Abstract
XRCC1 is a key component of DNA base excision repair, single strand break repair, and backup nonhomologous end-joining pathway. XRCC1 (X-ray repair cross-complementing gene 1) deficiency promotes genomic instability, increases cancer risk, and may have clinical application in breast cancer. We investigated XRCC1 expression in early breast cancers (n = 1,297) and validated in an independent cohort of estrogen receptor (ER)-α–negative breast cancers (n = 281). Preclinically, we evaluated XRCC1-deficient and -proficient Chinese hamster and human cancer cells for synthetic lethality application using double-strand break (DSB) repair inhibitors [KU55933 (ataxia telangectasia–mutated; ATM inhibitor) and NU7441 (DNA-PKcs inhibitor)]. In breast cancer, loss of XRCC1 (16%) was associated with high grade (P < 0.0001), loss of hormone receptors (P < 0.0001), triple-negative (P < 0.0001), and basal-like phenotypes (P = 0.001). Loss of XRCC1 was associated with a two-fold increase in risk of death (P < 0.0001) and independently with poor outcome (P < 0.0001). Preclinically, KU55933 [2-(4-Morpholinyl)-6-(1-thianthrenyl)-4H-pyran-4-one] and NU7441 [8-(4-Dibenzothienyl)-2-(4-morpholinyl)-4H-1-benzopyran-4-one] were synthetically lethal in XRCC1-deficient compared with proficient cells as evidenced by hypersensitivity to DSB repair inhibitors, accumulation of DNA DSBs, G2–M cell-cycle arrest, and induction of apoptosis. This is the first study to show that XRCC1 deficiency in breast cancer results in an aggressive phenotype and that XRCC1 deficiency could also be exploited for a novel synthetic lethality application using DSB repair inhibitors. Cancer Res; 73(5); 1621–34. ©2012 AACR.
Introduction
Impaired DNA repair is a driving force for carcinogenesis and may promote aggressive clinical behavior in breast cancer (1, 2). Base excision repair (BER) is required for the accurate removal of damaged DNA bases induced by oxidizing and alkylating agents. DNA single-strand breaks (SSB) induced by chemotherapeutics are processed by SSB repair (SSBR), a pathway related to BER. XRCC1 (X-ray repair cross-complementing gene 1) is a critical factor in BER, SSBR, and back-up nonhomologous end-joining pathway (B-NHEJ). XRCC1, a 70-kDa protein, has 3 functional domains: an N-terminal DNA-binding domain, a centrally located BRCT I domain, and a C-terminal BRCT II domain. Although XRCC1 has no known enzymatic activity (reviewed in refs. 3, 4), it functions as a molecular scaffold protein and is intimately involved in the coordination of DNA repair by interacting with several components of BER/SSBR pathway, such as DNA glycosylases, apurinic/apyrimidinic endonulcease (APE1), PARP-1, polynulceotide kinase (PNK), and ligase III (reviewed in refs. 3, 4).
Preclinically, XRCC1 deficiency delays SSB rejoining, induces mutations, and results in elevated levels of sister chromatid exchanges, a hallmark of genomic instability. XRCC1 deficiency results in hypersensitivity to ionizing radiation and chemotherapeutics (4, 5). In human association studies, germline polymorphisms in XRCC1 may influence cancer risk (6–9) and also response to radiotherapy (10, 11) and chemotherapy (12) in patients (13, 14).
Given the essential role of XRCC1 in DNA repair and carcinogenesis, we hypothesized that XRCC1 may be dysregulated in breast cancer. In the current study, we provide the first evidence that XRCC1 deficiency in tumors is an independent predictor of poor clinical outcome and is associated with an aggressive phenotype. Moreover, we also provide the first preclinical evidence that XRCC1 deficiency in cancer can be exploited for synthetic lethality application using DNA double-strand break (DSB) repair inhibitors. Our study has important clinical translational applications in patients.
Materials and Methods
Study patients
The study was conducted in a consecutive series of 1,650 patients with primary invasive breast carcinomas, who were diagnosed between 1986 and 1999 and entered into the Nottingham Tenovus Primary Breast Carcinoma series. All patients were treated uniformly in a single institution and have been investigated in a wide range of biomarker studies (15–17). Supplementary Table S1 summarizes patient demographics. Supplementary Treatment Data S1 summarizes various adjuvant treatments received by patients in this cohort.
We also evaluated an independent series of 281 estrogen receptor (ER)-α–negative invasive breast cancers diagnosed and managed at the Nottingham University Hospitals (Nottingham, United Kingdom) between 1999 and 2007. All patients were primarily treated with surgery, followed by radiotherapy and anthracycline chemotherapy. The characteristics of this cohort are summarized in Supplementary Table S2.
The Reporting Recommendations for Tumor Marker Prognostic Studies (REMARK) criteria, recommended by McShane and colleagues (18), were followed throughout this study. This work was approved by Nottingham Research Ethics Committee.
Tissue microarrays and immunohistochemistry.
Tumors were arrayed in tissue microarrays (TMA) constructed with 2 replicate 0.6-mm cores from the center and periphery of the tumors. The TMAs were immunohistochemically profiled for XRCC1 and other biologic antibodies (Supplementary Table S3) as previously described (19–22). Immunohistochemical staining for XRCC1 was carried out using the Bond Max automated staining machine and Leica Bond Refine Detection Kit (DS9800) according to manufacturer's instructions (Leica Microsystems). Pretreatment of TMA sections was conducted with citrate buffer (pH 6.0) antigen for 20 minutes. TMA sections were incubated for 15 minutes at room temperature with 1:200 anti-XRCC1 mouse monoclonal antibody (Ab-1, clone 33-2-5, Thermoscientific). HER2 expression was assessed according to the new American Society of Clinical Oncology/College of American Pathologists guidelines using immunohistochemistry and FISH (23).
To validate the use of TMAs for immunophenotyping, full-face sections of 40 cases were stained and protein expression levels of the different antibodies were compared. The concordance between TMAs and full-face sections was excellent (k = 0.8). Positive and negative [by omission of the primary antibody and immunoglobulin G (IgG)-matched serum] controls were included in each run.
Evaluation of immune staining.
The tumor cores were evaluated by 3 specialist pathologists blinded to the clinicopathologic characteristics of patients in 2 different settings. There was excellent intra- and interobserver agreements (k > 0.8; Cohen κ and multirater κ tests, respectively). Whole-field inspection of the core was scored and intensities of nuclear staining were grouped as follows: 0, no staining; 1, weak staining; 2, moderate staining; and 3, strong staining. The percentage of each category was estimated (0%–100%). H score (range, 0–300) was calculated by multiplying intensity of staining and percentage staining as previously described (19–22). Low/negative XRCC1 (XRCC1-) expression was defined by mean of H score of 100 or less. Not all cores within the TMA were suitable for immunohistochemical analysis due to missing cores or absence of tumor cells.
Statistical analysis.
Data analysis was conducted using SPSS (SPSS, version 17). Where appropriate, Pearson χ2, Fisher exact, Student t test, and ANOVA one-way tests were used. Cumulative survival probabilities were estimated using the Kaplan–Meier method, and differences between survival rates were tested for significance using the log-rank test. Multivariate analysis for survival was conducted using the Cox proportional hazard model. The proportional hazards assumption was tested using standard log–log plots. HR and 95% confidence intervals (95% CI) were estimated for each variable. All tests were two-sided with a 95% CI and a P value less than 0.05 was considered significant. For multiple comparisons, P values were adjusted according to Holm–Bonferroni correction method (24).
Sample size and power analysis.
Power of study, sample size, and effect size were determined using PASS (NCSS, version 11). Cox regression of the log HR on XRCC1 covariant (SD = 0.73), based on a sample of 1,163 observations, achieves 90% power at 0.05 significant level to detect a small regression coefficient equal to 0.25 and 0.28 for risk of recurrence and death, respectively. The sample size was adjusted as a multiple regression of the variable of interest on the other covariates in the Cox regression is expected to have an R2 of 0.3. The sample size was adjusted for recurrence and death event rate of 0.40 and 0.30, respectively.
Preclinical studies
Compounds and reagents.
Highly specific and potent Ataxia telangectasia–mutated (ATM) inhibitor [KU55933; 2-(4-morpholinyl)-6-(1-thianthrenyl)-4H-pyran-4-one], DNA-dependent protein kinase catalytic subunit (DNA-PKcs) inhibitors [NU7441 (8-(4-dibenzothienyl)-2-(4-morpholinyl)-4H-1-benzopyran-4-one) and NU7026 (2-(4-morpholinyl)-4H-naphthol[1,2-b]pyran-4-one)] were purchased from Tocris Bioscience. The compounds were dissolved in 100% dimethyl sulfoxide (DMSO) and stored at −20°C.
Cell lines and culture.
Previously well-characterized Chinese hamster (CH) ovary cells; CHO9 (wild-type), EM-C11 (XRCC1-mutant), EM-C12 (XRCC1-mutant) were provided by Prof. M.Z. Zdzienicka, Department of Molecular Cell Genetics, Collegium Medicum in Bydgoszcz, Nicolaus-Copernicus University in Torun (Bydgoszcz, Poland; ref. 25, 26). Cells were grown in Ham F-10 media (supplemented with 10% FBS and 1% penicillin/streptomycin; PAA).
XRCC1-deficient HeLa SilenciX cells and control XRCC1-proficient HeLa SilenciX cells were purchased from Tebu-Bio (www.tebu-bio.com). SilenciX cells were grown in Dulbecco's Modified Eagle's Medium (with l-glutamine 580 mg/L, 4,500 mg/L D19 glucose, with 110 mg/L sodium pyruvate) supplemented with 10% FBS, 1% penicillin/streptomycin, and 125 μg/mL hygromycin B.
MCF-7 breast cancer cells were grown in RPMI media (supplemented with 10% FBS and 1% penicillin/streptomycin; PAA).
XRCC1 knockdown using siRNAs.
Two XRCC1 siRNA constructs (sequences listed in supplementary Table S4), 1 negative scrambled control and siRNA for glyceraldehyde-3-phosphate dehydrogenase (GAPDH; positive control) were used in these studies. The siRNA constructs were purchased from Ambion Life Technologies. The transfection protocol was as described previously by Fan and colleagues (27). MCF-7 cells were plated in 6-well plates (2 mL medium/well without antibiotics). At 50% confluence, transfection was achieved using Lipofectamine 2000 (Invitrogen) according to the manufacturer's protocol. Briefly, siRNA (100 pmol) and Lipofectamine (5 μL) were each separately mixed with 250 μL Opti-MEM1 (GIBCO/Invitrogen) without FBS. After 5-minute incubation at room temperature, the siRNA and Lipofectamine solutions were combined and incubated for another 20 minutes at room temperature. This mixture was then added to plated cells, cultured at 37°C overnight and the medium was later replaced with fresh medium plus penicillin/streptomycin (1%). When the cells attained 100% confluence, they were trypsinized and subsequently transferred into 75 cm2 flasks for continued growth and/or treatment. XRCC1 knockdown was evaluated by Western blotting at various time points after transfection (days 3, 5, and 7).
Clonogenic survival assay.
Two hundred cells per well were seeded in 6-well plates. Cells were allowed to adhere for 4 hours. Compounds (KU55933 or NU7441) were added at the indicated concentrations. The plates were left in the incubator for 10 days for Chinese hamster cells and 14 days for SilenciX cells. For siRNA-transfected MCF-7 cells, 200 cells per well were seeded 3 days after transfection, treated with compound, and left in the incubator for 14 days. After incubation, the media was discarded, fixed (with methanol and acetic acid mixture), stained with crystal violet, and counted.
Surviving fraction = [no. of colonies formed/(no. of cells seeded × plating efficiency)] × 100. All clonogenic assays were done in triplicate.
Neutral COMET assay.
This assay was conducted as described previously (28). Briefly, subconfluent cells were exposed to DSB inhibitors [KU55933 (5 μmol/L)] or [NU7441 (1.5 μmol/L)]. At various time points after exposure (pretreatment, 24 and 48 hours), cells were extracted and comet assays were conducted. A total of 200 comet images were evaluated for olive tail moment for each time point (pretreatment, 24 and 48 hours).
γ-H2AX immunocytochemistry.
This assay was conducted as described previously (28). Briefly, cells were incubated in medium containing ATM inhibitor [KU55933 (5 μmol/L)] or DNA-PKCS inhibitor [NU7441 (1.5 μmol/L)] for 24 hours. The frequencies of cells containing γ-H2AX foci were determined in 100 cells per slide in 3 separate experiments. Nuclei containing more than 6 γ-H2AX foci were considered positive.
Flow-cytometric analyses for cell-cycle progression.
Cells grown to subconfluence were treated with ATM inhibitor [KU55933 (5 μmol/L)] or DNA-PKCS inhibitor [NU7441 (1.5 μmol/L)] for 24 hours and collected by trypsinization and centrifugation (1,000 rpm for 5 minutes). Fluorescence-activated cell sorting (FACS) assay was conducted as described previously (28).
Apoptosis detection by FITC–annexin V flow-cytometric analysis.
Cells were treated for 48 hours with ATM inhibitor [KU55933 (5 μmol/L)] or DNA-PKCS inhibitor [NU7441 (1.5 μmol/L)]. Cells were later collected by trypsinization and centrifugation (1,000 rpm for 5 minutes) and were washed twice with cold PBS and then resuspended cells in 1× binding buffer at a concentration of 1 × 106 cells/mL. Then 100 μL of the solution (1 × 105 cells) was transferred to a 5 mL culture tube and 5 μL of fluorescein isothiocyanate (FITC)–annexin V and 5 μL propidium iodide (PI) were added. The cells were then gently vortexed and incubated for 15 minutes at room temperature (25°C) in the dark. After the incubation, 400 μL of binding buffer was added to each tube and was analyzed by flow cytometry within 1 hour. Data were analyzed using FlowJo7.6.1 software.
Western blot analysis.
This assay was conducted as described previously (28). Primary antibodies used were mouse anti-XRCC1 (Thermo Fisher Scientific); rabbit anti-ATM (Cell Signaling Technology); and rabbit anti-DNA-PKcs (Novus Biologicals).
Results
Clinicopathologic significance of XRCC1 expression
A total of 1,297 tumors in the primary cohort were suitable for analysis of XRCC1 expression. A total of 1,093 of 1,297 (84%) of the tumors were positive for XRCC1 expression, and 204 of 1,297 (16%) tumors were negative for XRCC1 expression (Fig. 1A). Normal breast terminal duct lobular units (TDLU) showed moderate to strong nuclear XRCC1 expression throughout (Fig. 1A). In invasive breast cancer, loss of XRCC1 expression was highly significantly associated with aggressive clinicopathologic features (Table 1) including high histologic grade (P < 0.0001), pleomorphism (P < 0.0001), glandular dedifferentiation (P = 0.006), absence of hormonal receptors (ER−/PgR−/AR−; P < 0.0001), presence of basal-like phenotypes (P = 0.001), and triple-negative phenotypes (P < 0.0001; Table 1).
A, microphotographs of XRCC1 expression in normal (1) and breast cancer tissue (2–5; magnification, ×200). Kaplan–Meier curves showing breast cancer–specific survival (BCSS) in the whole cohort (B), lymph node–negative subgroup (C), and in lymph node–positive subgroup (D); low risk breast cancer (NPI < 3.4) received no adjuvant therapy (E); high-risk breast cancer (NPI ≥ 3.4) ER+ received endocrine therapy (F); high-risk breast cancer (NPI ≥ 3.4) received chemotherapy (G). See text for details.
Association between XRCC1 expression and clinicopathologic variables
Association between XRCC1 and biomarkers of DNA repair, cell-cycle progression and apoptosis
As shown in Table 1, loss of XRCC1 expression was significantly associated with loss of expression of other DNA repair proteins, such as BRCA1 (P < 0.0001) and TOP2A (P < 0.0001). In addition, abnormal expression of tumor suppressor proteins such as p53 (P = 0.03), p16 (P < 0.001), and FHIT (P < 0.0001) was more common in breast cancers with loss of XRCC1 expression as compared with XRCC1-positive tumors, reflecting a higher level of genomic instability in XRCC1-negative tumors. Notably, loss of XRCC1 expression was also significantly associated with low expression of p53 downstream genes that regulate cell-cycle progression and apoptosis such as p21 (P < 0.001), MDM2 (P < 0.001), MDM4 (P < 0.001), Bcl2 (P < 0.001), and Bax (P < 0.001).
Survival analyses
Loss of XRCC1 expression in tumors showed an adverse outcome at 10 years with a 2-fold increase in the risk of death (P < 0.0001; Fig. 1B), recurrence (P < 0.0001; Supplementary Fig. S1A1), and distant metastasis (P < 0.0001; Supplementary Fig. S1A2) compared with those tumors with positive XRCC1 expression. Investigating the clinical outcome of 678 patients with early-stage lymph node negative tumors revealed that tumors with loss of XRCC1 expression displayed a worse prognosis than cases with positive XRCC1 expression (Fig. 1C and Supplementary Fig. S1B1 and S1B2). Adverse clinical outcomes associated with loss of XRCC1 expression was also confirmed in 466 patients with locally advanced lymph node positive breast cancer (Fig. 1D and Supplemental Fig. S1C1 and S1C2).
Prognostic significance of XRCC1 expression in low risk breast cancer patients (Nottingham Prognostic Index < 3.4) who did not receive any adjuvant therapy
Patients with low risk breast cancer [Nottingham Prognostic Index (NPI) < 3.4] were treated with surgery followed by radiotherapy only (n = 368). At 10 years, breast cancer with loss of XRCC1 expression showed a 2.5- to 4.0-fold increase in the risk of death (P < 0.0001; Fig. 1E), recurrence (P < 0.0001; Supplementary Fig. S2A2), and distant metastasis (P = 0.0001; Supplementary Fig. S2A3) compared with those with positive XRCC1 expression.
Prognostic significance of XRCC1 expression in different adjuvant systemic therapy settings
Patients whose tumors were high risk (NPI ≥ 3.4), ER-α–positive, received endocrine therapy, and XRCC1-negative had a 2-fold increase in risk of death (P < 0.0001; Fig. 1F), recurrence (P < 0.0001; Supplementary Fig. S2B2), and distant metastasis (P < 0.0001; Supplementary Fig. S2B3) at 10 years compared with patients whose tumors were positive for XRCC1.
The poor outcome of breast cancer with loss of XRCC1 expression was further confirmed in high-risk breast cancer receiving adjuvant cyclophosphamide, methotrexate, 5-fluorouracil (CMF) chemotherapy. At 10 years, patients with XRCC1-negative tumors showed a 2-fold increase in the risk of death (P = 0.002; Fig. 1G), recurrence (P = 0.001; Supplementary Fig. S2C2), and distant metastasis (P = 0.02; Supplementary Fig. S2C3) compared with those whose tumors were XRCC1-positive.
Multivariate Cox regression analysis
In multivariate Cox regression analysis including other validated prognostic factors, such as lymph node stage, histologic grade, and tumor size (NPI components), XRCC1 expression was a powerful independent predictor for clinical outcome (P < 0.0001; Table 2).
Multivariate analysis using Cox regression analysis confirms that XRCC1 protein expression is independent prognostic factor
ER-α–negative invasive breast cancer (validation cohort)
To confirm the earlier findings in the context of modern adjuvant anthracycline-based chemotherapy, we validated XRCC1 expression in an independent cohort of 281 ER-α–negative breast cancers. The characteristics of this cohort are summarized in Supplementary Table S2. A total of 252 tumors were suitable for analyses of XRCC1 expression. A total of 212 of 252 (84%) of the tumors were positive for XRCC1 expression with 40 of 252 (16%) tumors negative for XRCC1 expression. Patients with XRCC1-negative tumors showed 2-fold increase in the risk of death (HR, 2.1; 95% CI, 1.2–3.7; P = 0.008; Fig. 2A), recurrence (HR, 2.1; 95% CI, 1.2–3.6; P = 0.007; Supplementary Fig. S3A2), and distant metastasis (HR, 2.2; 95% CI, 1.2–3.8; P = 0.006; Supplementary Fig. S3A3) compared with those whose tumors were XRCC1-positive.
Kaplan–Meier curves showing breast cancer specific–survival (BCSS) in the validation cohort. ER− high-risk breast cancer received anthracycline (A), triple-negative high risk that received chemotherapy (anthracycline or CMF; B), triple-negative high risk that received anthracycline chemotherapy only (C), triple-negative high risk that received CMF chemotherapy only (D). See text for details.
Prognostic significance of XRCC1 expression in triple-negative breast tumors
Subgroup analysis in all patients with triple-negative breast cancer was conducted (n = 455). Loss of XRCC1 expression was significantly associated with poor survival either after receiving CMF or anthracycline (HR, 2.3; 95% CI, 1.4–3.7; P = 0.0005; Fig. 2B), recurrence (HR, 2.5; 95% CI, 1.6–3.8; P = 0.00006; Supplementary Fig. S3B2), and distant metastasis (HR, 2.3; 95% CI, 1.4–3.7; P = 0.0005; Supplementary Fig. S3B3) compared with those whose tumors were XRCC1-positive. Prognostic significance was also confirmed when analyzed for the group receiving anthracycline chemotherapy only (HR, 2.9; 95% CI, 1.4–5.7; P = 0.003; Fig. 2C) or CMF only (HR, 1.9; 95% CI, 0.99–3.8; P = 0.05; Fig. 2D).
Preclinical Synthetic Lethality Evaluation
The clinical data confirm that XRCC1 loss is an independent poor prognostic marker. As XRCC1-deficient cells are BER/SSBR deficient and reliant on DSB repair pathways for maintaining genomic stability, we hypothesized that XRCC1-deficient cells are likely to be hypersensitive to DSB repair inhibitors and could be exploited for synthetic lethality applications. ATM is critical for DNA DSB sensing and signal transduction (29). DNA-PKcs is an essential component of NHEJ (30). We therefore investigated highly specific DSB repair small-molecule inhibitors for synthetic lethality. KU55933 is a potent, selective, and competitive ATM kinase inhibitor [Ki = 2.2 nmol/L; IC50 values are 13, 2,500, 9,300, 16,600, >100,000, and >100,000 nmol/L at ATM, DNA-PKcs, mTOR, phosphoinositide 3-kinase (PI3K), phosphoinositide 3-kinase (PI4K), and ATR, respectively; ref. 31]. NU7441 is a potent and selective DNA-PKcs inhibitor (IC50 values are 14, 1,700, 5,000, >100,000, and >100,000 nmol/L for DNA-PKcs, mTOR, PI3K, ATM, and ATR, respectively; ref. 32). NU7026 is a ATP-competitive inhibitor of DNA-PKcs. IC50 values are 0.23, 13.0, >100, and >100 μmol/L for DNA-PKcs, PI3K, ATM, and ATR, respectively (33). ATM and DNA-PKcs inhibitors were investigated in a panel of XRCC1-deficient and -proficient Chinese hamster and human cancer cells.
Chinese hamster cells
We first confirmed XRCC1 deficiency in EM-C11 and EM-C12 cells by Western blot analysis (Fig. 3A). EM-C11, EM-C12 as well as CHO9 are proficient in the expression ATM and DNA-PKcs (Fig. 3A). KU55933 as well as NU7441 induce reduced survival in EM-C11 and EM-C12 cells compared with CHO9 cells (Fig. 3B and C). Similar reduced survival was also seen with cells treated with NU7026 (another DNA-PKcs inhibitor; Supplementary Fig. S4A). The neutral COMET assay specifically detects DSBs in DNA. Figure 3D summarizes the results for EM-C11, EM-C12, and CHO9 cells treated with 5 μmol/L of KU55933. Compared with pretreatment samples, after 24 hours of exposure to KU55933, XRCC1-deficient cells accumulate significantly higher DSBs. Similar results were also seen in XRCC1-deficient cells treated with 1.5 μmol/L of NU7441 (Fig. 3E). DSBs induce phosphorylation of H2AX at serine 139 (γ-H2AX), and accumulation of γ-H2AX foci in the nucleus is a marker of DSBs. Nuclei containing more than 6 γ-H2AX foci were considered positive for DSB accumulation. Figure 3F and G show that EM-C11 and EM-C12 cells accumulate significantly more γ-H2AX foci in their nucleus compared with control cells after treatment with 5 μmol/L of KU55933. Similar results were also seen in XRCC1-deficient cells treated with 1.5 μmol/L with NU7441 (Fig. 3H). Accumulation of DSBs may delay cell-cycle progression. In EM-C11, EM-C12, and CHO9 cells exposed to DSB repair inhibitor for 24 hours cell-cycle progression was evaluated. Figure 4A and B summarizes the data for cells treated with 5 μmol/L of KU55933. At 24 hours, EM-C11 and EM-C12 were shown to be significantly arrested in G2–M phase of the cell cycle compared with CHO9 cells. Similar results were observed in XRCC1-deficient cells treated with 1.5 μmol/L with NU7441 (Fig. 4C). Increased early apoptosis at 48 hours is shown in EM-C11 and EM-C12 cells compared with CHO9 cells treated with 5 μmol/L of KU55933 or 1.5 μmol/L of NU7441 (Supplementary Fig. S4B).
A, Western blot analysis in Chinese hamster (CH) cells (CHO9, EM-C11, EM-C12). Clonogenic survival assays for Chinese hamster cells treated with KU55933 (B) and NU7441 (C) at indicated concentrations (see Materials and Methods for details). D, neutral COMET assays for Chinese hamster cells treated with KU55933. Compared with pretreatment samples, after 24 hours of exposure to ATM inhibitor, the mean tail moment was significantly higher in EM-C11 and EM-C12 cells at 24 hours (P = 0.01; P = 0.05, respectively) and at 48 hours (P = 0.05; P = 0.04, respectively) in comparison with CHO9 cells. E, neutral COMET assays for Chinese hamster cells treated with NU7441. The mean tail moment was significantly higher in EM-C11 and EM-C12 cells at 24 hours (P = 0.05; P = 0.04, respectively) and at 48 hours (P = 0.006; P = 0.01, respectively). F, microphotographs of γ-H2AX immunocytochemistry in Chinese hamster cells showing more than 6 γ-H2AX foci in EM-C11 and EM-C12 cells compared with CHO9 cells after 24 hours of treatment with KU55933. G, in CHO9 cells, the mean γ-H2AX–positive cells was 7 in pretreatment cells and increased to 15 after the 24-hour treatment with KU55933. In EM-C11 cells, the mean γ-H2AX–positive cells was 6 in pretreatment cells and increased to 36 after the 24-hour treatment (P = 0.01). H, γ-H2AX immunocytochemistry data for NU7441 is shown here. Cells were exposed to 1.5 μmol/L DNA-PKcs inhibitor for 24 hours and compared with control samples before compound treatment. In CHO9 cells, the mean γ-H2AX–positive cells was 8 in pretreatment cells and increased to 12 after the 24-hour treatment. In EM-C11 cells, the mean γ-H2AX–positive cells was 10 in pretreatment cells and increased to 26 after the 24-hour treatment (P = 0.06).
A, a typical FACS read out in Chinese hamster cells treated with 24 hours of KU55933 is shown here. Quantification of various phases of the cell cycle is shown for Chinese hamster cell treated with KU55933 (B) and NU7441 (C). D, Western blot analysis in human cancer cells (control silenciX and XRCC1-deficient silenciX cells). Clonogenic survival assays for human cells treated with KU55933 (E) and NU7441 (F) at indicated concentrations is shown here (see Materials and Methods for details). Neutral COMET assays for human cells treated with KU55933 (G) and NU7441 (H). I, γ-H2AX immunocytochemistry in human cells showing more than 6 γ-H2AX foci in XRCC1-deficient silenciX cells compared with control silenciX after 24 hours of treatment with KU55933. In XRCC1-proficient control SilenciX cells, the mean γ-H2AX–positive cells was 7 in pretreatment cells and increased to 10 after a 24-hour treatment with KU55933. In XRCC1-deficient SilenciX cells, the mean γ-H2AX–positive cells was 8 in pretreatment cells and increased to 16 after a 24-hour treatment (P = 0.04). J, γ-H2AX immunocytochemistry in cells treated with NU7441. In XRCC1-proficient control SilenciX cells, the mean γ-H2AX–positive cells was 7 in pretreatment cells and increased to 8 after a 24-hour treatment with NU7441. In XRCC1-deficient SilenciX cells, the mean γ-H2AX–positive cells was 8 in pretreatment cells and increased to 15 after a 24-hour treatment (P = 0.001; see Materials and Methods and Results section for details). K, quantification of various phases of the cell cycle in human cells treated with 24 hours of KU55933. L, quantification of various phases of the cell cycle in human cells treated with 24 hours of NU7441. M, quantification of human cells (control and XRCC1-deficient SilenciX) undergoing apoptosis after 48-hour treatment with KU55933 (5 μmol/L) and NU7441 (1.5 μmol/L).
Human HeLa SilenciX cells
Figure 4D confirms that XRCC1 SilenciX cells are XRCC1-deficient compared with control SilenciX cells. Both cell lines are proficient for ATM and DNA-PKcs protein expression. Figure 4E shows that KU55933 (5 μmol/L) is selectively toxic to XRCC1-deficient SilenciX cells in comparison with control XRCC1-proficient SilenciX cells. Similar results were also shown in XRCC1-deficient cells treated with 1.5 μmol/L with NU7441 (Fig. 4F). Neutral comet assays in cells treated with 5 μmol/L of KU55933 (Fig. 4G) show that the mean tail moment was significantly higher in XRCC1-deficient SilenciX cells at 24 hours (P = 0.04) and at 48 hours (P = 0.009) posttreatment in comparison with XRCC1-proficient control SilenciX cells. Similar results were also shown in XRCC1-deficient cells treated with 1.5 μmol/L with NU7441 (Fig. 4H). The mean tail moment was higher in XRCC1-deficient SilenciX cells at 24 hours (P = 0.03) and at 48 hours (P = 0.01) posttreatment in comparison with XRCC1-proficient control SilenciX cells. Figure 4I shows that XRCC1-deficient SilenciX cells accumulate significantly more γ-H2AX foci in their nucleus compared with control cells after treatment with 5 μmol/L of KU55933. Similar results were also seen in XRCC1-deficient cells treated with 1.5 μmol/L with NU7441 (Fig. 4J). Figure 4K and L shows that SilenciX cells were arrested at the G2–M phase of the cell cycle after treatment with 5 μmol/L of KU55933 or 1.5 μmol/L of NU7441, respectively. Figure 4M confirms substantial apoptosis at 48 hours in SilenciX cells treated with 5 μmol/L of KU55933 or 1.5 μmol/L of NU7441 compared with control cells.
MCF-7 breast cancer cells
We first confirmed XRCC1 knockdown using siRNA in MCF-7 cells. After transfection, cell lysates were sampled on days 3, 5, and 7 for XRCC1 knockdown by Western blot analysis. On day 3, there was robust XRCC1 knockdown (more than 80%) in cells transfected with construct 1 or 2 (Fig. 5A) compared with scrambled negative control. We used GAPDH as a positive control and also showed efficient knockdown of GAPDH in cells (Fig. 5A). KU55933 treatment reduced survival in XRCC1-deficient cells compared with proficient cells (Fig. 5B). Figure 5D summarizes the results for neutral comet assay with KU55933. Significant DSBs accumulation was observed in XRCC1-deficient cells at 24 hours [P = 0.001(construct 1); P = 0.003 (construct 2)] and 48 hours [P = 0.002 (construct 1); P = 0.05 (construct 2)] compared with XRCC1-proficient cells. This was associated with significant number of γ-H2AX foci [Fig. 5E; P = 0.05 (construct 1); P = 0.02 (construct 2)] and accumulation of cells in G2–M phase of the cell cycle [Fig. 5F; P = 0.05 (construct 1); P = 0.05 (construct 2)] in XRCC1-deficient cells compared with proficient cells. Similarly, NU7441 treatment was more toxic to XRCC1-deficient cells (Fig. 5C) and associated with accumulation of DSBs (Fig. 5G) at 24 hours [P = 0.002 (construct 1); P = 0.01 (construct 2)] and 48 hours [P = 0.01 (construct 1); P = 0.005 (construct 2)] and formation of more γ-H2AX foci [Fig. 5H; P = 0.01 (construct 1); P = 0.04 (construct 2)] and G2–M cell-cycle arrest [Fig. 5I; P = 0.002 (construct 1); P = 0.007 (construct 2)].
A, Western blot analysis confirming XRCC1 knockdown in MCF-7 cells transfected with siRNA constructs 1 and 2. Scrambled negative control as well as GAPDH-positive control is also shown here. B, clonogenic survival assay. XRCC1-deficient cells are hypersensitive to KU55933. C, clonogenic survival assay. NU7441 treatment was more toxic to XRCC1-deficient cells compared with proficient cells. D, neutral COMET assays. XRCC1 knockdown MCF-7 cells accumulate more DSBs after treatment with KU55933. E, after KU55933 treatment, XRCC1 knockdown cells accumulate significant number of γ-H2AX foci compared with proficient cells (see text for details). F, XRCC1 knockdown cells accumulate in G2–M phase of the cell cycle after treatment with KU55933. G, neutral COMET assays. XRCC1 knockdown cells MCF-7 cells accumulate more DSBs after treatment with NU7441. H, after NU7441 treatment, XRCC1 knockdown cells accumulate significant number of γ-H2AX foci compared with proficient cells (see text for details). I, XRCC1 knockdown cells accumulate in G2–M phase of the cell cycle after treatment with NU7441.
The data presented in human cancer cells as well as in Chinese hamster cells provide compelling evidence that in XRCC1-deficient cells DSB repair inhibitors induce synthetic lethality by inducing accumulation of DSB leading to cell-cycle arrest and death.
Discussion
XRCC1 is a key scaffolding protein intimately involved in BER, SSBR, and B-NHEJ. This is the first study to evaluate XRCC1 protein expression in breast cancer. Survival analyses confirmed that XRCC1 negativity is an independent predictor of poor clinical outcome. We validated our data in an independent cohort of ER− tumors and confirmed the negative prognostic significance. In triple-negative tumors, XRCC1 negativity remains significantly associated with poor clinical outcome compared with XRCC1-positive tumors. XRCC1 may be a useful tool in the stratification of patients with breast cancer for adjuvant therapy. Our data suggest that otherwise low risk (NPI < 3.4) XRCC1-negative tumors should be included within high-risk groups and would benefit from adjuvant systemic therapy. Moreover, poor clinical outcome with loss of XRCC1 expression was also confirmed in high-risk breast cancer (NPI ≥ 3.4), suggesting that alternative aggressive systemic chemotherapy may be warranted in this group. However, one limitation of this study is that these historical patients received CMF chemotherapy, and whether XRCC1 expression can stratify patients who receive a more modern anthracycline/taxol based chemotherapy remains to be established in follow-on studies.
Our clinical data are in contrast to preclinical studies wherein XRCC1-deficient cell lines have been shown to be sensitive to cytotoxic therapy (3, 4). Mutator phenotype is a well-recognized phenomenon where DNA repair deficiency leads to genomic instability, which further leads to increased accumulation of mutations and selection of aggressive malignant clones that may be resistant to therapy (2, 34). DNA polymerase-β, another key factor in DNA BER, has been shown to exhibit the mutator phenotype phenomenon (2, 34). We speculate that loss of XRCC1 leads to genomic instability and a mutator phenotype phenomenon may be operating in human cancer cells deficient in XRCC1 expression. To test this hypothesis in human tumors, we conducted correlative studies with markers of DNA repair, cell cycle, and apoptosis. XRCC1-negative tumors were biologically more aggressive as evidenced by high histologic grade, pleomorphism, glandular dedifferentiation, absence of hormonal receptors, and presence of basal-like phenotypes, triple-negative phenotypes, and loss of CK18. XRCC1 negativity was highly associated with loss of other DNA repair factors such as BRCA1 and TOP2A implying intrinsic genomic instability in these tumors. Mutant p53 (that promotes genomic instability) and loss of downstream markers (MDM2 and MDM4) was significantly associated with loss of XRCC1. p53-resistant tumors are known to be resistant to chemotherapy. In a previous study, we investigated the expression of inactive p53 in this cohort and showed resistance to therapy (20). The association with loss of Topo II expression and XRCC1 deficiency provides additional evidence for therapy resistance seen in this cohort. Whether gene silencing or posttranscriptional mechanisms operate to reduce XRCC1 protein expression remains unclear, although a recent preclinical study suggests that the PI3K-Akt/MAPK pathway, which is frequently dysregulated in breast cancer (35, 36), may be involved in regulating XRCC1 expression in cells (37).
Preclinically, we then investigated if XRCC1 deficiency could be exploited for stratified treatment strategy in breast cancer. The ability of PARP inhibitors to induce synthetic lethality in BRCA-deficient breast cancers (38) suggests that additional factors within BER/SSBR may be suitable for personalized approaches. XRCC1 is a key factor in BER and SSBR. As XRCC1-deficient cells may be reliant on DSB repair pathways to maintain cellular survival, we hypothesized that XRCC1 deficiency in somatic tumors could be exploited for synthetic lethality application by blocking the DSB repair pathway in cells. ATM and DNA-PKcs are members of the phosphatidylinositol-3-kinase-like protein kinase (PIKK) family. ATM is a key sensor and transducer of DNA damage signaling during homologous recombination in response to DSBs generated during replication (29). DNA-PKcs is an essential component of NHEJ (30). While NHEJ operates throughout the cell cycle, homologous recombination is limited to S and G2 phase. However, the mechanisms by which cells choose between NHEJ and homologous recombination remains to be established. In addition, the molecular complexities of interactions between DNA-PKcs, ATM, NHEJ, and homologous recombination are only beginning to emerge (39). For example, it seems that there may be more than 40 phophorylation sites in DNA-PKcs and how posttranslational modifications of DNA-PKcs impact homologous recombination is unknown (39). Moreover, novel interactions between XRCC1, ATM, and DNA-PKcs have also been recently reported (40, 41). ATM-Chk2–dependent phosphorylation of XRCC1 was been shown to promote BER by Chou and colleagues (40). In another study, DNA-PKcs was shown to phosphorylate BRCT1 domain of XRCC1 at serine 371 and cause XRCC1 dimer dissociation. In addition, XRCC1 was also shown to stimulate phosphorylation of p53-Ser 15 by DNA-PKcs in that study (41). XRCC1 is also a key component of the B-NHEJ pathway (42, 43). Therefore, given the interesting links, we investigated whether a synthetic lethality relationship exists between XRCC1, ATM, and DNA-PKcs. Detailed in vitro studies in XRCC1-deficient and -proficient cells treated with ATM inhibitor (31) or DNA-PKcs (32) inhibitors were conducted.
We show herein that highly specific and potent ATM and DNA-PKcs inhibitors are synthetically lethal in XRCC1-deficient cells. We have concluded synthetic lethality for the following reasons. First, in Chinese hamster cells deficient in XRCC1, we observed increased sensitivity to DSB repair inhibitors and confirmed this observation in XRCC1-deficient human cancer cells. We have recently shown a similar phenomenon in Chinese hamster cells as well as in human cancer cells deficient in APE1, another key factor in BER (28). Second, functional analyses in XRCC1-deficient Chinese hamster and human cells confirmed that DSB repair inhibitors led to an accumulation of DSBs, arrest of G2–M cell-cycle progression, and induction of apoptosis. We present a working model for DSB repair inhibition as a synthetic lethality strategy in XRCC1-deficient cells. In brief, XRCC1-deficient cells have elevated SSBs that are eventually converted to toxic DSBs. In cells deficient in DSB repair (such as by pharmacologic inhibition of ATM or DNA-PKcs), DSBs would persist and lead to the observed synthetic lethality. In cells that are proficient in DSB repair, DSBs would be repaired and cells would survive.
In conclusion, we provide the first clinical evidence that XRCC1 deficiency is an independent poor prognostic biomarker in patients with breast cancer. Preclinically, we have shown that XRCC1 deficiency can be exploited for synthetic lethality application in breast cancer by blocking DSB repair using ATM or DNA-PKcs inhibitors. Accelerated pharmaceutical development of DSB repair inhibitors is urgently warranted to expand synthetic lethality strategies in breast cancer.
Disclosure of Potential Conflicts of Interest
G. Ball is employed as Chief Technical Officer in CompanDX Ltd. No potential conflicts of interest were disclosed by the other authors.
Authors' Contributions
Conception and design: R. Sultana, T. Abdel-Fatah, S. Chan, I.O. Ellis, S. Madhusudan
Development of methodology: R. Sultana, T. Abdel-Fatah, C. Hawkes, C. Seedhouse, G. Ball, S. Chan, S. Madhusudan
Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): R. Sultana, T. Abdel-Fatah, R. Abbotts, S. Chan, I.O. Ellis
Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): R. Sultana, T. Abdel-Fatah, N. Albarakati, G. Ball, S. Chan, I.O. Ellis, S. Madhusudan
Writing, review, and/or revision of the manuscript: R. Sultana, T. Abdel-Fatah, C. Hawkes, C. Seedhouse, G. Ball, S. Chan, E.A. Rakha, I.O. Ellis, S. Madhusudan
Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): R. Sultana, T. Abdel-Fatah, S. Chan
Study supervision: T. Abdel-Fatah, S. Chan, S. Madhusudan
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Footnotes
Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).
- Received July 24, 2012.
- Revision received November 5, 2012.
- Accepted November 28, 2012.
- ©2012 American Association for Cancer Research.