Abstract
Background
Single-nucleotide polymorphisms in genes related to DNA repair capacity and ultraviolet exposure have not been well investigated in relation to skin lesions associated with arsenic exposure. This population based case–control study, of 600 cases and 600 controls, frequency matched on age and gender in Pabna, Bangladesh, in 2001–2002, investigated the association and potential effect modification between polymorphisms in Xeroderma Pigmentosum complementation group D (XPD) (Lys751Gln and Asp312Asn) genes, tendency to sunburn and arsenic-related skin lesions.
Methods
Unconditional logistic regression was used to estimate odds ratios (ORs) and 95% confidence intervals (CIs).
Result
No significant association was observed between skin lesions and the XPD 312 Asp/Asn (adjusted OR = 0.87, 95% CI = 0.65–1.15) Asn/Asn (adjusted OR = 0.76, 95% CI = 0.50–1.15) (referent Asp/Asp); XPD 751 Lys/Gln (adjusted OR = 0.92, 95% CI = 0.69–1.23) Gln/Gln (adjusted OR = 0.98, 95% CI = 0.66–1.45) (referent Lys/Lys). While we did not observe any evidence of effect modification of these polymorphisms on the association between well arsenic concentration and skin lesions, we did observe effect modification between these polymorphisms and sunburn tendency and arsenic-related skin lesions. Individuals with the heterozygote or homozygote variant forms (Asp/Asn or Asn/Asn) had half the risk of skin lesions (OR = 0.45, 95% CI = 0.29–0.68) compared with those with the wild-type XPDAsp312Asn genotype (Asp/Asp) and individuals with heterozygote or homozygote variant forms (Lys/Gln or Gln/Gln) had half the risk of skin lesions (OR = 0.47, 95% CI = 0.31–0.72) compared with those with the wild-type XPDLys751Gln genotype (Lys/Lys), within the least sensitive strata of sunburn severity. We observed effect modification on the multiplicative scale for XPD 751 and XPD 312.
Conclusion
XPD polymorphisms modified the relationship between tendency to sunburn and skin lesions in an arsenic exposed population. Further study is necessary to explore the effect of XPD polymorphisms and sun exposure on risk of arsenic-related skin lesions.
Introduction
Chronic arsenic exposure through drinking water impacts millions of people across the world, including Argentina, Bangladesh, Cambodia, Chile, China, Ghana, Hungary, Inner Mongolia, Mexico, Nepal, New Zealand, Philippines, Taiwan and Vietnam. Basal cell carcinoma and squamous cell carcinoma are the most commonly observed cancers associated with chronic elevated arsenic exposure through drinking water (1). The mechanism for arsenic carcinogenesis remains unclear; however, several proposed mechanisms do exist (2). Additionally the mechanism between arsenic-related skin cancer and pre-malignant arsenic-related skin lesions has not yet been elucidated. While the dose and magnitude of arsenic exposure necessary to result in skin lesions has not been established, the latency period is much shorter than for other cancer endpoints and limited evidence suggests that arsenic-related skin lesions may serve as markers of susceptibility for skin and other arsenic-related cancers (3). It has been documented that subjects with chronic arsenic exposure who develop basal or squamous cell non-melanoma skin cancers often develop these cancers at the site of keratoses (4). Skin lesions, hyperkeratosis and melanosis, are not required for the development of basal or squamous cell carcinomas.
Arsenic is well established as a human skin carcinogen, but its precise mechanism of action remains unknown and potential co-exposure to factors such as ultraviolet (UV) radiation contributes further to the complexity (2,5–8). One mechanism by which arsenic is thought to damage DNA is through an indirect pathway where it inhibits the DNA repair mechanisms necessary to repair DNA damage from other agents, resulting in increased mutations (9). It is also possible that arsenic acts as a co-carcinogen with another co-exposure by increasing positive growth signaling (8). Several experiments in mouse skin models, keratinocytes, Escherichia coli, Chinese hamster V79 cells and human lymphoblastoid TK6 cells have shown that arsenic and UV exposure are co-carcinogens (5,6,10–14). The dimethylarsinic acid metabolite of arsenic and UVB co-exposure resulted in carcinogenesis of the skin in mice (15). UV exposure, DNA repair genotype and arsenic exposure have been shown to be risk factors for non-melanoma skin cancer; however, these factors have not been investigated collectively in a population-based study (16–18).
DNA repair enzymes work in concert to protect the genome of the cell from carcinogenic exposure. The DNA repair Xeroderma Pig-mentosum complementation group D (XPD, also known as ERCC2) is a nucleotide excision repair (NER) helicase. The enzyme has a duel role in the NER pathway: (i) uncoiling the double helix at the site of DNA lesions and (ii) transcription (19). The NER pathway is the major pathway for the removal of bulky DNA lesions, in particular UV-induced photoproducts, bulky monoadducts, cross-links and products of oxidative damage (20). The XPD protein is essential for NER, and this has been demonstrated through point mutations in the XPD protein that results in diseases of DNA repair deficiency such xeroderma pigmentosum, trichothiodystrophy and Cockayne syndrome (21). It has been hypothesized that arsenic inhibits repair of UV-induced DNA damage, through inhibition of the NER pathway, and this may be a mechanism for the increase in mutations that result when cells are co-exposed to UV light and environmental arsenic exposure (9,22).
Several important single-nucleotide polymorphisms have been identified in the XPD locus. Among them, a G → A polymorphism in codon 312 of exon 10 results in an Asp → Asn substitution in evolutionarily conserved region and another polymorphism, C → A in codon 751 of exon 23, produce a Lys → Gln substitution (23). These variant phenotypes have been associated with lower DNA repair capacity or higher risk for carcinogenesis in population-based studies (23–26). Other studies have reported no association. The functional roles of the polymorphisms of XPD 751 and 312 are not clear (1,21). The impact of these polymorphisms on repair phenotype and cancer susceptibility is uncertain (27). Recent meta-analysis has reported that the XPD 312 Asp/Asn and Asn/Asn alleles were associated with an increased risk of breast and lung cancer compared with Asp/ Asp, whereas the XPD 751 Lys/Gln and Gln/Gln alleles were associated with increased risk of breast cancer compared with Lys/Lys (27).
To date, there are two studies related to XPD codon 751 that suggested a potential role for this gene in arsenic-related hyperkeratosis with limited sample size. One study, conducted in Bangladesh, reported that the risk of hyperkeratosis was elevated in individuals with the AA (Gln/Gln) genotype compared with the AC/CC (Lys/Gln, Lys/ Lys) genotypes; however, the association was not significant after the model was adjusted for arsenic exposure (1). A second study in West Bengal also reported that the Lys/Lys genotype was associated with an elevated risk of hyperkeratosis; however, this analysis was also not adjusted for arsenic exposure (28). No other studies have investigated the main or modifying effects of XPD (Asp312Asn) on arsenic-related skin lesions. It has been suggested that larger studies investigate this relationship with combinations of other DNA repair genes and other environmental exposures. To our knowledge, there have been no published studies on polymorphisms in XPD, severity of sunburn and pre-malignant skin lesions in an arsenic-exposed population.
In this study, we examined the role of XPD as a candidate susceptibility gene for arsenic-related skin lesions, in a population-based case–control study in Pabna, Bangladesh. We estimated the main effects for XPD codon 312 and codon 751 genotypes, as well as modification of odds ratios (ORs) for tendency to sunburn and arsenic-related skin lesions.
Materials and methods
Study population
This case–control study was conducted in the Pabna district of Bangladesh and has been described previously (29). The source population included the general population of the 23 villages of Pabna. Community meetings held in 2001–2002 were used to recruit subjects for this study in conjunction with Pabna Community Clinic, an affiliate of Dhaka Community Hospital. Pabna Community Clinic provides healthcare service in Pabna. Briefly, eligible cases were Pabna residents, at least 16 years of age, with one or more type of skin lesion: diffuse/spotted melanosis, diffuse/spotted keratosis, hyperkeratosis or leukomelanosis. Controls were healthy individuals diagnosed as free of arsenic-related disease randomly selected in a 1:1 ratio from Pabna, age of at least 16 years, living in the same village as cases but not sharing a tube well. Two trained physicians clinically diagnosed subjects by visual inspection, and treatment was provided at Pabna Community Clinic or Dhaka Community Hospital when necessary. Physicians were unaware of subjects’ exposure status. Individuals found to have arsenic exposure greater than 50 μg/l were advised to seek alternative drinking water sources. Remediation efforts in the Pabna district have been discussed elsewhere (30).
Cases and controls were frequency matched on gender and age (± 3 years). The participation rate was 98.0%; a total of 24 subjects from 1224 declined to participate. The population is ethnically homogenous, and similar to the population of Bangla (West Bengal), India. Informed consent was obtained from all study participants. The study protocol was approved by the Institutional Review Boards at Dhaka Community Hospital, Bangladesh and Harvard School of Public Health Boston, MA, USA.
Interviews and sample collection
In 2001–2002, 1200 subjects were recruited and interview and sample collection have been described previously (29,31,32). Physicians, blinded to exposure status, examined potential cases and controls. Trained interviewers administered the questionnaire regarding arsenic exposure, occupation/lifestyle factors and collected toenail and individual well water samples. Data were collected on liters of water/liquid ingested per day, disease history and residential history including identification of current primary water source (tube well), years of use and use of a previous tube well. Tendency to sunburn was assessed by self-report.
Two drops (0.2 ml) of pure nitric acid was added to each 100 ml water sample upon collection. The samples were stored in a cooler before storage in a 4°C refrigerated room. Environmental Protection Agency method 200.8 with inductively coupled plasma mass spectroscopy (Environmental Laboratory Services in North Syracuse, NY) (33). The method limit of detection was 1 μg arsenic/l.
Two 10 ml ethylenediaminetetraacetic acid tubes were used to collect blood and were stored in a cooler on ice until processed with cell lysis solution. Samples were sent to the Molecular Epidemiology Laboratory at Harvard School of Public Health for DNA extraction and genotyping.
Genotyping
DNA samples were stored at −80°C. Whole blood was collected from patients at the time of enrollment and DNA was extracted from these samples using the Puregene DNA Isolation kit (Gentra Systems, Minneapolis, MN). The genotyping methods for the ERCC2 Asp312Asp (exon 10) and Lys751Gln (exon 23) polymorphisms have been described in detail (34,35). Briefly, two separate polymerase chain reaction assays were used to detect the polymorphisms in exon 10 and exon 23 of ERCC2 using published primer sequences. DpnII and MspI (for exon 10) and Mbo II enzyme (for exon 23) digestion (New England BioLabs, Ipswich, MA) were used for restriction fragment analyses.
Genotyping for XPDAsp312Asn and XPDLys751Gln was completed for 1115 subjects. Laboratory personnel were blinded to case status, and a random 5% of the samples were repeated to validate genotyping procedures. Two authors independently reviewed all results with 100% concordance.
Statistical analysis
Analysis was restricted to subjects who reported using the same well greater than 6 months to minimize potential for temporal variability in arsenic exposure. As described by Rothman et al. (36,37), variables used in frequency matching of subjects (age and sex) were included in all regression models. Other covariates were considered as potential confounders and were included in the models if they changed the ORs by >10% (38). Arsenic concentration and volume of liquid consumed per day were not combined as a dose variable, since liquid volume included juice, milk, soup, tea and water. Data exploration using generalized additive models, implemented in R (version 1.8.1), suggested that the log odds of case identification varied linearly with the arsenic concentration of well water; consequently, untransformed arsenic concentration was used as a continuous predictor of case status. Generalized additive models suggested that the log odds of case status had a non-monotonic relationship with body mass index (BMI); therefore, a quadratic term for BMI was included. To facilitate numerical stability, both linear and quadratic terms for BMI were centered at the median BMI value, 19.1. Consolidated categories for educational status, tendency to sunburn and age were established. A cumulative sun exposure variable was created for number of hours spent in the sun for each job reported and number of years at the job and for recreational time. Subjects were asked about the numbers of hours spent in the sun during work and recreational time. A variable was created for the number of hours per year in the sun, per year of reported work/recreational time spent in the sun for each subject.
Univariate analyses were performed to describe population characteristics and to identify possible data errors and/or outliers. Continuous variables were summarized using means, medians, standard deviations and ranges, whereas categorical variables were described using percentages. Bivariate analyses (χ2 test or t-test, as appropriate) were conducted explore differences between cases and controls prior to multivariate modeling.
Multiple unconditional logistic regression was used to evaluate the associations between DNA repair polymorphisms and arsenic-related skin lesions. Data exploration using generalized additive models, implemented in R (version 1.8.1), suggested that the log odds of case identification varied linearly with the arsenic concentration of well water; consequently, untransformed arsenic concentration was used as a continuous predictor of case status. ORs and other model-based association parameters were obtained from the parameter estimates of regression models, as were their 95% confidence intervals (CIs). Regression models were fit using Statistical Analysis Systems (SAS Institute, Cary, NC) version 8.2.
To estimate interaction on the multiplicative scale, adjusted ORs and 95% CIs were estimated in logistic regression models and with interaction terms for each genotype and arsenic exposure as a continuous variable. Interaction on a multiplicative scale was assessed in the same way for tendency to sunburn as a categorical variable with each genotype. Interaction of XPD genotypes with each other, with arsenic exposure or with self-reported sunburn tendency was evaluated by adding cross product terms to the logistic regression models and were evaluated using the likelihood ratio test (LRT). Multiplicative interaction was assessed using the LRT by comparing models including the cross-product term with a model containing only the main effects. Stratified analysis by tendency to sunburn was also conducted for each genotype. Co-dominant models were used for XPDLys751Gln and XPDAsp312Asn.
Results
Characteristics of cases and controls were presented in Table I and have been presented previously (29). No significant departures were observed from Hardy–Weinberg equilibrium among cases and controls, XPD codon 751 SNP (A →C) (χ2; P = 0.45) and XPD codon 312 SNP (G →A) (χ2; P = 0.11) (Table II).
Table I.
Characteristics of skin lesion cases and population-based controls in Pabna, Bangladesh
| Controls
|
Cases
|
P | |||
|---|---|---|---|---|---|
| n | n | ||||
| Diffuse melanosis | 377 | ||||
| Leukomelanosis | 342 | ||||
| Spotted melanosis | 145 | ||||
| Diffuse keratosis | 117 | ||||
| Spotted keratosis | 73 | ||||
| Hyperkeratosis | 40 | ||||
| Mean age in years (SD) | 33.7(12.6) | 597 | 33.9 (12.7) | 592 | 0.58 |
| Mean BMI (kg/m2) (SD) | 20.4(3.1) | 597 | 20.1 (3.1) | 592 | 0.70 |
| % Male | 60.30% | 360 | 60.30% | 357 | 0.91 |
| Mean duration of present well use (years) (SD) | 10.1(9.0) | 592 | 8.0 (7.2) | 592 | 0.02 |
| % Reported a previous well | 2.84% | 17 | 7.77% | 45 | <0.0001 |
| Mean As level of current well (μg/I) (SD) | 66.2 (149.6) | 595 | 232.8 (315.7) | 592 | <0.0001 |
| Mean As level in nail sample (μg/g) (SD) | 2.8 (4.1) | 583 | 5.9 (7.4) | 589 | 0.0001 |
| Mean daily total water/liquid consumption (l) (SD) | 3.8(1.2) | 595 | 3.7 (1.1) | 592 | 0.84 |
| % Ever used betel nuts | 24.30% | 145 | 27.70% | 164 | 0.007 |
| Mean years of betel nut use (SD) | 10.8(8.9) | 143 | 11.0 (9.5) | 160 | 0.078 |
| Mean number of betel nuts chewed per day (SD) | 5.6(3.6) | 158 | 5.7 (3.8) | 149 | 0.53 |
| % Chew tobacco leaves | 16.40% | 587 | 17.10% | 590 | <0.0001 |
| Mean years of Tobacco leaves chewed (SD) | 9.9(9.1) | 95 | 10.9 (9.4) | 95 | 1.0 |
| % Smokes cigarettes currently | 30.50% | 597 | 26.70% | 592 | 0.03 |
| % Ever smoked | 31.00% | 597 | 28.70% | 592 | 0.0007 |
| Education level | 597 | 592 | 0.8 | ||
| % Illiterate | 17.40% | 104 | 22.90% | 136 | |
| % Literate (incomplete primary education) | 23.80% | 142 | 29.40% | 174 | |
| % Completed primary education | 11.70% | 79 | 11.80% | 70 | |
| % Completed middle school education | 31.90% | 191 | 23.50% | 139 | |
| % Completed secondary education or more | 13.60% | 81 | 26.00% | 154 | |
| Mean hours of sun exposure (SD) | 25937.35(22 821) | 597 | 26782.59 (22 343) | 592 | 0.51 |
Table II.
Adjusted ORs (95% CIs) of XPD genotypes on risk of skin lesions
| Genotype | Controls (N) (%) | Cases (N) (%) | Total | Crude OR | 95% CI | Adjusted OR | 95% CI | |
|---|---|---|---|---|---|---|---|---|
| XPD | ||||||||
| Asp312Asn | 1115 | |||||||
| Asp/Asp (GG) | 239 (42.7%) | 253 (45.7%) | 1.0 | 1.0 | ||||
| Asp/Asn (AG) | 244 (43.6%) | 227 (40.8%) | 0.91 | (0.69–1.19) | 0.87 | (0.65–1.15) | ||
| Asn/Asn (AA) | 77 (13.8%) | 75 (13.5%) | 0.86 | (0.58–1.28) | 0.76 | (0.50–1.15) | ||
| XPD | ||||||||
| Lys751Gln | 1115 | |||||||
| Lys/Lys (AA) | 86 (15.4%) | 88 (15.9%) | 1.0 | 1.0 | ||||
| Lys/Gln (AC) | 259 (46.3%) | 251 (45.2%) | 0.99 | (0.75–1.31) | 0.92 | (0.69–1.23) | ||
| Gln/Gln (CC) | 215 (38.4%) | 216 (38.9%) | 1.10 | (0.75–1.60) | 0.98 | (0.66–1.45) | ||
Crude OR adjusted for As (arsenic) concentration of well liters of liquid/.day.
Adjusted for sunburn, As concentration of well liters of liquid/day, educational status, smoking status, chewing tobacco use, betel nut use, sex, age and previous well use.
The results from the logistic regression models show that the XPD genotypes were not significantly associated with risk of skin lesions (Table II), nor was there evidence of effect modification between arsenic exposure and the genotypes on skin lesions [XPDLys751Gln (LRT, P = 0.10); XPDAsp312Asn (LRT, P = 0.73)]. There was no evidence of effect modification between tendency to sunburn and arsenic exposure (LRT, P = 0.18).
The adjusted main effects of the tendency to sunburn as associated with arsenic-related skin lesions are shown in Table III. A significant test for trend was observed with increasing sensitivity to sunburn; however, a significant association was only noted in individuals that reported painful burn or blisters after sun exposure (P-trend = 0.003). Those who reported the highest category of sun sensitivity had twice (OR = 2.17, 95% CI = 1.20–3.92) the odds of skin lesions than those who reported no tendency to sunburn after exposure.
Table III.
Adjusted OR for tendency to sunburn and arsenic-related skin lesions
| Cases (n) (%) | Controls (n) (%) | OR | 95% CI | |
|---|---|---|---|---|
| Practically no tendency | 207 (35%) | 256 (43%) | 1.0 (referent) | |
| Some redness only | 222 (37.5%) | 204 (34.2%) | 1.23 | (0.91–1.66) |
| Burn | 124 (21%) | 113 (19%) | 1.33 | (0.94–1.89) |
| Painful burn or blisters | 39 (6.5%) | 23 (3.8%) | 2.17 | (1.20–3.92) |
| P-trend | 0.003 |
Model adjusted for well arsenic concentration, total liquids per day, age, gender, previous well use, smoking, chewing tobacco and betel nut use.
We investigated whether the association for arsenic-related skin lesions and DNA repair genotypes differed by tendency to sunburn (Table IV). In stratified analysis, we found that in individuals who reported almost no tendency to sunburn, individuals with heterozygote or homozygote variant forms (Asp/Asn or Asn/Asn) had half the risk of skin lesions (OR = 0.45, 95% CI = 0.29–0.68) compared with those with the wild-type XPDAsp312Asn genotype (Asp/Asp). On the multiplicative scale, we found significance between the XPDAs-p312Asn genotype and tendency to sunburn (LRT, P = 0.03).
Table IV.
Adjusted ORs for skin lesions stratified by tendency to burn
| Tendency to Sunburn | Cases (n) (%) | Controls (n) (%) | OR | 95% CI | ||
|---|---|---|---|---|---|---|
| XPDAsp312Asn | ||||||
| Asp/Asp | (Asp/Asn + Asn/Asn) | Asp/Asp | (Asp/Asn + Asn/Asn) | |||
| Model 1 | ||||||
| Practically no tendency | 94 (37.2%) | 82 (27.2%) | 100 (41.8%) | 128 (39.9%) | 0.48 | (0.30–0.79) |
| Some redness only | 88 (34.8%) | 133 (44.0%) | 83 (34.7%) | 120 (37.4%) | 1.10 | (0.70–1.74) |
| Burn | 58 (22.9%) | 62 (20.5%) | 49 (20.5%) | 58 (18.1%) | 1.08 | (0.60–1.94) |
| Painful burn or blisters | 13 (5.1%) | 25 (8.3%) | 7 (2.9%) | 15 (4.7%) | 3.63 | (0.40–33.25) |
| XPDLys751Gln | ||||||
| Lys/Lys | (Lys/Gln + Gln/Gln) | Lys/Lys | (Lys/Gln + Gln/Gln) | |||
| Model 2 | ||||||
| Practically no tendency | 82 (38%) | 96 (28.4%) | 85 (39.5%) | 143 (41.5%) | 0.53 | (0.33–0.86) |
| Some redness only | 76 (35.1%) | 145 (42.9%) | 80 (37.2%) | 123 (35.7%) | 1.27 | (0.79–2.01) |
| Burn | 47 (21.8%) | 70 (20.7%) | 44 (20.5%) | 63 (18.3%) | 1.15 | (0.63–2.12) |
| Painful burn or Blisters | 11 (5.1%) | 27 (8.0%) | 6 (2.8%) | 16 (4.6%) | 2.84 | (0.32–25.37) |
Model 1 adjusted for well arsenic concentration, total liquids per day, XPD 312, age, gender, previous well use, smoking, chewing tobacco and betel nut use.
Model 2 adjusted for well arsenic concentration, total liquids per day, XPD 751, age, gender, previous well use, smoking, chewing tobacco and betel nut use.
In the stratified analysis, we also observed that individuals with heterozygote or homozygote variant forms (Lys/Gln or Gln/Gln) had half the risk of skin lesions (OR = 0.47, 95% CI = 0.31–0.72) compared with those with the wild-type XPDLys751Gln genotype (Lys/ Lys), within the least sensitive strata of sunburn severity (Table IV). When evaluated on the multiplicative scale, we found a significant association between the XPDLys751Gln genotype and tendency to sunburn (LRT, P = 0.05). When stratified, we observed that individuals with the combined XPD 751 Lys/Lys genotype and XPD 312 (Asp/Asn or Asn/Asn) genotype had a lower risk of skin lesions compared with individuals with wild type for both of these alleles (OR = 0.34, 95% CI = 0.16–0.73) (Table V). We did observe evidence of interaction between the XPD (Asp312Asn and Lys751Gln) genes (LRT, P = 0.03). We did not have adequate sample size to study interaction between both polymorphisms and arsenic or both polymorphisms and sunburn sensitivity on the risk of skin lesions.
Table V.
Estimated adjusted ORs and 95% CIs of XPD 751 and XPD 312 polymorphisms and skin lesions
| Genotype
| ||||||
|---|---|---|---|---|---|---|
| XPD 751 | XPD 312 | Controls (N) | Cases (N) | OR | 95% CI | P-value |
| Lys/Lys (wild type) | Asp/Asp (wild type) | 187 | 201 | 1.0 (referent) | ||
| Lys/Gln + Gln/Gln | Asp/Asp | 52 | 51 | 0.76 | (0.47–1.23) | 0.26 |
| Lys/Lys | Asp/Asn + Ans/Asn | 28 | 15 | 0.34 | (0.16–0.73) | 0.006 |
| Lys/Gln + Gln/Gln | Asp/Asn + Ans/Asn | 293 | 286 | 0.77 | (0.59–1.02) | 0.06 |
Model adjusted for well arsenic concentration, total liquids per day, age, gender, previous well use, smoking, chewing tobacco and betel nut use.
Discussion
Overall, the XPD 312 (Asp/Asn + Asn/Asn) and XPD 751 (Lys/Gln +Gln/Gln) were associated with a reduced risk of skin lesions in those individuals with no tendency to sunburn after 2 h of sun exposure. Since there are no previous results of modification by these genotypes of sunburn and pre-malignant hyperkeratosis and melanosis, we can compare our results with findings related to non-melanoma skin cancer. It is unclear whether individuals with the XPD Lys/Lys genotype have an increased or a decreased risk of non-melanoma skin cancer (39,40). Arsenic-related hyperkeratosis and melanosis are only associated with non-melanoma skin cancers; however, a study reported that the 751Gln allele was associated with melanoma and squamous cell carcinoma and that this relationship was modified by the number of sunburns and a measure of cumulative sun exposure (41). It was hypothesized by these authors that when challenged with high sun exposure melanocytes with the variant form of XPD 751 and impaired DNA repair capacity may accumulate more mutations and have a higher rate of apoptosis and thus a lower risk of melanoma (41). It is possible that a similar mechanism is responsible for the observed effects in our study.
While the functional role of XPD Lys/Lys (wild type) is not clear, it is thought that a polymorphism at 751 has the largest effect on enzyme activity (1,21). Other investigators have found that the Lys/Lys751 genotype was associated with a lower level of DNA repair activity (25). Individuals with the XPD Gln751Gln genotype also had 20% lower (P = 0.004) DNA repair capacity compared with the XPD Lys751Lys (wild type) phenotype measured by a Luciferase assay (42). Our findings are consistent with a previous study that found that modifications of the XPD gene also modify the effect of arsenic on hyperkeratosis, which is considered a pre-cursor to non-melanoma skin cancer. Prior to adjustment for arsenic exposure, a previous study observed that individuals with the AA genotype of XPD 751 had an increased risk of arsenic-related hyperkeratosis (1). While we did not observe a main effect of the genotype, we observed potential modification that would suggest those with the Lys/Lys genotype had an increased rate of skin lesions after adjustment for arsenic exposure and tendency to sunburn. We found that the prevalence of the Lys/Lys (AA) genotype was slightly lower in our Bangladesh study population (15.8%, 15.9%) than other populations (21 to almost 50%); however, our study was population based and this distribution is representative (1,43–45).
The functional role of XPD 312 has yet to be elucidated (21). While some studies have found no association with the Asp312Asn XPD polymorphism and DNA repair ability (25), other studies have reported that the Asp/Asp genotype of XPD 312 was associated with a decrease in apoptosis in response to UV exposure and this mechanism could be related to increased survival and cloning of cells with DNA damage (46). This finding is consistent with our finding that those with the (Asp/Asn + Asn/Asn) alleles had a reduced risk of skin lesions compared with the Asp/Asp genotype. It is possible that failure to undergo apoptosis allows cells to replicate and further mutate, first resulting in skin lesions and potentially later resulting in arsenic-related cancer. Additionally, it is possible that the effect is only observed among those with the least sensitivity to sunburn, because they may have continually experienced greater unprotected UV exposure since they have less tendency to sunburn. This information is not possible to assess from our data.
We found that increasing severity of sensitivity to sun exposure was associated with an increased risk of arsenic-related skin lesions among an arsenic exposed population. Increased sun exposure has been associated with an increased risk of non-melanoma skin cancer (47). Risk of arsenic-related skin lesions has been reported to be greater in men with excessive sun exposure (48). We did not find a significant association between cumulative sun exposure and risk of skin lesions. Cases and controls reported similarly high number of hours with sun exposure either working or in recreational activities (Table I). It is also possible that there is some recall bias; however, this study population was from a rural area and the majority of subjects reported occupations that required significant outdoor work.
The study measured tendency/severity of sunburn after sun exposure for two hours and hours spent in the sun for occupation and recreational activities by self-report. This measure does not take into account the darkness of skin pigmentation, which may be associated with the risk of sunburn and the dose of UV radiation to the dermis. Cumulative sun exposure was self-reported as was the tendency to sunburn; however, similar measures such as self-reported number of lifetime sunburns have been employed in other large population based studies (39,49,50). There is potential for recall bias; however, there is no established association between severity of sunburn history and arsenic-related skin lesions so there is no reason to believe that cases would have differentially reported tendency to sunburn differently than controls. A recent study validated that self-reported measures of sun sensitivity and self-reported history of non-melanoma skin cancer were reasonably reliable measures to use in screening subjects for studies and found that all groups were equally reliable when reporting sun sensitivity (51). We did not observe a significant difference between length of cumulative estimate of hours spent in the sun between cases and control subjects (P = 0.52), nor did we observe an association between cumulative sun exposure and risk of skin lesions (OR = 1.00, 95% CI = 1.00–1.0) in a multivariate adjusted model. It was possible that there was exposure misclassification in the number of hours spent working/recreationally outdoors; however, the majority of occupations in this rural area involved outdoor work. Any misclassification in sun exposure is likely to be non-differential and could bias results in either direction. Arsenic exposure was measured by individual well measurements in this study. Exposure misclassification was possible if current arsenic exposure from individual wells does not represent past exposure. Evidence exists that well arsenic concentration in Bangladesh is relatively stable over time, though shallow wells are somewhat more variable (52). Well depth was not measured by the field team in this study. In our population, the mean duration of current well use is 10.1 years for controls and 8.0 years for cases, which indicates that subjects have had a relatively long exposure period that is adequate for the development of skin lesions (3). It is possible that our findings are due to linkage to another significant polymorphism or by differential effects of the polymorphism by other modifying factors, such as skin pigmentation. Unmeasured confounders are possible. In order for the unmeasured confounder to explain the interaction results between genotype and sunburn tendency among an arsenic-exposed population, the confounder would have to be differentially distributed in the separate genotype–exposure strata. A strength of this study was that it was conducted on an ethnically homogenous population from Pabna, Bangladesh, and the study was of a large enough sample size to investigate gene–environment interactions. Reliability of genotyping and arsenic concentration of drinking water was high due to previously described quality control procedures. Hence, measurement error in genotyping and analysis of drinking water arsenic concentration was possible but unlikely. Arsenic concentration of wells was determined by a private lab in the USA without knowledge of case–control status. Genotyping was also conducted without indicators of study outcome status. Any measurement error would have been independent of case–control status, and any bias would have been non-differential. Limited numbers did not allow us to assess potential interaction between both XPD polymorphisms, co-exposure to arsenic and sunburn and skin lesions.
Polymorphisms in XPD 312 and 751 may play a role in the NER phenotype and therefore to genetic susceptibility to skin lesions that have been associated with increased risk for skin cancer. Further investigations are needed to examine of the effect of multiple XPD polymorphisms and co-exposure to arsenic and sun associated with risk of skin lesions.
Acknowledgments
This work was supported by grants ES 05947, ES 00002 and ES 11622 from the National Institutes of Health. K.M.M. was supported by 5T32CA 009330 from the National Cancer Institute, National Institutes of Health and by L30 CA124219-01 from the National Institutes of Health while the manuscript was prepared. We gratefully acknowledge Drs K.H. Kraemer, R. Wilson, H. Ahsan and for their expertise, the Dhaka Community Hospital field team for subject recruitment, sample collection and data entry, Janna Frelich for data management and programming assistance, Mike Wang and Tap Van Geel in the Harvard Molecular Epidemiology Laboratory and Lia Shimada for her expertise and assistance in Human Subjects approval.
Abbreviations
- BMI
body mass index
- CI
confidence interval
- LRT
likelihood ratio test
- NER
nucleotide excision repair
- OR
odds ratio
- UV
ultraviolet
- XPD
Xeroderma Pigmentosum complementation group D
Footnotes
Conflict of Interest statement: None declared.
References
- 1.Ahsan H, et al. DNA repair gene XPD and susceptibility to arsenic-induced hyperkeratosis. Toxicol Lett. 2003;143:123–131. doi: 10.1016/s0378-4274(03)00117-6. [DOI] [PubMed] [Google Scholar]
- 2.Kitchin KT. Recent advances in arsenic carcinogenesis: modes of action, animal model systems, and methylated arsenic metabolites. Toxicol Appl Pharmacol. 2001;172:249–261. doi: 10.1006/taap.2001.9157. [DOI] [PubMed] [Google Scholar]
- 3.National Research Council; Subcommittee on Arsenic in Drinking Water, C.o.T., Board on Environmental Studies and Toxicology, Commission on Life Sciences, National Research Council, editor. Arsenic in drinking water. National Academy Press; Washington, DC: 1999. p. 104. [Google Scholar]
- 4.Alain G, et al. Chronic arsenic toxicity. Int J Dermatol. 1993;32:899–901. doi: 10.1111/j.1365-4362.1993.tb01413.x. [DOI] [PubMed] [Google Scholar]
- 5.Rossman TG. Enhancement of UV-mutagenesis by low concentrations of arsenite in E. coli. Mutat Res. 1981;91:207–211. doi: 10.1016/0165-7992(81)90032-4. [DOI] [PubMed] [Google Scholar]
- 6.Rossman TG, et al. Arsenite is a cocarcinogen with solar ultraviolet radiation for mouse skin: an animal model for arsenic carcinogenesis. Toxicol Appl Pharmacol. 2001;176:64–71. doi: 10.1006/taap.2001.9277. [DOI] [PubMed] [Google Scholar]
- 7.Mudipalli A, et al. The effect of arsenicals on ultraviolet-radiation-induced growth arrest and related signaling events in human keratinocytes. Int J Oncol. 2005;27:769–778. [PubMed] [Google Scholar]
- 8.Rossman TG, et al. Arsenite cocarcinogenesis: an animal model derived from genetic toxicology studies. Environ Health Perspect. 2002;110:749–752. doi: 10.1289/ehp.02110s5749. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Angeline SA, et al. Decreased DNA repair gene expression among individuals exposed to arsenic in United States drinking water. 2003;104:263–268. doi: 10.1002/ijc.10968. [DOI] [PubMed] [Google Scholar]
- 10.Germolec DR, et al. Arsenic enhancement of skin neoplasia by chronic stimulation of growth factors. Am J Pathol. 1998;153:1775–1785. doi: 10.1016/S0002-9440(10)65692-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Burns FJ, et al. Arsenic-induced enhancement of ultraviolet radiation carcinogenesis in mouse skin: a dose-response study. Environ Health Perspect. 2004;112:599–603. doi: 10.1289/ehp.6655. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Chen PH, et al. Effects of arsenic and UVB on normal human cultured keratinocytes: impact on apoptosis and implication on photocarcinogenesis. Chem Res Toxicol. 2005;18:139–144. doi: 10.1021/tx049834b. [DOI] [PubMed] [Google Scholar]
- 13.Okui T, et al. Inhibition of human excision DNA repair by inorganic arsenic and the co-mutagenic effect in V79 Chinese hamster cells. Mutat Res. 1986;172:69–76. doi: 10.1016/0165-1218(86)90108-4. [DOI] [PubMed] [Google Scholar]
- 14.Li JH, et al. Comutagenesis of sodium arsenite with ultraviolet radiation in Chinese hamster V79 cells. Biol Met. 1991;4:197–200. doi: 10.1007/BF01141180. [DOI] [PubMed] [Google Scholar]
- 15.Yamanaka K, et al. The role of orally administered dimethylarsinic acid, a main metabolite of inorganic arsenics, in the promotion and progression of UVB-induced skin tumorigenesis in hairless mice. Cancer Lett. 2000;152:79–85. doi: 10.1016/s0304-3835(99)00440-1. [DOI] [PubMed] [Google Scholar]
- 16.Ramos J, et al. UV dose determines key characteristics of nonmelanoma skin cancer. Cancer Epidemiol Biomarkers Prev. 2004;13:2006–2011. [PubMed] [Google Scholar]
- 17.Maloney M. Arsenic in dermatology. Dermatol Surg. 1996;22:301–304. doi: 10.1111/j.1524-4725.1996.tb00322.x. [DOI] [PubMed] [Google Scholar]
- 18.Kraemer KH, et al. The role of sunlight and DNA repair in melanoma and nonmelanoma skin cancer. The xeroderma pigmentosum paradigm. Arch Dermatol. 1994;130:1018–1021. [PubMed] [Google Scholar]
- 19.Lehmann AR. The xeroderma pigmentosum group D (XPD) gene: one gene, two functions, three diseases. Genes Div. 2001;15(1):15–23. doi: 10.1101/gad.859501. [DOI] [PubMed] [Google Scholar]
- 20.Sancar A. DNA repair in humans. Annu Rev Genet. 1995;29:69–105. doi: 10.1146/annurev.ge.29.120195.000441. [DOI] [PubMed] [Google Scholar]
- 21.Benhamou S, et al. ERCC2/XPD Gene polymorphisms and lung cancer: a HuGE review. AJE. 2005;161(1):1–14. doi: 10.1093/aje/kwi018. [DOI] [PubMed] [Google Scholar]
- 22.Danaee H, et al. Low dose exposure to sodium arsenite synergistically interacts with UV radiation to induce mutations and alter DNA repair in human cells. Mutagenesis. 2004;19:143–148. doi: 10.1093/mutage/geh010. [DOI] [PubMed] [Google Scholar]
- 23.Shen MR, et al. Nonconservative amino acid substitution variants exist at polymorphic frequency in DNA repair genes in healthy humans. Cancer Res. 1998;58:604–608. [PubMed] [Google Scholar]
- 24.Hemminki K, et al. XPD exon 10 and 23 polymorphisms and DNA repair in human skin in situ. Carcinogenesis. 2001;22:1185–1188. doi: 10.1093/carcin/22.8.1185. [DOI] [PubMed] [Google Scholar]
- 25.Lunn RM, et al. XPD polymorphisms: effects on DNA repair proficiency. Carcinogenesis. 2000;21:551–555. doi: 10.1093/carcin/21.4.551. [DOI] [PubMed] [Google Scholar]
- 26.Au WW, et al. Functional characterization of polymorphisms in DNA repair genes using cytogenetic challenge assays. Environ Health Perspect. 2003;111:1843–1850. doi: 10.1289/ehp.6632. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Manuguerra M, et al. XRCC3 and XPD/ERCC2 Single nucleotide polymorphisms and the risk of cancer: a HuGE review. Am J Epidemiol. 2006;164:297–302. doi: 10.1093/aje/kwj189. [DOI] [PubMed] [Google Scholar]
- 28.Banerjee M, et al. Polymorphism in the ERCC2 codon 751 is associated with arsenic-induced premalignant hyperkeratosis and significant chromosome aberrations. Carcinogenesis. 2006;28:672–676. doi: 10.1093/carcin/bgl181. [DOI] [PubMed] [Google Scholar]
- 29.McCarty KM, et al. The impact of diet and betel nut use on skin lesions associated with drinking-water arsenic in Pabna, Bangladesh. Environ Health Perspect. 2006;114:334–340. doi: 10.1289/ehp.7916. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Joya SF, et al. One solution to the arsenic problem: a return to surface (improved dug) wells. J Health Popul Nutr. 2006;24:363–375. [PMC free article] [PubMed] [Google Scholar]
- 31.McCarty KM, et al. Arsenic methylation, GSTT1, GSTM1, GSTP1 polymorphisms and Skin Lesions. Environ Health Perspect. 2007;115:341–345. doi: 10.1289/ehp.9152. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.McCarty KM, et al. GSTT1, GSTM1, GSTP1 Polymorphisms, and arsenic related skin lesions. Environmental Health. 2007;6:5. doi: 10.1289/ehp.9152. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.U.S. Environmental and Protection Agency. Method 200.8, Determination of Trace Elements in Waters and Wastes by Inductively Coupled Plasma Mass Spectrometry. Cincinnati, OH: 1994. [Google Scholar]
- 34.Duell EJ, et al. Polymorphisms in the DNA repair genes XRCC1 and ERCC2 and biomarkers of DNA damage in human blood mononuclear cells. Carcinogenesis. 2000;21:965–971. doi: 10.1093/carcin/21.5.965. [DOI] [PubMed] [Google Scholar]
- 35.Zhou W, et al. Gene-environment interaction for the ERCC2 polymorphisms and cumulative cigarette smoking exposure in lung cancer. Cancer Res. 2002;62:1377–1381. [PubMed] [Google Scholar]
- 36.Rothman KJ, et al. Modern Epidemiology. Lippincott, Williams & Wilkins; New York: 1998. Chapter 10 matching. [Google Scholar]
- 37.Rothman KJ, et al. Modern Epidemiology. Lippincott, Williams & Wilkins; New York: 1998. Chapter 21 introduction to regression modeling. [Google Scholar]
- 38.Hosmer DW, et al. Applied Logistic Regression. John Wiley & Sons; New York: 1989. [Google Scholar]
- 39.Vogel U, et al. Polymorphisms of the DNA repair gene XPD: correlations with risk of basal cell carcinoma revisited. Carcinogenesis. 2001;22:899–904. doi: 10.1093/carcin/22.6.899. [DOI] [PubMed] [Google Scholar]
- 40.Dybdahl M, et al. Polymorphisms in the DNA repair gene XPD: correlations with risk and age at onset of basal cell carcinoma. Cancer Epidemiol Biomarkers Prev. 1999;8:77–81. [PubMed] [Google Scholar]
- 41.Han J, et al. Gentic variation in XPD, sun exposure and risk of skin cancer. Cancer Epidemiol Biomarkers Prev. 2005;14:1539–1544. doi: 10.1158/1055-9965.EPI-04-0846. [DOI] [PubMed] [Google Scholar]
- 42.Qiao Y, et al. Modulation of repair of ultraviolet damage in the host-cell reactivation assay by polymorphic XPC and XPD/ERCC2 genotypes. Carcinogenesis. 2002;23:295–299. doi: 10.1093/carcin/23.2.295. [DOI] [PubMed] [Google Scholar]
- 43.Terry MB, et al. Polymorphism in the DNA repair gene XPD, poly-cyclic aromatic hydrocarbon-DNA adducts, cigarette smoking, and breast cancer risk. Cancer Epidemiol Biomarkers Prev. 2004;13:2053–2058. [PubMed] [Google Scholar]
- 44.Justenhoven C, et al. ERCC2 Genotypes and a corresponding haplotype are linked with breast cancer risk in a German population. Cancer Epidemiol Biomarkers Prev. 2004;13:2059–2064. [PubMed] [Google Scholar]
- 45.Clarkson SG, et al. Polymorphisms in the human XPD (ERCC2) gene, DNA repair capacity and cancer susceptibility: an appraisal. DNA Repair. 2005;4:1068–1074. doi: 10.1016/j.dnarep.2005.07.001. [DOI] [PubMed] [Google Scholar]
- 46.Seker H, et al. Functional significance of XPD polymorphic variants: attenuated apoptosis in human lymphoblastoid cells with the XPD 312 Asp/Asp genotype. Cancer Res. 2001;61:7430–7434. [PubMed] [Google Scholar]
- 47.Ramos J, et al. UV dose determines key characteristics of nonmelanoma skin cancer. CEBP. 2004;13(12):2006–2011. [PubMed] [Google Scholar]
- 48.Chen Y, et al. Modification of risk of arsenic-induced skin lesions by sunlight exposure, smoking, and occupational exposures in Bangladesh. Epidemiology. 2006;17:459–467. doi: 10.1097/01.ede.0000220554.50837.7f. [DOI] [PubMed] [Google Scholar]
- 49.Han J, et al. Genetic variation in XRCC1, sun exposure, and risk of skin cancer. Br J Cancer. 2004;91:1604–1609. doi: 10.1038/sj.bjc.6602174. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Nelson HH, et al. The XRCC1 Arg399Gln polymorphism, sunburn, and non-melanoma skin cancer: evidence of gene-environment interaction. Cancer Res. 2002;62:152–155. [PubMed] [Google Scholar]
- 51.Clouser MC, et al. Risk group, skin lesion history, and sun sensitivity reliability in squamous cell skin cancer progression. Cancer Epidemiol Biomarkers Prev. 2006;15:2292–2297. doi: 10.1158/1055-9965.EPI-06-0405. [DOI] [PubMed] [Google Scholar]
- 52.Cheng Z, et al. Limited temporal variability of arsenic concentrations in 20 wells monitored for 3 years in Araihazar, Bangladesh. Environ Sci Technol. 2005;39:4759–4766. doi: 10.1021/es048065f. [DOI] [PubMed] [Google Scholar]