In this report, we confirmed our previous findings that deficiency in DNA damage-induced G2
/M arrest is associated with an increased risk of lung cancer in African Americans in this population (28
). Originally, we reported on 216 cases and 340 controls. The larger sample size in the present report allowed for increased statistical power to conduct stratified analyses by cross classification of race and gender, in which we found that the association for G2
/M arrest and lung cancer risk was restricted to African American women with an aOR of 2.6. Further, we assessed the role for genetic polymorphisms for affecting G2
/M arrest phenotypes and found that polymorphisms in ATM, CDC25C, CDKN1A, BRCA2, ERCC6, TP53,
genes were significantly associated with the γ-radiation-induced G2
/M arrest phenotype.
Lung cancer is more common in African Americans than in any other racial/ethnic group in the United States (33
), and cigarette smoking prevalence does not by itself provide a viable explanation for this observation (34
). Previous studies indicated that among smokers, African Americans were more susceptible to lung cancer than whites (35
). Among the smokers, the relative risk of lung cancer is significantly higher in both African American males (RR = 1.67) and African American females (RR = 1.20), compared to white males or females (35
). The increased risk is more prominent among light smokers (pack-years < 30)(35
). Among former smokers, the relative risk of lung cancer is 1.30 and 1.44 in African American men and African American women respectively, compared to white men and women (36
). The explanation for the observed racial variation in lung cancer susceptibility remains to be determined. One possible explanation for this disparity in lung cancer risk is that African American smokers are more susceptible to lung carcinogens from cigarette smoke than white smokers.
Genetically-mediated host factors may modulate the carcinogenic effect of tobacco smoke, and these host factors could distribute differently among racial groups. There is compelling evidence to indicate that the distribution of genetic polymorphisms throughout the human genome follows ethnic and/or racial composition of populations (37
) and these genetic variations may be associated with a difference in risk for disease among the racial groups (39
). For example, some studies have reported ethnic variation in blood levels of nicotine and cotinine after controlling for cigarette consumption (40
). CYP2A6, the primary nicotine metabolic enzyme, shows significant allelic variation among racial and ethnic groups (41
). Documented differences in allele frequencies between African Americans and whites for genes involved in DNA repair (42
) and hormone metabolism (43
) also have been shown to contribute to differences in lung cancer (44
), breast cancer (42
), and prostate cancer (46
) risk. Our group has previously reported that haplotypes of TP53
were significantly associated with lung cancer risk in African Americans, but not in Caucasians (47
). TP53 protein plays a key role in DNA repair and cell cycle control.
Cell cycle checkpoints regulate progression through the cell cycle, ensuring that each step takes place only once and in the right sequence. It has long been known that DNA-damaging agents induce a cell cycle arrest, buying time for repair, and thus, protecting the organism from the deleterious consequences of mutations (27
). Deficient or “leaky” cell cycle checkpoints could lead to significant accumulation of genetic mutations when the host is exposed to carcinogens, i.e., cigarette smoking, and consequently increasing cancer risk further. Our data suggest that deficiency in DNA damage-induced G2
/M checkpoint function may contribute to the increased susceptibility to lung cancer in African American women. If confirmed by other studies, these data provide some clues to the relatively high lung cancer incidence experienced by African American women.
We observed no significant association between the efficiency of DNA damage-induced G2
/M arrest and lung cancer risk in Caucasians. In a large case-control study of predominantly Caucasian subjects, Xing et al
reported that a less efficient G2
/M checkpoint was associated with a modestly increased lung cancer risk (OR = 1.28) in Caucasians and further analyses indicated that the case-control difference in the percent of G2
/M cell accumulation was only significant in Caucasian women, but not in Caucasian men (29
). They did not report the estimated odds ratio for Caucasian women. Although we did not find a significant association between G2
/M checkpoint and lung cancer risk in Caucasians, we did observe a borderline significant (P = 0.07) case-control difference of mean percent G2
/M arrest in Caucasian women.
The genotype-phenotype associations observed in this study support our hypothesis that polymorphisms in genes involving in DNA damage response affect the function of the cell cycle checkpoint. We found that the mean percent of DNA damage-induced G2/M arrest differs significantly by genotypes of 10 SNPs of 6 genes (ATM, BRCA2, CDC25C, CDKN1A, TP53, TP53BP1, ). Logistic regression analysis found that SNPs in ERCC6 were significantly associated with the DNA damage-induced G2/M arrest phenotype. However, the association between G2/M arrest and any individual SNP was modest. For example, the risk of having a less-efficient G2/M arrest was not significant when considering TP53 P72R or TP53BP1 IVS2+7G>A separately. When genotypes were combined and the effects were assessed by total number of adverse genotypes across a panel of genes, a consistent trend of strong associations were presented. The rationale for combining multiple SNPs for analysis is based on the concept that proteins of the cell cycle checkpoint and DNA repair pathways cooperate to carry out their highly coordinated functions. For example, it is possible that a less efficient variant TP53BP1 protein will have a diminished ability to recruit other proteins to activate p53, and a less efficient ERCC6 will not be able to coordinate competently the repair processes. Individuals carrying the multiple less efficient variant forms of these gene products will be at greater risk to have dysfunctional cell cycle control, hence increased cancer risk. Our data support this hypothesis.
Genetic polymorphisms have been intensively studied in terms of their associations with risk of various cancers (49
). The rationale behind these gene-cancer risk associations is that these genetic polymorphisms may result in alterations in gene products (i.e. protein structures) that affect the phenotypes (i.e. DNA repair capacity). However, the functions or phenotypes of the majority of these genetic polymorphisms are unknown and data on genotype-phenotype correlation are sparse. Two recent studies have examined the correlations between genetic variants in nucleotide excision repair pathway and BPDE-mutagen sensitivity and BPDE-induced DNA adducts level in cultured blood lymphocytes (50
). Lin et al
reported that several SNPs in XPC, XPA
were associated with mutagen sensitivity phenotype, and combined analysis of multiple SNPs revealed a significant dose-response relationship between increasing mutagen sensitivity with increasing number of adverse genotypes (51
). In a separate study, Zhao et al
reported that the genotypes and haplotypes of ERCC1
is significantly associated with level of BPDE-induced DNA adducts in cultured peripheral blood lymphocytes. Both mutagen sensitivity and BPDE-induced DNA adduct levels are considered reflective of the cellular DNA repair capacity. In the present study, we demonstrated that genetic polymorphisms in cell cycle control/DNA repair pathways are associated with the function of G2
/M checkpoint. Together, these data indicate that intermediate phenotypes of cancer susceptibility are useful tools to characterize potential function of SNPs, and to further the understanding of genetic contributions to lung cancer risk.
Our study has moderate sample size and this limited our ability to consider adjustment for multiple comparisons for the genotype-phenotype correlation analysis. There is a chance of reporting false positive association between the SNPs and G2/M arrest phenotype. In the present study, a total of 49 SNPs were examined using a p-value ≤0.05, thus the expected number of false positive SNPs is 3 (0.05 × 49 = 3). The observed number of positive SNPs (N = 13) with a p-value ≤ 0.05 is much larger than expected by chance (N = 3), suggesting that many of the identified SNPs are potentially true positive SNPs. The significant dose response relationship observed in the analysis of combined effects of 4 positive SNPs provided further evidence of true association (). Future larger studies are warranted to validate these new findings.
In summary, we have reported that a less-efficient G2/M checkpoint is significantly associated with lung cancer risk in African American women. Our data also suggest that genetic polymorphisms in ATM, BRCA2, CDC25C, CDKN1A, ERCC6, TP53 and TP53BP1 modulate the G2/M checkpoint function. Importantly, we found that the combination of multiple SNPs in the cell cycle control/DNA repair pathway is strongly associated with DNA damage-induced G2/M arrest phenotype. Future studies are warranted to further examine evidence supporting the hypothesis that genotypic and phenotypic differences underlie the observed disparities in lung cancer incidence between African Americans and Caucasians.