DOI:10.1158/1055-9965.

EPI-04-0846

Genetic Variation in XPD, Sun Exposure, and Risk of Skin Cancer

Jiali Han, Graham A. Colditz, Jun S. Liu, et al. Cancer Epidemiol Biomarkers Prev 2005;14:1539-1544. Published online June 7, 2005.

Updated Version

Access the most recent version of this article at: doi:10.1158/1055-9965.EPI-04-0846

Cited Articles Citing Articles

This article cites 32 articles, 21 of which you can access for free at: http://cebp.aacrjournals.org/content/14/6/1539.full.html#ref-list-1 This article has been cited by 15 HighWire-hosted articles. Access the articles at: http://cebp.aacrjournals.org/content/14/6/1539.full.html#related-urls

E-mail alerts Reprints and Subscriptions Permissions

Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at pubs@aacr.org. To request permission to re-use all or part of this article, contact the AACR Publications Department at permissions@aacr.org.

Downloaded from cebp.aacrjournals.org on November 18, 2012 Copyright 2005 American Association for Cancer Research

DOI:10.1158/1055-9965.EPI-04-0846
Cancer Epidemiology, Biomarkers & Prevention 1539

Genetic Variation in XPD , Sun Exposure, and Risk of Skin Cancer

Jiali Han,1,5 Graham A. Colditz,1,2,4 Jun S. Liu,3,6 and David J. Hunter 1,2,5
1 Channing Laboratory, Department of Medicine, Brigham and Womens Hospital and Harvard Medical School; 2Department of Epidemiology, 3Department of Biostatistics, 4Harvard Center for Cancer Prevention, and 5Program in Molecular and Genetic Epidemiology, Harvard School of Public Health, Boston, Massachusetts and 6Department of Statistics, Harvard University, Cambridge, Massachusetts

Abstract
The XPD gene is involved in the nucleotide excision repair pathway removing DNA photoproducts induced by UV radiation. Genetic variation in XPD may exert a subtle effect on DNA repair capacity. We assessed the associations between two common nonsynonymous polymorphisms (Asp312Asn and Lys751Gln) with skin cancer risk in a nested case-control study within the Nurses Health Study (219 melanoma, 286 squamous cell carcinoma, 300 basal cell carcinoma, and 874 controls) along with exploratory analysis on the haplotype structure of the XPD gene. There were inverse associations between the Lys751Gln and Asp312Asn polymorphisms and the risks of melanoma and squamous cell carcinoma. No association was observed between these two polymorphisms and basal cell carcinoma risk. We also observed that the association of the 751Gln allele with melanoma risk was modified by lifetime severe sunburns, cumulative sun exposure with a bathing suit, and constitutional susceptibility score (P for interaction = 0.03, 0.04, and 0.02 respectively). Similar interactions were also observed for the Asp312Asn. Our data suggest these two XPD nonsynonymous polymorphisms may be associated with skin cancer risk, especially for melanoma. (Cancer Epidemiol Biomarkers Prev 2005;14(6):1539 44)

Introduction
Skin cancer is the most common neoplasm in Caucasians in the United States. The genotoxic effect of sunlight exposure has been clearly shown in the etiology of both melanoma and nonmelanocytic skin cancer (1-3). One important defense mechanism against skin cancer is the ability to repair DNA damage induced by UV light. It has been suggested that reduced DNA repair capacity (DRC) is a susceptibility factor predisposing individuals to skin cancer (4-7). The predominant form of UV-induced DNA damage is DNA photoproducts caused by the direct absorption of UVB by DNA. Cyclobutane pyrimidine dimers and pyrimidine (6-4) pyrimidone photoproducts constitute the two major DNA photoproducts (2). DNA photoproducts are mainly removed by the nucleotide excision repair (NER). The NER is a versatile repair system to remove a variety of bulky, helix-distorting lesions, including UV photoproducts and bulky adducts (8, 9). Individuals with xeroderma pigmentosum, deficient in the NER, have a >1,000-fold increased risk of skin cancer. Human XPD maps to chromosome 19q13.3 and spans f54 kb. It comprises 23 exons and is 761 amino acids in length. The XPD gene encodes an ATP-dependent DNA helicase involved in the NER and in basal transcription as part of the transcription factor TFIIH. Disruption of the mouse Xpd gene results in preimplantation lethality (10). Mutations in the XPD gene lead to NER defects (11) and three clinical syndromes, Cockayne syndrome, xeroderma pigmentosum, and trichothiodystrophy, depending on the location of the mutation (9, 12). In addition, the XPD and p53 proteins can interact with each other to modulate apoptosis and the NER. The p53 binds and modulates the helicase activity of the TFIIH, and the repair of UV-induced dimers was attenuated in Li-Fraumeni syndrome cells (heterozygote p53 mutant; ref. 13). A deficiency in p53-mediated apoptosis was reported in XPD lymphoblastoid cell lines and fibroblasts from xeroderma pigmentosum patients with germ line mutations in the XPD gene (14, 15). We evaluated two common nonsynonymous XPD polymorphisms (Asp312Asn and Lys751Gln) in relation to skin cancer risk in a nested case-control study within the Nurses Health Study along with exploratory analysis on the haplotype structure of the XPD gene. We further investigated the hypothesis that XPD genetic variants modify the associations of sunlight-related risk factors with skin cancer risk.

Materials and Methods

Study Population. The Nurses Health Study was established in 1976, when 121,700 female registered nurses between ages 30 and 55 years completed a self-administered questionnaire on their medical histories and baseline health-related exposures. Updated information has been obtained by questionnaires every 2 years. Between 1989 and 1990, blood samples were collected from 32,826 of the cohort members. Eligible cases in this study consisted of women with incident skin cancer from the subcohort who gave a blood specimen, including squamous cell carcinoma (SCC) and basal cell carcinoma (BCC) cases with a diagnosis anytime after blood collection up to June 1, 1998 and melanoma cases (including in situ cases) up to June 1, 2000 with no previously diagnosed skin cancer. All available pathologically confirmed melanoma and SCC cases and 300 self-reported BCC cases randomly selected from f2,600 available self-reported BCC cases were included. The validity of self-report of BCC is high in this medically sophisticated population (90%; ref. 16). All the SCC and BCC cases had no history of melanoma diagnosis. A common control series (case/control, 1:1) was randomly selected from participants who gave a blood sample and were free of diagnosed skin cancer up to and including the

Received 11/16/04; revised 3/22/05; accepted 4/7/05. Grant support: NIH grants CA97746 and CA87969 and Harvard Specialized Programs of Research Excellence in Skin Cancer. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Requests for reprints: Jiali Han, Channing Laboratory, Harvard Medical School, 181 Longwood Avenue, Boston, MA 02115. Phone: 617-525-2098; Fax: 617-525-2008. E-mail: jiali.han@channing.harvard.edu Copyright D 2005 American Association for Cancer Research.

Cancer Epidemiol Biomarkers Prev 2005;14(6). June 2005 Downloaded from cebp.aacrjournals.org on November 18, 2012 Copyright 2005 American Association for Cancer Research

DOI:10.1158/1055-9965.EPI-04-0846
1540 XPD Gene, Sun Exposure, and Skin Cancer questionnaire cycle in which the case was diagnosed. One control was matched to each case by year of birth (F1 year) and self-reported race (Caucasian, Asian, Hispanic, African American, and unknown). More than 95% of cases and controls were Caucasian. At the time we selected cases and controls, 47 cases and 69 controls were deceased. To obtain additional information by supplementary questionnaires, we randomly selected a second matched living control when the first control was deceased and collected supplementary questionnaires from these second living controls. The nested case-control study consisted of 219 melanoma cases (including 77 in situ cases), 286 SCC cases, 300 BCC cases, and 874 matched controls. We mailed to 758 living cases and 804 living controls a supplementary questionnaire on lifetime sun exposure and other skin cancer risk factors; 695 cases responded, 15 cases refused to participate, and 48 cases did not respond after three mailings (participation rate, 92%). Among controls, 713 responded, 9 refused, and 82 did not respond (participation rate, 89%). The study protocol was approved by the Committee on Use of Human Subjects of the Brigham and Womens Hospital (Boston, MA). Exposure Data. Information regarding skin cancer risk factors was obtained from the prospective biennial questionnaires and the retrospective supplementary questionnaire. Information on natural hair color and childhood and adolescent tendency to sunburn or tan was asked in the 1982 prospective questionnaire (ethnic group in the 1992 questionnaire). The retrospective supplementary questionnaire consisted of questions in three major areas: (a ) pigmentation, constitutional, and susceptibility factors; (b ) history of residence (states and towns), sun exposure habits, and severe sunburns at different ages; and (c ) family history of skin cancer (father, mother, and siblings). In addition, the 11 states of residence of cohort members at baseline were grouped into three regions: Northeast (Connecticut, Massachusetts, Maryland, New Jersey, New York, and Pennsylvania), North Central (Michigan and Ohio), and West and South (California, Texas, and Florida). To estimate sunlight exposure for each subject, a UV database for 50 U.S. states was developed. The database used reports from the Climatic Atlas of the United States, which reported mean daily solar radiation (in Langleys) at the earths surface for weather stations around the country (17). The records of average annual solar radiation for January and July were extracted to represent winter and summer radiation, respectively. The mean solar radiation for each residence was derived from the UV values measured at the nearest weather station, and both summer and winter radiation indices were developed for each residence. A cumulative lifetime sun exposure was developed by combining the UV database and the information obtained from the supplementary questionnaires. Questions about sun exposure while wearing a bathing suit were used to define a cumulative lifetime intermittent (recreational) sun exposure variable for this behavior. Single Nucleotide Polymorphism Identification. There are two common nonsynonymous polymorphisms in XPD (Asp312Asn and Lys751Gln). Because these two single nucleotide polymorphisms (SNP) are not in complete linkage disequilibrium in the database and have potentially functional relevance, we genotyped both of them in our study. The XPD gene was resequenced by the National Institute of Environmental Health Sciences Environmental Genome Project at the University of Washington (http:/ /egp.gs.washington.edu/ data/ercc2/) on a subset of 90 samples from the NIH DNA Polymorphism Discovery Resource (18). Among the 136 polymorphisms identified across the genomic region of the XPD gene, we selected 86 polymorphisms with allele frequency over 1%. Based on these 86 polymorphisms, seven common haplotypes (>2% frequency) were inferred by the partition-ligation expectation-maximization algorithm of Qin et al. (19). The threshold of 2% was set to ascertain potential common (>5%) Caucasian-specific haplotypes from this mixed population (20). Five haplotype-tagging SNPs (htSNP) were selected by the BEST algorithm (21) to tag these seven common haplotypes and were genotyped in this study. Laboratory Assays. Genotyping was done by the 5V nuclease assay (TaqMan), using the ABI PRISM 7900HT Sequence Detection System (Applied Biosystems, Foster City, CA), in 384-well format. TaqMan primers and probes were designed using the Primer Express Oligo Design software version 2.0 (ABI PRISM). Laboratory personnel were blinded to casecontrol status and blinded quality-control samples were inserted to validate genotyping procedures; concordance for the blinded samples was 100%. Primers, probes, and conditions for genotyping assays are available on request. Statistical Methods. To avoid potential population stratification, we excluded one Asian melanoma case and one control, one Hispanic SCC case and two controls, and one African American control. We used a m2 test to assess whether the XPD genotypes were in Hardy-Weinberg equilibrium and to determine P s for differences in haplotype frequencies between cases and controls. Unconditional logistic regression was employed to calculate odds ratio (OR) and 95% confidence interval (95% CI) to assess the risk of skin cancer for XPD genotypes among all women. A test for trend was calculated across the three genotypes for each polymorphism. We used a likelihood ratio test to evaluate heterogeneity in the effects of the XPD genotypes on different types of skin cancer in polytomous logistic regression models (22). To summarize multiple variables, we constructed a multivariate confounder score to create a constitutional susceptibility score for skin cancer (23). Briefly, we applied the logistic regression coefficients from a multivariate model, including age, race, natural skin color, natural hair color, child or adolescent tendency to burn, and the number of palpably raised moles on arms, to each individuals values for the latter four of these variables and summed the values to compute a susceptibility risk score in the logit scale. We used median value of this score among controls to define women with low and high constitutional susceptibility. The number of severe lifetime sunburns and cumulative sun exposure with a bathing suit were categorized into tertiles based on the distribution of controls. To evaluate interactions between the environmental exposures and the XPD genotypes, we modeled them as ordinal variables to test significance of a single multiplicative interaction term. We modeled the genotype into three levels (homozygous wild-type, heterozygotes, and homozygous variants). For environmental risk factors, we modeled the number of lifetime severe sunburns into three levels, cumulative sun exposure with a bathing suit into three levels, and constitutional susceptibility score as a dichotomous variable. All statistical tests were two sided.

Results
Descriptive Characteristics of Cases and Controls. Detailed description of characteristics of cases and controls will be reported elsewhere.7 In brief, the mean age at diagnosis of melanoma cases was 63.4 years and that of SCC and BCC cases was 64.7 and 64.0 years, respectively. Women in the West and South regions were more likely to be diagnosed with SCC or BCC compared with those in Northeast. Cases of each type of skin cancer were more likely to have used sunlamps or

7 Han et al. Risk factors of skin cancer: a nested case-control study within the Nurses Health Study, in preparation.

Cancer Epidemiol Biomarkers Prev 2005;14(6). June 2005 Downloaded from cebp.aacrjournals.org on November 18, 2012 Copyright 2005 American Association for Cancer Research

DOI:10.1158/1055-9965.EPI-04-0846
Cancer Epidemiology, Biomarkers & Prevention 1541 attended tanning salons. A family history of skin cancer was a risk factor for the three types of skin cancer. Cases of each type of skin cancer were more likely to have had higher cumulative sun exposure with a bathing suit, more lifetime severe sunburns that blistered, and higher constitutional susceptibility risk score (Table 1). Associations of XPD Asp312Asn and Lys751Gln with Skin Cancer Risk. The genotype distributions of the two polymorphisms were in Hardy-Weinberg equilibrium among controls. The two polymorphisms were in linkage disequilibrium in controls (D V = 0.77, P = 8.7 1084; r 2 = 0.50, P = 8.4 1043). Genotype concordance between the two polymorphisms was 76.0% in controls, which is consistent with a previous report (24). The linkage disequilibrium statistics and genotype concordance in each case series was similar to those in controls. Inverse associations of the 751Gln and 312Asn alleles with the risks of melanoma and SCC were observed (Table 2). No association was observed between these two polymorphisms and BCC risk. There was no significant heterogeneity in the main effect of each polymorphism on the three types of skin cancer (Table 2). We also evaluated the combined effect of both polymorphisms. We grouped our population into three genotype categories, double wild-type (carriage of both Asp/Asp and Lys/Lys genotypes), double homozygous variants (carriage of both Asn/Asn and Gln/Gln genotypes), and others. The results of this combined analysis were similar to the results of the individual polymorphism analysis (Table 2). Interactions between XPD Polymorphisms and Risk Factors on Melanoma Risk. We evaluated gene-environment interactions between the two nonsynonymous polymorphisms and the number of lifetime severe sunburns, cumulative sun exposure while wearing a bathing suit, and constitutional susceptibility score on skin cancer risk (Table 3). An interaction was observed between the number of lifetime severe sunburns and the Lys751Gln polymorphism on melanoma risk (P for interaction = 0.03; Table 3). The number of lifetime severe sunburns was significantly associated with an increased risk of melanoma among women with the 751Lys/Lys genotype (z5 versus never, OR, 3.26; 95% CI, 1.39-7.63), and this excess risk was attenuated among those who carried the Gln allele. The significantly inverse association of the 751Gln allele with melanoma risk was limited to women with five or more lifetime sunburns (P for trend = 0.02), whereas no association between the 751Gln allele and melanoma risk was observed among women who had four or fewer lifetime sunburns. A similar interaction pattern was also seen between the number of lifetime severe sunburns and the polymorphism Asp312Asn on melanoma risk (P for interaction = 0.03). We observed a similar interaction pattern between cumulative sun exposure with a bathing suit and the Lys751Gln polymorphism on melanoma risk (Table 3). Cumulative sun exposure with a bathing suit was significantly associated with an increased risk of melanoma among women with the 751Lys/Lys genotype (highest tertile versus lowest tertile, OR, 4.84; 95% CI, 2.16-10.82), which was attenuated among those who carried the Gln allele. We observed an inverse association of the 751Gln allele with melanoma risk among the women in the highest tertile of sun exposure with a bathing suit (P for trend = 0.06) but not among those in the lowest or intermediate tertile (P for interaction = 0.04). No significant interaction was observed between the Asp312Asn polymorphism and cumulative sun exposure with a bathing suit on melanoma risk. We also observed interactions between the constitutional susceptibility score and both polymorphisms Lys751Gln and Asp312Asn on melanoma risk (Table 3). A significantly inverse trend between the 751Gln allele and melanoma risk (P for trend = 0.008) was only seen among women with low susceptibility score but not among those with high score (P for interaction = 0.02). A similar interaction pattern was observed for the Asp312Asn polymorphism (P for interaction = 0.08). We also evaluated the combined effect of both polymorphisms in the gene-environment interactions. Compared with the individual polymorphism, similar patterns of interactions were observed when we assessed the interactions of three genotype groups (double wild-type, double homozygous variants, and others) with history of sunburns ( P for interaction = 0.007), cumulative sun exposure while wearing a bathing suit (P for interaction = 0.11), and constitutional susceptibility score (P for interaction = 0.14). There was no significant difference in the Lys751Gln and Asp312Asn genotype distributions in different categories of exposures among controls (i.e., the interactions we observed were not driven by controls). In addition, we did not observe any interaction between the genotypes and geographic region on melanoma risk. No significant interactions were observed between the two polymorphisms and the above exposure risk factors on SCC or BCC risk. XPD Haplotypes and Skin Cancer Risk. We did exploratory analysis of the XPD haplotype structure. There appears two haplotype blocks in the XPD gene based on the resequencing data on 90 samples from the Environmental Genome Project. We evaluated the long-range haplotype across the two blocks. Five htSNPs were selected, two htSNPs in the first block (G1305C and G7988A), one in the second block (T19938C), and two in the linker region between the two blocks (T12733G and C12799T). Nine common inferred haplotypes with allele frequency >5% in controls accounted for 83.3% of the alleles of controls in the present study (Table 4). We observed that three different common haplotypes were less common in the cases of melanoma, SCC, and BCC than controls, respectively.

Table 1. Characteristics of skin cancer cases and controls in the nested case-control study
Characteristics Age at diagnosis (mean, years) Geographic region at baseline (%) Northeast North central West and South Sunlamp use or tanning salon attendance (%) Family history of skin cancer (%) Highest tertile of cumulative sun exposure with a bathing suit (%) No. lifetime severe sunburns (mean) Median above constitutional susceptibility risk score (%) Controls (n = 870) 64.5 55.2 23.4 21.4 10.0 25.1 33.4 5.4 49.9 Melanoma cases (n = 218) 63.4 58.0 16.9 25.1 19.2 36.5 53.3 9.6 75.8 SCC cases (n = 285) 64.7 51.7 17.1 31.1 14.3 35.7 46.1 7.8 72.0 BCC cases (n = 300) 64.0 49.3 20.3 30.3 14.7 42.7 42.6 8.2 68.3

Cancer Epidemiol Biomarkers Prev 2005;14(6). June 2005 Downloaded from cebp.aacrjournals.org on November 18, 2012 Copyright 2005 American Association for Cancer Research

DOI:10.1158/1055-9965.EPI-04-0846
1542 XPD Gene, Sun Exposure, and Skin Cancer Table 2. XPD genotypes and skin cancer risk
Genotype Controls Melanoma Cases OR* G7988A (Asp Asn) Asp/Asp Asp/Asn Asn/Asn P for trend Heterogeneityb A20337C (Lys Gln) Lys/Lys Lys/Gln Gln/Gln P for trend Heterogeneityb
751 312

SCC Cases OR* ORc

BCC Cases OR* ORc

ORc

342 373 121 0.27 295 415 134 0.07

88 1.00 1.00 128 1.00 1.00 104 1.00 1.00 99 1.03 (0.75-1.43) 1.02 (0.72-1.44) 115 0.82 (0.62-1.10) 0.80 (0.59-1.09) 149 1.32 (0.98-1.76) 1.26 (0.93-1.71) 19 0.61 (0.36-1.05) 0.65 (0.37-1.15) 37 0.82 (0.54-1.24) 0.81 (0.52-1.26) 32 0.86 (0.55-1.34) 0.88 (0.55-1.41) 0.19 0.27 0.21 0.19 0.84 0.87

81 1.00 1.00 126 1.00 1.00 98 1.00 1.00 99 0.88 (0.63-1.23) 0.89 (0.62-1.26) 112 0.63 (0.47-0.85) 0.62 (0.45-0.85) 141 1.03 (0.77-1.39) 1.00 (0.73-1.37) 23 0.63 (0.38-1.05) 0.67 (0.39-1.15) 42 0.73 (0.49-1.10) 0.73 (0.48-1.12) 47 1.06 (0.70-1.59) 1.06 (0.69-1.62) 0.09 0.16 0.02 0.03 0.78 0.84

Combined effect of Asp312Asn and Lys751Gln Asp/Asp + Lys/Lys 239 66 1.00 1.00 101 1.00 1.00 80 1.00 1.00 Others 487 115 0.85 (0.60-1.20) 0.86 (0.59-1.24) 144 0.70 (0.52-0.94) 0.68 (0.49-0.93) 169 1.04 (0.76-1.41) 1.03 (0.75-1.43) Asn/Asn + Gln/Gln 88 15 0.62 (0.34-1.15) 0.69 (0.36-1.32) 30 0.81 (0.50-1.30) 0.81 (0.49-1.34) 25 0.84 (0.50-1.41) 0.86 (0.50-1.47) P for trend 0.12 0.23 0.09 0.10 0.70 0.75 0.39 Heterogeneityb
NOTE: The number of participants does not sum to total women because of missing data on genotype. *Unconditional logistic regression adjusted for the matching variables: age and race (Caucasian or unknown). cUnconditional logistic regression adjusted for the matching variables, constitutional susceptibility score, family history of skin cancer, the number of lifetime severe sunburns which blistered (0, 1-5, 6-11, or >11), sunlamp use or tanning salon attendance (yes/no), cumulative sun exposure while wearing a bathing suit, and geographic region. bLikelihood ratio test to evaluate heterogeneity in the effects of the XPD genotypes on different types of skin cancer in polytomous logistic regression models adjusted for variables in the second multivariate model.

Discussion
In this nested case-control study, we observed that the XPD Gln allele was associated with decreased risks of melanoma and SCC along with finding that the effect of the 751Gln allele on melanoma risk was modified by the number of lifetime severe sunburns, cumulative sun exposure while wearing a bathing suit, and constitutional susceptibility score. Similar main effect and interactions were also observed for the Asp312Asn in melanoma risk. The nested case-control design, high follow-up rate, and high response rate for the retrospective supplementary questionnaire strengthen the validity of this study. Several studies have examined the functional significance of the two XPD nonsynonymous polymorphisms Asp312Asn and Lys751Gln using DNA adducts and UV-induced photoproducts, which are mainly repaired by the NER. The 312Asn and 751 Gln alleles were associated with decreased DRC for removal of UV photoproducts (25) and benzopyrene diol epoxide-DNA adducts (26) by host cell reactivation assays. Hemminki et al. (27) reported that the combined 312Asn/Asn and 751Gln/Gln genotype was associated with reduced DRC of UV-induced cyclobutane dimers in human skin in situ and the 751Gln allele was associated with reduced repair among individuals ages z50 years. Consistent with these observations, a positive association was observed of the 312Asn or 751Gln allele with increased aromatic DNA adduct level in peripheral lymphocytes (28). Seker et al. (29) reported that the 312Asn allele was associated with enhanced apoptotic response to UV. These authors did not detect any effect of the two polymorphisms in in vitro p53 binding or DRC. Nevertheless, the increased UVinduced apoptosis may reflect slightly impaired DRC (i.e., the subtle reduction in DRC may cause the accumulation of excess DNA damage and in turn trigger apoptosis). It was reported that unrepaired UV-induced damage on the transcribed strand blocks transcription, triggers the accumulation of p53, and induces apoptosis (30). We observed that the 751Gln allele was associated with a decreased risk of melanoma, and this decreased risk was more apparent among women with five or more severe lifetime
751

sunburns or those in the highest tertile of cumulative sun exposure with a bathing suit. Similar interactions were also observed for the 312Asn allele. A suggestive positive association of the variant allele with melanoma risk was observed among women who never had severe blistering sunburns or were in low category of sun exposure. Given previous data suggesting the reduced DRC of the 751Gln and 312Asn alleles, our data showed that the positive association of the variant allele and melanoma risk occurred in the context of low levels of DNA damage. However, when challenged by an overwhelmingly high dose of exposure measured as five or more sunburns or high level of cumulative sun exposure with a bathing suit, melanocytes with impaired DRC may accumulate excess DNA damage, inducing apoptosis and in turn decreasing the risk of melanoma. We also observed a significant interaction between the constitutional susceptibility score and the 312Asn and 751Gln variants. A significant association of the 751Gln or 312Asn allele with melanoma risk was seen among women with low score but not among those with high score. In addition, we also observed a decreased risk of SCC associated with the 751Gln allele. Although melanocytes are less sensitive to apoptosis than keratinocytes because of the lower levels of cell cycle and proliferation and the higher content of antiapoptotic proteins (31, 32), we observed that the 312Asn and 751Gln allele were associated with a decreased risk of melanoma potentially due to apoptosis. As suggested by the previous functional data (29), the two 312Asn and 751Gln alleles, which are in linkage disequilibrium, may enhance the cellular apoptotic response to UV exposure. This is consistent with our data that the decreased risk was stronger among women with high sun exposure as measured by lifetime sunburns and cumulative sun exposure while wearing a bathing suit. We did exploratory analysis of XPD long-range haplotype. Across two haplotype blocks, we inferred nine common haplotypes (>5% allele frequency) from five htSNPs. We used the data on 90 multiethnic samples from the Environmental Genome Project to infer haplotype structures. Two recent studies of haplotype variation in different ethnic groups (33, 34)

Cancer Epidemiol Biomarkers Prev 2005;14(6). June 2005 Downloaded from cebp.aacrjournals.org on November 18, 2012 Copyright 2005 American Association for Cancer Research

DOI:10.1158/1055-9965.EPI-04-0846
Cancer Epidemiology, Biomarkers & Prevention 1543 Table 3. Interaction between XPD polymorphisms and risk factors on melanoma risk
Cases/controls OR (95% CI) Cases/controls Lys Lys/Lys Never 1-4 z5 9/84 16/75 37/63 1.00 1.67 (0.67-4.16) 3.26 (1.39-7.63) Lys/Gln Lifetime severe sunburns* 12/105 1.13 (0.44-2.94) 22/115 1.54 (0.65-3.65) 42/104 2.23 (0.98-5.06) P for interaction = 0.03 OR (95% CI)
751

Cases/controls

OR (95% CI)

P for trend

Gln Gln/Gln 6/33 6/28 7/36 1.73 (0.54-5.57) 2.03 (0.63-6.53) 1.07 (0.35-3.27) 0.46 0.86 0.02

Low Intermediate High

10/88 17/74 39/66

Cumulative sun exposure with a bathing suitc 1.00 20/97 1.82 (0.78-4.24) 5/35 1.92 (0.80-4.59) 18/110 1.43 (0.61-3.39) 6/36 4.84 (2.16-10.82) 42/118 3.12 (1.43-6.82) 8/31 P for interaction = 0.04 1.00 1.82 (1.04-3.20) Constitutional susceptibility scoreb 23/217 0.50 (0.27-0.94) 76/198 2.03 (1.19-3.46) P for interaction = 0.02 Asp312Asn 3/67 20/67

1.53 (0.47-5.00) 1.42 (0.46-4.38) 2.13 (0.73-6.23)

0.24 0.41 0.06

Low High

25/139 56/156

0.23 (0.07-0.81) 1.69 (0.84-3.39)

0.008 0.92

Asp/Asp Never 1-4 z5 9/96 20/89 38/77 1.00 2.11 (0.88-5.07) 3.17 (1.37-7.31)

Asp/Asn Lifetime severe sunburn* 12/100 1.40 (0.54-3.63) 21/98 1.97 (0.83-4.69) 47/92 3.17 (1.41-7.15) P for interaction = 0.03

Asn/Asn 6/26 5/31 4/32 2.59 (0.78-8.59) 1.52 (0.45-5.16) 0.84 (0.23-3.09) 0.15 0.63 0.07

Low Intermediate High

13/92 17/95 42/79

Cumulative sun exposure with a bathing suitc 1.00 22/97 1.74 (0.80-3.80) 2/29 1.22 (0.54-2.74) 22/89 1.95 (0.89-4.30) 4/38 3.98 (1.91-8.32) 40/107 2.55 (1.23-5.30) 9/27 P for interaction = 0.26 1.00 2.82 (1.61-4.95) Constitutional susceptibility scoreb 24/187 0.83 (0.44-1.57) 75/186 3.09 (1.78-5.36) P for interaction = 0.08 Lys751Gln + Asp312Asn 2/68 17/53

0.70 (0.14-3.42) 0.67 (0.19-2.28) 2.72 (0.98-7.52)

0.81 0.79 0.18

Low High

22/171 66/171

0.21 (0.05-0.91) 2.64 (1.26-5.57)

0.06 0.88

312

Asp/Asp +

751

Lys/Lys 1.00 2.72 (0.88-8.42) 5.25 (1.82-15.14)

Others Lifetime severe sunburns* 15/128 1.83 (0.61-5.46) 27/134 2.51 (0.89-7.04) 50/117 3.65 (1.34-9.99) P for interaction = 0.007

312

Asn/Asn +

751

Gln/Gln 3.51 (0.85-14.57) 3.34 (0.76-14.76) 1.25 (0.26-6.00) 0.13 0.89 0.03

Never 1-4 z5

5/69 13/61 31/52

5/20 4/17 3/23

Low Intermediate High

9/71 11/63 32/53

Cumulative sun exposure with a bathing suitc 1.00 24/124 1.55 (0.66-3.63) 2/18 1.29 (0.48-3.45) 24/124 1.59 (0.67-3.75) 4/28 4.68 (1.96-11.18) 49/133 2.79 (1.24-6.27) 6/18 P for interaction = 0.11 1.00 2.46 (1.29-4.70) Constitutional susceptibility scoreb 29/249 0.68 (0.35-1.31) 86/238 2.30 (1.27-4.18) P for interaction = 0.14 2/47 13/41

1.45 (0.27-7.67) 0.99 (0.27-3.67) 2.87 (0.85-9.74)

0.36 0.82 0.17

Low High

17/115 49/124

0.26 (0.06-1.20) 2.31 (0.99-5.41)

0.08 0.88

NOTE: The number of participants does not sum to total women because of missing data on genotype. *Unconditional logistic regression adjusted for the matching variables, constitutional susceptibility score, family history of skin cancer, sunlamp use or tanning salon attendance (yes/no), cumulative sun exposure with a bathing suit (tertile), and geographic region. cUnconditional logistic regression adjusted for the matching variables, constitutional susceptibility score, family history of skin cancer, the number of lifetime severe sunburns which blistered (0, 1-5, 6-11, >11), sunlamp use or tanning salon attendance (yes/no), and geographic region. bUnconditional logistic regression adjusted for the matching variables, family history of skin cancer, the number of lifetime severe sunburns which blistered (0, 1-5, 611, >11), sunlamp use or tanning salon attendance (yes/no), cumulative sun exposure with a bathing suit (tertile), and geographic region.

reported substantial conservation of haplotypes among ethnicities with the fewest population-specific common haplotypes in Caucasians. In addition, using the BEST algorithm to tag the haplotypes of 105 genes, Sebastiani et al. showed that an average of 95% (and in most cases 100%) of the htSNPs in

the European American sample is a subset of the htSNPs of the African American sample (21). These suggest that use of the data from the Environmental Genome Project to infer haplotype in our study of mainly Caucasian individuals is appropriate to define common Caucasian haplotypes. We

Cancer Epidemiol Biomarkers Prev 2005;14(6). June 2005 Downloaded from cebp.aacrjournals.org on November 18, 2012 Copyright 2005 American Association for Cancer Research

DOI:10.1158/1055-9965.EPI-04-0846
1544 XPD Gene, Sun Exposure, and Skin Cancer Table 4. Frequencies of inferred XPD haplotypes in cases and controls
SNP 5 22 40 41 76 Allele frequency Melanoma cases (%), n = 348 9.2 9.5 10.0 8.3 8.6 21.9 5.2 6.0 5.6 P SCC cases (%), n = 458 16.5 10.5 6.5 6.9 6.5 18.1 5.5 7.1 6.8 P BCC cases (%), n = 460 13.9 7.1 8.9 10.7 8.9 18.2 5.9 5.5 3.4 P Common controls (%), n = 1,384 13.4 9.0 7.4 11.1 7.3 17.7 5.4 5.7 6.3

1 2 3 4 5 6 7 8 9

1 0 0 0 1 1 0 0 1

0 0 0 1 0 0 1 1 1

0 0 1 0 1 1 1 1 0

0 0 0 0 1 0 0 0 1

0 0 0 0 0 0 1 0 0

0.02 0.82 0.15 0.11 0.44 0.09 0.88 0.83 0.64

0.13 0.37 0.52 0.01 0.58 0.85 0.94 0.30 0.69

0.80 0.18 0.32 0.84 0.29 0.80 0.67 0.93 0.01

NOTE: Haplotype frequencies from the observed genotypes were estimated using partition-ligation expectation-maximization. The P s for differences in haplotype frequencies between cases and controls were determined by the two-sample proportion test incorporating SE from partition-ligation expectation-maximization output. 0, common allele; 1, rare allele.

observed that three different common haplotypes were less frequent in the cases of melanoma, SCC, and BCC risk than controls, respectively. Block-specific haplotype analysis is under investigation to better understand the associations of the polymorphisms in the regulatory regions of the XPD gene. In summary, this is the first report of interactions between two XPD nonsynonymous polymorphisms (Asp312Asn and Lys751Gln) and sun exposure on skin cancer risk. We observed an inverse association of the XPD 751Gln allele on melanoma and SCC risk and effect modification of the 751Gln allele on sun exposure related factors for melanoma risk along with the similar findings for the Asp312Asn. Additional functional data on XPD polymorphisms are warranted to elucidate the role of these variants in the development of skin cancer.

Acknowledgments
We thank Dr. Hardeep Ranu and Craig Labadie for their laboratory assistance, Rong Chen for her programming support, Dr. David Cox for his program to calculate linkage disequilibrium statistics, and the participants in the Nurses Health Study for their dedication and commitment.

References
Armstrong BK, Kricker A, English DR. Sun exposure and skin cancer. Australas J Dermatol 1997;38:S1 6. 2. Ravanat JL, Douki T, Cadet J. Direct and indirect effects of UV radiation on DNA and its components. J Photochem Photobiol B 2001;63:88 102. 3. English DR, Armstrong BK, Kricker A, Fleming C. Sunlight and cancer. Cancer Causes Control 1997;8:271 83. 4. Grossman L, Wei Q. DNA repair and epidemiology of basal cell carcinoma. Clin Chem 1995;41:1854 63. 5. Wei Q, Matanoski GM, Farmer ER, Hedayati MA, Grossman L. DNA repair and aging in basal cell carcinoma: a molecular epidemiology study. Proc Natl Acad Sci U S A 1993;90:1614 8. 6. Wei Q, Matanoski GM, Farmer ER, Hedayati MA, Grossman L. DNA repair capacity for ultraviolet light-induced damage is reduced in peripheral lymphocytes from patients with basal cell carcinoma. J Invest Dermatol 1995;104:933 6. 7. Wei Q, Matanoski GM, Farmer ER, Hedayati MA, Grossman L. DNA repair and susceptibility to basal cell carcinoma: a case-control study. Am J Epidemiol 1994;140:598 607. 8. Wood RD. Nucleotide excision repair in mammalian cells. J Biol Chem 1997; 272:23465 8. 9. de Boer J, Hoeijmakers JH. Nucleotide excision repair and human syndromes. Carcinogenesis 2000;21:453 60. 10. de Boer J, Donker I, de Wit J, Hoeijmakers JH, Weeda G. Disruption of the mouse xeroderma pigmentosum group D DNA repair/basal transcription gene results in preimplantation lethality. Cancer Res 1998;58:89 94. 11. Coin F, Marinoni JC, Rodolfo C, et al. Mutations in the XPD helicase gene result in XP and TTD phenotypes, preventing interaction between XPD and the p44 subunit of TFIIH. Nat Genet 1998;20:184 8. 12. Lehmann AR. The xeroderma pigmentosum group D (XPD) gene: one gene, two functions, three diseases. Genes Dev 2001;15:15 23. 1.

13. Wang XW, Yeh H, Schaeffer L, et al. p53 modulation of TFIIH-associated nucleotide excision repair activity. Nat Genet 1995;10:188 95. 14. Wang XW, Vermeulen W, Coursen JD, et al. The XPB and XPD DNA helicases are components of the p53-mediated apoptosis pathway. Genes Dev 1996;10:1219 32. 15. Robles AI, Wang XW, Harris CC. Drug-induced apoptosis is delayed and reduced in XPD lymphoblastoid cell lines: possible role of TFIIH in p53mediated apoptotic cell death. Oncogene 1999;18:4681 8. 16. Colditz GA, Martin P, Stampfer MJ, et al. Validation of questionnaire information on risk factors and disease outcomes in a prospective cohort study of women. Am J Epidemiol 1986;123:894 900. 17. National Oceanic and Atmospheric Administration. Climatic atlas of the United States. Washington (DC): National Oceanic and Atmospheric Administration; 1983. 18. Collins FS, Brooks LD, Chakravarti A. A DNA polymorphism discovery resource for research on human genetic variation. Genome Res 1998;8: 1229 31. 19. Qin ZS, Niu T, Liu JS. Partition-ligation-expectation-maximization algorithm for haplotype inference with single-nucleotide polymorphisms. Am J Hum Genet 2002;71:1242 7. 20. Han J, Hankinson SE, De Vivo I, et al. A prospective study of XRCC1 haplotypes and their interaction with plasma carotenoids on breast cancer risk. Cancer Res 2003;63:8536 41. 21. Sebastiani P, Lazarus R, Weiss ST, et al. Minimal haplotype tagging. Proc Natl Acad Sci U S A 2003;100:9900 5. 22. Marshall RJ, Chisholm EM. Hypothesis testing in the polychotomous logistic model with an application to detecting gastrointestinal cancer. Stat Med 1985;4:337 44. 23. Miettinen OS. Stratification by a multivariate confounder score. Am J Epidemiol 1976;104:609 20. 24. Zhou W, Liu G, Miller DP, et al. Gene-environment interaction for the ERCC2 polymorphisms and cumulative cigarette smoking exposure in lung cancer. Cancer Res 2002;62:1377 81. 25. Qiao Y, Spitz MR, Shen H, 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 9. 26. Spitz MR, Wu X, Wang Y, et al. Modulation of nucleotide excision repair capacity by XPD polymorphisms in lung cancer patients. Cancer Res 2001;61:1354 7. 27. Hemminki K, Xu G, Angelini S, et al. XPD exon 10 and 23 polymorphisms and DNA repair in human skin in situ . Carcinogenesis 2001; 22:1185 8. 28. Hou SM, Falt S, Angelini S, et al. The XPD variant alleles are associated with increased aromatic DNA adduct level and lung cancer risk. Carcinogenesis 2002;23:599 603. 29. Seker H, Butkiewicz D, Bowman ED, 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 4. 30. Ljungman M, Zhang F. Blockage of RNA polymerase as a possible trigger for U.V. light-induced apoptosis. Oncogene 1996;13:823 31. 31. Bowen AR, Hanks AN, Allen SM, et al. Apoptosis regulators and responses in human melanocytic and keratinocytic cells. J Invest Dermatol 2003;120:48 55. 32. Gilchrest BA, Eller MS, Geller AC, Yaar M. The pathogenesis of melanoma induced by ultraviolet radiation. N Engl J Med 1999;340:1341 8. 33. Gabriel SB, Schaffner SF, Nguyen H, et al. The structure of haplotype blocks in the human genome. Science 2002;296:2225 9. 34. Stephens JC, Schneider JA, Tanguay DA, et al. Haplotype variation and linkage disequilibrium in 313 human genes. Science 2001;293:489 93.

Cancer Epidemiol Biomarkers Prev 2005;14(6). June 2005 Downloaded from cebp.aacrjournals.org on November 18, 2012 Copyright 2005 American Association for Cancer Research