XPC polymorphisms and lung cancer risk
Version of Record online: 23 FEB 2005
Copyright © 2005 Wiley-Liss, Inc.
International Journal of Cancer
Volume 115, Issue 5, pages 807–813, 10 July 2005
How to Cite
Lee, G. Y., Jang, J.-S., Lee, S. Y., Jeon, H.-S., Kim, K. M., Choi, J. E., Park, J. M., Chae, M. H., Lee, W. K., Kam, S., Kim, I.-S., Lee, J.-T., Jung, T. H. and Park, J. Y. (2005), XPC polymorphisms and lung cancer risk. Int. J. Cancer, 115: 807–813. doi: 10.1002/ijc.20900
- Issue online: 11 MAY 2005
- Version of Record online: 23 FEB 2005
- Manuscript Accepted: 9 NOV 2004
- Manuscript Received: 18 JUL 2004
- National Cancer Control R & D Program 2003, Ministry of Health and Welfare (South Korea)
- lung cancer;
- genetic susceptibility
Polymorphisms in DNA repair genes may be associated with differences in the capacity to repair DNA damage and thereby influence an individual's susceptibility to smoking-related cancer. To test this hypothesis, we investigated the potential association of 7 XPC polymorphisms (–449GC, –371GA, –27GC, Val499Arg, PAT–/+, IVS11-5CA and Lys939Gln) and their haplotypes with lung cancer risk in a Korean population. XPC genotypes were determined in 432 lung cancer patients and 432 healthy controls frequency-matched for age and sex. XPC haplotypes were predicted using a Bayesian algorithm in the Phase program. The combined –27CG+CC genotype was associated with a significantly increased risk for overall lung cancer compared to the –27GG genotype (adjusted OR = 1.97, 95% CI 1.22–3.17, p = 0.005). The other 6 polymorphisms were not significantly associated with overall risk of lung cancer. When lung cancer cases were categorized by tumor histology, the –371AA genotype was associated with a significantly increased risk of squamous cell carcinoma compared to the combined –371GG and GA genotype (adjusted OR = 2.08, 95% CI 1.09–4.00, p = 0.03). The PAT–/+, IVS11-5CA and Lys939Gln polymorphisms were associated with a significantly decreased risk of small cell carcinoma (SM) under a dominant model for the polymorphic allele (adjusted OR = 0.49, 95% CI 0.29–0.82, p = 0.006; adjusted OR = 0.60, 95% CI 0.36–1.00, p = 0.05; and adjusted OR = 0.58, 95% CI 0.35–0.97, p = 0.04, respectively). Consistent with genotyping analyses, haplotype 4 (1112222) containing the PAT+/IVS11-5A/939Gln alleles was associated with a significantly decreased risk of SM (adjusted OR = 0.56, 95% CI 0.37–0.85, p = 0.007 and Bonferroni-corrected p = 0.049), whereas haplotype 5 (1122111) containing the –27C allele was associated with a significantly increased risk of SM (adjusted OR = 2.88, 95% CI 1.41–5.87, p = 0.004 and Bonferroni-corrected p = 0.028). These results suggest that XPC polymorphisms/haplotypes may contribute to genetic susceptibility for lung cancer. © 2005 Wiley-Liss, Inc.
Although cigarette smoking is the major cause of lung cancer, only a small fraction of smokers develop this disease, suggesting that genetic factors contribute to the interindividual variation in lung cancer risk. This genetic susceptibility may result from inherited polymorphisms in the genes controlling both carcinogen metabolism and repair of DNA damage.1, 2, 3
DNA repair systems play a critical role in protecting the genome from the insults of cancer-causing agents, such as those found in tobacco. In humans, more than 70 genes are involved in the 4 major DNA repair pathways: NER, base excision repair, mismatch repair and double strand break repair.4, 5 NER is a versatile repair pathway that can eliminate a wide variety of DNA lesions, including UV-induced photolesions and chemical carcinogen-induced bulky DNA adducts, such as benzo(a)pyrene-guanine adduct; and this pathway is composed of at least 2 subpathways, global genome repair and transcription-coupled repair.6, 7
Molecular epidemiologic studies have shown considerable interindividual variation for NER capacity in the general population. Individuals with suboptimal NER capacity are at increased risk of smoking-related cancers, such as lung cancer and squamous cell carcinoma of the head and neck.8, 9 Like many other phenotypic traits, variation in NER capacity may be the result of functional polymorphisms in the NER genes. Therefore, it has been hypothesized that inherited polymorphisms in the NER genes may modulate susceptibility to lung cancer. To test this hypothesis, we previously studied the contribution of polymorphisms in the NER genes XPA, XPD and XPG to the risk of lung cancer in a Korean population.10, 11, 12
The XPC protein binds tightly with HR23B (one of 2 human homologs of Saccharomyces cerevisiae NER factor RAD23), and this plays a central role in global genomic NER.13 The XPC–HR23B complex functions as an early damage detector and a molecular matchmaker for recruitment of other components of the repair apparatus to the damaged DNA in global genomic NER.14, 15 Mice defective in the XPC gene are highly prone to skin cancer following exposure to UV radiation and highly vulnerable to cancers of the internal organs, such as liver and lung, when exposed to chemical carcinogens,16, 17 suggesting that XPC is important for preventing carcinogenesis. Several polymorphisms in the XPC gene have been reported18, 19, 20, 21 and listed in public databases (e.g.,http://www.ncbi.nlm.nih.gov/SNP). While the functional effects of these polymorphisms have not been elucidated, we hypothesized that some of these variants, particularly their haplotypes, may have an effect on NER capacity, thereby modulating the susceptibility to lung cancer. To test this hypothesis, a case-control study was conducted to evaluate the association between XPC genotypes/haplotypes and lung cancer risk.
We evaluated the amino acid substitution variants Val499Arg and 21151TC in exon 8 and Lys939Gln and 33512AC in exon 15 (Genbank accession AY131066) since they are most likely to affect gene function. Among the candidate polymorphisms in the promoter region of the XPC gene, the –449GC, –371GA and –27GC (from the translation start site) polymorphisms (525GC, 603GA and 947GC, respectively; Genbank accession AY131066) were evaluated because they lead to changes in transcription factor binding sites that could possibly influence the transcription activity of the XPC promoter. Among the intronic variants, the IVS11-5CA polymorphism (30724CA, Genbank accession AY131066) in the XPC intron 11 splice acceptor site was evaluated since it exerts influence on the alternative splicing and function of XPC.20 The PAT–/+ polymorphism [24660_24664delGTAAC inspoly(AT), GenBank accession AY131066] in intron 9 was also evaluated since it has been associated with increased risk for SCC of the head and neck.21
Material and methods
Our case-control study included 432 lung cancer patients and 432 healthy controls. The method used for subject enrollment was the same as that in our previous studies.10, 12 Eligible cases included all patients (n = 447) newly diagnosed with primary lung cancer between January 2001 and February 2002 at Kyungpook National University Hospital. There were no age, gender, histologic or stage restrictions; but 15 patients with a prior history of cancer were excluded. The authors and a trained coordinator explained the objective and importance of the study, and we included all patients with histologically confirmed primary lung cancer. All patients signed informed consent for blood sample collection. Cases included 210 (48.6%) SCCs, 141 (32.6%) adenocarcinomas, 73 (16.9%) SMs and 8 (1.9%) large cell carcinomas. The demographics and clinical characteristics of cases were consistent with the statistical data of a nationwide lung cancer survey conducted in 1998 by the Korean Academy of Tuberculosis and Respiratory Disease.22 Controls were randomly selected from a pool of healthy volunteers who visited the general-health check-up center at Kyungpook National University Hospital during the same period. They were frequency-matched (1:1) to cases based on sex and age (±5 years). All cases and controls were ethnically Korean and resided in Daegu or the surrounding area. A detailed questionnaire was completed for each case and control by a trained interviewer. The questionnaire included information on average number of cigarettes smoked daily and number of years subjects had been smoking. For smoking status, a person who had smoked at least once a day for >1 year in his or her lifetime was regarded as a smoker. A former smoker was defined as one who had stopped smoking at least 1 year before diagnosis for patients and 1 year before the study began for controls. Cumulative cigarette dose (pack-years) was calculated by the following formula: pack-years = (packs per day) × (years smoked).
Genomic DNA was extracted from peripheral blood lymphocytes by proteinase K digestion and phenol/chloroform extraction. The XPC–449GC, –371GA, –27GC, Val499Arg and IVS11-5CA genotypes were determined using a PCR-RFLP assay. PCR primers were designed based on the Genbank reference sequence. PCR primers for –449GC, –371GA, –27GC, Val499Arg and IVS11-5CA polymorphisms were 5′-GGAATCCTGACAGCTCCATG-3′ (forward) and 5′-CACCCTGTCCTTTCCGTAG (mutated AG)C-3′ (reverse), 5′-CAGTTCCTTGTTTCCTTCAC-3′ (forward) and 5′-GTTTCCGGAGATTGACGTTG-3′ (reverse), 5′-ACAAGAGCAACGTCAATCT-3′ (forward) and 5′-TTGCTCTTGGCCTTGGATT-3′ (reverse), 5′-CGGCTCTGATTTTGAGCTCTCC-3′ (forward) and 5′-GCTTGAAGAGCTTGAGGATGG(mutated CG)C-3′ (reverse) and 5′-AATGCTGACTTGCTCACGC-3′ (forward) and 5′-GTAGCCGCCATGGAAATCAA-3′ (reverse), respectively. PCRs were performed in a 20 μl reaction volume containing 100 ng genomic DNA, 10 pM of each primer, 0.2 mM dNTPs, 8 mM TRIS HCl (pH 8.3), 0.5 μg/μl BSA, 2.5 mM MgCl2 and 1 unit of Taq polymerase (Takara Shuzo, Shiga, Japan). PCR cycle conditions consisted of an initial denaturation step at 94°C for 5 min, followed by 35 cycles of 30 sec at 94°C, 30 sec at 56°C for –449GC, 58°C for –371GA and –27GC, 55°C for Val499Arg and 57°C for IVS11-5CA; 30 sec at 72°C; and final elongation at 72°C for 5 min. PCR products were digested overnight with the appropriate restriction enzymes (New England Biolabs, Beverly, MA) at 37°C. Restriction enzymes for –449GC, –371GA, –27GC, Val499Arg and IVS11-5CA genotypes were HhaI, XbaI, NruI, HhaI and Hinp1I, respectively. Digested PCR products were resolved on 6% acrylamide gel (–449GC, –27GC, Val499Arg and IVS11-5CA) or 1.2% agarose gel (–371GA). The PAT–/+ and Lys939Gln genotypes were determined using the PCR and PCR-RFLP methods, respectively, as described previously.19 Genotyping analysis was performed blind with respect to case/control status. To avoid any incorrect assessment of genotype, 2 authors independently performed genotyping for all subjects, and the results were 100% concordant. To confirm the genotyping results, selected PCR-amplified DNA samples (n = 2 for each genotype) were examined by DNA sequencing.
Transcription activity analysis
Fragments of the XPC promoter region (from –973 to –1, the translation start site counted as +1) were synthesized by PCR using genomic DNA from donors carrying either the wild-type allele or the polymorphic allele at position –499, –372 or –27 from the translation start site. PCR primers for the XPC promoter were 5′-CAGAAGTCCTCTCTGAGGAG-3′ (forward) and 5′-GTTGCTTGTCTGGGCAAATT-3′ (reverse). PCR products were inserted upstream of the luciferase gene in pGL3-basic plasmid (Promega, Madison, WI), and the correct sequence of all clones was verified by DNA sequencing. Promoter activity was measured using the Luciferase Reporter Assay System (Promega). Chinese hamster ovary cells were grown in minimal essential medium supplemented with 10% FBS. Cells (1 × 105) were plated in a 6-well plate the day before transfection so that they were approximately 60% confluent by the next day. The pRL-SV40 and pGL3-basic plasmid with the synthesized fragment of the XPC promoter region were cotransfected using Lipofectin reagent (Invitrogen, Carlsbad, CA). The pRL-40 vector that provided constitutive expression of Renilla luciferase was used as an internal control to correct for differences in transfection and harvesting efficiency. Cells were collected 48 hr after transfection, and cell lysates were prepared according to Promega's instruction manual. Luciferase activity was measured using a Lumat LB953 luminometer (EG&G Berthold, Wildbad, Germany), and the results were normalized using the activity of Renilla luciferase. The experiment was performed 4 times in triplicate, and the results are reported as means ± SD.
Cases and controls were compared using Student's t-test for continuous variables and the χ2 test for categorical variables. Hardy-Weinberg equilibrium of alleles at individual loci was tested with a goodness-of-fit χ2 test, with 1 degree of freedom to compare the observed genotype frequencies with the expected genotype frequencies among subjects. We examined the widely used measure of LD between pairs of biallelic loci, Lewontin's D′.23 Haplotypes and their frequencies were estimated based on the Bayesian algorithm using the Phase program,24 (available at http://www.stat.washington.edu/stephens/phase.html). Unconditional logistic regression analysis was used to calculate ORs and 95% CIs, with adjustment for possible confounders (sex and smoking status, as nominal variables; age and pack-years, as continuous variables). Referent and 3 alternative models (codominant, dominant and recessive for the minor allele) were applied in the analyses. When multiple comparisons were made, pc values were also calculated for multiple testing using Bonferroni's inequality method. The interaction between the genotype and smoking was tested both by a logistic regression model including the interaction term between the genotype and smoking (pack-years of smoking or smoking exposure level) and by stratification analysis. For these analyses, subjects were categorized into 3 groups by smoking exposure: never-smokers, ever-smokers of <39 pack-years and ever-smokers of ≥39 pack-years (median pack-years of ever-smokers). The interaction term between the genotype and smoking was not statistically significant, and this was removed from the logistic regression model. All analyses were performed using Statistical Analysis Software for Windows, version 8.12 (SAS Institute, Cary, NC).
The demographics of cases and controls are shown in Table I. There were no significant differences in mean age and sex distribution between cases and controls, suggesting that the matching based on these 2 variables was adequate. The case group had a higher prevalence of current smokers than the control group (p < 0.001), and the number of pack-years for smokers was significantly higher among cases than controls (39.9 ± 17.9 vs. 34.4 ± 17.6 pack-years, p < 0.001). These differences were controlled in the later multivariate analyses.
|Variable||Cases (n = 432)||Controls (n = 432)|
|Age (years)||61.6 ± 9.0||60.9 ± 9.3|
|Male||352 (81.5)1||352 (81.5)|
|Female||80 (18.5)||80 (18.5)|
|Current||317 (73.4)||229 (53.0)|
|Former||39 (9.0)||98 (22.7)|
|Never||76 (17.6)||105 (24.3)|
|Pack-years3||39.9 ± 17.9||34.4 ± 17.6|
The genotype and polymorphic allele frequencies of the 7 XPC polymorphisms among controls and cases are shown in Table II. A total of 855 [99.0%, 426 of 432 cases (98.6%) and 429 of 432 controls (99.3%)] subjects were successfully genotyped for all XPC polymorphisms. The genotype distributions of the 7 polymorphisms among controls and cases were in Hardy-Weinberg equilibrium. The distribution of the –27GC genotypes and the frequency of the variant allele among the overall lung cancer cases were significantly different from those among controls (p = 0.002 and p = 0.01, respectively). For the other 6 polymorphisms, however, there was no significant difference in the distribution of genotypes between the overall lung cancer cases and control subjects. When lung cancer cases were categorized by tumor histology, the distributions of –27GC and PAT–/+ genotypes in SM cases significantly differed from controls (–27 GG, GC and CC genotypes, 77.5%, 22.5%, 0.0% vs. 93.0%, 6.5%, 0.5%, respectively, p < 0.001; and PAT –/–, –/+ and +/+genotypes, 56.2%, 37.0%, 6.8% vs. 38.7%, 48.1%, 13.2%, respectively, p = 0.015).
|Polymorphism||Variables||Genotype1||Polymorphic allele frequency||Number|
|−449G C||Controls||223 (51.7)||179 (41.5)||29 (6.7)||0.275||431|
|All cases||238 (55.1)||171 (39.6)||23 (5.3)||0.251||432|
|SCC||113 (53.8)||84 (40.0)||13 (6.2)||0.262||210|
|Adenoca.||79 (56.0)||58 (41.1)||4 (2.8)||0.234||141|
|Large cell ca.||7 (87.5)||1 (12.5)||0 (0.0)||0.063||8|
|SM||39 (53.4)||28 (38.4)||6 (8.2)||0.274||73|
|−371G A||Controls||243 (56.3)||166 (38.4)||23 (5.3)||0.245||432|
|All cases||236 (54.8)||163 (37.8)||32 (7.4)||0.263||431|
|SCC||111 (52.9)||78 (37.1)||21 (10.0)||0.286||210|
|Adenoca.||81 (57.4)||56 (39.7)||4 (2.8)||0.227||141|
|Large cell ca.||5 (62.5)||3 (37.5)||0 (0.0)||0.188||8|
|SM||39 (54.2)||26 (36.1)||7 (9.7)||0.278||72|
|−27G C||Controls||400 (93.0)||28 (6.5)||2 (0.5)||0.037||430|
|All cases||375 (87.4)||54 (12.6)||0 (0.0)2||0.0633||429|
|SCC||188 (90.0)||21 (10.0)||0 (0.0)||0.050||209|
|Adenoca.||124 (87.9)||17 (12.1)||0 (0.0)||0.060||141|
|Large cell ca.||8 (100.0)||0 (0.0)||0 (0.0)||0.000||8|
|SM||55 (77.5)||16 (22.5)||0 (0.0)4||0.1134||71|
|Val499Arg||Controls||31 (7.2)||187 (43.3)||214 (49.5)||0.712||432|
|(TC)||All cases||28 (6.5)||184 (42.6)||220 (50.9)||0.722||432|
|SCC||15 (7.1)||87 (41.4)||108 (51.4)||0.721||210|
|Adenoca.||6 (4.3)||65 (46.1)||70 (49.6)||0.727||141|
|Large cell ca.||0 (0.0)||3 (37.5)||5 (62.5)||0.813||8|
|SM||7 (9.6)||29 (39.7)||37 (50.7)||0.705||73|
|PAT−/+||Controls||167 (38.7)||208 (48.1)||57 (13.2)||0.373||432|
|All cases||192 (44.4)||179 (41.4)||61 (14.1)||0.348||432|
|SCC||91 (43.3)||93 (44.3)||26 (12.4)||0.345||210|
|Adenoca.||58 (41.1)||58 (41.1)||25 (17.7)||0.383||141|
|Large cell ca.||2 (25.0)||1 (12.5)||5 (62.5)4||0.6883||8|
|SM||41 (56.2)||27 (37.0)||5 (6.8)5||0.2536||73|
|IVS11-5CA||Controls||152 (35.2)||222 (51.4)||58 (13.4)||0.391||432|
|All cases||167 (38.7)||202 (46.9)||62 (14.4)||0.378||431|
|SCC||82 (39.2)||101 (48.3)||26 (12.4)||0.366||209|
|Adenoca.||50 (35.5)||65 (46.1)||26 (18.4)||0.415||141|
|Large cell ca.||1 (12.5)||2 (25.0)||5 (62.5)4||0.7507||8|
|SM||34 (46.6)||34 (46.6)||5 (6.8)||0.3018||73|
|Lys939Gln||Controls||150 (34.8)||222 (51.5)||59 (13.7)||0.394||431|
|(AC)||All cases||168 (39.0)||198 (45.9)||65 (15.1)||0.381||431|
|SCC||83 (39.5)||101 (48.1)||26 (12.4)||0.364||210|
|Adenoca.||50 (35.7)||63 (45.0)||27 (19.3)||0.418||140|
|Large cell ca.||1 (12.5)||2 (25.0)||5 (62.5)4||0.7507||8|
|SM||34 (46.6)||32 (43.8)||7 (9.6)||0.315||73|
Table III shows the lung cancer risk related to the 7 XPC polymorphisms. Adjusted ORs and 95% CIs were calculated using the more common genotype as the reference group. Compared to the combined –371 GG and GA genotypes, the –371AA genotype was associated with a significantly increased risk of SCC (adjusted OR = 2.08, 95% CI 1.09–4.00, p = 0.03) but not with risk of adenocarcinoma or SM. For the –27GC polymorphism, the combined –27CG+CC genotype was associated with a significantly increased risk for overall lung cancer compared to the –27GG genotype (adjusted OR = 1.97, 95% CI 1.22–3.17, p = 0.005). The risk effect of the combined –27CG+CC genotype on lung cancer was observed for 3 major histologic types, though this was statistically significant only for SM (adjusted OR = 3.98, 95% CI 2.01–7.88, p < 0.001). The PAT–/+, IVS11-5CA and Lys939Gln genotypes were significantly associated only with risk of SM. Individuals with at least one PAT+ allele were at a significantly decreased risk of SM compared to those with the PAT–/– genotype (adjusted OR = 0.49, 95% CI 0.29–0.82, p = 0.006). Individuals with at least one polymorphic allele of the IVS11-5CA and Lys939Gln polymorphisms were at borderline significantly decreased risk for SM compared to carriers of each homozygous wild-type allele (adjusted OR = 0.60, 95% CI 0.36–1.00, p = 0.05; adjusted OR = 0.58, 95% CI 0.35–0.97, p = 0.04, respectively).
|−449GC||All cases||1.0||0.90 (0.68–1.20)||0.76 (0.42–1.38)|
|SCC||1.0||0.87 (0.61–1.24)||0.87 (0.42–1.80)|
|Adenoca.||1.0||1.00 (0.67–1.50)||0.43 (0.14–1.32)|
|SM||1.0||0.82 (0.48–1.40)||1.25 (0.48–3.25)|
|−371GA||All cases||1.0||1.03 (0.77–1.37)||1.47 (0.83–2.62)|
|SCC||1.0||1.00 (0.70–1.44)||2.08 (1.09–4.00)3|
|Adenoca.||1.0||1.08 (0.72–1.62)||0.53 (0.18–1.62)|
|SM||1.0||0.99 (0.58–1.71)||1.84 (0.72–4.75)|
|−27GC4||All cases||1.0||1.97 (1.22–3.17)5|
|Val499Arg||All cases||0.90 (0.52–1.58)||0.95 (0.72–1.26)||1.0|
|(TC)||SCC||1.03 (0.52–2.05)||0.86 (0.60–1.23)||1.0|
|Adenoca.||0.64 (0.25–1.64)||1.16 (0.78–1.74)||1.0|
|SM||1.37 (0.55–3.38)||0.82 (0.48–1.39)||1.0|
|PAT−/+||All cases||1.0||0.72 (0.53–0.96)3||0.95 (0.62–1.45)|
|SCC||1.0||0.78 (0.54–1.12)||0.92 (0.53–1.59)|
|Adenoca.||1.0||0.76 (0.50–1.18)||1.18 (0.67–2.10)|
|SM||1.0||0.51 (0.30–0.87)7||0.39 (0.14–1.03)|
|IVS11-5CA||All cases||1.0||0.78 (0.58–1.05)||0.90 (0.52–1.58)|
|SCC||1.0||0.79 (0.54–1.14)||0.90 (0.52–1.58)|
|Adenoca.||1.0||0.85 (0.55–1.31)||1.25 (0.70–2.24)|
|SM||1.0||0.65 (0.38–1.10)||0.40 (0.15–1.08)|
|Lys939Gln||All cases||1.0||0.74 (0.55–1.00)9||0.97 (0.63–1.48)|
|(AC)||SCC||1.0||0.75 (0.52–1.09)||0.86 (0.49–1.50)|
|Adenoca.||1.0||0.82 (0.53–1.17)||1.27 (0.71–2.25)|
|SM||1.0||0.59 (0.35–1.01)||0.52 (0.22–1.28)|
The 6 polymorphisms, except –27GC, were in strong LD (Table IV), and we observed only 23 haplotypes out of the possible 128 (27). Sixteen haplotypes that had a frequency of <1% were excluded from further analysis, to avoid possible errors in either genotyping or the estimation process (data not shown). The remaining 7 haplotypes accounted for 97.0% of the chromosomes for the 864 subjects. The distributions of the XPC haplotypes among cases and controls are shown in Table V. There was no significant difference in the distribution of the haplotypes between the overall lung cancer cases and controls. When cases were categorized by tumor histology, however, the haplotype distribution in the SM group was significantly different from that among controls (p = 0.003). The adjusted OR and 95% CI for each haplotype were calculated by comparison to all the other haplotypes combined. Consistent with the results of genotyping analyses, haplotype 5 (1122111) containing the –27C allele was associated with a significantly increased risk of SM (adjusted OR = 2.88, 95% CI 1.41–5.87, p = 0.004, pc = 0.028), whereas haplotype 4 (1112222) containing the PAT+, IVS11-5A and 939Gln alleles was associated with a significantly decreased risk of SM (adjusted OR = 0.56, 95% CI 0.37–0.85, p = 0.007, pc = 0.049).
|Haplotype1||Controls (n = 847)||All cases (n = 830)||Histologic type of lung cancer2|
|Number (%)||Number (%)||OR (95% CI)3||SCC (n = 420)||Adenoca. (n = 282)||SM (n = 146)|
|Number (%)||OR (95% CI)||Number (%)||OR (95% CI)||Number (%)||OR (95% CI)|
|Ht1, 11111114||7 (0.8)||12 (1.4)||1.69 (0.65–4.39)||5 (1.2)||1.95 (0.59–6.48)||6 (2.2)||2.53 (0.81–7.93)||1 (0.7)||0.91 (0.11–7.85)|
|Ht2, 1112111||40 (4.7)||37 (4.5)||0.97 (0.61–1.55)||14 (3.4)||0.73 (0.39–1.39)||16 (5.9)||1.17 (0.63–2.16)||7 (5.1)||1.14 (0.50–2.63)|
|Ht3, 1112122||19 (2.2)||28 (3.4)||1.31 (0.72–2.41)||9 (2.2)||0.91 (0.39–2.13)||10 (3.7)||1.42 (0.63–3.18)||8 (5.8)||2.47 (1.02–5.98)5|
|Ht4, 1112222||315 (37.2)||283 (34.1)||0.87 (0.71–1.06)||142 (34.8)||0.91 (0.71–1.18)||98 (36.3)||0.94 (0.70–1.25)||34 (24.6)||0.56 (0.37–0.85)6|
|Ht5, 1122111||28 (3.3)||41 (4.9)||1.54 (0.94–2.54)||18 (4.4)||1.32 (0.71–2.46)||11 (4.1)||1.25 (0.60–2.58)||12 (8.7)||2.88 (1.41–5.87)7|
|Ht6, 1212111||205 (24.2)||221 (26.6)||1.15 (0.92–144)||115 (28.2)||1.23 (0.94–1.62)||64 (23.7)||1.01 (0.73–1.40)||39 (28.3)||1.24 (0.82–1.86)|
|Ht7, 2111111||233 (27.5)||208 (25.1)||0.89 (0.71–1.11)||105 (25.7)||0.89 (0.68–1.18)||65 (24.1)||0.88 (0.63–1.22)||37 (26.8)||0.94 (0.62–1.41)|
We assessed the potential interaction between genotype and smoking both by the logistic regression model, including the interaction term between genotype and smoking, and by stratification analysis. The interaction term between genotype and smoking was not statistically significant. There was no clear evidence that smoking modified the effect of the genotype or haplotype on lung cancer risk in the stratified analyses (data not shown).
DNA sequence variations in the XPC gene may lead to alteration of XPC production and/or activity, thereby causing interindividual differences in the susceptibility to lung cancer via action on DNA repair pathways. To test this hypothesis, we evaluated the potential association of 7 XPC polymorphisms (–449GC, –371GA, –27GC, Val499Arg, PAT–/+, IVS11-5CA and Lys939Gln) with lung cancer risk. In addition, we estimated XPC haplotypes of 7 polymorphisms and compared their frequency distributions in lung cancer cases and controls.
In the current study, we validated the presence of 7 XPC polymorphisms including 4 polymorphisms that had been previously reported (Val499Arg, PAT–/+, IVS11-5CA and Lys939Gln) in a Korean population. Frequencies of the 499Val and IVS11-5A alleles among healthy Koreans were 0.712 and 0.391, respectively, which were similar to those (0.768 and 0.423, respectively) of NIH donors in the United States.19, 20 Frequencies of the PAT+ and 939Gln alleles among healthy Koreans were 0.373 and 0.394, respectively, which were also similar to those of Caucasians (0.333 and 0.335, respectively).21, 25 In agreement with previous studies,19, 20, 26 we found that the Val499Arg, PAT–/+, IVS11-5CA and Lys939Gln polymorphisms were in LD. In addition, –449GC and –371GA polymorphisms were linked with these 4 polymorphisms. Khan et al.20 reported that haplotypes PAT–/IVS11-5C/939Lys and PAT+/IVS11-5A/939Gln accounted for approximately 60% and 40%, respectively, of the chromosome among NIH donors. Frequencies of the haplotypes of these 3 polymorphisms among controls in the current study were comparable with their report.20
DNA sequences upstream of the XPC transcription initiation site possess multiple putative transcription factor binding sites. The –449GC, –371GA and –27GC polymorphisms lead to changes in the putative transcription factor binding sites by computer prediction using the Alibaba2 program (http://www.alibaba2.com).27 The –449GC polymorphism leads to the creation of a Yin and Yang-1 binding site and one additional SP1 binding site, and it eliminates an EGFR-specific transcription factor binding site on the XPC promoter.27 The –371GA polymorphism eliminates an SP1 binding site, while the –27GC polymorphism leads to the creation of an SP1 binding site and a nuclear factor-1 binding site. Even though these are the predicted changes in the putative transcription factor binding sites, the –449GC polymorphism did not have an effect on the risk for lung cancer. However, the –371AA genotype was associated with a 2.1-fold increased risk for SCC compared to the –371GG genotype or the combined –371GG+GA genotype (a recessive model for the variant allele, data not shown). In accordance with the results of genotyping analyses, homozygotes of haplotype 6 containing the –371A showed a similar susceptibility effect for the risk of SCC (data not shown). These results suggest that the –371A allele has a recessive effect on the XPC phenotype, as is the case of the XPD Asp312Asn (23592GA of exon 10) polymorphism.28, 29 In the current study, we also found that the –27GC polymorphism was associated with a significantly increased risk of overall lung cancer. The risk effect of the –27C allele was statistically significant only for SM, though observed for the 3 major histologic types of lung cancer. Consistent with the results of genotype analyses, haplotype 5 (1122111) containing the –27C allele was associated with a significantly increased risk for SM.
A few studies have examined the association between polymorphisms in the exons and introns of the XPC gene and cancer risk. Shen et al.21 reported that the PAT+/+ genotype was associated with a 1.85-fold increased risk for SCC of the head and neck. Sanyal et al.25 reported that individuals carrying 2 variant alleles for the Lys939Gln polymorphism were at significantly increased risk of bladder cancer. In view of the reports that the PAT, IVS11-5CA and Lys939Gln polymorphisms are in LD,19, 20, 26 these 2 studies suggest that individuals homozygous for the PAT+/IVS11-5A/939Gln haplotype may have increased cancer susceptibility. In contrast to these 2 studies, haplotype 4 (1112222) containing the PAT+/IVS11-A/939Gln alleles was associated with a significantly decreased risk of SM in our current study. Although it is hard to decipher the reasons for the opposite association of the PAT+/IVS11-A/939Gln haplotype for different cancers in the previous and current studies, several genetic and environmental factors relevant to these polymorphisms, such as different carcinogen exposures in different populations and different types of DNA damage in the initiation of different cancers, might have caused the discrepancy. Moreover, since a protective haplotype in one population may increase the cancer risk in another population due to other linked polymorphisms that exhibit a stronger effect on the susceptibility to cancer, different genetic backgrounds of subjects also should be considered. In addition, inadequate study design, such as nonrandom sampling, limited sample size and unknown confounders, should be considered. Selection bias in a hospital-based case-control study might be a relevant issue. Given the fact that most of the lung cancer patients were treated at a university hospital in Korea, it is reasonable to assume that the case group represents the lung cancer cases in our community. Another selection bias may be derived from controls who did not participate in the study. However, because the age and sex distribution of nonparticipating controls were similar to those of the participating controls in the current study, a self-selection bias is unlikely. The fact that allele and genotype frequencies among our controls and cases are consistent with those derived from the Hardy-Weinberg equilibrium further supports the nonbiased sampling of our study. However, since the numbers in the subgroups were relatively small, larger studies will be needed to validate the genetic effects of the XPC polymorphisms on lung cancer.
In the current study, the –27GC polymorphism was associated with an increased risk of overall lung cancer. However, the XPC –371GA polymorphism was associated only with risk of SCC, whereas the PAT–/+, IVS11-5CA and Lys939Gln polymorphisms were related only to risk of SM. Although the reason for the observed histology-dependent difference in the genetic effect conferred by these XPC polymorphisms is unknown, it may be attributable to differences in the carcinogenesis pathways among the histologic types of lung cancer. Various lines of evidence have suggested that the histologic type of lung cancer may be determined by the particular initiating agent to which an individual is exposed.30, 31 Therefore, genetic factors involved in susceptibility could be different among the histologic subtypes of lung cancer.32, 33, 34
In conclusion, we found that the XPC polymorphisms and their estimated haplotypes are associated with lung cancer risk, suggesting that the XPC gene may be involved in the development of lung cancer. It is possible that our findings are attributable to chance because of the relatively small numbers in the subgroups, particularly the SM subgroup. Therefore, additional studies with larger sample sizes are required to confirm our findings. Future studies of other XPC sequence variants and their biologic function are also needed to understand the role of XPC polymorphisms in determining the risk of lung cancer. Moreover, since genetic polymorphisms often vary significantly between ethnic groups, further studies are warranted to clarify the association of the XPC polymorphisms with lung cancer in diverse ethnic populations.
- 27Alibaba2: context specific identification of transcription factor binding sites. In Silico Biol 2002; 2: S1–1..
- 31Cigarette smoking and adenocarcinoma of the lung: the relevance of nicotine-derived nitrosamines. J Smoking Relat Disord 1993; 4: 165–90., , , , , .