γ-Radiation-induced G₂ Delay, Apoptosis, and p53 Response as Potential Susceptibility Markers For Lung Cancer

Upon exposure to carcinogens, such as γ-radiation, a complex network of surveillance mechanisms or checkpoints interrupts cell cycle progression when damage to the genome is detected or when cells fail to complete DNA replication, thereby ensuring genetic integrity (1, 2). For example, most mammalian cells exhibit transient delays in the G₁ and G₂ phases of the cell cycle after exposure to γ-radiation (3, 4, 5). Cell cycle arrest prevents DNA replication and mitosis in the presence of unrepaired chromosome alteration and provides time for DNA repair. Disruption of the DNA damage checkpoint is strongly related to cancer progression (1, 6).

In cases where the damage is severe and cannot be repaired, the cells will go into apoptosis or programmed cell death. Apoptosis is a safe way for the organism to eliminate severely damaged cells rather than risk acquiring neoplastic autonomy (7). Defects in apoptotic pathways are now believed to contribute to a number of human diseases ranging from neurodegenerative disorders to malignancy (8). Studies in yeast and humans found that the DNA damage checkpoint was involved in the DNA repair process (9, 10), and that DNA damage-induced apoptosis is related to DNA damage checkpoint and DNA repair (11).

The tumor suppressor gene p53 functions as a guardian of genomic integrity and coordinates cellular responses to DNA damage by inducing either cell cycle arrest or programmed cell death (12, 13, 14, 15, 16). For example, after DNA damage, cells with disruption of the p53 or the p21 gene progress into mitosis before DNA repair is completed (12, 13, 14). Marked decreases in apoptosis after exposure to radiation correlate with the occurrence of p53 mutations in some transgenic mice (17).

We hypothesized that those individuals with inherited defects in cell cycle control, apoptosis, p53 function, or DNA repair might be susceptible to lung cancer development. The cells from susceptible individuals would have shorter cell cycle arrest, less apoptosis, suboptimal p53 function, and more DNA damage upon exposure to γ-radiation compared with cells from normal individuals. To test this hypothesis, we compared the levels of γ-radiation-induced cell cycle delay, apoptotic index, p53 protein expression levels, and chromatid breaks in lymphoblastoid cell lines from lung cancer patients with those from healthy controls. We also correlated these measurements to explore their potential relationships.

For our preliminary experiments, we used one lymphoblastoid AT³ cell line (GM01526c; Coriell Institute of Medical Research, Camden, NJ). The two other cell lines that were used (3590P, from a healthy blood donor; 3640P, from a patient with multiple primary cancer) were transformed with EBV in Dr. Hsu’s laboratory. The AT and 3640P cell lines are sensitive to γ-radiation, whereas the 3590P cell line is resistant to γ-radiation (18).

For our primary experiments, we used lymphoblastoid cell lines, which were obtained from 30 lung cancer patients and from 22 healthy controls and then transformed by the EBV. Information on age was available for all of the patients and 12 healthy controls. The mean age was 58 years for the patients and 36 years for the controls.

Cells in the logarithmic phase of growth were plated 24 h before irradiation at a concentration of 1×10⁵/ml in 25-cm² flasks to achieve optimal cell cycling. Standardization experiments were carried out at relative γ-radiation doses using a ¹³⁷Cs source at room temperature. Unirradiated samples were also harvested at each time point. Two sets of experiments, each repeated three times, were conducted. First, γ-radiation treatment was assessed by harvesting cells after 1, 3, 5, 8, 10, 16, 20, 24, 28, and 48 h of γ-radiation treatment to determine the optimal duration of treatment with various doses. Second, γ-radiation dose response was measured in cultures treated with γ-radiation doses of 0, 1, 2.5, 5.0, and 7.5 Gy. As the optimal time points, we selected 10 h after exposure to γ-radiation for cell cycle analysis and 48 h after exposure to γ-radiation for the apoptosis assay because these time points provided the best results in distinguishing between the 3590 and 3640 cell lines. Next, we chose 2.5 Gy as the optimal dose because it provides a clear discrimination between the 3590 and 3640 cell lines, and most of the cells were still in good condition. Thirty cell lines from lung cancer patients and 22 cell lines from healthy controls were treated with γ-radiation (2.5 Gy) after 10 h and 48 h, and then harvested.

After harvesting, cells were first centrifuged, then washed twice with 2 ml of PBS and, finally, fixed with ethanol. The samples were analyzed with a FACScan (Becton Dickinson Immunocytometry Systems, San Jose, CA). Immediately before analysis, cell suspensions were filtered through a 37-μm filter to remove debris. DNA staining was performed by adding 1 ml of a solution of propidium iodide in PBS (propidium iodide solution, 250 μg/ml; RNase, 5 mg/ml) to 0.2 ml of cell suspension immediately before analysis. A minimum of 2 × 10⁴ cells were analyzed for each sample. Cell cycle-phase distributions were determined using Lysis 2 software (Becton Dickinson Immunocytometry Systems). The differences in G₂ cell percentages between treated cells and untreated cells were recorded as the G₂ cell cycle delay percentage.

Two methods were used to detect γ-radiation-induced apoptosis: the ethanol fixation method and the TUNEL method, which is the percentage of TUNEL signal-positive cells. Cell preparation by the ethanol fixation-based method was similar to the method used in the cell cycle analysis. The difference in percentage of apoptotic cells between treated and untreated cells was recorded as the apoptotic index. Apoptosis-mediated, endonuclease-induced DNA strand breaks were detected by the addition of fluorescence-labeled dUTPs by the TUNEL assay (the in situ cell death detection kit TUNEL assay; Boehringer Mannheim, Mannheim, Germany). On designated days, suspended cells in the irradiated and control flasks were fixed with 4% paraformaldehyde for 1 h, rinsed with PBS, and then permeabilized with Triton X-100. Cells were rinsed twice with PBS, then 50 μl of TUNEL reaction mixtures were added, and finally the cells were analyzed by the flow cytometry method.

p53 protein level was measured by the ELISA method using the Pantropic p53 Rapid Format ELISA kit (Oncogene, San Diego, CA) and quantified by micro-BCA method (Pierce, Rockford, IL). First, cell pellets from treated cells and untreated controls were lysed. Next, samples and the detector antibody were added to wells and incubated for 4 h at 37°C in a humidified incubator. After incubation, the wells were washed three times, then streptavidin conjugate was added into each well, and finally the wells were incubated for 30 min at room temperature. After incubation, the wells were washed three times and the substrate solution was added to each well, and then the wells were incubated in the dark at room temperature for 30 min. Next, stop solution was added to each well, and the absorbance in each well was measured by using Vmas Microplate Reader (Molecular Devices, Sunnyvale, CA) at dual wavelengths of 450/595 nm. This ELISA was repeated three times for each sample.

Cells in the logarithmic phase of growth were plated 24 h before irradiation at a concentration of 5 ×10⁵/ml in 25-cm² flasks to achieve optimal cell cycling. Standardization experiments were carried out at 1.25-Gy of γ-radiation dose using a ¹³⁷Cs source at room temperature. After γ-radiation of the treated samples, both the treated and untreated samples were harvested for 5 h. Cell harvesting followed the standard procedure (hypotonic KCl pretreatment, fixation in Carnoy’s solution, washing, and air-drying). The slides were subjected to Giemsa staining without banding. The number of breaks in 50 metaphases/sample was counted and expressed as the average number of breaks/cell.

The results from healthy controls and lung cancer patients’ cell lines were analyzed to determine a possible relationship between inherited defects and lung cancer risk. These comparisons were made by a standard two-sample Student t test (two-tailed) at a significant level of 0.05. The Pearson correlation was used to calculate the relationship among these markers.

Fig. 1 a summarizes the mean percentage of cells in G₂ phase after 2.5 Gy of γ-radiation at various time points in lymphoblastoid cells from a healthy donor (3590), a multiple primary cancer patient (3640), and an AT patient. Each value was the average of three separate experiments. The optimal treatment duration for estimating differential γ-radiation sensitivity in the different cell lines was 10 h for G₂ delay. The mean value was 33.3% for 3590 cells, 16.4% for 3640 cells, and 24.2% for AT cells at 10 h.

Fig. 1 b summarizes the mean percentage of apoptotic index after 2.5 Gy of γ-radiation at various time points in lymphoblastoid cells from the three cell lines mentioned above. Each value was the average of three separate experiments. The optimal treatment duration for estimating differential γ-radiation sensitivity in the different cell lines was 48 h for apoptosis. The mean value at 48 h was 31.2% for 3590 cells, 18.5% for 3640 cells, and 12.7% for AT cells.

Fig. 2,a shows the dose-response effect of γ-radiation at the 10-h time point for G₂ delay. When the γ-radiation dose was increased from 1 to 7.5 Gy, the percentage of cells in G₂ phase increased proportionally in all three cell lines. At 2.5 Gy, the mean value was 37.2% for 3590 cells, 21.6% for 3640 cells, and 18.4% for AT cells. Fig. 2 b shows the dose-response effect of γ-radiation at the 48-h time point for apoptosis. At 2.5 Gy, the mean value was 47.2% for 3590 cells, 21.5% for 3640 cells, and 18.7% for AT cells.

The time point of 10 h after γ-radiation and the γ-radiation dose of 2.5 Gy were chosen as optimal conditions to test their effect on G₂ delay. γ-Radiation-induced G₂-phase accumulation was performed on lymphocyte cell lines from 30 lung cancer patients and 22 healthy controls. The results demonstrate that the mean γ-radiation-induced G₂-phase accumulation in healthy cell lines was 22.5% ± 10.5%, significantly higher than the value for cases whose mean was 14.7% ± 8.8% (Fig. 3). The mean difference between the two groups was 7.8%, which was statistically significant (P = 0.0076).

Ethanol fixation and TUNEL methods were used to measure the γ-radiation-induced apoptosis. The time point of 48 h after γ-radiation and the γ-radiation dose of 2.5 Gy were chosen as the optimal conditions. The γ-radiation-induced apoptosis was performed on lymphocyte cell lines from 30 lung cancer patients and 22 healthy controls. Using the ethanol fixation method, the mean apoptotic index was 20.4% ± 11.7% in control cell lines compared with 14.3% ± 7.8% in case cell lines (Fig. 4,a). The difference between controls and cases was 6.1% and this difference was statistically significant (P = 0.042). Using the TUNEL assay, the mean apoptotic index for the controls was 9.0% ± 7.4% compared with 4.8% ± 7.8% for the cases (Fig. 4 b). The mean difference of 4.2% was statistically significant (P = 0.040). Both of these assays yielded similar results.

The time point of 10 h after γ-radiation and the γ-radiation dose of 2.5 Gy were chosen as optimal conditions for assaying p53 protein expression in lymphocyte cell lines from 30 lung cancer patients and 22 healthy controls (Fig. 5). The mean p53 protein expression at baseline was higher in cases than controls (0.40 ± 0.29 μg/ml versus 0.20 ± 0.063 μg/ml; P = 0.011). After treatment, the mean was 0.72 ± 0.48 μg/ml for cases and 0.55 ± 0.42 μg/ml for controls, however, the results were not statistically significant (P = 0.24). The ratio of treated:untreated was significantly higher in controls than in cases (2.86 ± 1.87 and 1.99 ± 0.75, respectively; P = 0.03).

The time point of 5 h after γ-radiation and the γ-radiation dose of 1.25 Gy were chosen as optimal conditions for the mutagen sensitivity assay. The chromatid breaks were measured in lymphocyte cell lines from 17 lung cancer patients and 17 healthy controls (Table 1). The results demonstrate that the mean chromatid breaks in the control cell lines were significantly lower than those of the cases (0.26 ± 0.06 versus 0.38 ± 0.10; P < 0.001). The difference between the two means was statistically significant (mean difference, 0.12 ± 0.03; P < 0.001).

To investigate further the relationship between these susceptibility biomarkers, we performed correlation analysis (Table 2). An increasing p53 ratio was positively correlated with cell cycle G₂ delay (γ = 0.413; P = 0.002) and negatively correlated with chromatid breaks (γ = 0.384; P = 0.028). We did not find any other statistically significant correlation between other biomarkers.

The contribution of genetic instability to cancer development is exemplified by a set of rare chromosome instability syndromes such as Bloom syndrome, Fanconi anemia, AT, Werner syndrome, xeroderma pigmentosum, and Li-Fraumeni familial cancer syndrome. These patients are characterized by in vivo and in vitro chromosomal breakage, defective DNA repair systems, abnormal cell cycle control checkpoints, and increased cancer risk (19, 20, 21, 22, 23, 24). Our results clearly demonstrated a difference in γ-radiation-induced chromatid breaks in lymphoblastoid cell lines between lung cancer patients and healthy controls. The control cell lines showed fewer chromatid breaks than the case cell lines, indicating that patients with lung cancer have accumulated more DNA damage than healthy individuals. This suggests that DNA repair capacity might be a risk factor for lung cancer development.

Increased γ-radiation-induced DNA damage may be caused by the disruption of genes controlling cell cycle checkpoints (25). These checkpoints play an essential part in DNA repair. By delaying cell cycle progression, the checkpoints provide more time for DNA repair before entering the phases of DNA replication and mitosis. Loss or attenuation of these checkpoint functions may increase spontaneous and/or induced gene mutations and chromosomal aberrations by reducing the efficiency of DNA repair. Mutations in checkpoint control genes may, therefore, contribute to the genetic instability that seems to drive neoplastic evolution. There are several cell cycle checkpoints, namely G₁, S, and G₂ checkpoints. G₁ and S checkpoints protect the cell from DNA replication errors. When damaged DNA is detected, the G₁ checkpoint delays the progression of G₁ cells into the S phase, which causes DNA replication inhibition. The G₂ checkpoint keeps the DNA damaged cells in the G₂ phase, and this, in turn, protects the cell from mitotic errors until the damage has been repaired. The AT gene also contributes to checkpoint control (26). Low doses of radiation inhibit 50% of DNA synthesis in normal cells but have little effect on DNA synthesis in AT cells (27). Cells in the G₂ phase are extremely sensitive to ionizing radiation-induced DNA damage, and the ensuing G₂ delay after irradiation has been recognized for many years (28). We found that γ-radiation-induced G₂ delay has a dose-response relationship. The G₂ delay was more evident in cell lines from healthy controls (3590) than in cell lines from cancer patients with multiple primaries (3640) and in AT patients. Furthermore, G₂ delay was significantly higher in cell lines from controls than in those of lung cancer patients. Olivieri and Micheli (27) showed that G₂ cells failing to delay mitosis after irradiation displayed the highest level of induced chromosome aberrations. The high levels of chromosomal aberrations in immortal fibroblasts from patients with the Li-Fraumeni familial cancer syndrome further suggest that these cells have lost control of the mechanism that prevents cells from entering mitosis with damaged chromosomes (29). In addition, defects in cell cycle G₂ delay are strongly related to human carcinogenesis (30). Therefore, individuals with inherited defects in the G₂ checkpoint may be predisposed to cancer development. Contrary to our findings, Lavin et al. (31) reported that breast cancer patients have a longer G₂ delay than healthy controls.

When the damage is severe and cannot be repaired, DNA damage can trigger apoptosis to eliminate the damaged cells. Studies in transgenic and knockout mice provide direct evidence that disruption of apoptosis can promote tumor development (17, 32). We found that the control cell lines showed higher apoptotic indices than the case cell lines when using both the ethanol fixation and TUNEL methods; these methods generated similar findings. Therefore, disruption of apoptosis may also contribute to lung tumor development.

We also noted that the extent of the cell cycle delay and apoptosis depend on DNA damage and how well the cells respond to the damage. When the cell lines from multiple primary and AT patients, and controls were exposed to γ-radiation, cell cycle delay occurred in all three cell lines. At early time points (<8 h), cell lines from controls demonstrated a lower percentage of G₂ delay compared with the cell lines from the patients. Conversely, at later time points (≥8 h), the control cell lines showed a significantly higher percentage of G₂ delay than both multiple primary and AT cell lines. We believe that at early time points, DNA damage triggers cell cycle arrest to repair the DNA and, usually, most of the DNA damage will be repaired quickly. However, if damaged DNA still exists, then the cell will continue to remain at cell cycle arrest. In the case of multiple primary and AT cell lines, although they exhibited more DNA damage than healthy controls, they may not be able to sense the damaged DNA to initiate cell cycle arrest. At time points earlier than 16 h, cell lines from multiple primary and AT patients had a slightly higher apoptotic index than those of the controls. Conversely, after 16 h, the multiple primary and AT cell lines had a significantly lower apoptotic index than the control cell lines. A possible explanation for these results may be attributed to the inability of multiple primary and AT patient cell lines to detect existing damaged DNA. We established that 10 h for cell cycle G₂ delay and 48 h for apoptosis after exposure to γ-radiation were the optimal time points to differentiate cell lines from multiple primaries, AT patients, and controls. Because apoptosis is a gene-directed program to eliminate cells with irreparable DNA damage, it is a late event. As a result, late time points are required to observe the difference in apoptosis among the cell lines.

The p53 tumor suppressor gene is a guardian of genomic integrity. Activation of p53 can occur in response to a number of cellular stresses, including DNA damage, hypoxia, and nucleotide deprivation. In response to γ-radiation exposure, p53 may respond in the following manner: (a) temporarily halt the cycle of cell division to allow time for DNA repair before DNA replication; (b) trigger severely damaged cells to self-destruct; or (c) directly and indirectly stimulate the DNA repair machinery (33). Under normal conditions, p53 proteins are maintained at low levels because of their extremely short half-life. Moreover, these proteins usually exist in a largely inactive state and are relatively inefficient at binding to DNA to activate transcription (34). We found that p53 protein levels at baseline without treatment with γ-radiation were significantly higher in lung cancer case cell lines than in healthy control cell lines. We believe that cells from lung cancer patients are genetically more unstable than cells from controls, resulting in a greater accumulation of p53 in the cells. In response to γ-radiation, the level of p53 proteins rises dramatically and is stabilized by posttranslational modification (35). We found that after exposure to γ-radiation, both case and control cell lines responded to DNA damage by increasing the level of p53 proteins, however, the response of the case cell lines was less efficient than the control cell lines. The ratio of p53 protein response was statistically significant between the case and the control cell lines. Using the p53 response ratio is more informative than simply relying on p53 levels in case cell lines to assess the level of DNA damage because of interindividual differences in p53 response.

The activation of p53 leads to the activation of a number of genes whose products trigger cell cycle arrest, apoptosis, or DNA repair (36). Consequently, the differences in the p53 response ratio might partly contribute to the differences in cell cycle G₂ delay, apoptosis, and DNA damage seen in the cases and controls. Indeed, in the correlation analyses, we found a positive correlation between p53 response ratio and cell cycle G₂ delay and a negative correlation between p53 response ratio and chromatid breaks. These results suggest that p53 plays an important role in cell cycle G₂ delay and DNA repair after exposure to γ-radiation. The correlation between apoptosis and p53 response ratio is not statistically significant at the 5% level. However, a weak positive correlation was detected between apoptosis and p53 response in both ethanol and TUNEL assays (γ = 0.232; P = 0.098 and γ = 0.238; P = 0.089, respectively). A similar relationship was observed between cell cycle G₂ delay and chromatid breaks (γ = −0.325; P = 0.061). As expected, there was no correlation between apoptosis and cell cycle G₂ delay. In addition, we tested the correlation between age and apoptosis as well as G₂ delay. We found no correlation between age and apoptosis in either ethanol or TUNEL assays (γ = −0.387; P = 0.214; γ = 0.000; P = 0.999 for controls and γ = −0.003; P = 0.989; γ = −0.033; P = 0.863 for cases, respectively). We also found no correlation between age and G₂ delay (γ = −0.128; P = 0.692 for controls and γ = 0.235; P = 0.212 for cases; data not shown).

In summary, our results clearly demonstrated the difference in γ-radiation-induced G₂ delay, apoptosis, and chromatid breaks between lymphoblastoid cell lines from lung cancer patients and healthy controls. The control cell lines showed greater G₂ delay, apoptotic index and p53 response ratio and fewer chromatid breaks than the case cell lines. Our data suggest that the apoptosis assays, cell cycle delay and p53 response ratio might be potentially useful in identifying lung cancer risk subgroups. This is the first study to address the difference of inherited cell cycle delay, apoptosis, and p53 protein level between lung cancer patients and normal controls. These biomarkers and assays may be validated and applied in future, larger epidemiological studies.