Abstract
Background: Chronic arsenic exposure has been associated with human cancer. The objective of this study was to investigate the effects of arsenic exposure on a DNA nucleotide excision repair gene, ERCC1, expression in human blood cells. Patients and Methods: Water and toenail samples were collected from a total of 327 Inner Mongolian residents for arsenic analysis. Blood samples were collected to determine mRNA expression levels by real-time PCR. Results: The mRNA levels of ERCC1 expression were positively associated with water arsenic concentration (slope=0.313, p=0.0043) and nail arsenic concentration (slope=0.474, p=0.0073). mRNA levels of ERCC1 expression were significantly correlated with those of OGG1, a base pair excision repair gene (r=0.275, p<0.0001). Conclusion: The results showed that mRNA levels of ERCC1 expression were significantly associated with arsenic concentrations in drinking water, implicating the DNA repair response was induced by arsenic exposure.
- Arsenic
- ERCC1
- DNA repair
- drinking water
- NER
- skin hyperkeratosis
Abbreviations: ERCC1, Excision repair cross complementation group-1; NER, nucleotide excision repair; BER, base excision repair; OGG1, 8-oxo-guanine DNA glycosylase; PCR, polymerase chain reaction; ROS, reactive oxygen species.
Chronic arsenic exposure by ingestion of contaminated drinking water occurs worldwide and has become a major public health concern. Areas of China, Bangladesh, India, Chile and parts of the United States such as New Hampshire, Nevada and California are known for the arsenic-contaminated water in groundwater (1-3). Epidemiological studies show that long-term arsenic exposure is associated with wide range of health effects, including high risk of lung, bladder and skin cancer (4, 5). The molecular mechanisms of arsenic toxicity and carcinogenesis are still poorly understood.
One widely accepted mode of arsenic action is its role in oxidative stress, inducing genomic instability (6). Arsenic is known to generate reactive oxygen species (ROS) such as hydrogen peroxide, hydroxyl radical and superoxide anion, which induce a variety of oxidative DNA adducts and DNA protein cross-links, and single-strand DNA and double-strand DNA breaks (7, 8). Arsenic-induced oxidative stress and oxidative DNA damage are evident in various tissue culture cells, animals and humans chronically exposed to arsenic (9, 10). Elevated levels of oxidative stress DNA damage product 8-hydroxy-2′ -deoxyguanosine were found in the skin lesions of arsenic-exposed individuals (9). When these oxidative DNA lesions are not removed by DNA repair mechanisms, in time they can become self-perpetuating mutations that contribute to aging and human degenerative diseases such as cancer and other chronic diseases (11).
DNA damages is repaired by different DNA repair pathways depending on the nature of damage. The nucleotide excision repair (NER) pathway repairs a wide spectrum of structurally unrelated lesions such as UV-induced photoproducts, bulky chemical adducts and certain DNA cross-links (12), whereas base excision repair (BER) typically repairs oxidation and alkylation of nucleobases. Excision repair cross complementation group-1 (ERCC1) is an essential gene in the NER pathway encoding a subunit of endonucleases which make the 5′ incision of DNA damage in NER (13). ERCC1 expression has been shown to correlate with the capacity for repair of DNA damage (14). It has been reported that ERCC1 expression is significantly elevated by the cisplatin treatment of A2780/CP70 human ovarian cancer cell line (15). ERCC1 mRNA levels in liver tissue increased 2-fold in mice treated with inorganic arsenicals (16). These studies indicate that ERCC1 is inducible by DNA-damaging agents such as arsenic.
The present study was conducted in the Bayingnormen (Ba Men) region of Inner Mongolia, China, where the residents have been chronically exposed to arsenic via drinking water (17). The residents of Ba Men are mostly farmers and have been exposed to a wide range of arsenic levels up to 1.8 mg/l for more than 20 years. Arsenic exposure has been associated with health effects including dermal, neurological and cardiovascular effects (17, 18). We selected the Ba Men region as our study site to investigate the health effects of arsenic for the following reasons. More than 80% of Ba Men residents have their own wells. Therefore, it is possible to assess arsenic exposure on an individual basis. The Ba Men residents grow their own grain and livestock and have good nutrition. Ba Men is located in an inland area of China. The residents occasionally consume freshwater fish from the Huang He River and very seldom consume seafood from the ocean (19). A previous report showed that the arsenic levels of the water from the Huang He River are below the Chinese arsenic standard for drinking water (<50 μg/l) (17). We previously reported that elevated 8-oxo-guanine DNA glycosylase (OGG1) mRNA levels in blood cells of the exposed individuals are associated with arsenic concentrations in well water in this population (19). It is known that OGG1 and ERCC1 work in a coordinated fashion and play important roles in maintaining genomic integrity (20). It was reported that the NER pathway is coordinated with the BER pathway to repair DNA damage induced by oxidative stress and sunlight in humans (21). To evaluate the role of NER in arsenic-induced carcinogenesis, we investigated the relationship between arsenic concentrations in well water and ERCC1 expression in the blood of the Ba Men residents exposed to a wide range of arsenic exposure levels. It has been reported that arsenic concentrations in toenails reflect long-term arsenic exposure up to 12 months (22). In our previous study in the same population, arsenic concentrations in well water and toenails were highly correlated (19). In this study, we also analyzed the relationship between ERCC1 expression levels and toenail arsenic concentrations. The relationship between the levels of ERCC1 and OGG1 gene expression was also investigated.
Materials and Methods
Study participants. A total of 327 residents living in the sub-villages of Wulan, Jianshe, Fengchan and Xinyao in Sha Hai Village, Hangjin Hou County and the sub-villages of Miaohao, Xigelian in Sheng Feng Village, Wu Yuan County participated in this study. The residents in these sub-villages have been exposed to a wide range of arsenic levels in drinking water from the artesian wells for more than 10 years. Prior to participant selection, the well water samples of the homes from these sub-villages were collected and analyzed for arsenic concentrations. The study participants were selected according to the following criteria set for study design focusing on the arsenic effects on ERCC1 expression at arsenic levels equal to or lower than 200 μg/l. The criteria for selection were: (i) 70% or more of individuals with arsenic exposure levels from non-detectable to 200 μg/l and 30% with arsenic exposure levels >200 μg/L, (ii) approximately equal number of males and females and 30% of smokers and 70% nonsmokers, and (iii) age from 11-65 years with exposure of at least 5 years' duration. Questionnaires were administered to all participants to obtain demographic information, history of well use, diet, smoking, occupation, pesticide use and medical information. All participants gave written informed consent to participate in this study. This study was conducted according to the recommendations of the World Medical Association Declaration of Helsinki for International Health Research (23). The research protocol met the requirements for protection of human subject certification by the U.S. Environmental Protection Agency.
Water and human sample collection and analysis. Water samples were collected from each participant's home well and sent to the United States on ice packs and stored at -80°C until analysis. Quantification of inorganic arsenic in well water was performed as reported previously (19). Briefly, water samples were acidified to pH 1.4-1.6 with HNO3 and then diluted with 1% HNO3 5-10 times for arsenic analysis using inductively coupled plasma mass spectrometry (ICPMS) as described by Gong et al. (24). The detection limit for arsenic by ICPMS was 0.1 μg/l.
Toenail samples were analyzed for arsenic content by instrumental neutron activation analysis at the Nuclear Services Department, North Carolina State University, Raleigh, North Carolina, USA (25). The detection limit by INAA was 0.012 μg/g for arsenic.
Real-time quantitative PCR. A 2.5 ml peripheral blood sample was collected into a PAXgene blood RNA tube (Qiagen, Valencia, CA, USA), kept at room temperature for two hours and then stored at -40°C until shipping. The blood samples were shipped to the USA in dry ice and stored at -80°C until RNA isolation. Total RNA was isolated from whole blood using PAXgene blood RNA kits according to the manufacturer's instructions (Qiagen). The total amount of RNA was quantified by Beckman spectrophotometer. The ratios of 260/280 of all isolated RNA were between 1.90-2.10. The integrity of RNA was checked by Bioanalyzer 2100 (Agilent Technologies, Walbronn, Germany). There was no degradation of RNA indicated by 18S and 28S RNA in Bioanalyzer analysis. The first stranded-cDNA synthesis was conducted in a reaction mixture of total volume 50 μl containing 1X buffer, 2.5 μM random primers, 5 mM DTT, 500 μM dNTP, 20 unit RNase inhibitor, 200 unit SuperScript III reverse transcriptase (Invitrogen, CA, USA) and 500 ng total RNA. A total of 2 μl of cDNA were used for each real-time PCR measurement in total volume of 50 μl. ERCC1 and β -actin cDNA were PCR amplified in the same tube using ABI PRISM 7700 Sequence Detection System (Applied Biosystems, Foster, CA, USA). ERCC1 (assay ID: HS00157415_m1) and β-actin (assay ID: 4326315E) primers and probes were obtained from Applied Biosystems. Real-time PCR was performed in duplicate with 2X TaqMan Universal PCR Master Mix. ERCC1 mRNA levels were determined by a relative standard curve method using HT1080 cell RNA as standard (User Bulletin #2; Applied Biosystems) and normalized to β-actin mRNA amount assayed in the same tube. β-Actin was chosen as an endogenous control because β-actin mRNA levels showed the least variation amongst 11 candidate genes examined. Pooled blood cDNAs and HT1080 cell cDNA were used as positive control in all assays. RNase-free water and blood cell RNA were used as negative controls in the experiments. A standard curve was included in each real-time PCR assay. To avoid amplification of the contaminating residual genomic DNA, primers and probes were designed around the junction region of two exons so that they are mRNA specific. Ten percent of blood cDNA samples were repeated. The standard deviation on repeated measurements of same blood cDNA in the repeated experiments was <10%, indicating the day to day variation of the assay.
Statistical analysis. Mean ERCC1 expression was compared in subgroups using t-test or F-tests from analysis of variance models (ANOVA). Linear regression models and Spearman and Pearson correlation coefficients were used to evaluate the association between continuous dependent variables (ERCC1 expression) and independent predictor variables. Logistic regression models were used to evaluate the association between binary dependent variables (skin hyperkeratosis) and independent predictor variables. Interaction terms between key covariates and predictors were included in regression models. Multivariate models were constructed, controlling for variables that were related to the outcome or predictor variable. To test for interaction or effect modification, multiplicative interaction terms were between water and nail arsenic and sex, age category and smoking. Wald tests were conducted to test the significance of the interaction terms. To assess the effect of non-linearity, two- and four-degree fractional polynomials were fitted and compared to a linear model using a deviance difference test (26). Statistical analyses were conducted using Stata version 9.2 (StataCorp 2005, College Station, TX, USA).
Results
Study participants. The information on study participants is shown in Table I. The study participants were mostly farmers (79%). Among the 327 participants, 164 were male and 163 were female, with ages ranged from 11 to 61 years (median age of 38). Thirty three percent were smokers and 67% were nonsmokers. The arsenic concentrations of the well water used by the study participants ranged from 0.34 to 826 μg/l, and with a mean exposure of 13 years. Seventy-six percent of participants had been exposed to arsenic at the concentrations of 0.34 to 200 μg/l and 24% to 201-826 μg/l. Most participants (81%) did not drink alcohol or drank less than twice a week. Forty-eight percent reported having used pesticides in the previous 5 years. Most participants (99%) reported eating meat and dairy products, and most consumed freshwater fish occasionally. Regarding skin abnormalities, 31% of study participants showed skin hyperkeratosis, 12% showed depigmentation and 3% showed hyperpigmentation.
ERCC1 expression versus water and nail arsenic. ERCC1 gene expression was positively associated with the arsenic level in drinking water (slope=0.313, p=0.0043) (Figure 1). Non-linear fractional polynomial models did not show improvement over the linear model (data not shown). Arsenic concentrations in well water and toenails were highly correlated (Spearman r=0.8817, p<0.0001). As shown in Figure 2, ERCC1 expression was also positively associated with toenail arsenic concentrations (slope=0.474, p=0.0073). We also evaluated the effects of other factors, including age, sex, smoking, alcohol and pesticide use on ERCC1 gene expression. ERCC1 mRNA expression was associated with water arsenic in the age group of 19-50 years (slope=0.366, p=0.0065) and showed borderline significance in the age group of 50 years and older (slope=0.438, p=0.069). The 11-18 years age group did not show statistical significance (p=0.85). There was significant association between ERCC1 expression and water arsenic exposure (slope=0.38, p=0.0264) in males, and females showed a border-line significant association (slope=0.24, p=0.082). The association was significant for nonsmokers (slope=0.342, p=0.0107). The trend was positive among smokers, although the association was not statistically significant, possibly due in part to lower numbers (N=107; slope=0.242, p=0.203). Alcohol and pesticides use did not affect ERCC1 expression in the study participants. There were no interactions between the arsenic effect on ERCC1 mRNA expression and age, gender or smoking.
ERCC1 versus OGG1 expression. ERCC1 mRNA levels were positively correlated with OGG1 mRNA levels in the study participants (r=0.275, Spearman r=0.275, p<0.0001) (Figure 3). Data for OGG1 mRNA levels were obtained from the previous study (19).
Discussion
In this study, we investigated effects of chronic arsenic exposure on ERCC1 mRNA expression of blood cells in a Chinese population and also linked ERCC1 gene expression to toenail arsenic concentrations of the study participants. The results showed that ERCC1 mRNA levels significantly increased with water and toenail arsenic levels. We also investigated other factors that affect ERCC1 expression. There were no significant differences in arsenic induction of ERCC1 expression due to tobacco smoke or pesticide use. There were some differences in the relationship between arsenic exposure and induction of ERCC1 expression by age, showing a stronger effect in two older groups. This may be due to the older groups ingesting larger quantities of the arsenic-contaminated water and thus being exposed to greater oxidative stress and the associated DNA repair responses than the younger groups. It has been reported that sunlight radiation increases ERCC1 expression in humans (20). In this study, we also found that the male participants showed higher rate of ERCC1 expression increase associated with arsenic exposure than did females. Males in Ba Men usually engaged in more outdoor physical activity than females and consumed more arsenic-contaminated drinking water, as well as being exposed to more UV radiation from sun. The higher exposures in males may induce greater oxidative stress and DNA repair response than in females.
It has been reported that the protective response to environmental toxicants increases gene transcription, protein levels and enzyme activity of DNA repair and oxidative stress response genes (27). This protective response may be responsible for the increased ERCC1 mRNA expression due to arsenic exposure in this study. Arsenic in the range of 0.1-2 μM induces a protective response in human cultured cells (28), whereas it induces down-regulation of DNA repair and oxidative stress genes at levels of arsenic exposure >10 μM (29, 30). Mice treated with arsenic showed elevated expression of DNA repair response genes including ERCC1 mRNA levels (16). Wu et al. reported that ERCC1 mRNA levels were up-regulated in lymphocytes from arsenic-exposed humans with blood arsenic levels between 9.60-46.5 μg/l in Taiwan (31). Andrew et al. reported that arsenic exposure was associated with decreased expression of ERCC1 mRNA and protein in lymphocytes in arsenic-exposed individuals in New Hampshire, U.S.A. and Mexico (29, 32). The disagreement may be due to the differences in water arsenic concentrations in these studies comparing with ours and Taiwanese studies.
The molecular mechanisms of ERCC1 gene induction in arsenic-exposed individuals are not completely known. The transcription activator responsible for the induction of ERCC1 gene expression has not been well established. There is evidence showing that the induction of ERCC1 expression appears to be mediated through transcription factor AP-1, which is a heterodimer of c-Jun and c-Fos protein. AP-1 DNA binding sites are the cis-acting elements of ERCC1. There is an AP-1-like binding site in the 5′-flanking region of ERCC1 (33). It has been reported that cisplatin induces an AP-1 mediated increase in ERCC1 mRNA expression in human ovarian tumor cells. Suppression of c-Jun and c-fos expression by cyclosporine A and herbimycin A block-d the cisplatin-induced increase in ERCC1 mRNA levels (15). It has been reported that arsenite induced c-Jun and c-fos expression and increased AP-1 transactivation activity in tissue culture cells (34). Arsenic has been reported to induce activation of transcription factor AP-1 in vitro and in vivo (35). Furthermore, ROS induced by low concentrations of arsenic have been shown to increase the transcription of AP-1, which results in cell signaling, transcription factor binding to DNA, and subsequent stimulation of cell proliferation (36). Therefore, the observed increase of ERCC1 expression in this study may result from arsenic-induced oxidative stress and followed by increased expression of c-Jun and c-fos and AP-1 activation, which is then followed by enhanced ERCC1 mRNA expression.
Our previous study showed that arsenic exposure increased mRNA levels of the BER repair gene OGG1 in blood cells from this population (19). In this study, we found that the mRNA levels of these two DNA repair genes were significantly correlated. NER and mismatch repair in yeast have been shown to play a role in the repairing of 8-hydroxy-2′-deoxyguanosine (8-OHdG), the main oxidation product of oxidative stress (37). The NER and BER up-regulation mechanisms could mediate global repair generated by the oxidative stress-induced DNA damage caused by arsenic and maintain tolerable levels of DNA damage. In concert with OGG1, the elevated ERCC1 mRNA levels could be interpreted as enhanced DNA repair capacity for eliminating the oxidative stress-induced DNA damage in arsenic-exposed individuals. A similar close association has also been shown between ERCC1 and OGG1 in healthy humans with solar radiation and patients with colorectal adenomas and carcinomas (20, 38). The coordinated expression of ERCC1 and OGG1 indicates that both genes are subjected to common transcription regulation or the same mode of action.
In conclusion, the positive association of ERCC1 expression with chronic arsenic exposure in a human population was shown in the present study, suggesting that arsenic exposure could induce DNA damage repair response. ERCC1 mRNA expression was positively correlated with OGG1 mRNA levels, suggesting a similar mode of action. The present study also demonstrated that the technology of real-time PCR for quantitative gene expression is a valuable tool for human epidemiological studies in assessing health effects of environmental contaminants.
Acknowledgements
We greatly appreciate the staff members of the Inner Mongolia Center for Endemic Disease Control and Research and Ba Men Anti-epidemic Station who provided technical and administrative support during the field study in Ba Men. The research described here has been reviewed by the National Health and Environmental Effects Research Laboratory, U.S. Environmental Protection Agency, and approved for publication. Approval does not signify that the contents necessarily reflect the views and policies of the agency, nor does mention of trade names or commercial products constitute endorsement or recommendation for use.
- Received March 18, 2009.
- Revision received June 3, 2009.
- Accepted June 11, 2009.
- Copyright© 2009 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved