Skip Navigation


Carcinogenesis Advance Access originally published online on November 21, 2006
Carcinogenesis 2007 28(5):988-994; doi:10.1093/carcin/bgl225
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
28/5/988    most recent
bgl225v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (1)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Lu, J.
Right arrow Articles by Wei, Q.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lu, J.
Right arrow Articles by Wei, Q.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2006. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

172G>T variant in the 5' untranslated region of DNA repair gene RAD51 reduces risk of squamous cell carcinoma of the head and neck and interacts with a P53 codon 72 variant

Jiachun Lu1, Li-E Wang1, Ping Xiong1, Erich M. Sturgis1,2, Margaret R. Spitz1 and Qingyi Wei1,*

1 Department of Epidemiology
2 Department Head and Neck Surgery, The University of Texas M. D. Anderson Cancer Center, Houston, TX 77030, USA

* To whom correspondence should be addressed. Tel: +1 713 792 3020; Fax: +1 713 563 0999; Email: qwei{at}mdanderson.org


   Abstract
Top
 Abstract
Introduction
 Materials and methods
 Results
 Discussion
 References
 
RAD51 participates in homologous recombination (HR) repair of double-stranded DNA breaks (DSBs) that may cause genomic instability and cancer. Two single-nucleotide polymorphisms (SNPs) and three P53 binding sites have been found in the RAD51 promoter and 5' untranslated region. We hypothesized that RAD51 and P53 SNPs may interact and alter risk of squamous cell carcinoma of the head and neck (SCCHN) and we genotyped for RAD51 135G>C and 172G>T and P53 Arg72Pro SNPs in 716 SCCHN patients and 719 matched controls (all non-Hispanic whites) and evaluated their effects on gamma radiation-induced mutagen sensitivity. We found that RAD51 172TT homozygotes had a significantly decreased risk [adjusted odds ratio (OR) = 0.66, 95% confidence interval (CI) = 0.50–0.87] of SCCHN, compared with carriers of other genotypes, particularly in P53 Arg72Arg homozygotes (adjusted OR = 0.60, 95% CI = 0.41–0.89) (homogeneity test P = 0.047), although no alterations in the risk were associated with the RAD51 135G>C and P53 Arg72Pro SNPs. Consistent with a protective effect of the 172TT genotype, significantly fewer gamma radiation-induced chromatid breaks per cell were present in 172TT homozygotes (mean ± SD = 0.36 ± 0.13) than in subjects with other genotypes (mean ± SD = 0.46 ± 0.13, P < 0.001) among 148 control subjects we tested. The finding that the functional RAD51 172G>T SNP, particularly in the presence of the P53 Arg72Arg genotype, may be a marker of susceptibility to SCCHN needs to be validated by larger studies of different ethnic populations.

Key Words: Case-control study • DNA repair • Genetic susceptibility • Molecular epidemiology • Mutagen sensitivity

Abbreviations: AIC, Akaike's information criterion; b/c, breaks per cell; CI, confidence interval; DSB, DNA double-strand break; HR, homologous recombination; HWE, Hardy–Weinberg equilibrium; LD, linkage disequilibrium; MAF, minor allele frequency; NIH, National Institutes of Health; OR, odds ratio; PCR, polymerase chain reaction; SNP, single-nucleotide polymorphism; SCCHN, squamous cell carcinoma of the head and neck; UTR, untranslated region


   Introduction
Top
 Abstract
 Introduction
Materials and methods
 Results
 Discussion
 References
 
Homologous recombination (HR) repair is a fundamental process whereby cells maintain genomic stability. In HR repair, cells respond to DNA double-stranded breaks (DSBs) by using the undamaged sister DNA molecule (i.e. homologous chromatid) as a template for reconstruction of the broken DNA strands. When the protein complex, MRE11–RAD50–NBS1, recognizes the site of initial damage, the central HR protein, RAD51, recognizes the homology, catalyzes the 3'-end strand, invading the unbroken homologue, and mediates homologous pairing of two DNA duplexes (1). The role of RAD51 in HR and genomic integrity is exemplified by the RAD51-knockout mouse (RAD51–/–) model, which is characterized by early embryonic lethality (2). Loss of RAD51 expression predisposes cells to chromosome breaks or aberrations, mutagenesis and cell death (3,4). Even minor changes in or deregulation of RAD51 may lead to genetic instability and cancer (5,6).

The RAD51 gene, a homologue of recA in Escherichia coli, has been mapped to chromosome 15q15.1 in humans (7). It spans >39 kb, contains 10 exons and encodes a 339 amino acid protein (genomic accession no: NM_133487 [GenBank] ). The first exon and part of the second exon of RAD51 are located at its 5' untranslated region (5'UTR), which may regulate gene expression. RAD51 is highly polymorphic: 143 single-nucleotide polymorphisms (SNPs) of RAD51 are listed in the Environmental Genome Project SNP database of the National Institute of Environmental Health Sciences (http://egp.gs.washington.edu), and 98 of these SNPs are also reported in the GenBank dbSNP database (http://www.ncbi.nlm.nih.gov). However, only a few SNPs are located in the coding regions of RAD51, none of which has a minor allele frequency (MAF) >5%, suggesting that its protein structure may be genetically conservative.

Two common (MAF > 0.05) RAD51 SNPs, 135G>C (rs1801320) and 172G>T (rs1801321) in the 5'UTR have been reported to be associated with altered gene transcription (8). Other studies have shown that the RAD51 135C variant allele was associated with an increased risk of female breast cancer in carriers of BRCA2 mutations, but not in carriers of BRCA1 mutations (9,10) or non-mutation carriers (11). In contrast, a recent study of BRCA1 mutation carriers in a Polish population reported that the RAD51 135C variant allele was associated with a reduced risk of breast cancer in women (12). In addition, the variant T allele of the RAD51 172G>T SNP was shown to be associated with a non-significantly decreased risk of sporadic breast cancer in women (13,14). However, no reported studies have investigated the role of RAD51 SNPs in the etiology of squamous cell carcinoma of the head and neck (SCCHN).

In the 5'UTR and nearby promoter region of RAD51, three P53 binding sites exist (i.e. –463 to –444 bp, –162 to –118 bp and +40 to +51 bp) (8,15). The P53 tumor suppressor gene is multifunctional, influencing repair of DNA damage, the cell cycle, apoptosis and maintenance of genomic stability. For example, P53 modulates HR by both transcriptional (15) and non-transcriptional (i.e. direct protein–protein interaction) (16,17) regulation of the RAD51 gene. Somatic mutations of the P53 gene have been found in at least half of all human tumors, suggesting its importance in the etiology and molecular pathogenesis of cancer (18). The Arg72Pro SNP (rs1042522), which causes a change of arginine to proline in a proline-rich region of P53, is known to be associated with P53-mediated apoptosis (19). However, no reported studies have investigated the interaction between SNPs of P53 and RAD51 in cancer risk.

We hypothesized that genetic variations in RAD51 are associated with cancer susceptibility and that this susceptibility may be modulated by the P53 Arg72Pro SNP. We tested this hypothesis in a hospital-based case-control study of 716 patients with newly diagnosed SCCHN and 719 sex- and age frequency-matched cancer-free control subjects. To extend the previously reported in vitro data on the functional relevance of the RAD51 SNPs (8), we also analyzed the SNPs' effects on phenotypic modulation of sensitivity to gamma radiation-induced chromosomal aberrations in 148 controls.


   Materials and methods
Top
 Abstract
 Introduction
 Materials and methods
Results
 Discussion
 References
 
Study subjects and data collection
The recruitment of study subjects was described previously (20). Briefly, the study population included 716 patients with newly diagnosed SCCHN and 719 cancer-free control subjects recruited from May 1995 through September 2004. Approximately 95% of the eligible patients (all registered with The University of Texas M. D. Anderson Cancer Center) contacted chose to participate in this study. Only non-Hispanic white patients were included in this analysis because few patients of ethnic minority groups were recruited and because genotype frequencies can vary between ethnic groups. Patients with second SCCHN primary tumors, primary tumors of the nasopharynx or sinonasal tract, primary tumors outside the upper aerodigestive tract, cervical metastases of unknown origin or histopathologic diagnoses other than squamous cell carcinoma were excluded. For the cancer-free controls, subjects were recruited from visitors to the hospital who were genetically unrelated to patients with SCCHN. We first surveyed potential control subjects at the clinics by using a short questionnaire to determine their willingness to participate in research studies and to obtain demographic information. Among the respondents we contacted for recruitment, the response rate was ~85%. We frequency matched the controls to the patients by age (±5 years) and sex. After signing an informed consent form, each eligible subject was interviewed to provide data on tobacco smoking, alcohol use and other risk factors, and each donated a venous blood sample of ~30 ml, 1 ml of which was used for genomic DNA extraction. The fresh blood samples of 148 additional control subjects were also used in the gamma radiation sensitivity assay (21). M. D. Anderson Cancer Center's institutional review board approved the research protocol.

Genotyping analysis
From each blood sample, the leukocyte cell pellet was obtained from the buffy coat by centrifugation of 1 ml of whole blood. Genomic DNA was extracted from the cell pellet using the DNA Blood Mini Kit (Qiagen, Valencia, CA), according to the manufacturer's instructions. DNA purity and concentrations were determined by spectrophotometric measurement of absorbance at 260 and 280 nm.

The RAD51 135G>C and 172G>T SNPs were genotyped by means of the modified polymerase chain reaction (PCR)–restriction fragment length polymorphism method (22), using the following primers to amplify the SNP-containing fragment: 5'-TGG GAA CTG CAA CTC ATC TGG-3' (forward) and 5'-GCT CCG ACT TCA CCC CGC CGG-3' (reverse). PCR was performed in 20 µl reaction systems containing 4 mM MgCl2, 0.08 mM deoxynucleotide triphosphates, 2.0 U Taq polymerase and the manufacturer's buffer [20 mM Tris–HCI (pH 8.4) and 50 mM KCl]. The PCR profile consisted of an initial melting step at 95°C for 5 min; 30 cycles of 95°C for 30 s, 65°C for 45 s and 72°C for 50 s and a final extension step of 72°C for 10 min. The amplified fragments were separated into two tubes. One group of fragments was digested with BstNI (New England BioLabs, Beverly, MA) overnight to generate RAD51 135G>C polymorphisms, and the other group was digested with NgoMIV (New England BioLabs) overnight to generate 172G>T polymorphisms. The products were separated in 3% MetaPhor agarose gel (Cambrex, East Rutherford, NJ). The 135GG genotype produced two bands (71 and 60 bp), whereas the 135GC genotype produced three bands (131, 71 and 60 bp) and the 135CC heterozygote displayed only one band (131 bp). The 172GG genotype produced two bands (110 and 21 bp), whereas the 172TT genotype produced only one band (131 bp) and the 172GT heterozygote displayed all three bands (131, 110 and 21 bp). The P53 Arg72Pro polymorphism was genotyped according to a previously published method (23).

The PCRs were performed with a PTC-200 DNA Engine Peltier Thermal Cycler (MJ Research, Waltham, MA), and the results were evaluated by two of us who were blinded to the subjects' patient or control status. The genotype patterns were confirmed by direct sequencing. We randomly selected 10% samples for each of the three SNPs to perform repeat assays, and the results were 100% concordant.

Mutagen sensitivity assay
To explore the functional relevance of the SNPs to be tested, we evaluated gamma radiation sensitivity in 148 additional control subjects, following previously published protocols (21). Briefly, two parallel short-term cultures of each blood sample were established from each individual by inoculating each of the two T-25 plastic culture flasks with 1 ml of whole blood and 9 ml of RPMI 1640 medium (Life Technologies, Grand Island, NY) supplemented with 20% fetal bovine serum and a final concentration of 112.5 µg/ml phytohemagglutinin (Remel, Compton, CA) to stimulate lymphocyte growth. After 67 h of culture, one of the two cultures was irradiated with 1.5 Gy of incident gamma rays at a dose rate of 13.18 Gy/min for 6.9 s from a 137Cs source (Cesium Irradiator Mark 1, Model 30; J. L. Shepherd and Associates, Glendale, CA); the other culture was left untreated. Both cultures were incubated for another 4 h and treated with a final concentration of 0.06 µg/ml colcemid (GIBCO BRL, Carlsbad, CA) to induce mitotic arrest. After 1 h, the cells were harvested by conventional chromosome harvesting procedures: The cells were treated for 15 min with 60 mM KCl hypotonic solution and fixed for 15 min with freshly prepared methanol:acetic acid (3:1 vol:vol); this was followed by preparation of air-dried slides, which were then stained with 4% Giemsa (Biomedical Specialties, Santa Monica, CA) for 7 min. Each slide was examined for chromosomal aberrations using a Labphoto-2 photomicroscope (Nikon, Instrument Group, Melville, NY). The number of simple chromatid breaks was scored from 50 well-spread metaphases for both treated and untreated samples from each subject. A chromatid break was scored as one break, and each isochromatid break set and each exchange figure (or interstitial deletion) were scored as two breaks. Gaps were not included in the analyses.

Statistical analysis
To compare the distributions of demographic variables and selected risk factors between patients and controls, {chi}2 tests were used. The Hardy–Weinberg equilibrium (HWE; p2 + 2pq + q2 = 1, where p is the frequency of the variant allele and q = 1 – p) was tested by a goodness-of-fit {chi}2 test to compare the observed genotype frequencies with expected genotype frequencies in cancer-free controls. The association between case-control status and each individual SNP, measured by the odds ratio (OR) and its corresponding 95% confidence interval (CI), was estimated using an unconditional logistic regression model, both with and without adjustment for age, sex, smoking status, alcohol use and P53 genotype and RAD51 genotype. Logistic regression modeling was also used for the trend test. The effect of each SNP was assumed to be co-dominant, dominant and recessive in separate analyses. In the co-dominant model, the common homozygote genotype in the controls was defined as the reference group; the rare homozygous and heterozygous genotypes were variants, and their effects were individually estimated by comparison with the reference. In the dominant model, the rare homozygote and the heterozygote were combined to form a variant group, whose effect was then estimated and compared with that of the common homozygote. In the recessive model, only the rare homozygote was defined as the variant; the other two genotypes were combined to define the reference. Akaike's information criterion (AIC) was used to select the best genetic effect model for each SNP (i.e. that in which the AIC value was smallest) (24). Considering the effect of covariates, we calculated the AIC in an unconditional logistic regression model with adjustment for age, sex, smoking and alcohol usage. The data were further stratified by age, sex, smoking status, alcohol use and polymorphisms of P53 codon 72 to evaluate the stratum variable-related ORs among the RAD51 genotypes. Homogeneity between stratum variable-related ORs was tested (25). The potential multiplicative and additive interactions among gene–gene and gene–environmental factors were also evaluated by means of logistic regression analysis (26). For the gamma radiation sensitivity data, the number of chromatid breaks per cell (b/c) was analyzed as a continuous variable. Student's t-test was used to compare the mean number of chromatid b/c between groups. We used linear regression model to evaluate the main effect of RAD51 or P53 genotypes on the chromatid b/c with adjustment for age, sex, smoking and alcohol usage. All statistical tests were two-sided, and P < 0.05 was considered statistically significant. The analysis of additive interaction was performed with the bootstrapping test using Stata software (version 8.0, StataCorp, College Station, TX). All other analyses were performed using Statistical Analysis System software (version 9.1, SAS Institute, Cary, NC).


   Results
Top
 Abstract
 Introduction
 Materials and methods
 Results
Discussion
 References
 
Demographic characteristics of the study population
The study population consisted of 716 patients with primary SCCHN and 719 cancer-free controls. The frequency distributions of selected characteristics of the patients and controls are presented in Table I. The patients and controls appeared to be adequately matched with respect to age and sex: the mean ± SD age was 55.8 ± 11.3 years for the patients (range = 18–90 years) and 56.1 ± 11.6 years for the controls (range = 20–87 years, P = 0.108), and 73.6% of the patients and 74.3% of the controls were men (P = 0.774). There were significantly more current smokers and drinkers among the patients (35.2 and 50.3%, respectively) than among the controls (16.8 and 41.5%, respectively; P < 0.001 for both exposures) (Table I). Therefore, these variables were further adjusted for in the multivariate logistic regression model to control for possible confounding of the main effects of the studied SNPs.


View this table:
[in this window]
[in a new window]

  		Table I. Frequency distributions of selected variables in SCCHN patients and cancer-free controls

 
Distributions of RAD51 and P53 genotypes
The genotype and allele distributions of the RAD51 135G>C and 172G>T and the P53 Arg72Pro SNPs in the patients and controls are summarized in Table II. The observed genotype frequencies of these three SNPs in the control subjects were all in agreement with the HWE (P = 0.170, 0.169 and 0.216, respectively). The linkage disequilibrium (LD) analyses in the controls showed that the two RAD51 SNPs, 135G>C and 172G>T, were in incomplete LD (D' = 0.656 and r2 = 0.024, Table II), suggesting that each may have an independent effect on risk of SCCHN.


View this table:
[in this window]
[in a new window]

  		Table II. Distribution of polymorphisms in RAD51 and P53 and results of logistic regression analysis for associations with risk of SCCHN

 
Although the distributions of the RAD51 135G>C genotype and variant C allele frequencies did not differ between the patients and the controls (P = 0.973 and P = 0.730, respectively), the distribution of RAD51 172G>T genotype frequencies did differ significantly between the patients and controls (P = 0.021, Table II). Specifically, the frequency of the 172TT variant homozygous genotype in the patients (14.5%) was significantly lower than that in the controls (20.0%), and the variant T allele frequency was also significantly lower in the patients (0.390) than in the controls (0.433) (P = 0.020). However, for the P53 Arg72Pro SNP, the genotype and variant Pro allele frequency distributions did not differ between the patients and controls (P = 0.605 and 0.686, respectively, Table II).

Risk estimates for RAD51 and P53 polymorphisms
As shown in Table II, compared with the individuals carrying the RAD51 172GG genotype, those with the 172TT variant homozygote had a significantly decreased risk of SCCHN (OR = 0.64, 95% CI = 0.47–0.88), but those with the 172GT heterozygote did not; this pattern best fit the recessive genetic model, as indicated by the AIC value (1920.92) being the smallest among all possible assumed models (data not shown). Under the recessive model of inheritance, the 172TT genotype was associated with a significantly decreased risk of SCCHN (OR = 0.66, 95% CI = 0.50–0.87), as compared with other genotypes. For RAD51 135G>C and P53 Arg72Pro SNPs, however, none of the variant genotypes was associated with a significantly altered risk of SCCHN under any of the assumed genetic models of inheritance (data not shown). Only one patient and one control carried the RAD51 135CC genotype.

Stratification and interaction analysis of RAD51 genotypes and risk of SCCHN
The risks associated with the tested SNPs were subjected to stratification analysis, but the data for only the RAD51 172G>T SNP are presented in Table III because the others had no further significant findings. In stratification analysis, the RAD51 172TT variant genotype was associated with a significantly decreased risk of SCCHN in the subgroups of subjects at least 55 years old (adjusted OR = 0.60, 95% CI = 0.41–0.88), women (adjusted OR = 0.53, 95% CI = 0.30–0.94), ever-smokers (adjusted OR = 0.60, 95% CI = 0.41–0.88), ever-drinkers (adjusted OR = 0.67, 95% CI = 0.47–0.95) and subjects with cancer of the oral cavity (adjusted OR = 0.49, 95% CI = 0.30–0.79). Tests for homogeneity of the stratum-related risks did not show any evidence that these risks differed between the strata (Table III). However, those subjects carrying both the RAD51 172TT variant genotype and the P53 Arg72Arg genotype did have a significantly decreased risk of SCCHN (adjusted OR = 0.60, 95% CI = 0.41–0.89, homogeneity test P = 0.047), suggesting that the P53 codon 72 SNP modulates the association between the RAD51 172G>T SNP and the risk of SCCHN (Table III). We further examined possible interactions between the RAD51 genotypes and subject age, sex, smoking or drinking status and P53 codon 72 SNP status in regard to cancer risk, but no evidence of either a multiplicative or additive interaction was found (data not shown).


View this table:
[in this window]
[in a new window]

  		Table III. Stratification analysis of the RAD51 172G>T genotypes by selected variables in SCCHN patients and controls

 
Gamma radiation-induced chromatid breaks
To further determine the functional relevance of the tested SNPs on cellular response to DNA damage, we analyzed the phenotype of gamma radiation-induced chromatid b/c in 148 control subjects and investigated the modulating effect of SNPs of RAD51 and P53 on the sensitivity phenotype. Because the mean ± SD spontaneous b/c value derived from 50 metaphases of untreated cells was 0.02 ± 0.02, which was >20-fold less than that of gamma-irradiated cells (0.44 ± 0.38), we used gamma radiation-induced b/c values only for statistical comparisons, as recommended by Hsu et al. (27). We found that the mean ± SD b/c value for the RAD51 172TT carriers was 0.36 ± 0.13 (median = 0.35, range = 0.06–0.86), which was >20% lower than that for 172GT + GG carriers (mean ± SD = 0.46 ± 0.13, median = 0.44, range = 0.18–0.96, Student's t-test: P < 0.001, linear regression test: P < 0.001). Consistently, we found that the b/c values in those who carried both RAD51 172TT and P53 Arg72Arg genotypes (mean ± SD = 0.38 ± 0.17, median = 0.36, range = 0.06–0.86) were significantly lower than those in carriers of all other genotypes (mean ± SD = 0.45 ± 0.13, median = 0.44, range = 0.18–0.96, Student's t-test: P = 0.050, linear regression test: P = 0.051). The b/c values did not differ between RAD51 135C variant carriers (mean ± SD = 0.46 ± 0.16, median = 0.44, range = 0.23–0.86) and 135GG wild-type genotype carriers (mean ± SD = 0.44 ± 0.13, median = 0.41, range = 0.06–0.96) or between P53 Arg72Arg carriers (mean ± SD = 0.45 ± 0.13, median = 0.44, range = 0.06–0.86) and Arg/Pro or Pro/Pro carriers (mean ± SD = 0.44 ± 0.15, median = 0.40, range = 0.18–0.96, Student's t-test: P = 0.402 and 0.700, linear regression test: P = 0.479 and 0.645, respectively) (Figure 1).


Figure 1
View larger version (14K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 1. Effect of RAD51 and P53 SNPs on gamma radiation-induced chromatid b/c in peripheral blood lymphocytes from 148 control subjects. P values were obtained from Student's t-tests. (A) Comparison of b/c between RAD51 135GC carriers and 135GG carriers (P = 0.402). (B) Comparison of b/c between RAD51 172TT carriers and 172GG + GT carriers (P < 0.001). (C) Comparison of b/c between P53 Arg72Arg carriers and Arg72Pro + Pro/Pro carriers (P = 0.700). (D) Comparison of b/c between both RAD51 172TT and P53 Arg72Arg genotypes carriers and carriers of other combinations of genotypes (P = 0.050).

 

   Discussion
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
References
 
In this hospital-based case-control study, we investigated the association between three putatively functional SNPs (135G>C and 172G>T of RAD51 and Arg72Pro of P53) and risk of SCCHN in non-Hispanic whites. When we evaluated each SNP individually, we found that the RAD51 172TT homozygous variant genotype was associated with a reduced risk of SCCHN before and after adjustment for age, sex, smoking and drinking status and P53 genotype; stratification analysis, however, showed that this protection remained significant only in the presence of the P53 Arg72Arg genotype, suggesting a possible interaction between RAD51 172G>T and P53 Arg72Arg SNPs in the etiology of SCCHN. Furthermore, this protection remained significant only in the presence of environmental exposure known to increase risk, such as ever-smoking or ever-drinking, although the power of the tests for interactions was limited by the study sample size. Our findings appear to be supported by the previously reported effect of enhanced promoter activities by the RAD51 172T allele (8) and by our present finding that the RAD51 172TT genotype protected against gamma radiation-induced chromosomal aberrations.

Few studies have investigated the association between the RAD51 172G>T SNP and risk of cancer. In a large European case-control study of 2205 patients with breast cancer and 1805 controls, the 172T variant genotypes of RAD51 were found to be associated with a non-significantly reduced risk of breast cancer (13). Similar results were reported in a Korean case-control study of breast cancer (14). Conversely, in a recent case-control study of epithelial ovarian cancer, none of the 135G>C and 172G>T variants of RAD51 were associated with a reduction in risk (28). However, no reported studies have investigated the role of these RAD51 SNPs in smoking-related cancers, such as those of the lung and head and neck. In the present study, we found that the RAD51 172TT homozygous variant genotype was associated with a significantly reduced risk of SCCHN. Therefore, whether the reported association of RAD51 SNP with cancer risk is cancer-specific needs to be validated in additional studies of cancers at different cancer sites in more diverse ethnic groups.

Our finding of a reduced risk of SCCHN associated with the RAD51 172TT homozygous variant genotype appears to be biologically plausible. Bioinformatics analysis showed that the 172T allele is located in a binding site for the transcription factor P300/CBP, a cofactor of nuclear receptor signaling that possesses strong histone acetyltransferase activities and functions as an adaptor protein that enhances transcription of the genes to which the ligand is bound (29). In contrast, the RAD51 172G allele does not form a binding site of cis-transcriptional elements for P300/CBP. Thus, the 172T allele may have a greater effect on RAD51 gene expression. One study using the luciferase reporter gene assay found that compared with the 172G allele, the 172T variant allele was associated with significantly greater activity of the RAD51 promoter (8). It was reported that the knockout and loss of RAD51 expression can be lethal (2,3), and inactivation of the RAD51 gene can lead to mutagenesis (4). In contrast, overexpression of RAD51 was reported in several cancer lines (30), including pancreatic cancers (31) and lung cancer (32). However, the injection of cells overexpressing RAD51 into nude mice did not induce tumorigenesis, whereas cells carrying dominant-negative RAD51 induced centrosome fragmentation, ploidy and tumorigenesis (33). The finding that overexpression of RAD51 in cell lines and Drosophila could increase their apoptotic potential (34,35) might explain the protective role of the RAD51 172TT genotype on cancer risk: apoptotic cells would have a low probability of tumorigenesis. Taken together, these findings predict an enhanced role for the 172TT genotype in HR activities, a hypothesis that is consistent with the protective effect it had on risk of SCCHN in the present study. Although we did not examine the levels of RAD51 mRNA in target tissues that contained different 172G>T genotypes, it is likely that the exact mechanisms by which the 172T variant regulates RAD51 transcription activity in vivo are not the same as those observed in cell lines in vitro, which warrants additional in vivo mechanistic studies.

The biological plausibility of a role for the RAD51 172TT variant genotype in reducing risk of SCCHN is also supported by our finding that the cells from carriers of this variant genotype had fewer chromosomal aberrations induced by ionizing radiation than cells from carriers of other genotypes. Radiation can induce both single- and DNA DSBs (36) that may lead to chromosomal breaks. The high frequency of chromatid breaks in peripheral blood lymphocytes after in vitro exposure to gamma radiation, as measured by the mutagen sensitivity assay (21), has been reported to be associated with an increased risk of various cancers, including glioma (37), breast cancer (38) and thyroid cancer (39). It is likely that deficient repair of DNA DSBs could result in an abnormally high frequency of chromosomal aberrations after exposure to radiation, conferring a predisposition to cancer (40). In our present analysis of gamma radiation-induced chromatid b/c in 148 control subjects, we found that the frequency of chromatid breaks was significantly lower in the 172TT carriers than in the 172GT + GG carriers of the RAD51 gene. This finding implies that people carrying the RAD51 172TT genotype may have a greater cellular capacity for DSB repair or have a lower sensitivity to gamma radiation than carriers of other genotypes, supporting our observation that the RAD51 172TT genotype has a protective effect on risk of SCCHN.

Although some evidence suggests that the 135G>C SNP of RAD51 is involved in modifying risk of breast cancer in BRCA2 mutation carriers (9), but not in BRCA1 mutation carriers (10) or non-mutation carriers (11), we did not observe any significant association between the presence of the RAD51 135G>C SNP and risk of SCCHN in the present study. However, the current result is consistent with our previous finding that this mutation had no association with risk of glioma (22), in which both exposure to ionizing radiation and gamma radiation-induced mutagen sensitivity were risk factors (37). One possible reason for our failure to detect any effect of the RAD51 135G>C SNP on risk of SCCHN is that the present study may not have been large enough to have a sufficient power to detect a weak effect of the RAD51 135G>C SNP on risk of SCCHN.

Similarly, we found that the codon 72 SNP of the P53 gene was not associated with risk of SCCHN, which is consistent with findings in previous reports (23,41). However, we found that carriers of the RAD51 172TT variant genotype had a significantly decreased risk of SCCHN in those who also carried the P53 Arg72Arg genotype, not in those with other P53 genotypes. Consistently, as observed in the gamma radiation mutagen sensitivity assay, the frequency of chromatid breaks was significantly lower in those who carried both the RAD51 172TT and P53 Arg72Arg genotypes than in those who carried other combinations of genotypes. These results suggest that the P53 Arg72Pro SNP might modulate risk of SCCHN associated with the RAD51 172TT genotype. However, this finding was of only borderline significance and therefore must be validated by larger studies involving ethnically diverse subjects.

It is possible that our findings were obtained by chance due to selection bias inherent in hospital-based case-control studies. This study population may have had a distribution of observed phenotypes different from that in the general population, particularly among the controls. We do not believe that such bias was encountered, however, because the RAD51 genotype frequency distributions in our study population were similar to those reported for similar populations. For example, the frequencies of minor alleles (i.e. the RAD51 135C and 172T and P53 72Pro alleles) among our 719 Texas non-Hispanic white controls were 0.068, 0.467 and 0.259, respectively; similar frequencies were reported for Caucasians included in the International HapMap Project (http://www.hapmap.org) (0.067, 0.467 and 0.233, respectively) and for Caucasians included in Cancer Genome Anatomy Project SNP500 Cancer Database (http://snp500cancer.nci.nih.gov) (0.100, 0.433 and 0.274, respectively). Insufficient study power may be another concern in regard to our negative findings, particularly for SNPs with a rare variant allele, such as the RAD51 135C allele. For the RAD51 172G>T SNP, which had a relatively high MAF, we achieved 83.8% study power (two-sided test, {alpha} = 0.05) to detect an OR of 0.66 for the 172TT genotype (which occurred at a frequency of 20.0% in the controls) compared with the 172GG + GT genotypes, assuming a recessive genetic model. Therefore, it appears that our finding that the 172TT genotype was associated with a decrease in risk of SCCHN is unlikely to have been obtained by chance. One final concern is that because we restricted the subjects to non-Hispanic whites, it is uncertain whether our findings can be generalized to other populations.

In conclusion, we found a significant association between the 172TT genotype of RAD51 and a decreased risk of SCCHN in this relatively large, hospital-based case-control study. Our data also suggest that the 172TT genotype may confer protection by enhancing DNA repair capacity or reducing radiation mutagen sensitivity. It is likely that the Arg72Pro SNP of P53 may modulate the RAD51 172TT genotype's effect on risk of SCCHN. Therefore, we conclude that the RAD51 172G>T SNP may be a biomarker of reduced susceptibility to SCCHN but our findings in this study need to be validated by larger studies, particularly in different ethnic populations.


   Acknowledgments
 
We thank Margaret Lung, Kathryn Patterson and Leanel Fairly for their assistance in recruiting the subjects; Xiaodong Zhai and Yawei Qiao for technical assistance; Jianzhong He, Kejin Xu and Yinyan Li for their laboratory assistance; Monica Domingue for manuscript preparation and Kathryn Carnes and Beth Notzon for scientific editing. This study was supported by the National Institutes of Health (NIH) grant ES 11740 (Q.W.) and in part by the NIH grants CA100264 (Q.W.) and CA16672 (The University of Texas M. D. Anderson Cancer Center).

Conflict of Interest Statement: None declared.


   References
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 

  1. Jackson SP. Sensing and repairing DNA double-strand breaks. Carcinogenesis (2002) 23:687–696.[Abstract/Free Full Text]
  2. Tsuzuki T, et al. Targeted disruption of the RAD51 gene leads to lethality in embryonic mice. Proc. Natl Acad. Sci. USA (1996) 93:6236–6240.[Abstract/Free Full Text]
  3. Sonoda E, et al. RAD51-deficient vertebrate cells accumulate chromosomal breaks prior to cell death. EMBO J. (1998) 17:598–608.[CrossRef][ISI][Medline]
  4. Lambert S, et al. Inactivation of the RAD51 recombination pathway stimulates UV-induced mutagenesis in mammalian cells. Oncogene (2002) 21:4065–4069.[CrossRef][ISI][Medline]
  5. Thacker J. The RAD51 gene family, genetic instability and cancer. Cancer Lett. (2005) 219:125–135.[CrossRef][ISI][Medline]
  6. Richardson C. RAD51, genomic stability, and tumorigenesis. Cancer Lett. (2005) 218:127–139.[CrossRef][ISI][Medline]
  7. Shinohara A, et al. Cloning of human, mouse and fission yeast recombination genes homologous to RAD51 and recA. Nat. Genet. (1993) 4:239–243.[CrossRef][ISI][Medline]
  8. Hasselbach L, et al. Characterisation of the promoter region of the human DNA-repair gene RAD51. Eur. J. Gynaecol. Oncol. (2005) 26:589–598.[ISI][Medline]
  9. Wang WW, Spurdle AB, Kolachana P, et al. A single nucleotide polymorphism in the 5' untranslated region of RAD51 and risk of cancer among BRCA1/2 mutation carriers. Cancer Epidemiol. Biomarkers Prev. (2001) 10:955–960.[Abstract/Free Full Text]
  10. Levy-Lahad E, Lahad A, Eisenberg S, et al. A single nucleotide polymorphism in the RAD51 gene modifies cancer risk in BRCA2 but not BRCA1 carriers. Proc. Natl Acad. Sci. USA (2001) 98:3232–3236.[Abstract/Free Full Text]
  11. Kadouri L, et al. A single-nucleotide polymorphism in the RAD51 gene modifies breast cancer risk in BRCA2 carriers, but not in BRCA1 carriers or noncarriers. Br. J. Cancer (2004) 90:2002–2005.[CrossRef][ISI][Medline]
  12. Jakubowska A, et al. Breast cancer risk reduction associated with the RAD51 polymorphism among carriers of the BRCA1 5382insC mutation in Poland. Cancer Epidemiol. Biomarkers Prev. (2003) 12:457–459.[Abstract/Free Full Text]
  13. Kuschel B, et al. Variants in DNA double-strand break repair genes and breast cancer susceptibility. Hum. Mol. Genet. (2002) 11:1399–1407.[Abstract/Free Full Text]
  14. Lee KM, et al. Genetic polymorphisms of selected DNA repair genes, estrogen and progesterone receptor status, and breast cancer risk. Clin. Cancer Res. (2005) 11:4620–4626.[Abstract/Free Full Text]
  15. Arias-Lopez C, et al. p53 modulates homologous recombination by transcriptional regulation of the RAD51 gene. EMBO Rep. (2006) 7:219–224.[CrossRef][ISI][Medline]
  16. Buchhop S, et al. Interaction of p53 with the human RAD51 protein. Nucleic Acids Res. (1997) 25:3868–3874.[Abstract/Free Full Text]
  17. Linke SP, et al. p53 interacts with hRAD51 and hRAD54, and directly modulates homologous recombination. Cancer Res. (2003) 63:2596–2605.[Abstract/Free Full Text]
  18. Harris CC. p53: at the crossroads of molecular carcinogenesis and risk assessment. Science (1993) 262:1980–1981.[Free Full Text]
  19. Dumont P, et al. The codon 72 polymorphic variants of p53 have markedly different apoptotic potential. Nat. Genet. (2003) 33:357–365.[CrossRef][ISI][Medline]
  20. Liu Z, et al. Polymorphisms of vitamin D receptor gene protect against the risk of head and neck cancer. Pharmacogenet. Genomics (2005) 15:159–165.[ISI][Medline]
  21. Wang LE, et al. Gender difference in smoking effect on chromosome sensitivity to gamma radiation in a healthy population. Radiat. Res. (2000) 154:20–27.[CrossRef][ISI][Medline]
  22. Wang LE, et al. Polymorphisms of DNA repair genes and risk of glioma. Cancer Res. (2004) 64:5560–5563.[Abstract/Free Full Text]
  23. Shen H, et al. P53 codon 72 polymorphism and risk of squamous cell carcinoma of the head and neck: a case-control study. Cancer Lett. (2002) 183:123–130.[CrossRef][ISI][Medline]
  24. Akaike H. A new look at statistical model identification. IEEE Trans. Autom. Control (1974) 19:716–723.[CrossRef]
  25. Hong Y, et al. The role of P53 and MDM2 polymorphisms in the risk of esophageal squamous cell carcinoma. Cancer Res. (2005) 65:9582–9587.[Abstract/Free Full Text]
  26. Wang LE, et al. In vitro sensitivity to ultraviolet B light and skin cancer risk: a case-control analysis. J. Natl Cancer Inst. (2005) 97:1822–1831.[Abstract/Free Full Text]
  27. Hsu TC, et al. Sensitivity to genotoxic effects of bleomycin in humans: possible relationship to environmental carcinogenesis. Int. J. Cancer (1989) 43:403–409.[ISI][Medline]
  28. Auranen A, et al. Polymorphisms in DNA repair genes and epithelial ovarian cancer risk. Int. J. Cancer (2005) 117:611–618.[CrossRef][ISI][Medline]
  29. Chakravarti D, et al. Role of CBP/P300 in nuclear receptor signalling. Nature (1996) 383:99–103.[CrossRef][Medline]
  30. Raderschall E, et al. Elevated levels of RAD51 recombination protein in tumor cells. Cancer Res. (2002) 62:219–225.[Abstract/Free Full Text]
  31. Maacke H, et al. DNA repair and recombination factor RAD51 is over-expressed in human pancreatic adenocarcinoma. Oncogene (2000) 19:2791–2795.[CrossRef][ISI][Medline]
  32. Qiao GB, et al. High-level expression of RAD51 is an independent prognostic marker of survival in non-small-cell lung cancer patients. Br. J. Cancer (2005) 93:137–143.[CrossRef][ISI][Medline]
  33. Bertrand P, et al. Overexpression of mammalian RAD51 does not stimulate tumorigenesis while a dominant-negative RAD51 affects centrosome fragmentation, ploidy and stimulates tumorigenesis, in p53-defective CHO cells. Oncogene (2003) 22:7587–7592.[CrossRef][ISI][Medline]
  34. Yoo S, et al. Overexpression of Drosophila RAD51 protein (DmRAD51) disrupts cell cycle progression and leads to apoptosis. Chromosoma (2004) 113:92–101.[ISI][Medline]
  35. Flygare J, et al. Effects of HsRAD51 overexpression on cell proliferation, cell cycle progression, and apoptosis. Exp. Cell Res. (2001) 268:61–69.[CrossRef][ISI][Medline]
  36. Little JB. Radiation carcinogenesis. Carcinogenesis (2000) 21:397–404.[Abstract/Free Full Text]
  37. Bondy ML, et al. Gamma-radiation sensitivity and risk of glioma. J. Natl Cancer Inst. (2001) 93:1553–1557.[Abstract/Free Full Text]
  38. Buchholz TA, et al. Radiation-induced chromatid breaks as a predictor of breast cancer risk. Int. J. Radiat. Oncol. Biol. Phys. (2001) 49:533–537.[CrossRef][ISI][Medline]
  39. Xiong P, et al. A pilot case-control study of gamma-radiation sensitivity and risk of papillary thyroid cancer. Thyroid (2005) 15:94–99.[CrossRef][ISI][Medline]
  40. Parshad R, et al. Radiation-induced chromatid breaks and deficient DNA repair in cancer predisposition. Crit. Rev. Oncol. Hematol. (2001) 37:87–96.[ISI][Medline]
  41. Hamel N, et al. No association between P53 codon 72 polymorphism and risk of squamous cell carcinoma of the head and neck. Br. J. Cancer (2000) 82:757–759.[CrossRef][ISI][Medline]
Received September 4, 2006; revised November 9, 2006; accepted November 11, 2006.


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
28/5/988    most recent
bgl225v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (1)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Lu, J.
Right arrow Articles by Wei, Q.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lu, J.
Right arrow Articles by Wei, Q.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?