Deficient G₂-M and S Checkpoints are Associated with Increased Lung Cancer Risk: A Case-Control Analysis

Mammalian cells have evolved a complex and well-regulated defense system consisting of DNA repair, chromatin remodeling, cell cycle checkpoints, and apoptosis to maintain genomic integrity (1). Among these defenses, cell cycle checkpoints play a critical role in response to damaged DNA or impaired chromosome function by eliciting a multifaceted cellular response that delays cell cycle progression and allows time for DNA repair (2). Loss or attenuation of checkpoint function can compromise the fidelity of DNA repair.

Mammalian cell cycle checkpoints include the G₁-S, intra-S, and G₂-M checkpoints (3-5). The G₁-S checkpoint mainly prevents the cell from replicating damaged DNA by inducing sustained and sometimes even permanent G₁ arrest. The intra-S phase checkpoint protects the integrity of the stalled replication forks by reversibly inhibiting the firing from the DNA replication origins that have not yet been initiated. The G₂-M checkpoint prevents cells from initiating mitosis when they have experienced DNA damage during the G₂ phase or when they progress into the G₂ phase with unrepaired damage.

It has long been known that the exposure of cells to ionizing radiation delays the normal progression through the cell cycle (6). Although a delay in G₂ phase is the most evident, significant delays also occur in G₁, as well as throughout the S phase. Although the initial acute components of the G₁ checkpoint may use ubiquitin/proteasome-mediated protein degradation to ensure fast response of preventing a cell irradiated in G₁ from entering S phase with a full load of DNA damage, the G₁ checkpoint function is of no benefit to a cell irradiated during S phase or G₂-M phase. Additionally, cells entering into S phase with some unrepaired DNA lesions may activate S phase checkpoints to provide more time for DNA repair. Similarly, it is reasonable to speculate that G₂-M checkpoints will play a more pivotal role in keeping genomic integrity. In early studies, a transient inhibition of cell cycle progression after exposure to ionizing radiation was found in S and G₂-M phases (7). Also, Fournier and Taucher-Scholz (8) reported that more pronounced delays in S and G₂-M phase have been observed with increasing LET radiation.

Previous studies have suggested that defects in cell cycle checkpoints promote carcinogenesis (9). Cells with an intact DNA-damage response pathway frequently undergo arrest or die in reaction to DNA damage, thus reducing the likelihood of progression to malignancy. Aberrations in cell cycle checkpoint pathways, however, may permit the survival or continued growth of cells with abnormalities, thereby enhancing the chance of malignant transformation. Mutations in cell cycle control genes, such as p53, p21, and ATM have been directly linked to chromosomal aberrations and genomic instability (10-12). Amplification or overexpression of the CCND1 gene plays a pivotal role in the development of several human cancers including breast cancer, colon cancer, prostate cancer, and lung cancer (13-15). In addition, CCNE overexpression has also been frequently reported in colon and breast cancers (16, 17).

Our previous study (18) showed that G₂-M phase delay induced by γ-radiation in lymphoblastoid cell lines was more prominent in healthy controls than in patients with lung cancer. This finding was further supported by a pilot study using cultured lymphocytes (19), which suggested that lower percentages of γ-radiation–induced cells in the S and G₂ phases were associated with an increased risk of lung cancer. More recently, Zheng et al. (20) reported that a less-efficient G₂-M checkpoint was associated with an increased risk of lung cancer in African Americans but not in Caucasians. In the present study, we aimed to confirm the association of G₂-M and S cell cycle checkpoints and lung cancer risk. To our knowledge, this is by far the largest case-control study reporting the effect of cell cycle checkpoint functions on lung cancer risk.

The patients with lung cancer were consecutively recruited from The University of Texas M. D. Anderson Cancer Center in Houston, TX. All cases were histologically confirmed, and had received no prior chemotherapy or radiotherapy before enrollment. There were no age, sex, or stage restrictions on recruitment. This analysis was restricted to Caucasians due to small sample size in other ethnicity. Healthy controls without previous cancer history were recruited from the Kelsey-Seybold Clinics, the largest multispecialty physician group in the Houston metropolitan area. The controls were frequency-matched to the cases on the basis of age (±5 years), sex, ethnicity and smoking status. A total of 747 patients and 745 healthy controls were included in this analysis.

All participants were interviewed by trained M. D. Anderson Cancer Center staff interviewers to collect information regarding demographics, smoking history, alcohol consumption, family history of cancer, occupational exposures, and dietary patterns. After the interview, 40 mL of blood was drawn into coded heparinized tubes and delivered to the laboratory for analysis. Laboratory personnel were blinded to the case-control status of the samples. All relevant review boards approved this research, and signed informed consent was obtained from each individual.

The percentage of cultured lymphocytes in different cell cycle phases was determined by flow cytometry as previously described (18). In brief, two blood samples were cultured for each subject. In each culture, 1 mL of whole blood was mixed with 9 mL of RPMI 1640 (JRM Biosciences) supplemented with 20% FCS and 112.5 μg/mL of phytohemagglutinin (Wellcome Research Laboratories). After 67 h of incubation at 37°C and 5% CO₂, one cell culture from each subject was irradiated with a 137Cs source (Cesium Irradiator Mark 1, model 30; J.L. Shepherd and Associates). The optimal dose of γ-radiation was 2.5 Gy, as determined in a previous study (18). The cell cultures were then incubated for another 10 h before being harvested. The unirradiated cells were harvested at the same time. After harvesting, cells were centrifuged, washed twice with 2 mL of PBS, and fixed with ethanol. Immediately before analysis, cell suspensions were filtered through a 37 μm filter to remove debris. DNA was stained by mixing 1 mL of PBS containing 250 μg/mL of propidium iodide and 5 mg/mL of RNase with 0.2 mL of cell suspension. Samples were then immediately analyzed with a FACScan (Becton Dickinson Immunocytometry Systems), with a minimum of 2 × 10⁴ cells analyzed for each sample. Cell cycle distribution was determined using Lysis 2 software (Becton Dickinson Immunocytometry Systems). The γ-radiation–induced cell accumulation index was calculated to evaluate cell cycle checkpoint functions as previously described (20, 21). The cell accumulation index at G₂-M or S phase was defined as: (percentage of cells in G₂-M with radiation exposure − percentage of cells in G₂-M or S phase without radiation exposure) / (percentage of cells in G₂-M or S phase without radiation exposure). Cells both at G₂ and M phases contain twice the amount of DNA (i.e., a tetraploid chromosomal DNA content) and are inseparable by flow cytometry. Therefore, we used the G₂-M phase accumulation index as an indicator instead of G₂ delay as reported previously (22, 23). However, these two terms reflect the same measurements because the M phase generally has a very short time duration in the cell cycle compared with G₂ phase and a majority of cells with twice the amount of DNA are at G₂ phase.

All statistical analyses were done using the Stata 8.0 statistical software package (Stata Corporation). We used the Pearson's χ² test to examine the distribution of sex and smoking status between cases and controls. The Student's t test was used to analyze continuous variables with normal distribution, such as age and pack-years. Due to the nonnormal distribution of γ-radiation–induced cell accumulation index, we used the rank-sum test to investigate the difference(s) between cases and controls. Additionally, the cell accumulation index was also dichotomized based on the median value in controls or categorized into tertiles based on its distribution in controls. The association between the cell accumulation index at S phase or G₂-M phase and the risk of developing lung cancer was estimated by odds ratios (OR) along with their corresponding 95% confidence intervals (95% CI). To adjust for the confounding effects of age, sex, and smoking status, unconditional logistic regression analysis with multiple covariates was done. Age was used for adjustment as a continual variable. Smoking status was characterized as never, former, and current categories. A never smoker was defined as a person who had never smoked or smoked <100 cigarettes in his or her lifetime. A former smoker had stopped at least 1 year before the diagnosis of cancer (for cases) or 1 year before the interview (for controls). A current smoker was someone who continued smoking or who had stopped smoking <1 year prior to the diagnosis of cancer (cases) or prior to the interview (controls). The number of pack-years was calculated as the average number of cigarettes smoked per day divided by 20 cigarettes and then multiplied by smoking years.

In addition, risk assessment was done in the subgroups of lung cancer patients defined by histology and stage. Histology and clinical stage of lung cancer cases were obtained from the clinical records. The stage of non–small cell lung cancer (NSCLC) was categorized according to the 1997 revision of the International System for Staging Lung Cancer (24). For the purposes of analysis, early-stage NSCLCs (stages I and II) were combined. The small number of small cell lung cancer (SCLC) cases precluded meaningful analyses of SCLC by stage. The differences of the cell accumulation index at S phase or G₂-M phase among subgroups stratified by host characteristics were also investigated. All P values were based on two-sided tests. A probability level of 0.05 was used as the criterion for statistical significance.

The characteristics of the 747 patients with lung cancer and the 745 controls are summarized in Table 1. There were no significant differences between cases and controls according to age (P = 0.166) or sex (P = 0.178). By study design, cases and controls were also well-matched on smoking status (P = 0.300). However, the ever smokers among the patients with lung cancer reported a significantly higher level of tobacco consumption (40.26 ± 33.45 pack-years) than those in the controls (35.16 ± 33.21 pack-years; P = 0.005). In the patients with lung cancer, histologic data from 716 patients and stage data from 636 NSCLCs were available for statistical analysis. The tumors of those patients included 356 adenocarcinomas (49.7%), 174 squamous cell carcinomas (24.3%), 37 SCLCs (5.2%), and 149 others (20.8%). There were 211 NSCLCs at stage I and stage II (33.2%), 219 at stage III (36.7%), and 206 at stage IV (32.4%).

The baseline and γ-radiation–induced cell percentages at S and G₂-M phases in the cultured peripheral blood lymphocytes were measured by flow cytometry (Table 2). The cell percentages at the G₂-M phase were significantly lower in the cases than in the controls both at baseline (cases versus controls, 6.82 ± 4.75% versus 7.64 ± 3.55%; P < 0.001) and after radiation (cases versus controls, 12.42 ± 6.68% versus 14.26 ± 6.23%; P < 0.001). Similarly, the cases also exhibited significantly lower cell percentages at S phase than did the controls both at baseline (cases versus controls, 23.97 ± 12.44% versus 27.81 ± 10.51%; P < 0.001) and after radiation (cases versus controls, 25.59 ± 12.21% versus 31.24 ± 10.88%; P < 0.001). As additional indicators of cell cycle checkpoint functions, we reported γ-radiation–induced cell accumulation indices at S and G₂-M (Table 2). The cell accumulation index at G₂-M phase was significantly lower in the cases (median, 0.774; 95% range, −0.182 to 3.420) than in the controls (median, 0.882; 95% range, −0.053 to 3.175; P = 0.002). Similar results were also obtained for the indices at S phase (cases versus controls, 0.226 versus 0.243; P = 0.001). In addition, γ-radiation–induced cell arrest was higher in G₂-M than in S phase for both cases (0.774 versus 0.226, P < 0.001) and controls (0.882 versus 0.243, P < 0.001).

We next analyzed the association between the γ-radiation–induced cell accumulation indices at G₂-M phase and S phase and overall lung cancer risk (Tables 3 and 4). Using the median value of the cell accumulation index at G₂-M phase in the controls as the cutoff point, we found that a lower G₂-M phase cell accumulation index was associated with a 1.28-fold increased overall lung cancer risk (adjusted OR, 1.28; 95% CI, 1.04-1.58). Stratified analysis according to histology showed that the associations were statistically significant for the adenocarcinoma (adjusted OR, 1.28; 95% CI, 1.00-1.66) and squamous cell carcinoma (adjusted OR, 1.74; 95% CI, 1.21-2.51), but not for SCLC (adjusted OR, 0.92; 95% CI, 0.46-1.84). Additionally, the results of Pearson's χ² test also indicated that there was a significant difference of the cell accumulation index at G₂-M phase among the histologic subgroups (P = 0.023). Stratified analysis according to stage showed that the increased risk of lung cancer existed in all clinical stages. Also, no significant difference of the cell accumulation index at G₂-M phase among the three subgroups with different clinical stages was indicated by Pearson's χ² test (P = 0.811). Similar results were obtained for the cell accumulation index at the S phase. A 1.30-fold increased overall lung cancer risk was observed for the lower S phase index (adjusted OR, 1.30; 95% CI, 1.06-1.61). The increased risks were found to be statistically significant for the adenocarcinoma (adjusted OR, 1.40; 95% CI, 1.07-1.82) and squamous cell carcinoma (adjusted OR, 1.49; 95% CI, 1.04-2.14), but not for SCLC (adjusted OR, 1.07; 95% CI, 0.54-2.15). Obvious increased risks of lung cancer were displayed in all clinical stages and no significant difference of S phase index among the three subgroups of clinical stages was found by Pearson's χ² test (P = 0.111). We further categorized subjects into tertiles of cell accumulation index based on its distribution among controls. A significant reverse dose-response was observed between the cell accumulation index and lung cancer risk at both G₂-M (P for trend = 0.001) and S (P for trend = 0.001) phases. Using the highest tertile as a reference, the adjusted OR (95% CI) for individuals in the medium and lowest tertiles of the cell accumulation index was 0.87 (0.67-1.13) and 1.49 (1.17-1.91) at G₂-M phase and 1.00 (0.77-1.30) and 1.49 (1.16-1.92) at S phase, respectively (Tables 3 and 4).

In subgroup analyses (Table 5), the case-control difference for the γ-radiation–induced cell accumulation index at the G₂-M phase was statistically significant in women (cases versus controls, 0.767 versus 0.972; P = 0.001) but not in men (cases versus controls, 0.776 versus 0.824; P = 0.221), in persons who were older than 65 years (cases versus controls, 0.667 versus 0.872; P < 0.001) but not in persons who were between 58 and 65 or younger than 58 years (cases versus controls, 0.823 versus 0.921; P = 0.127 or 0.856 versus 0.852; P = 0.664, respectively). There was a significant trend for an increased case-control difference of γ-radiation–induced cell accumulation index at the G₂-M phase as age increased (P for trend < 0.001). The difference also existed in both ever smokers (cases versus controls, 0.773 versus 0.874; P = 0.011) and never smokers (cases versus controls, 0.775 versus 0.891; P = 0.060). For the cell accumulation index at S phase, the case-control differences were statistically significant in persons who were younger than 58 years (cases versus controls, 0.105 versus 0.189; P = 0.013) but not in persons who were between 58 and 65 or older than 65 years (cases versus controls, 0.103 versus 0.119; P = 0.253 or 0.120 versus 0.165; P = 0.210, respectively). There was a significant reverse trend between case-control difference and age: as the age increased, the case-control difference of γ-radiation–induced cell accumulation index at the S phase decreased (P for trend < 0.001). This difference also existed in ever smokers (cases versus controls, 0.105 versus 0.162; P < 0.001) but not in never smokers (cases versus controls, 0.069 versus 0.109; P = 0.818), and in both men (cases versus control, 0.092 versus 0.142; P = 0.025) and women (cases versus controls, 0.113 versus 0.167; P = 0.018).

In this large case-control study, we applied an in vitro assay to evaluate the differences in cell cycle checkpoint functions between patients with lung cancer and healthy controls. We found that cell accumulation was clearly induced by γ-radiation in G₂-M and S phases for both cases and controls, which is consistent with previous observations, indicating that cell cycle arrest is an ionizing radiation–inducible function (6, 25-27). Furthermore, our data showed that the cell accumulation indices at both the G₂-M and S phases of the cell cycle were significantly lower in patients with lung cancer than in controls, suggesting that attenuated G₂-M and S cell cycle checkpoints may be risk factors for lung cancer.

Our findings showed that the lower G₂-M phase cell accumulation index was significantly associated with a 1.28-fold increased lung cancer risk, suggesting a less-efficient G₂-M checkpoint in patients with lung cancer. In a recent study, Zheng et al. showed that a less-efficient G₂-M checkpoint was associated with an increased risk of lung cancer in African Americans (20). However, they found no significant association between the G₂-M checkpoint and lung cancer risk in Caucasians. This discrepancy may be due to the relatively small sample size (154 cases and 194 controls) or the different approach for assessing the G₂-M checkpoint functions in their study. Scott et al. (28) reported that the G2 arrest of irradiated cells derived from breast cancer patients was less than that of irradiated cells from healthy controls. Olivieri and Micheli (29) also showed that cells with deficient radiation-induced G₂ arrest displayed a high level of induced chromosome aberrations. In addition, it has been shown that individuals with inherited mutations in genes involved in the G₂-M checkpoint pathways such as ATR, ATM, NBS1, BRAC1, and 14-3-3ε were more susceptible to the development of various cancers (30-34). All these lines of evidence support the hypothesis that attenuated function of the G₂-M checkpoint may result in genomic instability, and therefore, an elevated cancer risk. Our data also shows a 1.30-fold increased risk of lung cancer associated with S phase cell accumulation, indicating that S phase checkpoint activation might be impaired in cases of lung cancer following ionizing radiation exposure. Previous studies provide strong support for this finding. For example, deficient ionizing radiation–induced S phase checkpoint arrest has been observed in patients with cancer-prone disorders such as ataxia-telangiectasia syndrome and Nijmegen breakage syndrome (35, 36). These findings clearly suggest that a less-efficient S checkpoint predisposes to lung cancer development.

In stratified analysis according to histology and clinical stage, our data showed that the lower cell accumulation indices at both G₂-M and S phases were significantly associated with increased risk for adenocarcinoma and squamous cell carcinoma, but not for SCLCs. These findings suggest that the effects of reduced G₂-M and S checkpoints on lung cancer development might be histology-specific. However, we must be cautious in drawing our conclusions. On the one hand, the small sample size for the SCLC subgroup (37 subjects) decreased the statistical power so that meaningful results might not be assured. On the other hand, we did not include the other groups due to high heterogeneity of histology, which might impair the accuracy of stratified analysis in this subgroup. Additionally, no significant association was found between the G₂-M or S phase cell accumulation indices and clinical stage of patients with lung cancer, suggesting that reduced G₂-M and S checkpoints are more likely to be the cause of disease, rather than the effect of disease.

In our study, stratified analysis was also done according to host characteristics. Our findings showed that the case-control difference of G₂-M phase cell accumulation index was significant only in female individuals, suggesting that women might be more liable to suffer from less-efficient G₂-M cell cycle checkpoints. This finding is consistent with the recent report by Zheng et al. (20), in which women exhibited lower γ-radiation–induced G₂-M arrest in lung cancer cases than in healthy controls, but men did not. However, the underlying mechanism is still under investigation. Our results also showed that the case-control difference of G₂-M phase cell accumulation index was evident only in older individuals (≥65 years), suggesting that age might interact with other risk factors to affect G₂-M cell cycle checkpoint function. Furthermore, smoking status was not found to be associated with γ-radiation–induced G₂-M phase cell accumulation index, which is also in line with the report by Zheng et al. (20). In comparison, the case-control differences for the S phase cell accumulation index were significant only in younger individuals (<58 years old), suggesting that the deficiency of S phase checkpoint might be associated with the development of early onset lung cancer. Our data also showed that the significant case-control difference for the S phase cell accumulation index only existed in ever smokers, indicating a possible role of the gene-environment interaction. The observations in our study could, if true, point towards the difference in the effects of some host characteristics on G₂-M phase and S phase cell cycle checkpoints. The difference is possibly due to different signal transduction pathways. We also could not rule out the possibility of chance findings due to limited size. Further investigations are needed to confirm our findings.

In our study, we found that the percentages of G₂-M and S phases at the baseline were significantly lower in cases than in controls. A possible explanation is that cases may have a spontaneous deficiency of G₂-M and S phase cell cycle checkpoints. In addition, our data showed that cell accumulation of G₀/G₁ phase was higher in cases than in controls (data not shown). However, we could not separate cells at G₁ phase from those at G₀ phase, due to the same DNA amount in cells at G₀ and G₁ phase. Therefore, the G₁ checkpoint function could not accurately be evaluated in our present study.

Our study has several strengths. First, the sample size was the largest to date for studying the association of cell cycle checkpoints with lung cancer risk in Caucasians. Second, covariables (age, sex, and smoking status) were well matched between cases and controls. Therefore, these confounding effects were, to a large extent, eliminated. Moreover, because all samples were collected before any cancer therapy, the assessment of cell cycle checkpoints was not confounded by treatment. Our study also has some potential limitations. Whether the cell cycle checkpoints measured in lymphocytes are representative of those of lung epithelial cells needs to be determined. Also, the cause-effect relationship between the function of cell cycle checkpoints and lung cancer development could not be shown by this retrospective analysis. Further confirmation using a prospective study design is warranted.

In conclusion, this is the largest epidemiologic study to date that addresses the roles of cell cycle checkpoints in lung cancer development. Our data suggest that attenuated γ-radiation–induced G₂-M and S checkpoints are associated with increased lung cancer risks among Caucasians.