.

Knockdown of AS3MT inhibit cell proliferation and promote cell apoptosis
The AS3MT was silenced by RFect siRNA transfection reagent. qRT-PCR and western blot analysis results show siAS3MT successfully knocked down AS3MT, and the morphology of the cells were normal (Figure 2A-C). Results imply that these cells suitable for subsequent experiments.
After 72 hours of transfection, the data of CCK-8 assay demonstrated that the cell viability was lower in the siAS3MT groups than in the siCtrl group( Figure 3A), which imply AS3MT knockdown may inhibit cell proliferation and promote cell apoptosis. Furthermore, Edu staining, HO/PI double staining and JC-1 assay were performed to verify the inferences. The results of EdU staining reveal that knockdown of AS3MT evidently inhibite cell proliferation in A549 and 16HBE cells ( Figure 3B-E). Additionally, the number of apoptosis and necrosis signi cantly increase in siAS3MT cells compare to the siCtrl cells, but the late apoptotic cells were clearly observed in the 16HBE cells ( Figure 3F-I). Morover, the mitochondria transmembrane potentials was decreased in siAS3MT groups compared to the siCtrl group ( Figure 3J). Taken together, the low AS3MT inhibit cell proliferation and promotes apoptosis signi cantly.
3.9 AS3MT inhibit p53 phosphorylation through p38 MAPK The p53 were examined by qRT-PCR and western blot in siCrtl and siAS3MT cells. Results show p53 mRNA and protein was higher in siAS3MT cells( Figure 4A). Further experiments found that silencd AS3MT activate p53 downstream genes, such as p21, MDM2, Fas, Puma and Bax ( Figure 4C-D). Then, the phosphorelationd state of p53 Ser15 and Ser392 site was assessed.
Western blot show that only Ser392 site was changed in siAS3MT cells. The p38 MAPK of p53 protein kinase was remarkably increased after silencing of AS3MT ( Figure 4B). It can be speculated that knockdown of AS3MT can promote p53 Ser392 site phosphorylation through activating p38 MAPK.

AS3MT inhibit p53 transcription through competitive binding with c-Fos
The Co-Immunoprecipitation assay was done to study the mechanism that AS3MT affect p53 transcription through regulating AP-1 related proteins, and verify the interaction of two proteins. Results show AS3MT, c-Jun and c-Fos were coprecipitated by each antibody in A549 and 16HBE cells ( Figure 5A). Then, the interaction between AS3MT and c-Jun, c-Fos proteins were examined, which show AS3MT and c-Fos could be reciprocally co-immunoprecipitated in A549 and 16HBE cell lysates. It is almost impossible for AS3MT binding to c-Jun ( Figure 5B). These results suggest AS3MT and c-Jun competitively bond to c-Fos, and inhibit the formation of AP-1 complex. Then, protein binding ability assay after AS3MT silencing were performed to identify whether knockdown of AS3MT increase the binding ability between c-Jun and c-Fos in A549 and 16HBE cells. As shown in Figure 5C, with c-Jun as the internal reference gene, c-Jun could combine more c-Fos in siAS3MT cells than in the siCtrl cells. c-Fos could combine more c-Jun in siAS3MT groups with c-Fos as the internal reference gene. Finally, the transcriptional activity of p53 was measured by the luciferase reporter gene assay, which demonstrat that AS3MT was able to inhibit p53 transcriptional activity ( Figure 5D). In summary, AS3MT can affect the formation of AP1 dimer by competitively binding c-Fos, leading to reduce the transcriptional activity of p53.

Discussion
There exist great individual differences for metabolism transformation and genotoxicity of arsenic (20). In this study, many methods are used to analyze the relationships along various arsenic species and a series of genotoxic markers. There are closely relationships between DMA or PMI or SMI and many genotoxic markers, but between iAs or MMA and genotoxic markers only for base modi cations of p53. It suggests that iAs and MMA became DMA will related to many genotoxic changes. This probably related to AS3MT closely. This study show there may no dose-response relationships between arsenic exposure and most relative RNAs. The content and proportion of arsenic compounds may play limited roles.
In this study, the effect of arsenic on AS3MT protein was examined by western blot to identify the relationship between arsenic and AS3MT. A549 and16HBE cells treated with various doses of NaAsO2 indicate that arsenic has increased AS3MT. It is consistent with previous epidemiological conclusion. AS3MT handled by MMA and DMA was evidently lower, compared with NaAsO2. It suggests there exert different toxic effects. Results imply various arsenic species have special metabolic rate in human, and iAs, MMA, and DMA may induce different expression models of AS3MT.
Then, the relationship between the mRNA of AS3MT and relative indicators of genotoxicity were analyzed. The hypothesis is that arsenic and its metabolism compounds are the initial inducing factors for mRNA of AS3MT, but not the main in uencing factors. There are some regulatory networks along p53, relative ncRNAs and mRNAs, which play special and important roles in arsenic inducing health hazards. A certain dose of inorganic arsenic can induce the change of AS3MT. This dose may vary greatly among individuals, and the vast majority of the selected population has reached this dose. The mRNA of AS3MT is mainly affected by factors other than arsenic, but such as the above RNAs regulatory networks, and changes in base modi cations of particular fragments of p53.
It is found that mRNA of AS3MT was closely related to a series of genotoxic indexes. It is particularly noteworthy that the correlation coe cient r is greater than 0.3, even 0.7. There is a great possibility of causality. Combined with other reports, the mRNA of AS3MT probably play a novel and very important role in arsenic inducing genotoxicity and tumorigenesis (3,21).
The miR-548 and miR-190 may regulate the mRNA of AS3MT through base complementary pairing. They are signi cantly negatively correlated with mRNA of AS3MT. Adaptive changes of speci c base in these genes need attention. The results show that miR-190 is signi cantly negatively correlated with most selected mRNAs which related to miRNA production and maturation, and some selected mRNAs which related to p53 and tumor formation and development. MiR-548 is signi cantly negatively correlated with many selected lncRNAs and some selected mRNAs which related to miRNA production and maturation. Comprehensive analysis show that both of them are important factors in the regulation of AS3MT mRNA, but there are obvious differences in the regulation mechanism. The low conservation and high evolution for miR-548 need paid great attention (19,22).
In this study, most of the genes selected are related to p53, even belonged to p53 family (20,23,24). It is found that the changes of base modi cation in different exons of p53 are closely related to the genetic toxicity and health hazards in workers exposed to arsenic (25)(26)(27). Here it show AS3MT RNA play a great role in the genotoxicity and carcinogenesis which started by arsenic, but in uenced by many other factors greatly, such as p53, relative ncRNAs and mRNAs.
In previous study, AS3MT mRNA was over expressed in lung cancer patients who have not exposed to arsenic (28). How AS3MT induced by arsenic could directly act on the cell and affect tumorigenesis and progression need to study. Then, the novel role of AS3MT was explored by knocking down AS3MT in A549 and 16HBE cells. Results indicate that the cell viability and proliferation rate of A549 and 16HBE was depressed after knockdown of AS3MT. Conversely, the number of karyopyknosis and/or late age apoptosis was increased signi cantly in siAS3MT cells. In addition, higher levels of cleaved-caspase3 and cleaved-caspase-7 were  (29). Furthermore, the low dose of NaAsO2 activated the ERKs signal pathways, which depressed p53 activation and blocked apoptosis in human keratinocytes cells (30). Interestingly, the present results show that p53 and its down-stream gene p21, Bax, Puma, and Fas were over expressed in A549and 16HBE cells after AS3MT silencing. P21 is a negative regulator of the cell cycle, as p53 binds to the p21 promoter and activates its transcription, which inhibit DNA replication and impede entry into mitosis by binding to PCNA (31,32). However, previous research revealed that PCNA and CDKs were signi cantly low expression in siAS3MT groups (28). So, it been speculated that AS3MT is involved in regulation p53/p21 axis, and promote cell proliferation. Similarly, p53 positively regulates the expression of Bax, Puma and Fas, and promote cells apoptosis (33).These results imply that AS3MT, as a new proto-oncogene, can promote cell proliferation and inhibit apoptosis by p53 signaling pathway.
In mammalian cells, p53, as a "cellular gatekeeper", is involved in the regulation of cell cycle arrest, metabolism, apoptosis, proliferation, or senescence (34). Studies indicated that more than 50% of malignant tumors patients have the p53 mutation or de ciency, and the p53 is regulated mainly through post-translational mechanisms and/or transcription regulation (11,12). For instance, γ-irradiation and UV light enhance the phosphorylation levels of p53 ser15 through ATR, which can activate p53 and promote cell apoptosis (35); Homeodomain-interacting Protein Kinase 2 phosphorylated p53 Ser46 and induced p53 Lys382 acetylation (36).This study found that p38 MAPK and p53 Ser392 were obviously activated in siAS3MT cells. The up regulation of phosphorylated p53Ser392 require p38MAPK pathway activation, which increase its sequence-speci c DNA-binding activity and regulate the balance between proliferation and apoptosis (37). It could be concluded from these results that AS3MT can activate p53 Ser392 by p38 MAPK and directly involving in cell proliferation and apoptosis regulation. In addition, the transcriptional activity of p53 was regulated by a large number of transcription factors, including YB1 (38), RREB1(39), E2F1 (40) and AP-1(13). Among them, the expression level of AP-1 was decreased in cervical cancer cells with As 2 O 3 treatment(41). In mammalian cells, AP-1 is mainly composed of Jun and Fos. Based on this, the Co-Immunoprecipitation assay was done to verify the interaction of two proteins. The results showed that AS3MT can speci cally bind with c-Fos, and knockdown of AS3MT facilitate the binding ability between c-Fos and c-Jun. These results indicate that AS3MT can speci cally bind with c-Fos and suppress the activity of AP-1, which inhibit p53 signaling pathway and involve in arsenic carcinogenesis.

Conclusions
AS3MT may serve as a novel and very important protooncogene, which may induced by arsenic, but in uenced by p53, relative mRNAs and ncRNAs greatly. Further study found that AS3MT inhibit p53 signaling pathway through inhibiting the activation of p38 MAPK and competitive binding with c-Fos, which mean AS3MT can directly act on cell, and affect cell proliferation and apoptosis.

Consent for publication
All authors are aware of and agree to the content of the paper and their being listed as a co-author of the paper.

Availability of data and material
The data and material used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.   Two-tailed Student's t test, compared with the siCtrl group * *p< 0.01. (C)The silence e ciency of AS3MT protein was estimated by Western blot with β-tubulin as internal control. All experiments were repeated three times independently with similar results, and the data of one representative experiment were used to analyse.  Cleaved-Caspase3and Cleaved-Caspase7 were examined using western blot in A549 and 16HBE cells after AS3MT silencing.