|Home | About | Journals | Submit | Contact Us | Français|
The double-strand break (DSB) repair capacity has been implicated in the survival of patients in several cancer types. However, little is known about the prognostic importance of the key DSB repair genes—ataxia-telangiectasia mutated (ATM), DNA-dependent protein kinase catalytic subunit (DNA-PKcs), and the Ku heterodimeric regulatory complex 86-kD subunit (Ku80)—in nonsmall cell lung cancer (NSCLC). To address this issue, the authors determined the messenger RNA (mRNA) expression of these genes in patients NSCLC and assessed their prognostic relevance.
mRNA expression levels of ATM, DNA-PKcs, and Ku80 were measured in tumor and adjacent normal tissues from 140 patients with NSCLC by using quantitative real-time polymerase chain reaction analysis. Then, a Cox proportional hazards regression model and Kaplan-Meier plots were used to evaluate the association between the tumor:normal (T/N) expression ratios of the 3 genes and the overall survival rate and duration in patients with NSCLC.
mRNA expression of ATM and DNA-PKcs, but not of Ku80, was significantly higher in tumor tissues than in adjacent normal tissues (P = .003 and P < .001, respectively). The high T/N expression ratios of ATM and DNA-PKcs were associated significantly with a 1.82-fold increased risk of death (95% confidence interval, 1.05–2.70) and a 2.13-fold increased risk of death (95% confidence interval, 1.21–3.76), respectively. However, no significant association with risk was observed for Ku80. Kaplan-Meier analyses revealed that patients with high T/N expression ratios of ATM or DNA-PKcs had notably shorter median survival than patients with low ratios.
The current findings suggested that the T/N expression ratios of ATM and DNA-PKcs may be useful for identifying NSCLC patients with a poor prognosis who may benefit from more aggressive therapy.
Lung cancer is the leading cause of cancer death in both men and women in the United States with an estimated 213,380 new cancer cases and 160,390 deaths in 2007.1 The current 5-year survival rates are <15%, a rate that has improved little over the past 2 decades.2 Recently, efforts have been made to identify new prognostic markers for lung cancer on the basis of biologic characteristics, such as gene expression profiles and specific molecular abnormalities. These markers may allow clinicians to identify appropriate candidates for specific therapies, thus improving the outcome of patients with lung cancer.
DNA double-strand breaks (DSBs) are regarded as the most lethal of all DNA lesions.3 Prompt and efficient repair of DNA DSBs is critical for maintaining genomic integrity. Defects in the DSB repair pathway may cause genetic alterations, chromosome instability, and, ultimately, malignant transformation.4,5 Conversely, tumor cells with increased DNA DSB repair capacity would be more likely to survive and proliferate, leading to the poor prognosis in cancer patients. Thus, the messenger RNA (mRNA) expression levels of key genes in the DSB repair pathway may be used as prognostic markers for cancer patients.
The ataxia-telangiectasia mutated (ATM) protein has been identified as the principal activator and master controller of cellular response to DSBs. ATM phosphorylates the key genes of the DNA damage response network and, thus, initiates cell cycle arrest, apoptosis, and DNA repair.6,7 Individuals with inherited mutations or polymorphisms in the ATM gene (ATM) are more susceptible to various cancers, including lung cancer, breast cancer, leukemia, and lymphoma.8-11 In addition, altered ATM expression levels have been identified in colon and pancreatic tumor tissues.12,13 Accordingly, the prognostic value of ATM expression in colorectal and breast cancer has been investigated.12,14 However, to our knowledge, its significance in lung cancer has not been studied.
The nonhomologous end-joining (NHEJ) pathway is the dominant mechanism for the repair of DSBs in mammalian cells. A key element of the NHEJ pathway is the DNA-dependent protein kinase (DNA-PK), which consists of a 465-kD catalytic subunit, DNA-PKcs, and a heterodimeric regulatory complex, Ku, which includes a 70-kD subunit (Ku70) and a 86-kD subunit (Ku80).15 It has been reported that cells lacking DNA-PKcs or Ku80 have defective DNA DSB repair.16,17 Recent reports10,18 suggest that a positive relation exists between Ku and cancer development. Higher levels of Ku or DNA-PKcs expression have been identified in bladder,19 colon,20 nasopharyngeal,21 and cervical22 tumor tissues compared with corresponding normal tissues. The expression patterns of Ku and DNA-PKcs also reportedly have been associated with tumor radiosensitivity and survival in patients with rectal,23 nasopharynx,21 and cervical22 carcinomas. Nonetheless, to our knowledge, no similar analyses of Ku80 and DNA-PKcs expression have been conducted to date in patients with nonsmall cell lung cancer (NSCLC).
In this study, we determined the mRNA expression levels of the ATM, DNA-PKcs, and Ku80 genes in tissue samples from 140 patients with NSCLC and compared those levels with patients’ overall survival durations and rates to determine their prognostic value. To our knowledge, this is the first study of the association between ATM, DNA-PKcs, and Ku80 mRNA expression and survival in patients with NSCLC.
In total, 140 patients were recruited for this study, all of whom had histologically confirmed NSCLC and underwent curative surgery at the University of Texas M. D. Anderson Cancer Center in Houston, Texas (M. D. Anderson) between 1993 and 1997. There were no age, sex, ethnicity, or tumor stage restrictions on patient enrollment. All patients gave written informed consent, and the study protocol was approved by the M. D. Anderson Institutional Review Board. Patients’ demographic variables and smoking history were obtained by chart review. Clinical data were collected from the computerized tumor registry at M. D. Anderson. Patients were followed until June 2005. Tissue samples (tumor and adjacent normal tissues) were obtained and snap-frozen in liquid nitrogen immediately after excision and stored at −80°C until further use. The World Health Organization classification system was used for histopathologic typing and grading of tumors, and the 1997 revisions of the International System for Staging Lung Cancer guidelines were used for staging.24
Total RNA was extracted from frozen tissues by using the EZNA RNA kit (Omega Bio-tek, Doraville, Ga) in accordance with the manufacturer’s protocol. The RNA concentration was measured by using a spectro-photometer (Beckman Coulter, Fullerton, Calif). The quality of RNA samples was determined by 1% agarose gel electrophoresis and ethidium bromide staining. Then, the complementary DNA (cDNA) was synthesized by using a reverse transcription kit (Applied Biosystems, Branchburg, NJ) in a final volume of 20 μL containing 0.5 μg of total RNA, 5 mM MgCl2, 250 μM of each deoxyribonucleotide triphosphate (dNTP), 20 U of RNase inhibitor, 50 U of multiscribe reverse transcriptase (RT), 2.5 μM random hexamers, and 1 × RT buffer. The reaction was carried out under the following thermal cycler conditions: 10 minutes at 25°C, 30 minutes at 42°C, and 5 minutes at 99°C. All cDNA products were stored at −30°C until they were used in the real-time polymerase chain reaction (PCR) analysis.
The primers and probes that were used for real-time PCR were designed by using Primer Express software (version 2.0; Applied Biosystems). We searched the Genebank database to confirm the specificity of primers and probes and the absence of single nucleotide polymorphisms. To avoid amplifying contaminated genomic DNA, 1 of the 2 primers was designed to span 2 exons. The sequences of all primers and probes are listed in Table 1. Real-time PCR was performed with a 384-well optic plate on the ABI Prism 7900 sequence detection system (Applied Biosystems). The PCR was carried out in a total volume of 10 μL, which consisted of 1 × Taqman buffer A, 3.4 mM MgCl2, 100 μM of each dNTP, 0.2 μM of each primer, 0.1 μM probe, and 0.02 U of AmpliTaq Gold DNA polymerase. The following reaction conditions were used: 10 minutes at 95°C to activate the Ampli-Taq Gold enzyme, then 15 seconds at 95°C and 1 minute at 60°C for 40 cycles. For each tested gene, the standard curve was generated by using commercial human total RNA (Stratagene, La Jolla, Calif) to quantify each transcript in unknown samples. Three housekeeping genes, glyceraldehyde-3-phosphate dehydrogenase (GAPDH), 18S ribosomal RNA (18S rRNA), and β-actin, were used as endogenous controls to normalize the RNA input amount as well as the reverse transcription efficiency and RNA quality on the basis of the geometric mean values of their mRNA expression. The PCR reaction for each sample was performed in duplicate to determine the consistency of the results.
At the end of the study, 2 patients had recurrences only, 37 patients had metastasis only, and 4 patients had both recurrences and metastasis. Because the assessment of time to recurrence or metastasis after surgery in patients with NSCLC is less reliable and less precise than that of overall survival, we only evaluated overall survival as an endpoint in this study. The overall survival duration was defined as the time from lung cancer diagnosis to patient death or last follow-up. Smoking status was categorized as 1 of the following: never-smoker, an individual who had smoked fewer than 100 cigarettes in his or her lifetime; former smoker, an individual who had stopped smoking ≥1 year before being diagnosed with cancer; and current smoker, an individual who continued smoking or who had stopped smoking <1 year before being diagnosed with cancer. Pack-years were calculated as the mean number of cigarettes smoked per day divided by 20 and then multiplied by years of smoking. We evaluated the distribution of patient characteristics by survival status (alive or dead) using the Pearson chi-square test or the Fisher exact test for categorical variables and the Student t test for continuous variables. The difference in mRNA expression level between tumor and adjacent normal tissues was compared by using the Wilcoxon rank-sum test. The Spearman correlation coefficient test was used to assess the correlation between the mRNA expression of ATM and DNA-PKcs, ATM and Ku80, and Ku80 and DNA-PKcs in both tumor tissues and normal tissues. The tumor:normal (T/N) expression ratios of all 3 genes were dichotomized as high or low, and the median T/N expression values were used as the cutoff points. A Cox proportional-hazards regression model was used to determine the hazard ratios (HRs) and 95% confidence intervals (95% CIs) and was adjusted for age, sex, ethnicity, smoking status, tumor grade, clinical disease stage, and treatment. The association between the overall survival duration and the T/N expression ratios of the 3 genes was evaluated by using Kaplan-Meier plots and the log-rank test. All analyses were performed by using STATA statistical software (version 8.0; STATA Corp., College Station, Tex). All statistical tests were 2-sided, and P values <.05 were considered statistically significant.
One hundred forty patients with NSCLC (mean age, 65 years) were included in this study (Table 2), including 133 white patients (95%) and 122 ever-smokers (former smokers plus current smokers; 87%). Seventy-four patients (53%) had adenocarcinoma, and 36 patients (26%) had squamous cell carcinoma. Sixty-six patients (47%) had stage I disease, 23 patients (16%) had stage II disease, and 33 patients (24%) had stage III disease. The degree of differentiation was moderate (grade 2) in 55 patients (39%) and poor (grade 3) in 56 patients (40%). Fifty-eight patients (41%) survived for ≥5 years (median survival for all patients, 39.7 months), and 44 patients (31%) remained alive at the end of the study period. High tumor stage and grade at diagnosis, as expected, were associated significantly with a higher risk of death (P = .02 and P = .03, respectively). Among 130 patients with complete treatment information, 71 patients (55%) underwent surgery alone, and 59 patients (45%) underwent surgery and received adjuvant chemotherapy or radiotherapy. Compared with patients who underwent surgery alone, patients who underwent surgery and received adjuvant therapy had a significantly higher death rate (P < .001); mainly because adjuvant therapy as a secondary treatment was received mostly by patients who had high-stage disease and were at high risk of recurrence and metastasis. No significant differences were observed in the distribution of patients according to age (P = .37), sex (P = .50), ethnicity (P = .58), smoking status (P = .34), pack-years (P = .77), or tumor histologic subtype (P = .81) between the patients who remained alive and the patients who died (Table 2).
The median mRNA expression levels were significantly higher in tumor tissues than in adjacent normal tissues for ATM (0.080 vs 0.058; P = .003) and DNA-PKcs (0.073 vs 0.051; P < .001) but not for Ku80 (0.256 vs 0.222; P = .491) (Table 3). The mRNA expression levels of ATM, DNA-PKcs, and Ku80 were correlated with one another in both tumor tissues and adjacent normal tissues. The correlation coefficients for ATM and DNA-PKcs, ATM and Ku80, and Ku80 and DNA-PKcs were 0.71 (P < .001), 0.80 (P < .001), and 0.72 (P < .001), respectively, in tumor tissues and 0.64 (P < .001), 0.66 (P < .001), and 0.51 (P < .001), respectively, in normal tissues (Table 4). The mRNA expression levels of ATM, DNA-PKcs, and Ku80 in tumor tissues were not associated significantly with age, sex, tumor stage, grade, or histologic type (data not shown).
We dichotomized the T/N expression level of ATM, DNA-PKcs, and Ku80 into high and low groups by using the median value as the cutoff point. After adjustment for age, sex, ethnicity, smoking status, tumor stage, clinical disease stage, and treatment, the multivariate Cox proportional-hazards model revealed that the T/N expression ratios of ATM and DNA-PKcswere associated significantly with patients’ overall survival rates (Table 5). Patients who had high T/N expression ratios of ATM and DNA-PKcs had 1.82-fold (95% CI, 1.05–2.70; P = .032) and 2.13-fold (95% CI, 1.21–3.76; P = .009) increased risks of death, respectively, compared with patients who had low T/N expression ratios. No statistically significant association was observed between the T/N expression ratio of Ku80 and overall survival (HR, 1.32; 95% CI, 0.79–2.21; P = .283).
Kaplan-Meier analysis revealed that patients who had high T/N expression ratios of DNA-PKcs had significantly shorter median survival time (MST) than patients who had low T/N expression ratios (34.8 months vs 64.9 months; P = .013; log-rank test) (Fig. 1). Similar results were obtained for patients who had high and low T/N expression ratios of ATM (MST: 37.8 months vs 60.9 months; P = .137, log-rank test), although the difference did not reach statistical significance. The MST for patients who had high and low expression ratios of Ku80 was 38.83 and 47.63 months, respectively (log-rank test; P = .70).
Stratified analyses revealed associations between the T/N expression ratio of ATM and the risk of death in younger patients (aged <66 years; HR, 3.71; 95% CI, 1.29-10.72; P 5 .015) but not in older patients (aged ≥66 years; HR, 1.07; 95% CI, 0.48–2.37; P = .875), in women (HR, 3.79; 95% 5 CI, 1.16-12.39; P = .027) but not in men (HR, 1.05; 95% CI, 0.47–2.38; P = .901), and in light smokers (HR, 8.04; 95% CI, 2.09-30.96; P = .002) but not in heavy smokers (HR, 1.36; 95% CI, 0.63–2.96; P = .434) (Table 5). Similarly, high T/N expression ratios of DNA-PKcs also were associated significantly with a low survival rate in younger patients (aged <66 years; HR, 3.75; 95% CI, 1.35–10.47; P = .011) but not in older patients (aged ≥66 years; HR, 1.71; 95% CI, 0.80–3.65; P = .164), in women (HR, 4.78; 95% CI, 1.59–14.41; P = .005) but not in men (HR, 1.70; 95% CI, 0.75–3.88; P = .207), and in light smokers (HR, 3.24; 95% CI, 1.17–8.98; P = .024) but not in heavy smokers (HR, 2.09; 95% CI, 0.89-4.92; P = .092). In addition, the risk of death conferred by high T/N ratios appeared to be evident in patients with adenocarcinoma but not in patients with squamous cell carcinoma for both ATM and DNA-PKcs. We also performed stratified analysis by adjuvant therapy and observed that the association between the T/N expression ratio of ATM and the risk of death was evident in patients who underwent surgery alone (HR, 2.71; 95% CI, 1.11–6.64; P = .029) but not in patients who underwent surgery and received adjuvant therapy (HR, 1.00; 95% CI, 0.47–2.10; P = .989), whereas the association for DNA-PKcs was stronger in patients who underwent surgery and received adjuvant therapy (HR, 2.52; 95% CI, 1.13–5.65l P = .024) than in patients who underwent surgery alone (HR, 1.78; 95% CI, 0.80-3.97; P = .158). No significant association was observed between the T/N expression ratio of Ku80 and the risk of death for either treatment group.
In the current study, we determined the mRNA expression of 3 key components of the DNA DSB repair pathway in patients with NSCLC and observed significantly higher expression in tumor tissues compared with normal tissues for ATM and DNA-PKcs but not Ku80. The high T/N expression ratios of ATM and DNA-PKcs, but not Ku80, were associated significantly with increased risks of death and notably shorter MST compared with the low T/N expression ratios. To our knowledge, this is the first study to measure mRNA expression of these 3 genes in NSCLC tissue and to report that ATM and DNA-PKcs may be useful prognostic markers in NSCLC.
There have been some studies evaluating the expression of these genes in other cancer tissues. Consistent with our results, Hosoi et al.20 observed an elevated mRNA levels of DNA-PKcs in colorectal cancer tumor tissues compared with normal tissues. Studies of ATM and Ku80 expression in tumor tissues have yielded inconsistent findings across different cancer types, with some reporting higher expression, some reporting lower expression, and others reporting no changes.12,19,20,25-27 The inconsistency in ATM and Ku80 regulation in different tumor types suggests a remarkable difference on the molecular alterations related to the progression of cancer at different sites. In addition, the methodology of measuring mRNA expression and the selection of control genes also may contribute to the inconsistency of results. In this study, we used the geometric mean of the expression of 3 housekeeping genes (GAPDH, 18S rRNA, and β-actin) as internal controls to normalize target gene expression. Vandesompele et al.28 provided compelling evidence to support the finding that geometric averaging of multiple internal control genes offers more accurate normalization of real-time quantitative RT-PCR data than single internal control genes. This method can eliminate the expression variability of a single control gene in response to various factors. In the current study, we also compared the use of single and multiple housekeeping genes and observed that the variations were much smaller when the data were normalized to the geometric mean of 3 housekeeping genes than any individual gene (data not shown). Most previous studies used a single gene as an internal control; therefore, it is difficult to compare the results across different studies.
The mechanisms of ATM, DNA-PKcs, and Ku80 regulation remain to be elucidated. However, exogenous and endogenous factors may play a role in their regulations. Jiang et al.25 observed that the benzo (a)pyrene diol epoxide in tobacco smoke was responsible for increased ATM expression levels in premalignant and malignant esophageal tissues, and Hosoi et al.20 demonstrated that overexpression of the transcription factor SP1 contributes to elevated levels of DNA-PKcs and Ku70/Ku80. A correlation between Ku80 and DNA-PKcs expression has been reported in patients with colorectal cancer and head and neck cancer.20,29 The promoter regions of both Ku80 and DNA-PKcs contain consensus SP1 recognition elements; thus, DNA-PKcs and Ku80 may be regulated transcriptionally by SP1. A connection also may exist between the levels of DNA-PKcs and ATM.30 Peng et al.31 demonstrated that the amount or activity of DNA-PKcs influenced ATM expression. In our study, the mRNA levels of ATM, DNA-PKcs, and Ku80 were correlated with one another. These findings suggest that there are highly coordinated efforts of the cellular defense system in repairing DNA DSBs.
Most noteworthy, we observed that high expression of ATM and DNA-PKcs was associated with poor overall survival in patients with NSCLC. Safar et al.32 reported that ATM hypermethylation was associated with good overall survival in patients with NSCLC, and Bosken et al.33 observed that patients with NSCLC who had effective systemic nucleotide excision repair had poorer survival than patients who had suboptimal nucleotide excision repair. Those results are consistent with our findings. However, there also have been some inconsistent results in terms of the prognostic value of these 3 genes in other cancer types. For example, low ATM expression has been associated with an increased risk of death in colorectal and breast cancers.12,14 Again, these inconsistencies may be the results of many factors, including different cancer types, different histology and tumor characteristics, and accuracy in measuring mRNA expression. In addition, all of the literature on the prognostic value of ATM, DNA-PKcs, and Ku80 discussed above were based on gene expression in tumor tissues. However, in our study, we attempted to assess the possibility of using the T/N expression ratios of 3 DNA repair genes as prognostic markers. We compared the prognostic values between T/N expression ratios and tumor expression levels of the 3 genes by using a multivariate Cox proportional-hazards model. Patients with a high T/N expression ratios of ATM and DNA-PKcs exhibited significantly increased risk of death, as discussed above. However, no significant associations were observed between survival and tumor expression levels of ATM (HR, 0.92; 95% CI, 0.53–1.60; P = .766), Ku80 (HR, 1.48; 95% CI, 0.90–2.43; P = .120), or DNA-PKcs (HR, 1.17; 95% CI, 0.71–1.92; P = .530). Therefore, we believe that, for patients with cancer, the T/N expression ratio, which reflects alteration of the gene expression level, may be a better predictor of prognosis, although the requirement of having normal tissue available for analysis is a limitation for measurement of the T/N expression ratio.
The molecular mechanism of cell survival modulation by ATM and DNA-PKcs remains to be elucidated. Harima et al.22 suggested that tumor cells with a greater ability to repair damaged DNA are more likely to survive and proliferate. In a comprehensive review, Kaina also reported that cells that exhibit a defect in various DNA repair pathways are liable to suffer from DNA damage-triggered apoptosis.34 Therefore, we speculate that high expression of ATM and DNA-PKcs leads to increased DNA DSB repair capacity in tumor cells and that tumor cells with higher DSB repair capacity would be more likely to survive, proliferate, and metastasize. In addition, the expression of ATM modulates patients’ survival by triggering growth factor-mediated pathways35; for example, ATM may increase the transcriptional expression of the insulin-like growth factor-1 receptor, a cell surface molecule with tyrosine kinase activity that can mediate mitogenesis, cell transformation, and apoptosis inhibition. Other functional roles of ATM and DNA-PKcs in survival modulation are under investigation.
In the current study, no significant correlation was observed between the mRNA expression of ATM, Ku80, and DNA-PKcs in tumor tissues and patients’ age, sex, tumor grade, stage, and histologic type, further suggesting that the effect of the 3 DNA repair genes on patient survival is independent of these clinical characteristics. In addition, the results of our stratified analyses indicate that the prognostic values of ATM and DNA-PKcs are significant only in younger patients (aged <66 years), in women, in light smokers, and in patients with adenocarcinoma, suggesting that age, sex, smoking, and histologic type may play critical roles in modifying the prognostic importance of ATM and DNA-PKcs expression in patients with NSCLC. In addition, adjuvant therapy also appeared to have different modifying effects on the prognostic value of ATM and DNA-PKcs expression. This result may have been obtained because of the different functions of ATM and DNA-PKcs in DNA repair and cell proliferation. However, the results from these stratified analyses should be interpreted with caution given the small sample size in each stratum. Large studies are warranted to confirm these findings.
In summary, we observed that the expression levels of ATM and DNA-PKcs were higher in NSCLC tumor tissues than in normal tissues and that high T/N expression ratios of ATM and DNA-PKcs were associated significantly with poor survival in patients with NSCLC. The current results indicate that ATM and DNA-PKcs may be useful in predicting the clinical outcome of patients with NSCLC, thus helping physicians to determine optimal therapies for individual patients.
Supported by grants CA 70907, CA 55769, and CA 111646 from the National Cancer Institute and by grant DAMD 17-02-1-0706 from the U.S. Department of Defense.