Abstract
“Replicative stress” is one of the main factors underlying neoplasia from its early stages. Genes involved in DNA synthesis may therefore represent an underexplored source of potential prognostic markers for cancer. To this aim, we generated gene expression profiles from two independent cohorts (France, n = 206; United Kingdom, n = 117) of patients with previously untreated primary breast cancers. We report here that among the 13 human nuclear DNA polymerase genes, DNA Polymerase θ (POLQ) is the only one significantly up-regulated in breast cancer compared with normal breast tissues. Importantly, POLQ up-regulation significantly correlates with poor clinical outcome (4.3-fold increased risk of death in patients with high POLQ expression), and this correlation is independent of Cyclin E expression or the number of positive nodes, which are currently considered as markers for poor outcome. POLQ expression provides thus an additional indicator for the survival outcome of patients with high Cyclin E tumor expression or high number of positive lymph nodes. Furthermore, to decipher the molecular consequences of POLQ up-regulation in breast cancer, we generated human MRC5-SV cell lines that stably overexpress POLQ. Strong POLQ expression was directly associated with defective DNA replication fork progression and chromosomal damage. Therefore, POLQ overexpression may be a promising genetic instability and prognostic marker for breast cancer.
Keywords: specialized DNA replication, prognosis marker, S-phase checkpoint
Besides the “replicative” DNA polymerases POLA, POLD, and POLE, which are involved in conventional DNA replication of the undamaged genome, mammalian nuclei contain 10 additional specialized DNA polymerases that play a role in replication, repair, and recombination of damaged DNA (1, 2) and thus may be of paramount importance to preserve the integrity of the genome.
Specialized DNA polymerases are frequently deregulated in neoplasia (3–10). Indeed, the intracellular balance between the error-free, replicative polymerases POLA, POLD, and POLE and the error-prone, specialized DNA polymerases (POLH, POLL, POLM, POLN, POLK, POLB, POLI, POLQ, POLZ/REV3L, and REV1) appears to be of great importance for the maintenance of genome stability (11–14). Here, we wondered whether misregulation of DNA polymerases could be a signature of breast cancer progression. Indeed, beside the standard classification used by pathologists, there is a clear lack of tools to accurately predict the clinical outcome of many patients.
We specifically measured the expression levels of the 13 human replicative and specialized DNA polymerases in 206 breast carcinomas. We report that, differently from the replicative and the other specialized DNA polymerases, POLQ was significantly up-regulated in most of the breast tumors analyzed. Such up-regulation was associated with poor clinical outcome.
POLQ is an error-prone, specialized DNA polymerase that might operate during “normal” genomic replication because it bypasses some endogenous DNA lesions and replicates undamaged DNA (15–17). POLQ is also allelic to the murine chaos1 gene (for chromosome aberrations occurring spontaneously 1) (18). The observation that chaos1 mutant mice display enhanced micronuclei in red blood cells further suggests a role of POLQ in maintaining genomic stability in the absence of any exogenous stress. In addition it seems to be involved in tolerance of bulky adducts or in DNA repair pathways such as the base-excision repair (BER), DNA interstrand cross-link (ICL), and DNA break repair pathways (19–23). To investigate the molecular consequences of POLQ up-regulation, we generated human MRC5-SV cells that stably overexpressed the POLQ protein. We found that in such cells the elongation of the replication fork was slower than in control cells. Moreover, MRC5-SV cells that stably overexpress POLQ showed increased formation of γ-H2AX foci (the phosphorylated form of histone H2AX that accumulates in cells containing damaged DNA) and higher frequency of chromosomal abnormalities in comparison with control cells. These results suggest that POLQ overexpression, through the induction of genetic instability, could favor the emergence and survival of proliferating cancer cells.
Results
POLQ is the most up-regulated DNA polymerase gene in breast tumors. To evaluate the gene expression profiles of both replicative (POLA, POLD, and POLE) and specialized (POLH, POLL, POLM, POLN, POLK, POLB, POLI, POLQ, POLZ/REV3L, and REV1) DNA polymerases in breast carcinoma, real-time PCR assays were performed by using a first set (n = 101; Table S1) of tumor (T) and nontumor (N) breast tissues from a French cohort of patients with breast carcinoma. The relative expression levels of the 13 DNA polymerases in each tumor were normalized to their median expression level in seven nontumor breast tissues (Fig. 1). Among all of the polymerases analyzed, POLQ showed the highest T/N ratios. The expression of POLQ was 3- to 26-fold higher in tumor samples than in normal tissues. In 70 of the 101 tumors (69.3%) POLQ expression was up-regulated by 5-fold or more. Expression of POLN, which is the polymerase most closely related to POLQ (24), was also significantly higher in tumors. Very similar data were found when an independent analysis (i.e., using an independent batch of low-density arrays) was performed by using a second set of tumor samples from the cohort of French patients (n = 105; Table S1 and Fig. S1), with the exception of the POLB, REV1, POLI, and POLK T/N ratios, which were mostly <1 in this second group.
Fig. 1.
Relative expression of replicative and specialized DNA polymerases in breast tumors from set 1 of the French cohort (n = 101). Normalized mRNA expression ratios for tumor (T) and normal (N) breast samples (T/N) were calculated. T/N > 1 and < 1 indicate higher and lower expression levels in tumors compared with pooled normal tissues, respectively. The individual T/N ratios were normalized to the expression of HMBS and IPO8, the two most stable control genes. T/N expression ratios were transformed into binomial data: 0 if T/N < 1; 1 if T/N > 1. The new variables then followed a binomial distribution according to the parameters n and p0, which represent the number of patients and the probability that a gene is overexpressed, respectively. To test the significance of a gene over- or underexpression, we checked if the p0 parameter of this binomial distribution was different from 0.5 by using bilateral exact binomial tests (null hypothesis: p0 = 0.5). Because of multiple testing, the P values from these tests were compared with a significance level obtained by using the Benjamini et al. procedure (39) for an overall false discovery rate (FDR) of 0.05. The patients’ samples on the x axis are classified in the same order in each panel. + and – stand for higher and lower expression in the tumor (T) compared with the normal (N) tissues, respectively. For graph representation T/N values <1 were transformed into the inverse N/T values.
Then the nonparametric Spearman’s correlation coefficients were calculated to assess the relationship between the expression data of the different polymerases. We found that in the first tumor set (n = 101), the expressions of the specialized polymerases POLN, POLM, POLK, POLL, and POLI were significantly correlated (Fig. S2; all Spearman’s correlation coefficients > 0.5). Conversely, the POLQ expression profile was not clustered in this pattern, indicating that POLQ up-regulation in tumors is independent from that of the other specialized polymerases.
Up-Regulation of POLQ in Breast Cancer Is Associated with Poor Clinical Outcome.
To investigate whether the expression profile of the different DNA polymerases in breast tumors could be related to patients’ survival, a log-rank test was carried out (Fig. 2). For the French cohort, high expression of POLQ was significantly associated with poor survival (P = 0.0001) (Fig. 2A). This observation was supported by the results we obtained by using a second independent United Kingdom cohort of 117 patients with breast tumors (P = 0.035) (Table S1, Fig. 2B). Results for both French and United Kingdom cohorts were consistent, strongly supporting the association of POLQ up-regulation with poorer survival in breast cancer patients.
Fig. 2.
Relationship between POLQ expression level to cancer-specific survival. In the French cohort, POLQ expression levels ≤0.063 were defined as “low expression.” For the British cohort, POLQ expression ≤0.31 was defined as low expression. These cutoffs were defined in relation to survival to identify two statistically different populations of patients. ROC (Receiver Operating Characteristics) curves were used to determine the optimal cutoff point that best segregated patients in terms of survival. Cancer-specific survival was defined as the interval between the date of breast surgery and the date of death or of the last-follow-up (censored data). Patients who died from another cause were considered as censored observations. Survival rates were estimated according to the Kaplan–Meier method. (A and B) Kaplan–Meier survival curves of breast cancer patients according to the level of POLQ expression in the primary tumor. For these graphical representations, P values were given according to the log-rank test. A P value <0.05 was considered statistically significant. (A) Patients from the French cohort (n = 203 instead of 206, because 3 samples were discarded because POLQ expression could not be determined). (B) Patients from the British cohort (n = 117). (C and D) Kaplan–Meier survival curves of pairwise comparison between POLQ gene expression in the primary breast tumor and the number of positive lymph nodes. We used a multivariate Cox model adjusted for POLQ expression and the number of positive lymph nodes. This model detected an effect of POLQ adjusted for the number of lymph nodes on the survival in both the French (P = 0.0001) and the British cohort (P = 0.002). (C) Patients from the French cohort (n = 203). (D) Patients from the British cohort (n = 117). Ln stands for positive lymph node. A low number of metastatic lymph nodes was defined as a lymph node count equal to 1 or less, distinct from patients with 2 or more metastatic axillary nodes.
Lymph node metastasis is also known to be associated with poor survival in breast cancer (25). Results from both cohorts (Fig. 2 C and D) indicated that patients with a high number of positive lymph nodes (i.e., nodes with metastatic breast cancer) had significantly poorer survival rate when POLQ also was up-regulated (pairwise comparison, French cohort, P = 0.0001; British cohort, P = 0.002). Specifically, patients with high number of positive lymph nodes and low POLQ expression had a survival rate comparable with that of patients with low number of positive lymph nodes and high POLQ expression (pairwise comparison, French cohort, P = 0.793; British cohort, P = 0.882). Finally, patients with high number of positive lymph nodes and high POLQ expression showed very poor survival.
We then investigated the correlation between POLQ expression and the level of Cyclin E, a powerful predictor together with the number of positive lymph nodes of the outcome of breast cancer (refs. 25–27; Fig. S3A). A Kaplan-Meier survival analysis (Fig. S3A) indicated that the expression level of POLQ and Cyclin E constituted independent prognostic factors, even if up-regulation of POLQ was significantly more frequent in tumors that overexpressed Cyclin E (Spearman ρ = 0.71). To further confirm that POLQ and Cyclin E up-regulation are two independent cellular events, we depleted Cyclin E in MCF7 breast cancer cells, which overexpress Cyclin E, by RNA interference and could show that POLQ expression was not affected (Fig. S3 B).
We then used a Cox proportional-hazards regression model to further examine the relationship between survival distribution and these different covariates. Univariate analysis confirmed that patients with tumors that strongly overexpressed POLQ had a 4.3-fold higher risk of death than patients with tumors that presented normal levels of POLQ (Table 1). Moreover, multivariate analysis indicated that a close correlation between POLQ expression and survival still remained after adjusting the survival data by taking into account the expression of Cyclin E or the number of positive nodes (Table 1; P = 0.006, P = 0.04, and P = 0.006, P = 0.006, respectively).
Table 1.
Cox regression analyses
Variables | HR | Confidence interval 95% | P (Wald) |
Univariate analysis | |||
POLQ (n = 203) | |||
Low | 1 | ||
High | 4.28 | [2.03;9.01] | 0.0001 |
Cyclin E (n = 206) | |||
Low | 1 | ||
High | 3.60 | [1.76;7.53] | 0.0005 |
Lymph Nodes (n = 206) | |||
≤1 | 1 | ||
>1 | 5.22 | [1.58;17.22] | 0.007 |
Multivariate analysis | |||
POLQ (n = 203) | |||
Low | 1 | ||
High | 3.10 | [1.37;6.97] | 0.006 |
Cyclin E (n = 206) | |||
Low | 1 | ||
High | 2.27 | [1.05;5.00] | 0.04 |
Lymph nodes (n = 206) | |||
≤1 | 1 | ||
>1 | 5.25 | [1.59;17.32] | 0.006 |
Finally, a significant statistical association was found between POLQ expression and a number of prognostic indicators of breast cancer, such as histological grade III (P < 0.0001), tumor size (P = 0.0237), estrogen receptor (ER) status (P < 0.0001), progesterone receptor (PgR) status (P = 0.0034), Ki67 expression (P < 0.0001), HER2 status (P = 0.0032), and again the number of positive lymph nodes (P = 0.0172) (Table S2). Specifically, “triple-negative” tumors (i.e., ER negative, PgR negative, and HER2 negative) were more frequently associated with high levels of POLQ expression (P = 0.0422).
Ectopic Expression of POLQ Affects Cell Cycle and Proliferation.
The previous data suggest that POLQ is either a bystander prognostic marker or that it might actively contribute to tumor progression. To differentiate between these two hypotheses, we decided to evaluate the specific impact of POLQ overexpression on genetic stability and cell proliferation. To this aim, we generated three clonal MRC5-SV cell lines (Q1, Q2, and Q3) that stably express increasing amounts of POLQ protein compared with the endogenous level displayed by the isogenic controls (Fig. S4A). First, we examined the effect of POLQ overexpression on cell cycle progression. Quantitative FACS analysis after DNA staining by propidium iodide (Fig. S4B) revealed that MRC5-SV Q1, Q2, and especially Q3 (the cell line with the highest POLQ expression level) cells significantly accumulated in S phase compared with controls (cells transfected with the empty pcDNA 3.1 vector). MRC5-SV Q3 cells significantly accumulated in G2/M phase as well. To determine the length of the S phase, BrdU-pulsed MRC5-SV Q1 and Q3 cells were chased with thymidine, blocked at G2/M with nocodazole and collected at various times for FACS analysis (Fig. S5). At 8 h, ≈25% of MRC5-SV Q1 and 31% Q3 cells were still in S-phase, whereas ≈89% of control cells had reached the G2/M phase, showing that excess POLQ affects the global duration of S phase. Collectively, these results indicate that up-regulation of POLQ in human cells affects cell cycle progression in the absence of external stress.
POLQ Overexpression Induces the DNA Damage Response.
We next investigated the impact of POLQ overexpression on genomic stability by performing conventional and large-scale immunofluorescence assays to detect γ-H2AX, the phosphorylated form of histone H2AX that accumulates in cells containing damaged DNA. MRC5-SV Q1, Q2, and Q3 cells that overexpress POLQ showed a significant 2- to 3-fold increase in γ-H2AX foci formation compared with the isogenic controls (Fig. 3 A and C). Immunoblotting analysis using anti-γ-H2AX antibodies (Fig. S6A) confirmed these data. We then determined the level of activation of CHK1 and CHK2, two central transducer kinases of the DNA damage response. Although the level of phosphorylated CHK1 did not differ between MRC5-SV Q1, Q2, and Q3 and control cells (Fig. S6B), the number of phosphorylated CHK2 foci was significantly higher in MRC5-SV Q1, Q2, and Q3 cells (Q1, P < 0.0018; Q2, P < 0.002; Q3, P < 0.01) than in isogenic control cells (Fig. 3B). Confocal analysis showed that, in Q1, Q2, and Q3 cells, most of the PT68-CHK2 foci colocalized with the γ-H2AX foci (Fig. 3C), indicating that activated CHK2 was localized to sites of DNA damage. Accumulation of γ-H2AX was directly linked to POLQ overexpression because siRNA-mediated depletion of POLQ in MRC5-SV Q3 fibroblasts (Fig. 3D) and in MCF7 breast cancer cells (Fig. 3E) significantly reduced the number of γ-H2AX-positive cells. This result suggests that overexpression of POLQ is associated with activation of the γ-H2AX-ATM-CHK2 DNA damage checkpoint.
Fig. 3.
DNA damage and DNA damage response in POLQ overexpressing MRC5-SV clones. Immortalized human lung fibroblasts (MRC5-SV) were transfected with POLQ cDNA subcloned in the pcDNA3.1/Hygro(+) expression vector and three clones (Q1, Q2, and Q3) that show progressively increasing expression of POLQ were selected for stable expression. Average number of POLQ overexpressing cells positive for γ-H2AX (A) and PT68-CHK2 (B) immunolabeling. Control (CTL) clones (MRC5-SV cells stably transfected with empty pcDNA3.1 vector) treated with UV (10 J·m−2) were used as positive controls. (C) Representative confocal microscopy images of MRC5-SV cells overexpressing POLQ (Q1, Q2, Q3) show colocalization of γ-H2AX and PT68-CHK2 foci. (D) Western blots of whole-cell lysates prepared from POLQ overexpressing MRC5-SV cells (Q3 clone) transfected with either control Luciferase (C) or POLQ (Q) siRNAs at 24, 48, or 72 h after transfection. Quantification by Arrayscan analysis of γ-H2AX expression in control or Q3 MRC5-SV cells transfected with Luciferase or POLQ siRNAs. NT, nontransfected; *, aspecific signal. (E) Normalized real-time PCR quantification of POLQ expression in human mammary epithelial cell (HMEC) nontumoral controls and MCF7 breast cancer cells transfected with control Luciferase or POLQ siRNAs (n = 3). Western blotting did not allow detection of POLQ protein expression in MCF7 cells. Quantification by Arrayscan analysis of γ-H2AX signal. For each analysis, a minimum of 200 MRC5 or 1,000 MCF7 cells were analyzed in at least three independent experiments. MRC5 and MCF7 positive cells contained more than 1 and 5 foci, respectively. P values were calculated by using the Student’s t test.
POLQ Overexpressing Cells Display Spontaneous Chromosome Abnormalities.
To further investigate the consequences of POLQ overexpression on chromosomal stability, we analyzed metaphase spreads of MRC5-SV Q1, Q2, and Q3 cells and isogenic controls. POLQ overexpression was associated with a significant increase in the frequency of chromosomal abnormalities, mostly end-to-end fusions and chromatid breaks (Table S3). These data indicate that chromosome instability occurs in POLQ-overexpressing cells.
Replication Dynamics Is Perturbed in POLQ Overexpressing Cells.
The previous results led us to examine replication dynamics in POLQ overexpressing cells. To assess whether POLQ overexpression affected the rate of replication fork progression, dynamic molecular combing was performed in MRC5-SV Q2 and Q3 as well as control cells. This method allowed us to determine the polarity of replication forks in vivo at the level of individual replicating DNA molecules and the distribution of fork velocities (Fig. 4). The median speed was 1.699 kb/min in control MRC5-SV cells (empty pcDNA3 vector) (n = 144), 1.403 kb/min (n = 113) in Q2, and 1.402 kb/min (n = 200) in Q3 cells.
Fig. 4.
DNA replication fork velocity in MRC5-SV cells overexpressing POLQ. (A) Representative image of combed and immunostained DNA fibers in control and Q2 and Q3 (which stably overexpress POLQ) MRC5-SV cells. (B) Replication fork velocity distribution in control (CTL2, cell stably transfected with empty vector), Q2, and Q3 MRC5-SV cells. (C) Total length (in megabases) is the sum of all DNA fibers that were studied for each clone; “n” is the number of tracks of IdU and CldU scored for each clone. The median value of the population is given in kilobases per min. The uncertainty of the median replication fork velocity is given in units of the median absolute deviation. The Mann–Whitney test was used to compare Q2 and Q3 data with the CTL2 control.
Because POLQ has been proposed to play a role in BER and translesion synthesis (TLS) (19, 21, 22, 25), we hypothesized that overexpression of POLQ could interfere with the cellular tolerance to DNA damage (including damage generated by reactive oxygen species), thus leading to reduced kinetics of replication fork progression. Therefore, we measured the sensitivity of the POLQ overexpressing MRC5-SV cells to increasing doses of methyl-methane-sulfonate (MMS) and N-Nitroso-N-Methylurea (MNU) (Fig. S7), two agents known to induce DNA damage that is mostly repaired by BER. Survival of Q1, Q2, and Q3 cells was slightly affected upon MNU treatment and even less after incubation with MMS, suggesting that BER may be less effective or that bypass of unrepaired lesions is less effective in POLQ overexpressing cells. This result was confirmed by the analysis of two other POLQ overexpressing MRC5-SV clones (Q4 and Q5) in which POLQ expression was similar to that of Q3.
In conclusion, these data show that excess POLQ reduces DNA fork speed and induces DNA damage. Cell cycle progression abnormalities, genetic instability, and activation of the DNA damage checkpoint could then be consequences of defective DNA synthesis or tolerance to endogenous DNA damage in POLQ overexpressing cells.
Discussion
Our analysis of the gene expression profiles of untreated primary breast cancer samples from two independent cohorts of patients identify POLQ as the specialized DNA polymerase with the greatest change in expression in breast cancer compared with normal breast tissue. Our data differ from those of a previous study (6), probably because we normalized all of the expression data to those of IPO8/HMBS, two genes which show very stable expression (Materials and Methods). Specifically, after normalization, breast tumor samples showed higher expression of the genes encoding the replicative DNA polymerases, and the T/N ratio for POLQ remained the highest in our analysis.
This result expands the range of human solid tumors in which up-regulation of POLQ has been reported (10). The frequent up-regulation of POLQ expression in a wide range of tumors suggests that this event is unlikely to be related to a particular carcinogen and may play a more general role favoring the survival and growth of unstressed, proliferating cells. We also show that POLQ up-regulation in breast cancer is independent from the expression of other specialized DNA polymerase genes. This data suggests that, in breast cancer, POLQ may be involved in a very specific pathway of DNA metabolism and that its up-regulation might have a distinct physiological impact.
The most important finding we report here is that POLQ up-regulation is significantly associated with poor clinical outcomes in two independent cohorts of patients suffering from breast cancer. To our knowledge, this is the first report in which the expression of a DNA polymerase is closely associated with the survival distribution of breast cancer patients. Statistical analysis of the two independent cohorts from two different countries show that among patients with metastatic lymph nodes, only the individuals with a high T/N ratio for POLQ expression had poor survival. Because the tumor, node, metastasis staging classification is a poor predictor of patients’ survival, our results indicate that up-regulation of POLQ could be a useful prognostic factor to better forecast the potential outcome and to determine which patients merit more aggressive therapeutic measures. Furthermore, the correlation between high POLQ expression and the triple-negative phenotype (ER negative, PgR negative, HER2 negative phenotype, linked to the basal-like molecular subtype of breast cancer) supports the suggestion that genetic instability is an important feature of this specific subgroup of breast cancer with poorer prognosis (28). Finally, we found that the association between POLQ up-regulation and survival was independent from the level of Cyclin E expression and the number of positive lymph nodes. The association between POLQ expression and outcome was the same in patients with tumors that had either high or low Cyclin E expression levels per number of positive nodes. We also show in MCF7 breast cancer cells that depletion of Cyclin E did not alter POLQ expression, indicating that POLQ expression is not stimulated in response to the increased replication stress caused by the Cyclin E oncogene (29). These data reveal that POLQ up-regulation is a unique and independent predictor of poor outcome in patients with breast cancer.
Our study also shows that overexpression of POLQ is associated with reduced DNA replication fork speed and induction of the DNA damage response as determined by phosphorylation of H2AX. The involvement of POLQ appears to be direct, because depletion of POLQ in MRC5-SV cells that overexpress POLQ or in MCF7 breast cancer cells reduces the γ-H2AX response.
These findings suggest that, in breast cancer, cells in which POLQ is up-regulated might be selected during early tumorigenesis, a stage in which the chronic activation of the replication and the DNA damage checkpoints has been shown to play a major role (30, 31). Nevertheless, we cannot rule out the possibility that mechanisms other than DNA replication could be involved. Excess POLQ could titrate cofactors that are involved in TLS or DNA repair pathways, such as alternative nonhomologous end joining, ICL repair, or BER, where this polymerase might be involved (19–23). Shorter DNA replication tracts and a modest sensitivity to alkylating agents in POLQ-overexpressing cells could therefore result from an accumulation of DNA damage due to these defective DNA transactions. Such deficiencies could lead to accumulation of single-strand gaps and, finally, to reduced kinetics of replication fork progression.
Cells expressing high amounts of POLQ also display gross cancer-associated chromosomal changes such as dicentric chromosomes or end-to-end fusions, which suggest a possible defect in telomere formation. The frequency of abnormal metaphases in POLQ overexpressing and chaos1 cells (32) are comparable, supporting the idea of the necessity of a tight regulation of POLQ expression.
Oncogene deregulation has been shown to modify the replicative program, thus inducing DNA damage during S phase and creating genetic instability in the absence of functional checkpoints (29–31). The present work shows also that unbalanced expression of DNA polymerases can mimic such an oncogenic effect by perturbing the course of replication forks and enhancing genetic instability. Further investigations are now required to better understand the role of POLQ deregulation in breast cancer and to define the best treatment for patients with cancers presenting POLQ gene up-regulation, lymph node metastases, and/or Cyclin E overexpression to improve their survival rate. Indeed, inhibition of the expression/activity of POLQ, a DNA repair protein, may increase responsiveness to conventional genotoxic treatments.
Materials and Methods
Study Design, Patient and Tumor Samples, Differential Gene Expression, and Statistical Analysis.
A first cohort of tumor samples (n = 206) were obtained from French patients with untreated (at the moment of the biopsy) primary breast cancer and who represented a subset of the women enrolled in an adjuvant multicentric phase III clinical trial (PACS01 trial). The characteristics of the patients and the results of this clinical trial have been published (33). Tumor samples from this cohort were divided in two sets for gene expression analysis (n = 101 and n = 105). Normal breast tissues were obtained from the Claudius Regaud Institute (Toulouse, France) and taken from the surgical specimens, at more than 3 cm of distance from the breast cancer. A second cohort of tumor samples from randomly selected breast tumor patients (n = 117) was obtained from the Tayside Tissue Bank (Dundee, UK). The characteristics of the patients and tumors for both cohorts are described in Table S1. Sample preparation and analysis of differential gene expression were as described (ref. 34; SI Materials and Methods). SI Materials and Methods also include the primer sequences and an extensive description of the statistical analyses.
Subcloning of POLQ.
The design and construction of the POLQ-expression vector are described in SI Materials and Methods. To check the wild-type status of the exogenous POLQ, genomic DNA was extracted from the three MRC5-SV clones (Q1, Q2, and Q3) that stably overexpress POLQ, PCR-amplified and sequenced. The whole 7.7-kb cDNA sequence did not show any mutations (SI Materials and Methods). Genomic DNA was amplified by using primers that were complementary to different exons to prevent amplification of endogenous POLQ. POLQ DNA from CTL2 control cells was not amplified by using these primers, showing that the amplified sequences came from the transgene and not from endogenous POLQ DNA.
Cell Analysis and Transfections.
MRC5-SV fibroblasts were transfected as described (35). POLQ overexpression was immuno-detected as described (36), using purified, recombinant POLQ as size control. Cell cycle analysis and cytotoxicity studies were performed as reported (13). To analyze S phase, cells were pulse-labeled with 10 μM BrdU for 15 min (Sigma-Aldrich), washed then incubated in 30 μM thymidine and 1 μM nocodazole (Sigma-Aldrich), collected, and fixed at various times in cold 70% ethanol. BrdU labeling and FACS analysis were then performed as described (35). Immunodetection of γH2AX and PT68-CHK2 was carried out according to Rao et al. (37). High capacity acquisition of fluorescent cell images was obtained by using an ArrayScan HCS with a ×20 objective lens reader, and image analysis was carried out by using the Cellomics analysis software (Thermo Scientific). Chromosomal aberrations were analyzed by spreading metaphases as described in ref. 38 and using a Nikon microscope (TE300, with ×100 objective). Where indicated, cells were treated at 37 °C with methyl methanesulfonate (MMS; Sigma-Aldrich) or continuously with MNU (Sigma-Aldrich) for 1 h. Additional details can be found in SI Materials and Methods.
DNA Combing and Statistical Analysis.
Experiments were performed as described (35). Cells were sequentially labeled for 20 min with 100 μM IdU and then with 100 μM CldU (Sigma-Aldrich). Additional details can be found in SI Materials and Methods.
Supplementary Material
Acknowledgments
We thank M. Spielman (G Roussy Institute, Paris), T. Delozier (F Baclesse Center, Caen, France), and P. Fumoleau (R Gauducheau Center, Nantes, France) for their contribution to PACS 01. We also thank A Groulet-Martinec (Ipsogen, Marseille, France), S. Bray (Ninewells Hospital), P. Quinlan and L. Jordan (Tayside Tissue Bank), J.J. Maoret (Q-PCR platform; Institut Federatif de Recherche 31, Toulouse), B. Lepage (Tiersmips platform; Toulouse), and our collaborator C. Didier from the Cancer Cell Imaging Platform (INCa and Région Midi-Pyrénées, Laboratoire d’ Hématologie, Centre Hospitalier Universitaire Purpan, Toulouse). Finally, we acknowledge F. Viala (Institut de Pharmacologie et de Biologie Structurale, Toulouse) for iconography assistance. This work was supported by National Institute of Cancer INCa GSO Grant ACI 3R, the 2RITC Foundation, Ligue Nationale Contre le Cancer CC/JSH “équipe labellisée”, Région Midi-Pyrénées, the French Federation of Comprehensive Cancer Centres, Breast Cancer Research Scotland, “Epigenetic profiling in hematological and breast cancer” ARECA network from ARC, and the M. D. Anderson Research Trust.
Footnotes
The authors declare no conflict of interest.
*This Direct Submission article had a prearranged editor.
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.0910759107/-/DCSupplemental.
References
- 1.Sweasy JB, Lauper JM, Eckert KA. DNA polymerases and human diseases. Radiat Res. 2006;166:693–714. doi: 10.1667/RR0706.1. [DOI] [PubMed] [Google Scholar]
- 2.Hanawalt PC. Paradigms for the three rs: DNA replication, recombination, and repair. Mol Cell. 2007;28:702–707. doi: 10.1016/j.molcel.2007.11.014. [DOI] [PubMed] [Google Scholar]
- 3.Bavoux C, et al. Up-regulation of the error-prone DNA polymerase kappa promotes pleiotropic genetic alterations and tumorigenesis. Cancer Res. 2005;65:325–330. [PubMed] [Google Scholar]
- 4.Bergoglio V, et al. Deregulated DNA polymerase beta induces chromosome instability and tumorigenesis. Cancer Res. 2002;62:3511–3514. [PubMed] [Google Scholar]
- 5.Lemée F, et al. Characterization of promoter regulatory elements involved in downexpression of the DNA polymerase kappa in colorectal cancer. Oncogene. 2007;26:3387–3394. doi: 10.1038/sj.onc.1210116. [DOI] [PubMed] [Google Scholar]
- 6.Albertella MR, Lau AA, O'Connor MJ. The overexpression of specialized DNA polymerases in cancer. DNA Repair (Amst) 2005;4:583–593. doi: 10.1016/j.dnarep.2005.01.005. [DOI] [PubMed] [Google Scholar]
- 7.Kunkel TA. Considering the cancer consequences of altered DNA polymerase function. Cancer Cell. 2003;3:105–110. doi: 10.1016/s1535-6108(03)00027-8. [DOI] [PubMed] [Google Scholar]
- 8.Winnepenninckx V, et al. Gene expression profiling of primary cutaneous melanoma and clinical outcome. J Natl Cancer Inst. 2006;98:472–482. doi: 10.1093/jnci/djj103. [DOI] [PubMed] [Google Scholar]
- 9.Pan Q, Fang Y, Xu Y, Zhang K, Hu X. Down-regulation of DNA polymerases kappa, eta, iota, and zeta in human lung, stomach, and colorectal cancers. Cancer Lett. 2005;217:139–147. doi: 10.1016/j.canlet.2004.07.021. [DOI] [PubMed] [Google Scholar]
- 10.Kawamura K, et al. DNA polymerase theta is preferentially expressed in lymphoid tissues and upregulated in human cancers. Int J Cancer. 2004;109:9–16. doi: 10.1002/ijc.11666. [DOI] [PubMed] [Google Scholar]
- 11.Servant L, Bieth A, Hayakawa H, Cazaux C, Hoffmann JS. Involvement of DNA polymerase beta in DNA replication and mutagenic consequences. J Mol Biol. 2002;315:1039–1047. doi: 10.1006/jmbi.2001.5307. [DOI] [PubMed] [Google Scholar]
- 12.Canitrot Y, et al. Overexpression of DNA polymerase beta in cell results in a mutator phenotype and a decreased sensitivity to anticancer drugs. Proc Natl Acad Sci USA. 1998;95:12586–12590. doi: 10.1073/pnas.95.21.12586. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Bergoglio V, Bavoux C, Verbiest V, Hoffmann JS, Cazaux C. Localisation of human DNA polymerase kappa to replication foci. J Cell Sci. 2002;115:4413–4418. doi: 10.1242/jcs.00162. [DOI] [PubMed] [Google Scholar]
- 14.Rey L, et al. Human DNA polymerase eta is required for common fragile site stability during unperturbed DNA replication. Mol Cell Biol. 2009;29:3344–3354. doi: 10.1128/MCB.00115-09. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Arana ME, Seki M, Wood RD, Rogozin IB, Kunkel TA. Low-fidelity DNA synthesis by human DNA polymerase theta. Nucleic Acids Res. 2008;36:3847–3856. doi: 10.1093/nar/gkn310. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Seki M, Wood RD. DNA polymerase theta (POLQ) can extend from mismatches and from bases opposite a (6-4) photoproduct. DNA Repair (Amst) 2008;7:119–127. doi: 10.1016/j.dnarep.2007.08.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Seki M, et al. High-efficiency bypass of DNA damage by human DNA polymerase Q. EMBO J. 2004;23:4484–4494. doi: 10.1038/sj.emboj.7600424. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Shima N, et al. Phenotype-based identification of mouse chromosome instability mutants. Genetics. 2003;163:1031–1040. doi: 10.1093/genetics/163.3.1031. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Yoshimura M, et al. Vertebrate POLQ and POLbeta cooperate in base excision repair of oxidative DNA damage. Mol Cell. 2006;24:115–125. doi: 10.1016/j.molcel.2006.07.032. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.McVey M, Lee SE. MMEJ repair of double-strand breaks (director’s cut): Deleted sequences and alternative endings. Trends Genet. 2008;24:529–538. doi: 10.1016/j.tig.2008.08.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Ukai A, et al. Role of DNA polymerase theta in tolerance of endogenous and exogenous DNA damage in mouse B cells. Genes Cells. 2006;11:111–121. doi: 10.1111/j.1365-2443.2006.00922.x. [DOI] [PubMed] [Google Scholar]
- 22.Prasad R, et al. Human DNA polymerase theta possesses 5′-dRP lyase activity and functions in single-nucleotide base excision repair in vitro. Nucleic Acids Res. 2009;37:1868–1877. doi: 10.1093/nar/gkp035. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Muzzini DM, Plevani P, Boulton SJ, Cassata G, Marini F. Caenorhabditis elegans POLQ-1 and HEL-308 function in two distinct DNA interstrand cross-link repair pathways. DNA Repair (Amst) 2008;7:941–950. doi: 10.1016/j.dnarep.2008.03.021. [DOI] [PubMed] [Google Scholar]
- 24.Marini F, Kim N, Schuffert A, Wood RD. POLN, a nuclear PolA family DNA polymerase homologous to the DNA cross-link sensitivity protein Mus308. J Biol Chem. 2003;278:32014–32019. doi: 10.1074/jbc.M305646200. [DOI] [PubMed] [Google Scholar]
- 25.de Boer M, van Dijck JA, Bult P, Borm GF, Tjan-Heijnen VC. Breast cancer prognosis and occult lymph node metastases, isolated tumor cells, and micrometastases. J Natl Cancer Inst. 2010;102:410–425. doi: 10.1093/jnci/djq008. [DOI] [PubMed] [Google Scholar]
- 26.Keyomarsi K, et al. Cyclin E and survival in patients with breast cancer. N Engl J Med. 2002;347:1566–1575. doi: 10.1056/NEJMoa021153. [DOI] [PubMed] [Google Scholar]
- 27.Sutherland RL, Musgrove EA. Cyclin E and prognosis in patients with breast cancer. N Engl J Med. 2002;347:1546–1547. doi: 10.1056/NEJMNEJMp020124. [DOI] [PubMed] [Google Scholar]
- 28.Hicks J, et al. Novel patterns of genome rearrangement and their association with survival in breast cancer. Genome Res. 2006;16:1465–1479. doi: 10.1101/gr.5460106. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Bartkova J, et al. Oncogene-induced senescence is part of the tumorigenesis barrier imposed by DNA damage checkpoints. Nature. 2006;444:633–637. doi: 10.1038/nature05268. [DOI] [PubMed] [Google Scholar]
- 30.Gorgoulis VG, et al. Activation of the DNA damage checkpoint and genomic instability in human precancerous lesions. Nature. 2005;434:907–913. doi: 10.1038/nature03485. [DOI] [PubMed] [Google Scholar]
- 31.Bartkova J, et al. DNA damage response as a candidate anti-cancer barrier in early human tumorigenesis. Nature. 2005;434:864–870. doi: 10.1038/nature03482. [DOI] [PubMed] [Google Scholar]
- 32.Shima N, Munroe RJ, Schimenti JC. The mouse genomic instability mutation chaos1 is an allele of Polq that exhibits genetic interaction with Atm. Mol Cell Biol. 2004;24:10381–10389. doi: 10.1128/MCB.24.23.10381-10389.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Roché H, et al. Sequential adjuvant epirubicin-based and docetaxel chemotherapy for node-positive breast cancer patients: The FNCLCC PACS 01 Trial. J Clin Oncol. 2006;24:5664–5671. doi: 10.1200/JCO.2006.07.3916. [DOI] [PubMed] [Google Scholar]
- 34.Chirgwin JM, Przybyla AE, MacDonald RJ, Rutter WJ. Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochemistry. 1979;18:5294–5299. doi: 10.1021/bi00591a005. [DOI] [PubMed] [Google Scholar]
- 35.Pillaire MJ, et al. Upregulation of error-prone DNA polymerases beta and kappa slows down fork progression without activating the replication checkpoint. Cell Cycle. 2007;6:471–477. doi: 10.4161/cc.6.4.3857. [DOI] [PubMed] [Google Scholar]
- 36.Seki M, Marini F, Wood RD. POLQ (Pol theta), a DNA polymerase and DNA-dependent ATPase in human cells. Nucleic Acids Res. 2003;31:6117–6126. doi: 10.1093/nar/gkg814. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Rao VA, et al. Endogenous gamma-H2AX-ATM-Chk2 checkpoint activation in Bloom’s syndrome helicase deficient cells is related to DNA replication arrested forks. Mol Cancer Res. 2007;5:713–724. doi: 10.1158/1541-7786.MCR-07-0028. [DOI] [PubMed] [Google Scholar]
- 38.Capp JP, et al. The DNA polymerase lambda is required for the repair of non-compatible DNA double strand breaks by NHEJ in mammalian cells. Nucleic Acids Res. 2006;34:2998–3007. doi: 10.1093/nar/gkl380. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Benjamini Y, Drai D, Elmer G, Kafkafi N, Golani I. Controlling the false discovery rate in behavior genetics research. Behav Brain Res. 2001;125:279–284. doi: 10.1016/s0166-4328(01)00297-2. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.