Decreased DNA repair gene expression among individuals exposed to arsenic in United States drinking water

Abstract

Arsenic is well established as a human carcinogen, but its precise mechanism of action remains unknown. Arsenic does not directly damage DNA, but may act as a carcinogen through inhibition of DNA repair mechanisms, leading indirectly to increased mutations from other DNA damaging agents. The molecular mechanism underlying arsenic inhibition of nucleotide excision repair after UV irradiation (Hartwig et al., Carcinogenesis 1997;18:399–405) is unknown, but could be due to decreased expression of critical genes involved in nucleotide excision repair of damaged DNA. This hypothesis was tested by analyzing expression of repair genes and arsenic exposure in a subset of 16 individuals enrolled in a population based case-control study investigating arsenic exposure and cancer risk in New Hampshire. Toenail arsenic levels were inversely correlated with expression of critical members of the nucleotide excision repair complex, ERCC1 (r² = 0.82, p < 0.0001), XPF (r² = 0.56, p < 0.002), and XPB (r² = 0.75, p < 0.0001). The internal dose marker, toenail arsenic level, was more strongly associated with changes in expression of these genes than drinking water arsenic concentration. Our findings, based on human exposure to arsenic in a US population, show an association between biomarkers of arsenic exposure and expression of DNA repair genes. Although our findings need verification in a larger study group, they are consistent with the hypothesis that inhibition of DNA repair capacity is a potential mechanism for the co-carcinogenic activity of arsenic. © 2003 Wiley-Liss, Inc.

Arsenic is well established as a human carcinogen, but the level at which it poses a measurable health risk has been the topic of considerable debate and its precise mechanism of action remains unknown. International studies suggest that exposure to arsenic in drinking water increases the incidence of several cancers.2 Risk assessments using low-dose extrapolation of these and other data predicted that there is a lower but appreciable risk of bladder and other cancers at the more modestly elevated levels of arsenic found in regions of the US. Based on these estimations, the US EPA recently lowered the US drinking water maximum contaminant level (MCL) from 50–10 ppb. This has been controversial, however, as data on arsenic, cancer risk and carcinogenic mechanisms in the US population are sparse. Recent studies, however, suggest a possibly increased cancer risk at low levels of arsenic exposure.3, 4

The precise mechanisms by which arsenic increases cancer risk are not known. We hypothesize that arsenic may act as a human carcinogen, at least in part, through inhibition of DNA repair mechanisms. Cell culture studies have demonstrated that arsenite can reduce nucleotide excision repair capacity, and incision frequency in particular, after ultraviolet radiation.1 Although other DNA repair pathways may be involved, several important forms of DNA damage caused by environmental agents are removed primarily by the nucleotide excision repair pathway.5, 6 These lesions include DNA crosslinks, UV photolesions, and bulky chemical adducts. Arsenic modifies transcription factor expression and activity as well as the methylation status of the cell, affecting expression of a variety of genes.7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18 One plausible explanation for the observed decreases in nucleotide excision repair capacity in arsenic exposed cells is modification of the expression of genes critical to this repair pathway.

To elucidate a potential mechanism for the decreased nucleotide excision repair capacity associated with arsenic exposure, we isolated RNA from the cryo-preserved lymphocytes of individuals in a population based study investigating arsenic exposure and cancer risk,19, 20 and used RT-PCR to determine the mRNA expression of specific DNA repair genes in these samples.

Specifically, we examined samples from a subset of 16 individuals (6 cases, 10 controls) enrolled in a case-control study that identified cases of bladder cancer diagnosed among New Hampshire residents, ages 25–74 years, from July 1, 1994 to June 30, 1998 from the state cancer registry.20 Controls aged 65 and younger were selected using population lists obtained from the New Hampshire Department of Transportation, while controls 65 years of age and older were chosen from data files provided by the Centers for Medicare & Medicaid Services (CMS) based on exposure to extreme levels of arsenic (low or high) in drinking water and availability of a cryo-preserved lymphocyte sample. Through an in-person interview, subjects provided information on demography, lifestyle factors, including smoking, and medical history, current and past household drinking water supplies, daily volume of water consumed from the household drinking water supply as well as biologic specimens (toenail clippings and blood) and household drinking water supply samples.

MATERIAL AND METHODS

Water, toenail and cryo-preserved lymphocyte samples were used to examine the relationship between arsenic exposure and expression of nucleotide excision repair genes. Samples of drinking water were analyzed for arsenic concentration using a Finnigan MAT Corp. ELEMENT High Resolution Inductively Coupled Mass Spectrometer (HR-ICP-MS) equipped with a Precision Glass Blowing Corp. VS-1 Membrane Gas Liquid Separator in the Trace Metal Analysis Core Facility at Dartmouth. All laboratory QA/QC procedures were strictly followed to assure analytical accuracy.

Toenail clipping samples were analyzed for arsenic and other trace elements by Instrumental Neutron Activation Analysis (INAA) at the University of Missouri Research Reactor, using a standard comparison approach as described previously.4 With proper treatment nails are not susceptible to external contamination.4 Toenail arsenic levels were not available for two individuals involved in the study. Therefore, these 2 individuals were excluded from the toenail arsenic analyses (n = 14).

A venous blood sample of approximately 20 ml was drawn from each subject into 2 10-ml heparinized tubes. No later than 24 hr after the blood draw, the lymphocytes were isolated using standard buoyant density centrifugation methods. Lymphocytes were cryo-preserved (−120°C) as needed using freezing media at a controlled rate of 1°C/min. This procedure has previously been demonstrated to ensure approximately 90% viability after thawing.21

Total cellular RNA was harvested from lymphocytes taken from a subset of individuals enrolled in the study using Triazol reagent (Gibco/BRL, Life Technologies, Gaithersburg, MD) according to the manufacturer's instructions and quantitated by spectrophotometric absorbance at 260 nm. Reverse transcription polymerase chain reaction (RT-PCR) was carried out using specific primers and reagents from Ambion (Austin, TX), Promega (Madison, WI) and Amersham Pharmacia Biotech (Piscataway, NJ). Gene specific primers were used to examine expression of nucleotide excision repair genes including: XPA, XPB, XPF, XPG and ERCC1 (Clontech, Palo Alto, CA). Briefly, total RNA (0.5 μg) was reverse transcribed using 100 U M-MLV reverse transcriptase in a mixture with oligo-dT and dNTPs. Samples were reverse transcribed in a MJ Research PTC-100 thermocycler (MJ Research, Watertown, MA) for 60 min at 44°C and the reaction terminated by heating to 95°C for 10 min. cDNAs were then amplified by PCR using Taq polymerase (Promega) and gene specific primers. The PCR was carried out in a volume of 50 μl in the MJ Research Thermocycler under the following conditions: denaturation at 94°C for 20 sec; annealing at 58°C for 30 sec; extension at 72°C for 40 sec. Gene specific primers were tested to determine the optimal number of PCR cycles required to be within the linear range of amplification. Based on this testing, the following number of PCR cycles were chosen: XPA 30 cycles, XPB 28 cycles, XPF 36 cycles, XPG 28 cycles, ERCC1 28 cycles and GAPDH 28 cycles. PCR products were run on 2% agarose gels stained with ethidium bromide. Densitometry was carried out on ethidium bromide stained gels electronically using the NIH Image gel plotting macro. The nucleotide excision repair gene mRNA expression was normalized to the housekeeping gene GAPDH by taking the ratios of the gene band density to GAPDH band density as determined by NIH Image.

We grouped subjects by arsenic exposure into higher and lower exposure categories (e.g., drinking water concentration ≥10 μg/L vs. <10 μg/L), by bladder cancer case-control status, or by smoking status (reported ever smoking more than 100 cigarettes in lifetime vs. non-smokers) and normalized nucleotide excision repair gene expression levels assessed by densitometric analysis were compared by Student's t-test. The arsenic cut point was chosen because 10 μg/L arsenic is the revised EPA maximum contaminant level (MCL) for drinking water. Previous analysis comparing drinking water concentrations and toenail arsenic levels in our study indicated that subjects who consumed drinking water containing 10 μg/L arsenic had toenail arsenic levels of about 0.2 μg/g of toenail tissue.22 We used linear regression analysis to evaluate whether ingestion of arsenic in drinking water was associated with decreased expression of DNA repair genes. In this analysis, arsenic exposure was a continuous variable using toenail concentration, drinking water concentration, or daily arsenic intake. Daily arsenic consumption (μg/day) was calculated by multiplying the volume of household drinking water the subject reported consuming per day (L/day) by the arsenic concentration of that water source (μg/L).

RESULTS

We started by comparing 3 different measures of arsenic intake to demonstrate the validity of our arsenic exposure assessment: 1) the concentration of arsenic in drinking water; 2) estimated daily arsenic consumption; and 3) toenail arsenic level for each individual enrolled in the study. Arsenic levels were measured in toenail clippings using instrumental neutron activation analysis, and in drinking water samples using high resolution inductively coupled plasma mass spectrometry (ICP-MS) with hydride generation. Arsenic consumption was calculated based on the daily volume of water consumed from a specified water source and the arsenic concentration in that water. In agreement with previous studies relating toenail to drinking water arsenic,23 Figure 1 demonstrates the correlation between the toenail arsenic level, drinking water arsenic concentration, and daily arsenic consumption in these subjects.

Figure 1. Elevated toenail arsenic concentration is associated with increased drinking water arsenic levels and arsenic consumption. Toenail and drinking water samples were collected from individuals enrolled in a population based, case-control study of bladder cancer in New Hampshire. Arsenic concentrations were determined in toenail clippings using instrumental neutron activation analysis and drinking water samples using high resolution ICP-MS with hydride generation. Arsenic consumption was calculated by multiplying the drinking water arsenic concentration by the amount of water consumed from that source each day. Linear regression analysis indicates a significant rise in toenail arsenic concentration with increasing drinking water and arsenic consumption levels (drinking water r² = 0.67, p < 0.0004, consumption r² = 0.71, p < 0.0002, n = 14).

Using RT-PCR determination of gene expression, individuals exposed to high arsenic levels (drinking water arsenic levels ≥10 μg/L and toenail arsenic concentrations ≥0.2 μg/g) had noticeably diminished expression of the nucleotide excision repair genes ERCC1, XPF, and XPB compared to those with low levels of arsenic in their drinking water (Fig. 2a). In contrast, expression of XPG and XPA did not vary according to arsenic exposure level.

Figure 2. Exposure to high levels of arsenic (drinking water ≥10 μg/L, toenail ≥0.2 μg/g) was significantly associated with decreased expression of the nucleotide excision repair genes XPB, XPF and ERCC1. a: Total RNA was collected from cryopreserved lymphocytes taken from individuals enrolled in a case-control study of bladder cancer in New Hampshire. mRNA levels of the indicated nucleotide excision repair genes, as well as GAPDH was determined by RT-PCR. Subjects were divided into 2 groups based on whether their toenail arsenic concentration was ≥0.2 μg/g or <0.2 μg/g. b: The ratio of nucleotide excision repair gene to GAPDH mean ± SD density of ethidium bromide-stained bands of PCR product in UV-transluminated 2% agarose gels for individuals with high arsenic exposure were compared to those with low exposure (b includes the densitometric analysis of a). High arsenic exposure is defined as a drinking water arsenic level ≥10 μg/L. The graph represents data from a total of 16 individuals (6 with high vs. 10 with low arsenic exposure levels) and significance is designated by ***p < 0.001, **p < 0.01, *p < 0.05).

Densitometric analysis was used to semi-quantitively assess relative expression of nucleotide excision repair genes normalized to the housekeeping gene GAPDH. According to this analysis, the differences in band densities representing expression of ERCC1, XPF and XPB between the high and low arsenic exposure groups was statistically significant (p < 0.0005, 0.005, 0.0001 respectively) (Fig. 2b). In a linear regression analysis, toenail arsenic levels, drinking water arsenic concentrations, and arsenic consumption were inversely correlated with expression of critical members of the nucleotide excision repair incision complex, ERCC1 and XPF, as well as the TFIIH DNA helicase XPB (Fig. 3, Table I). In contrast, expression of XPA and XPG was not significantly correlated to any of the measures of arsenic exposure examined in our study (Table I). Furthermore, bladder cancer case status was not related significantly to expression of any of the nucleotide excision repair genes examined (Fig. 4a). XPG expression was significantly elevated among subjects who reported ever smoking more than 100 cigarettes in their lifetime when compared to non-smokers (p < 0.04), whereas expression of ERCC1, XPF, XPB and XPA were unrelated to smoking status (Fig. 4b).

Figure 3. Expression of ERCC1 is inversely correlated with toenail arsenic level, drinking water arsenic concentration, and arsenic consumption. Total RNA was collected from cryopreserved lymphocytes taken from individuals enrolled in a case-control study of bladder cancer in New Hampshire. mRNA levels of ERCC1, as well as GAPDH was determined by RT-PCR. The ratio of ERCC1 to GAPDH mean ± SD density of ethidium bromide-stained bands of PCR product in UV-transluminated 2% agarose gels was compared to: (a) toenail arsenic concentration (μg/g); (b) drinking water arsenic concentration (μg/L) (log scale); or (c) daily arsenic consumption (μg/day) (log scale) by linear regression (toenail r² = 0.82, p < 0.0001, n = 14, drinking water r² = 0.40, p < 0.008, n = 16, consumption r² = 0.40, p < 0.009, n = 16).

Table I. Summary of r² and p Values Obtained from Linear Regression Analysis of Arsenic Exposure and Expression of the Specified Nucleotide Excision Repair Gene Normalized to GAPDH

Gene	Toenail arsenic (n = 14)		Drinking water arsenic (n = 16)		Arsenic consumption (n = 16)
	r^2	p <	r^2	p <	r^2	p <
XPA	0.29	0.05	0.14	0.16	0.12	0.19
XPB	0.75	0.0001	0.54	0.001	0.52	0.002
XPF	0.56	0.002	0.26	0.04	0.29	0.03
XPG	0.09	0.30	0.0	0.91	0.0	0.85
ERCC1	0.82	0.0001	0.40	0.008	0.40	0.009

Figure 4. Case-control and smoking status in relation to expression of nucleotide excision repair genes. Total RNA was collected from cryopreserved lymphocytes taken from individuals enrolled in a case-control study of bladder cancer in New Hampshire. mRNA levels of the indicated nucleotide excision repair genes, as well as GAPDH was determined by RT-PCR. The ratio of nucleotide excision repair gene to GAPDH mean ± SD density of ethidium bromide-stained bands of PCR product in UV-transluminated 2% agarose gels was compared to (a) bladder cancer case-control status (n = 16) and (b) smoking status (ever smoked more than 100 cigarettes in lifetime vs. never smoker). Significance is designated by *p < 0.05 (n = 16).

DISCUSSION

The decreased nucleotide excision repair gene expression seen in individuals exposed to elevated levels of arsenic (Fig. 2a) may provide a mechanism to explain the impaired DNA repair activity observed previously in response to arsenic treatment in cell culture. Arsenic increased the persistence of DNA lesions induced by mutagens including ultraviolet radiation, methyl methanesulphonate, and benzo(a)pyrene.24, 25 In another study, human fibroblasts exposed to low (2.5 μM) concentrations of arsenite had reduced nucleotide excision repair capacity, particularly incision frequency, after ultraviolet radiation treatment.1

Nucleotide excision repair is a major DNA repair pathway that removes DNA lesions including certain DNA crosslinks, UV photolesions, and bulky chemical adducts.6, 26 The nucleotide excision repair system requires the cooperative function of many gene products for damage recognition, incision, excision, elongation, and ligation to restore DNA structure.26 Although the precise timing and role of each gene product in nucleotide excision repair remain unclear, data shown in Figure 2b indicate that arsenic exposure is associated with significant decreases in expression of genes that are important for the DNA incision and helicase activities. The DNA helicase XPB (3′–5′ polarity) is required for the TFIIH complex to unwind the damaged DNA.26 In addition to its involvement in DNA repair, XPB may also be a critical factor in the p53 mediated apoptosis pathway. Cells taken from individuals with mutations in XPD or XPB show decreased apoptotic response to DNA damage.27 The structure specific endonucleases, XPG (3′ incision) and ERCC1-XPF (5′ incision) cut near the junction of single and double stranded DNA, releasing the damaged oligonucleotide.26 The results of the current study indicate that arsenic exposure is associated with decreased ERCC1 and XPF expression, potentially inhibiting the cleavage process. Other studies have shown that arsenic also inhibits DNA ligase, an enzyme involved in sealing the sugar-phosphate backbone during the final step of nucleotide excision repair.28

There are several possible mechanistic explanations for decreases in gene expression observed in the current study. Exposure to arsenic modifies a variety of transcription factors and signal transduction pathways that are highly dependent on the dose and time course of exposure, including NF-κB, tyrosine kinases, MAP kinases, extracellular regulated kinase (ERK) and p38.10, 14, 15, 16, 17, 18 Sequence analysis of the promoters region of nucleotide excision repair genes shows a high level of homology with several different potential transcription factor binding sites, including XPG: AP-1, CRE-BP1; ERCC1: AML, USF, Cdxa, c-Rel; XPA: SRY, MZF1; XPB: AP-1, AML-1A, GATA-1, NF-κB; XPF: Cdxa, E4BP4, SRY, S8 (www.tfsearch.com). It is unclear, however, whether transcription factors actually bind to these sites and elicit biologically functional changes in gene expression. Previous work in the laboratory indicates that gene expression is the result of the integration of multiple signals in the promoter and cannot often be explained by the effects of a single transcription factor.8 Further studies are needed to determine which transcription factors or co-factors are required to explain the observed decreases in nucleotide excision repair gene expression.

DNA methylation is an important mechanism of transcriptional control because hypermethylation of CpG sites located in the promoter region prevents expression of the adjacent gene, probably by restricting the access of transcription factors to DNA.29, 30, 31, 32 Hyper- or hypo-methylation of the promoter region of certain genes (e.g., p16, p53) occurs after arsenic exposure and could potentially occur in nucleotide excision repair genes, explaining the observed gene expression changes.7, 11, 12, 13 Decreases in mRNA stability and increased mRNA turnover rates are other possible explanations for the observed changes. Because we observed a specific association between arsenic exposure and modification of certain genes, but not others, generalized changes in mRNA transcription, stability and turnover seem less likely.

Nevertheless, association between arsenic exposure and decreased expression of genes involved in nucleotide excision repair provides a plausible mechanism for the previously observed inhibition of nucleotide excision repair capacity by arsenic,1 and may also contribute to its carcinogenic activity. Such a mechanism would predict that the carcinogenic effects of arsenic would be most evident in combination with a DNA damaging agent such as UV irradiation, much as DNA repair deficiencies such as Xeroderma Pigmentosum predispose individuals to increased skin cancer from sunlight exposure. In support of this, inorganic arsenic alone has been reported to be negative as a carcinogen in animal studies, whereas arsenic was recently reported to be strongly co-carcinogenic in the presence of UV irradiation and other mutagen-carcinogens.33, 34

Decreased DNA repair has been associated previously with increased susceptibility to cancer.35 For example, decreased DNA repair activity increases skin cancer susceptibility.21 High stage bladder tumors show decreased mismatch repair activity that correlates with decreased expression of mismatch repair genes in comparison to normal.36 Moreover, decreased expression of repair genes in peripheral lymphocytes has been correlated with prostate, colorectal, head and neck cancer, and glioma incidence in humans.37, 38, 39, 40 In this preliminary analysis, bladder cancer case status was unrelated to the expression of the nucleotide excision repair genes examined among the subset of subjects used for our study (Fig. 4). Future studies will further examine the relationship between both case-control and smoking status and expression of nucleotide excision repair genes among a larger group of study subjects.

It is possible that the ability of arsenic to decrease DNA repair gene expression could be used advantageously in combination with DNA damaging agents for chemotherapy treatment of existing cancers. In support of this idea, expression levels of nucleotide excision repair genes in lymphocytes closely parallel their expression in other proliferating tissues41 indicating that arsenic could impact nucleotide excision repair gene expression in other organs, as well. It is interesting to note that cell lines deficient in ERCC1 and XPF gene expression are extremely sensitive to treatment with DNA crosslinking agents,42 however, the sensitivity of human cancer cell lines to the alkylating agents, cisplatin and melphalan, did not correlate with the mRNA levels of nucleotide excision repair genes in those cells.43

The current study provides evidence from human subjects in the US of an association between drinking water arsenic exposure in the controversial range of 10–50 ppb and molecular changes associated with cancer susceptibility. Results shown in Figure 3 indicate that using toenail arsenic levels as an internal marker of arsenic dose may be a more accurate method of predicting this type of biological response than external measures of arsenic exposure based on the concentration of arsenic in home drinking water.

In conclusion, this and other studies support the hypothesis that arsenic may impair nucleotide excision repair. Further investigation is needed, however, to determine whether the observed decreases in mRNA expression levels result in changes at the protein or enzyme activity level. Although decreased DNA repair may contribute to cancer susceptibility, this may be one of several mechanisms by which arsenic increases cancer risk.