Cancer Research Cancer Epigenetics Jordan
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
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 PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kyoizumi, S.
Right arrow Articles by Nakachi, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kyoizumi, S.
Right arrow Articles by Nakachi, K.
[Cancer Research 65, 5462-5469, June 15, 2005]
© 2005 American Association for Cancer Research

Epidemiology

Individual Variation of Somatic Gene Mutability in Relation to Cancer Susceptibility: Prospective Study on Erythrocyte Glycophorin A Gene Mutations of Atomic Bomb Survivors

Seishi Kyoizumi1, Yoichiro Kusunoki1, Tomonori Hayashi1, Masayuki Hakoda2, John B. Cologne3 and Kei Nakachi1

Departments of 1 Radiobiology/Molecular Epidemiology, 2 Clinical Studies, and 3 Statistics, Radiation Effects Research Foundation, Hiroshima, Japan

Requests for reprints: Yoichiro Kusunoki, Department of Radiobiology/Molecular Epidemiology, Radiation Effects Research Foundation, 5-2 Hijiyama Park, Minami-ku, 732-0815 Hiroshima, Japan. Phone: 81-82-261-3131; Fax: 81-82-261-3170; E-mail: ykusunok{at}rerf.or.jp.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
It has previously been reported that hemizygous mutant fraction (Mf) at the glycophorin A (GPA) locus in erythrocytes increased with radiation dose in heterozygotes among Hiroshima and Nagasaki atomic bomb survivors. In the present study, we analyzed the relationship between GPA Mf and cancer risk using newly developed cancers among previously cancer-free subjects whose GPA Mf had been measured between 1988 and 1996. Among 1,723 survivors (1,117 in Hiroshima and 606 in Nagasaki), we identified 186 subjects who developed a first cancer by the end of 2000. We compared the radiation dose responses of GPA Mf between cancer and cancer-free groups using a linear-quadratic model fit by multiple regression analysis in combination with age, sex, and city. The slope of the GPA Mf dose-response curve was significantly higher in the cancer group than in the cancer-free group among Hiroshima subjects. Moreover, no significant difference of GPA Mf between cancer and cancer-free groups was found in unexposed controls in the two cities. The same conclusions were obtained using a linear dose-response model and by further analysis using Cox regression of cancer incidence. These findings suggest that there might be interindividual variation in mutability of somatic genes and that Hiroshima survivors who have higher mutability in response to radiation exposure would be expected to have a higher probability of suffering radiation-related cancer.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Interindividual variability in human responses to mutagen exposures, including ionizing radiation, is believed to be a critical element in determining individual risk of cancer as well as the incidence of cancer in a population. At least a part of such interindividual variability of cancer susceptibility may be attributed to capacity of responses to oxidative DNA damage generated by mutagens (1). Multistaged defense mechanisms may exist in the responses to oxidative DNA damage, involving the initial defense against reactive oxygen species by superoxide dismutase and catalase, inhibition of incorporating oxidized bases into DNA by hydrolase, and repair of DNA damage (i.e., base excision repair, transcription-coupled repair, global genome repair, mismatch repair, translesion synthesis, homologous recombination, and nonhomologous end joining). Among them, much attention has been paid to several DNA repair genes. There is increasing evidence that mild reductions in DNA repair capacity, assumed to be the consequence of common genetic variation, affect cancer predisposition (2, 3). Currently, molecular epidemiologic studies are being conducted in many laboratories to define the roles that polymorphisms in DNA repair genes play in individual cancer susceptibility (3–5). In contrast to such genetic markers, phenotypic markers of DNA repair capacity and cancer susceptibility comprise both genetically and environmentally determined fractions and express integrated effects of complicated processes where a number of gene products are involved. Thus, phenotypic markers have often played vital roles in cancer research, specifically in prospective cohort studies, assessing the exposure levels (biodosimetry) as well as cancer risk (2, 6).

Many phenotyping assays have been developed using blood cells and skin fibroblasts for quantifying in vivo somatic mutations and in vitro DNA repair capacity (6). The erythrocyte glycophorin A (GPA) mutation assay, which can enumerate hemizygous mutants at the GPA locus in long-lived hematopoietic stem cells of heterozygous donors, provides one useful phenotypical end point for the assessment of cancer risk. This was supported by the findings that highly elevated GPA mutant fractions (Mf) were detected in patients with cancer-prone diseases, such as ataxia telangiectasia (7), Bloom's syndrome (8, 9), Fanconi's anemia (8, 10), and Werner syndrome (11, 12). These patients have defects in genes that are involved in several pathways of DNA repair mechanisms. It was also reported that GPA Mfs can be used as an assessment marker for the development of secondary induced leukemia in patients treated for childhood acute lymphoblastic leukemia (13). These findings suggest that the GPA Mf may, in some way, reflect individual repair capacity and cancer risk.

To clarify the association between radiation-induced mutation and cancer risk, prospective studies are critical to exclude the role of cancer itself in the association, such as through chemotherapy and radiation therapy. Because the atomic bomb (A-bomb) survivor population is an epidemiologically well-controlled cohort in terms of dose estimation (14) and cancer follow-up (15), such an analysis is feasible in this population. We previously measured hemizygous GPA Mf in ~1,200 heterozygous A-bomb survivors in Hiroshima and Nagasaki between 1988 and 1993 and analyzed the dose response of GPA Mf and the relationship between GPA Mf and cancer risk (16). It was found that the doubling dose of GPA Mf was similar to that of solid-cancer incidence in A-bomb survivors. Furthermore, the dose response was significantly higher in persons who had been diagnosed with cancer than in cancer-free individuals among Hiroshima survivors. This suggests an earlier onset of cancer due to enhanced mutagenesis or a higher radiation sensitivity in the cancer group. However, although we attempted to exclude all survivors who had undergone chemotherapy and radiotherapy, we may have missed some of them due to incompleteness of the medical records. Thus, because we could not completely exclude the possible effect of the therapies on GPA Mf, a prospective study was desired.

We have extended the GPA Mf measurements to ~1,900 survivors in total as of 1996 and followed them until 2000 to identify newly developed cancers among the previously cancer-free subjects. Based on these prospective data, we reanalyzed the relationship between GPA Mf and cancer development. In the present report, we show the reproducibility of the previous findings and discuss interindividual variation of susceptibility to radiation-induced mutagenesis, which may be associated with subsequent cancer risk.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Study subjects. Blood samples were obtained randomly from 1,902 survivors whose MN blood types were heterozygous by the hemagglutination test, who were participating in the Radiation Effects Research Foundation (RERF) Adult Health Study from June 1988 to August 1996. We excluded 179 survivors who were diagnosed with cancer before the GPA measurements and observed subsequent cancer development. Subject ages ranged from 43 to 100; mean ages were 63 for males and 67 for females.

Survivors in Hiroshima and Nagasaki, who have been diagnosed with malignant tumors (n = 186; 118 in Hiroshima; 68 in Nagasaki) through December 2000, were identified from the RERF tumor registry (17). Diagnoses and medical treatment histories for these survivors were also confirmed from the Adult Health Study medical charts. Identified malignant tumors included stomach (n = 32), colon (n = 31), lung (n = 19), liver (n = 18), breast (n = 10), rectum (n = 10), pancreas (n = 8), prostate (n = 7), gall bladder (n = 6), esophagus (n = 5), thyroid (n = 5), and other (n = 35) cancers.

The distribution of subjects by DS86 bone marrow doses (14), sex, and city are shown in Table 1. This distribution is similar to that of the total Adult Health Study population. The estimated dose includes both neutron and {gamma}-ray components. The analyses described in the present report were based on weighted bone marrow doses computed as the {gamma} dose plus 10 times the neutron dose and adjusted for the effect of imprecision on regression analyses (18). The weighting factor will be called the relative biological effectiveness of neutrons, and weighted doses are expressed in sieverts (Sv). This study population consists of Hiroshima and Nagasaki survivors who were exposed to significant radiation doses of ≥0.004 Sv because of their location within 2 km of the hypocenter plus a second group whose exposures were at distances in excess of 3 km from the hypocenter and as a result led to them receiving radiation doses of <0.004 Sv (i.e., doses that are indistinguishable from background). The latter group of distally exposed survivors includes the most appropriate controls for all of our studies of the effects of A-bomb radiation exposures, including the present one.


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

  		Table 1. Distribution of subjects excluding persons who had malignant cancer before GPA Mf measurement

 
Measurement of glycophorin A mutation frequency. Using the GPA mutation assay, four types of mutant cells, M{phi}, N{phi}, MM, and NN cells, can be detected among the erythrocyte populations of MN heterozygous donors. Hemizygous M{phi} and N{phi} cells are caused by deactivation of N or M alleles of the GPA gene, respectively. Homozygous MM and NN cells may be induced by somatic recombination between the two chromosomes on which the M and N alleles reside. Among these four types of mutants, the reproducibility of NN cells was low, probably due to carbohydrate modification of the GPA molecules (19, 20). Also, MM mutant frequency is significantly affected by overlapping of M{phi} mutants in the MM mutant window on the flow cytogram, particularly for the high-dose exposed who have high M{phi} Mfs. Thus, in this report, statistical analysis was undertaken for the mean of M{phi} and N{phi} hemizygous Mf (GPA Mf).

The detailed method for the flow cytometric measurement of mutant erythrocytes has been described previously (20). Briefly, using a single-beam cell sorter, FACStar (Becton Dickinson Immunocytometry System, San Jose, CA), four types of variants lacking the expression of one GPA allele were distinguished from normal MN heterozygous cells. By two-color staining with the GPA (M + N)–specific monoclonal antibody (mAb) 10F7 and the GPA (M)–specific mAb 6A7, two mutant cell types, hemizygous N{phi} and homozygous NN cells, from MN heterozygous donors can be detected simultaneously. By combining the GPA (M)–specific mAb 9A3 and the GPA (N)–specific mAb NN3, hemizygous M{phi} and homozygous MM cells can be measured simultaneously. 10F7 and 9A3 mAbs were directly labeled with fluorescein and mAbs 6A7 and NN3 were conjugated with biotin followed by labeling with streptavidin-conjugated phycoerythrin (Biomeda, Foster City, CA). Mutant cells displaying a hemizygous or homozygous phenotype were sorted onto a glass slide. Cells showing typical erythrocyte morphology with fluorescein fluorescence matched for the mutant phenotype were counted under a fluorescence microscope. Typically, ~106 total erythrocytes were assayed per sample.

Statistical analysis. Dose responses for GPA Mf and incidence proportion of cancer were fit using ordinary (least squares) regression. Weighted, adjusted bone marrow dose was used as described above. Age at examination was centered at its mean (65 years). City, sex, and cancer status were treated as indicator variables. Application of least-squares regression to the GPA Mf and radiation dose-response data for purposes of statistical testing would necessitate logarithmic transformation of both the GPA Mf and radiation dose variables (16) to achieve approximate normality and constant variance of the response variable (log GPA Mf) and approximately uniform distribution of the predictor variable (log radiation dose), but we desired to mimic standard radiobiological practice and fit linear or linear-quadratic dose responses. Thus, we did not transform the variables for the purpose of estimating the radiation dose response and instead verified the fit of the least-squares regression using nonparametric curve-fitting methods. As a further check on adequacy of the dose-response fit, individual observations with large influence on the regression analysis were identified through single-deletion regression diagnostics and regression models were refit after excluding such points. As a result, the small number of subjects with GPA Mf values >400 was not used in estimating the GPA Mf dose response. For fitting the cancer incidence proportion to radiation dose, least squares regression was applied to the binary indicator of cancer status. Approximately homogeneous variance and fit to the data were confirmed by comparing the fitted regression to a plot of binomial proportions grouped on radiation dose with approximately equal numbers of subjects (Fig. 1).



View larger version (18K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. Dose response of cancer incidence proportion subsequent to GPA Mf measurement adjusted for city, sex, and age at examination. The total number of subjects is 1,723. Points, crude proportions in 10 dose groups with approximately equal numbers of subjects (n = 115) except for unexposed controls (n = 691); bars, SE of the estimated proportions and the quartiles of dose.

 
Follow-up for incident cancer subsequent to GPA measurement was analyzed using Cox regression with age as the time scale and adjustment for year of birth and age at examination. The effect of GPA Mf on cancer incidence was assessed using either the logarithm of continuous GPA Mf to reduce the influence of the small number of points with large values of GPA Mf, or the untransformed GPA Mf excluding the subjects with values larger than 400. For graphic presentation of the results of the Cox regression, summary plots of cumulative incidence (proportion of subjects who were free of cancer) were produced by dividing subjects into two strata—low and high values of GPA Mf—based on the median GPA Mf among cancer cases.

All analyses were conducted using S-plus version 2000 (Mathsoft, Inc., Seattle, WA).


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Cancer prevalence in the study cohort. One hundred eighty-six subjects developed a first cancer between the GPA Mf measurement and the end of 2000. There was a statistically significant increase in cancer incidence proportion with dose after adjustment for city, sex, and age (Fig. 1). This suggests that the subjects of the present study developed cancers in a dose-dependent manner even >40 years after exposure.

Effects of cancer status and city difference on dose responses of glycophorin A mutant fraction. Figure 2 shows nonparametric curves for the GPA Mf (the mean of hemizygous M{phi} and N{phi} Mf) values according to bone marrow dose among all study subjects (n = 1,902), including cancer cases diagnosed before GPA Mf measurements. The plots suggest that the GPA radiation dose response is steeper among cancer patients in Hiroshima, particularly among those whose cancer was diagnosed subsequent to GPA measurement (Fig. 2, left). Persons who had cancer diagnosed before the time of GPA measurement may not be representative, because some individuals with cancer might have been censored—debilitated or deceased—and unable to attend the Adult Health Study examination. In contrast, no apparent differences were observed between cancer and noncancer groups among Nagasaki survivors (Fig. 2, right).



View larger version (17K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2. Dose response of GPA Mf based on nonparametric curve fitting for cancer (either before or after GPA Mf measurement) and noncancer groups in Hiroshima and Nagasaki. Symbols indicate subjects who were diagnosed with cancer before GPA Mf measurement ({blacksquare}), those who were diagnosed with cancer after GPA Mf measurement ({blacktriangleup}), and those who were not diagnosed with cancer ({circ}). Lines denote fitted dose-response curves for cancer and noncancer groups.

 
To evaluate further the possible difference in cancer-related GPA dose response between the two cities, we did standard regression on GPA Mf using bone marrow dose, cancer status, sex, and age at examination as covariates, excluding subjects with values of GPA Mf >400 and those who had cancer diagnoses before GPA measurement (Table 2). Background GPA Mf (estimate for 0 Gy exposure) was lower in Nagasaki than in Hiroshima and lower in females than in males. An increase in GPA Mf with age of the subject at examination was only marginally significant, probably because age-dependent increase in GPA Mf reaches a plateau after about 50 years of age (21). There was an initial increase in the GPA response with bone marrow dose, followed by an attenuation in the slope (negative quadratic term). Although the quadratic term was statistically significant (P = 0.033), there was little quantitative difference in the results with or without the quadratic term except for a slightly lower dose-response slope without the quadratic term. There was no significant difference in background GPA Mf between male subjects who subsequently developed cancer and those who did not, but among females the background GPA Mf was lower among those with cancer.


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

  		Table 2. Regression analysis of GPA Mf using linear and linear-quadratic dose-response models

 
The fitted linear-quadratic regression models for each of the two cities, adjusted for sex and age, are shown in Fig. 3. The dose response of GPA Mf was significantly higher in Hiroshima subjects who subsequently developed cancer than in those who did not. Whether there was a difference in dose response by cancer status depended significantly on city (a three-way interaction between city, dose, and cancer status; P = .0081), with no difference in Nagasaki. Seven points were identified that had a high influence on the value of the city x cancer x dose three-way interaction term. Upon deleting these points, the value of the interaction term decreased somewhat but remained statistically significant. Among Nagasaki subjects, the initial slope of the GPA dose response in the linear-quadratic model adjusted for the average values of the other factors was 22.6 in cancer-free subjects and 22.3 in subjects with subsequent cancer (P = 0.29). In Hiroshima, the similarly adjusted initial slopes were 35.8 in cancer-free subjects and 51.3 in subjects with subsequent cancer (P = 0.0039). There was no evidence that the quadratic term of the dose response differed according to cancer status.



View larger version (17K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3. Dose response of GPA Mf among cancer (post-GPA Mf measurement) and noncancer groups based on the linear-quadratic regression model in Hiroshima and Nagasaki. The number of subjects in each group is listed in Table 1. Symbols indicate subjects who were diagnosed with cancer after GPA Mf measurement ({blacktriangleup}) and subjects without cancer ({circ}). Lines denote fitted dose-response curves for cancer and noncancer groups. P values represent the statistical significance of difference of dose response (quadratic term) between cancer post-GPA Mf measurement and noncancer groups.

 
Difference of glycophorin A mutant fraction between cancer and noncancer groups among high dose–exposed subjects. Because the GPA Mf values are highly skewed, even when restricted to ≤400 x 10–6, the results of statistical tests of parameters from the regression analysis may not be reliable. Therefore, we compared log-transformed GPA Mf values between high dose and unexposed persons, thereby avoiding assuming any particular dose-response model, using either a t test (Fig. 4) or regression analysis with adjustment for sex and age (Table 3). Among unexposed subjects, there was no difference in log GPA Mf between persons with and without cancer in either city. The log GPA Mf was significantly higher in subjects with subsequent cancer than without cancer among the heavily exposed (>1.5 Sv) subjects in Hiroshima (t test P = 0.012, regression P = 0.0057) but was not significantly different between subjects with and without cancer in Nagasaki (t test P = 0.52, regression P = 0.21).



View larger version (18K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4. Comparison of GPA Mf between cancer and noncancer groups within unexposed and high dose exposed (>1.5 Sv), Hiroshima and Nagasaki. {diamondsuit}, mean of GPA Mf; bars, SE. P values represent the statistical significance of difference of GPA Mfs between cancer and noncancer groups within each dose group.

 

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

  		Table 3. Relationship between log GPA Mf and cancer status in high dose–exposed survivors

 
The analysis of cancer onset rates during the follow-up period by Cox regression confirmed the finding that cancer risk was related to GPA Mf level among high-dose–exposed subjects in Hiroshima but not in Nagasaki (Fig. 5). Whereas there was no association between log GPA Mf level and cancer onset rate among unexposed persons or Nagasaki high-dose–exposed persons, there was significantly higher risk of cancer with higher GPA Mf value in Hiroshima high-dose–exposed persons (P = 0.043). The estimated relative risk of cancer for a 10-unit difference in GPA Mf (i.e., a 50% increase over the median) was 1.13 (95% confidence interval, 1.00-1.27; Table 4). Relative risks of cancer for GPA for all individual dose groups by city are shown in Table 4. There was no change in the significance of the results when log GPA was used.



View larger version (18K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5. Cox regression analyses of cancer incidence following GPA Mf measurement within unexposed and high dose exposed (>1.5 Sv) in Hiroshima and Nagasaki. Dashed line, low GPA Mf; solid line, high GPA Mf; low and high are defined as below or above the median GPA Mf level among cancer cases. Numbers of cases of cancer are as shown in Fig. 4. P values represent the statistical significance of difference of cancer incidence between high and low GPA Mf groups within each dose group.

 

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

  		Table 4. Relative risk of cancer for GPA (relative risk for 10-unit difference in GPA Mf)

 
The t tests and Cox regression analyses were repeated after excluding persons whose follow-up was <1 year. There was no change in the pattern of results. The association among high dose and cancer on log GPA Mf was significant in Hiroshima (P = 0.024) but not in Nagasaki (P = 0.41). The effect of GPA Mf on cancer incidence was significant in high-dose persons in Hiroshima (P = 0.041) but not in Nagasaki (P = 0.66).


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
It is increasingly accepted that the accumulation of multiple abnormalities in cancer-associated genes of a target cell is required for the development of cancer (22–24). Although the mechanism of radiation carcinogenesis is still unknown, some evidence has been presented that A-bomb radiation seems to reduce the number of gene changes needed for cancer induction, thereby inducing earlier onset of cancer in the exposed, compared with the unexposed (25). If GPA Mf reflects nonspecific mutability of all somatic cells in an individual, it can be presumed that GPA Mf may also reflect the prevalence of mutations at cancer-associated genes. Among survivors exposed to the same dose, those who have a higher GPA Mf would be expected to have a higher probability of suffering cancer at any given point in time. In fact, as shown in the present study, the dose response of GPA Mf in the cancer group among Hiroshima survivors was found to be significantly higher than that in the noncancer group. The dose response among persons with cancer diagnosis before GPA measurement was intermediate to that of the groups with no cancer or cancer after GPA measurement, which is further consistent with a higher cancer risk for high GPA Mf radiation–exposed subjects because persons at high risk of cancer would more likely be censored (unobserved due to death or debility) before GPA measurement. The individual difference of GPA Mf in the same dose group might be explained by individual variation in the capacity to repair radiation-induced DNA damage. We discuss below the validity and feasibility of these hypotheses regarding individual differences in DNA repair capacity among A-bomb survivors.

A potentially important source of interindividual variability in relation to cancer risk is DNA repair capacity, including the DNA repair–defective cancer-prone diseases, such as ataxia telangiectasia, Bloom's syndrome, Fanconi's anemia, and Werner syndrome. Apart from these rare and extreme familial cases, there is increasing evidence that a moderate reduction in DNA repair capacity contributes to the sporadic incidence of cancer in the general population (2, 3). Conventional phenotype assays have detected considerable interindividual variation in DNA repair capacity (2, 26, 27). These reduced repair capacity phenotypes have been associated with an increased risk of cancer (2, 28, 29). Evidence of the importance of moderate reduction in DNA repair capacity is also accumulating from mouse models, which have provided results regarding cancer risk increased by heterozygous knock-out in DNA repair genes (30, 31) and those regarding strain differences in cancer susceptibility (32, 33). Furthermore, a number of molecular epidemiologic studies have been initiated using the data from systematic screening of populations for common variants in DNA repair genes (3, 34). Associations of common variants in several repair genes with increased cancer risk have been reported in case-control studies (5). In general human populations, it has been suggested that individual differences in peripheral blood T cell chromosome aberration frequencies may be associated with individual differences in cancer susceptibility (35, 36). These accumulating data are consistent with the hypothesis that interindividual variation in DNA repair capacity has an impact on cancer risk.

Statistical analysis in this study showed that there was a city difference (i.e., significant interaction among radiation, cancer, and city in their association with GPA MF; Table 2). This does not necessarily imply that there is a significant correlation between GPA Mf dose response and cancer in Hiroshima subjects but not in Nagasaki subjects. Small numbers of cases in Nagasaki make it difficult to clearly state the apparent negative finding. However, the observed city difference may be, at least in part, due to possible differences in ethnic background between Hiroshima and Nagasaki, which were suggested by the previous biochemical genetic study of A-bomb survivors and their children (37). We might presume that the proportion of individuals who have higher mutability of somatic genes is somewhat larger in Hiroshima than in Nagasaki. Interestingly, the background GPA Mf of Hiroshima is significantly higher than that of Nagasaki, as shown in the present study. Such a city difference was also reported for the background solid-tumor mortality (38) and chronic myelogenous leukemia (CML) of survivors (39). City differences of dose response (lower in Nagasaki) were also suggested for the solid tumor mortality (15), the chromosome aberration frequencies in lymphocytes (40), and the incidence of CML (39). These city differences have been attributed to dose estimation errors and/or to a qualitative difference in the radiation produced by the bombs (i.e., the difference in the amount of neutron and {gamma}-ray components), assuming that there is no city difference in radiation sensitivity of A-bomb survivors (15, 40). Because recent genomic analyses have shown extensive interindividual—including ethnic—variations in gene polymorphisms, as mentioned above, this assumption should be reassessed.

The following possible caveats of the present study should be kept in mind. We conducted a prospective study, which is critical to exclude possible effects of chemotherapy and/or radiotherapy on in vivo somatic mutations. Nevertheless, because the period between GPA Mf measurement and cancer diagnosis is rather short (average: about 3.7 years; range: 22 days-9.4 years; n = 187), it is possible that tumors of preclinical size had already developed before GPA Mf measurement. Tumor burden, even with a very small lesion, might increase somatic mutations through high metabolic rate and excessive endogenously generated oxidative stress. However, this may not be the case because GPA Mf values were nearly constant in the cancer subjects (n = 29) whose Mf were measured more than twice during the 8-year examination period before cancer diagnosis (data not shown). Further, the radiation dose responses of GPA Mf between cancer and cancer-free groups did not change after excluding all persons whose follow-up ended within 1 year of GPA measurement (data not shown). Another factor complicating the interpretation is uncertainty in dose estimation, because persons with radiation-related outcomes, such as cancer, are more likely to have underestimated dose than persons without. However, our comparison between high-dose–exposed and nonexposed persons, which is unaffected by dose uncertainties, did not change the conclusions.

Although our hypothesis is valid, it is far from proved based on the present study alone. We believe that the evolving ability to study polymorphisms in DNA repair genes can contribute to understanding about the relationship between DNA repair capacity and cancer risk in A-bomb survivors. Because the difference in GPA Mf between cancer and cancer-free groups is significant only for high dose–exposed survivors, a candidate polymorphic gene affecting interindividual variability could be one involved in repair of DNA double-strand breaks induced by high-dose irradiation. Double-strand breaks are potentially cytotoxic to cells and mutagenic. At least two molecular mechanisms are involved in the pathway of double-strand break repair: homologous recombination repair and nonhomologous end joining (3, 5, 41, 42). Homologous recombination repair occurs predominantly in the S or G2 phase of cell division and exchanges DNA strands between the damaged chromatid and the intact sister chromatid. Nonhomologous end joining repair involves direct ligation of the two double-strand break ends. These two pathways are thought to involve numerous molecules. Although extensive polymorphic variation in double-strand breaks repair genes has been reported, only a few common polymorphisms of these genes have been examined in epidemiologic studies for their association with cancer risk (43–48). We are planning to study the A-bomb survivors further to elucidate the relationship between DNA repair gene polymorphisms and risk of radiation-induced cancer. The phenotypical data, such as those obtained in the present study, will provide valuable information toward drawing conclusions from genotype-cancer association analyses in future studies.


   Acknowledgments
 
Grant support: Radiation Effects Research Foundation (RERF), Hiroshima and Nagasaki, Japan. RERF is a private nonprofit foundation funded equally by the Japanese Ministry of Health, Labour, and Welfare and the U.S. Department of Energy, the latter through the National Academy of Sciences. This publication was based on RERF Research Protocol 3-87 and supported in part by grants-in-aid for Scientific Research from the Ministry of Health, Labour, and Welfare.

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.

We thank Kazunori Kodama, Masazumi Akahoshi, and other clinical staff at RERF for providing blood samples; Mika Yonezawa for manuscript preparation; and Kazumi Tanabe and Yoshiko Kubo for excellent technical help.


   Footnotes
 
Note: S. Kyoizumi is currently at Yasuda Women's University, Hiroshima, Japan.

Received 4/ 5/04. Revised 2/ 8/05. Accepted 3/24/05.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Collins A, Harrington V. Repair of oxidative DNA damage: assessing its contribution to cancer prevention. Mutagenesis 2002;17:489–93.[Abstract/Free Full Text]
  2. Berwick M, Vineis P. Markers of DNA repair and susceptibility to cancer in humans: an epidemiologic review. J Natl Cancer Inst 2000;92:874–97.[Abstract/Free Full Text]
  3. Mohrenweiser HW, Wilson DM III, Jones IM. Challenges and complexities in estimating both the functional impact and the disease risk associated with the extensive genetic variation in human DNA repair genes. Mutat Res 2003;526:93–125.[Medline]
  4. Miller MC III, Mohrenweiser HW, Bell DA. Genetic variability in susceptibility and response to toxicants. Toxicol Lett 2001;120:269–80.[CrossRef][Medline]
  5. Goode EL, Ulrich CM, Potter JD. Polymorphisms in DNA repair genes and associations with cancer risk. Cancer Epidemiol Biomarkers Prev 2002;11:1513–30.[Abstract/Free Full Text]
  6. Bonassi S, Au WW. Biomarkers in molecular epidemiology studies for health risk prediction. Mutat Res 2002;511:73–86.[CrossRef][Medline]
  7. Bigbee WL, Langlois RG, Swift M, et al. Evidence for an elevated frequency of in vivo somatic cell mutations in ataxia telangiectasia. Am J Hum Genet 1989;44:402–8.[Medline]
  8. Kyoizumi S, Nakamura N, Takebe H, et al. Frequency of variant erythrocytes at the glycophorin-A locus in two Bloom's syndrome patients. Mutat Res 1989;214:215–22.[Medline]
  9. Langlois RG, Bigbee WL, Jensen RH, et al. Evidence for increased in vivo mutation and somatic recombination in Bloom's syndrome. Proc Natl Acad Sci U S A 1989;86:670–4.[Abstract/Free Full Text]
  10. Sala-Trepat M, Boyse J, Richard P, et al. Frequencies of HPRT-lymphocytes and glycophorin A variants erythrocytes in Fanconi anemia patients, their parents and control donors. Mutat Res 1993;289:115–26.[Medline]
  11. Kyoizumi S, Kusunoki Y, Seyama T, et al. In vivo somatic mutations in Werner's syndrome. Hum Genet 1998;103:405–10.[CrossRef][Medline]
  12. Moser MJ, Bigbee WL, Grant SG, et al. Genetic instability and hematologic disease risk in Werner syndrome patients and heterozygotes. Cancer Res 2000;60:2492–6.[Abstract/Free Full Text]
  13. Hewitt M, Mott MG. The assessment of in vivo somatic mutations in survivors of childhood malignancy. Br J Cancer 1992;66:143–7.[Medline]
  14. Roesch WC, editor. U.S.-Japan joint reassessment of atomic bomb radiation dosimetry in Hiroshima and Nagasaki, final report, vols. 1 and 2. Hiroshima: Radiation Effects Research Foundation; 1987.
  15. Pierce DA, Shimizu Y, Preston DL, et al. Studies of the mortality of atomic bomb survivors. Report 12, Part I. Cancer: 1950-1990. Radiat Res 1996;146:1–27.[Medline]
  16. Kyoizumi S, Akiyama M, Cologne JB, et al. Somatic cell mutations at the glycophorin A locus in erythrocytes of atomic bomb survivors: implications for radiation carcinogenesis. Radiat Res 1996;146:43–52.[CrossRef][Medline]
  17. Mabuchi K, Soda M, Ron E, et al. Cancer incidence in atomic bomb survivors. Part I: Use of the tumor registries in Hiroshima and Nagasaki for incidence studies. Radiat Res 1994;137:S1–16.[CrossRef][Medline]
  18. Pierce DA, Stram DO, Vaeth M. Allowing for random errors in radiation dose estimates for the atomic bomb survivor data. Radiat Res 1990;123:275–84.[CrossRef][Medline]
  19. Langlois RG, Bigbee WL, Kyoizumi S, et al. Evidence for increased somatic cell mutations at the glycophorin A locus in atomic bomb survivors. Science 1987;236:445–8.[Abstract/Free Full Text]
  20. Kyoizumi S, Nakamura N, Hakoda M, et al. Detection of somatic mutations at the glycophorin A locus in erythrocytes of atomic bomb survivors using a single beam flow sorter. Cancer Res 1989;49:581–8.[Abstract/Free Full Text]
  21. Akiyama M, Kyoizumi S, Hirai Y, et al. Mutation frequency in human blood cells increases with age. Mutat Res 1995;338:141–9.[CrossRef][Medline]
  22. Peto R, Roe FJ, Lee PN, et al. Cancer and aging in mice and men. Br J Cancer 1975;32:411–26.[Medline]
  23. Balmain A, Gray J, Ponder B. The genetics and genomics of cancer. Nat Genet 2003;33 Suppl:238–44.
  24. Thilly WG. Have environmental mutagens caused oncomutations in people? Nat Genet 2003;34:255–9.[CrossRef][Medline]
  25. Pierce DA, Mendelsohn ML. A model for radiation-related cancer suggested by atomic bomb survivor data. Radiat Res 1999;152:642–54.[CrossRef][Medline]
  26. Cloos J, Nieuwenhuis EJ, Boomsma DI, et al. Inherited susceptibility to bleomycin-induced chromatid breaks in cultured peripheral blood lymphocytes. J Natl Cancer Inst 1999;91:1125–30.[Abstract/Free Full Text]
  27. Roberts SA, Spreadborough AR, Bulman B, et al. Heritability of cellular radiosensitivity: a marker of low-penetrance predisposition genes in breast cancer? Am J Hum Genet 1999;65:784–94.[CrossRef][Medline]
  28. Wu X, Gu J, Hong WK, et al. Benzoapyrene diol epoxide and bleomycin sensitivity and susceptibility to cancer of upper aerodigestive tract. J Natl Cancer Inst 1998;90:1393–9.[Abstract/Free Full Text]
  29. Wu X, Gu J, Patt Y, et al. Mutagen sensitivity as a susceptibility marker for human hepatocellular carcinoma. Cancer Epidemiol Biomarkers Prev 1998;7:567–70.[Abstract]
  30. Goss KH, Risinger MA, Kordich JJ, et al. Enhanced tumor formation in mice heterozygous for Blm mutation. Science 2002;297:2051–3.[Abstract/Free Full Text]
  31. Spring K, Ahangari F, Scott SP, et al. Mice heterozygous for mutation in Atm, the gene involved in ataxia-telangiectasia, have heightened susceptibility to cancer. Nat Genet 2002;32:185–90.[CrossRef][Medline]
  32. Shoemaker AR, Moser AR, Midgley CA, et al. A resistant genetic background leading to incomplete penetrance of intestinal neoplasia and reduced loss of heterozygosity in ApcMin/+ mice. Proc Natl Acad Sci U S A 1998;95:10826–31.[Abstract/Free Full Text]
  33. Tripodis N, Hart AA, Fijneman RJ, et al. Complexity of lung cancer modifiers: mapping of thirty genes and twenty-five interactions in half of the mouse genome. J Natl Cancer Inst 2001;93:1484–91.[Abstract/Free Full Text]
  34. de Boer JG. Polymorphisms in DNA repair and environmental interactions. Mutat Res 2002;509:201–10.[Medline]
  35. Hagmar L, Bonassi S, Stromberg U, et al. Chromosomal aberrations in lymphocytes predict human cancer: a report from the European Study Group on Cytogenetic Biomarkers and Health (ESCH). Cancer Res 1998;58:4117–21.[Abstract/Free Full Text]
  36. Hagmar L, Stromberg U, Bonassi S, et al. Impact of types of lymphocyte chromosomal aberrations on human cancer risk: results from Nordic and Italian cohorts. Cancer Res 2004;64:2258–63.[Abstract/Free Full Text]
  37. Neel JV, Satoh C, Smouse P, et al. Protein variants in Hiroshima and Nagasaki: tales of two cities. Am J Hum Genet 1988;43:870–93.[Medline]
  38. Shimizu Y, Kato H, Schull WJ. Studies of the mortality of A-bomb survivors. 9. Mortality, 1950-1985: Part 2. Cancer mortality based on the recently revised doses (DS86). Radiat Res 1990;121:120–41.[CrossRef][Medline]
  39. Preston DL, Kusumi S, Tomonaga M, et al. Cancer incidence in atomic bomb survivors. Part III. Leukemia, lymphoma and multiple myeloma, 1950–1987. Radiat Res 1994;137:S68–97.[Medline]
  40. Kodama Y, Pawel D, Nakamura N, et al. Stable chromosome aberrations in atomic bomb survivors: results from 25 years of investigation. Radiat Res 2001;156:337–46.[CrossRef][Medline]
  41. Haber JE. Partners and pathways repairing a double-strand break. Trends Genet 2000;16:259–64.[CrossRef][Medline]
  42. Pierce AJ, Stark JM, Araujo FD, et al. Double-strand breaks and tumorigenesis. Trends Cell Biol 2001;11:S52–9.[Medline]
  43. Healey CS, Dunning AM, Teare MD, et al. A common variant in BRCA2 is associated with both breast cancer risk and prenatal viability. Nat Genet 2000;26:362–4.[CrossRef][Medline]
  44. Winsey SL, Haldar NA, Marsh HP, et al. A variant within the DNA repair gene XRCC3 is associated with the development of melanoma skin cancer. Cancer Res 2000;60:5612–6.[Abstract/Free Full Text]
  45. David-Beabes GL, Lunn RM, London SJ. No association between the XPD (Lys751G1n) polymorphism or the XRCC3 (Thr241Met) polymorphism and lung cancer risk. Cancer Epidemiol Biomarkers Prev 2001;10:911–2.[Free Full Text]
  46. Butkiewicz D, Rusin M, Enewold L, et al. Genetic polymorphisms in DNA repair genes and risk of lung cancer. Carcinogenesis 2001;22:593–7.[Abstract/Free Full Text]
  47. Matullo G, Guarrera S, Carturan S, et al. DNA repair gene polymorphisms, bulky DNA adducts in white blood cells and bladder cancer in a case-control study. Int J Cancer 2001;92:562–7.[CrossRef][Medline]
  48. Fu YP, Yu JC, Cheng TC, et al. Breast cancer risk associated with genotypic polymorphism of the nonhomologous end-joining genes: a multigenic study on cancer susceptibility. Cancer Res 2003;63:2440–6.[Abstract/Free Full Text]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
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 PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kyoizumi, S.
Right arrow Articles by Nakachi, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kyoizumi, S.
Right arrow Articles by Nakachi, K.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online