Introduction
Although considerable information has been amassed concerning potential risk factors and the genetic background of breast cancer, the etiology of the disease is still poorly understood [
1]. Recent evidence led to the proposal that normal stem cells may be the key cells in a tissue or organ that undergo mutation and transformation giving rise to 'cancer stem cells' [
2‐
5]. As normal stem cells are long-lived cells and the precursors to differentiated cells, DNA repair and mutation avoidance in these cells should be critical. Mutation and transformation of normal stem cells are most likely the result of DNA damage arising from exogenous and endogenous agents, including oxidative free radicals and dietary and environmental factors [
6,
7]. To counter such damage, cells possess a variety of multi-protein DNA repair pathways, each responsible for handling a class of DNA lesions [
8].
Until recently, evidence to support a direct role for an altered DNA repair response in normal and cancer stem cells was limited and mostly confined to hematopoietic cells [
9‐
11]. A study of bone marrow-derived mesenchymal stem cells, for example, identified a more efficient reactive oxygen species (ROS) scavenging capacity in these cells. Moreover, these stem cells exhibit active homologous recombination (HR) and nonhomologous end-joining (NHEJ) in the repair of double-stranded breaks to facilitate their radio-resistance [
11]. Data from studies with murine embryonic stem cells indicate that these cells efficiently repair DNA damage [
12‐
14], and that repair in embryonic stem cells may even be superior to that in differentiated embryoid bodies or embryonic fibroblasts [
13]. The spontaneous mutation frequency in murine embryonic stem cells is also significantly lower than that in differentiated embryonic fibroblasts [
15].
For cancer stem cells, mutation avoidance may be less crucial, but cell survival should be a dominant characteristic and may result in enhanced resistance to radiation and chemotherapeutic agents [
4,
16]. Thus, there is a clear need to identify the mechanisms that are involved in the maintenance of genome stability and cell survival in cancer stem cells. It has proven extremely difficult to culture sufficient numbers of stem cells from fresh solid tumor material for such studies. However, many established tumor cell lines possess a small fraction of self-renewing tumor-initiating (stem) cells that can form tumors from very few cells [
5,
17‐
19]. Studies with such glioma [
20] and breast tumor-initiating cells [
16,
21,
22], cultured as neurospheres and mammospheres, respectively, have provided some insights. In both cases the stem cells were observed to have elevated resistance to ionizing radiation, and additional data indicated enhanced DNA damage checkpoint activation [
20] and repair capacity, as well as a marked reduction in the measured level of ROS following exposure to ionizing radiation [
21], in comparison to the monolayer cell populations from which the stem cells were isolated.
To further explore the role of factors that safeguard genomic integrity and cell survival, we carried out a detailed comparison of single-strand break repair (SSBR) and double-strand break repair (DSBR), telomerase activity, and cell death in MCF-7 breast cancer cells and tumor-initiating cells isolated from the MCF-7 population as mammospheres [
21,
23]. To date most attention has focused on the apoptotic response of cancer stem cells exposed to ionizing radiation and other cytotoxic agents. Here we show for the first time that solid tumor cancer cell 'stemness' can also influence the senescence pathway.
Materials and methods
Cell culture and mammosphere preparation
The MCF-7 breast cancer cell line was obtained from the American Type Culture Collection (Manassas, VA, USA). The cells were cultured in 1:1 DMEM/nutrient mixture F12 supplemented with 10% fetal bovine serum, 50 units/ml penicillin, 50 µg/ml streptomycin, 2 mM L-glutamine, 0.1 mM nonessential amino acids and 1 mM sodium pyruvate, all from Gibco/BRL-Invitrogen (Carlsbad, CA, USA), and maintained at 37°C under 5% carbon dioxide in a humidified incubator. When cells approached confluency, the supernatant was collected and plated on ultralow attachment plates (Corning, Acton, MA, USA) in serum-free medium supplemented with basic fibroblast growth factor (Millipore, Billerica, MA, USA), epithelial growth factor and insulin (Sigma, St. Louis, MO, USA) as described previously [
23]. Mammospheres, appearing as spheres of floating viable cells, were collected by centrifugation and dissociated with 0.05% trypsin followed by mechanical dissociation with a Pasteur pipette. Cells were then sieved sequentially through a 100 µm and a 40 µm cell strainer (Falcon, San Jose, CA, USA) and plated at low density.
Flow cytometry and cell cycle analysis
Purified mammospheres were characterized by analysis of appropriate cell surface markers, CD24 and CD44, on a Coulter Epics flow cytometer (Becton Dickinson, San Jose, CA, USA). Briefly, 0.5 × 106 cells were mixed with fluorescently-tagged antibodies to the surface antigens (R-phycoerythrin-labelled anti-CD24 and fluorescein isothiocyante-labelled anti-CD44 antibodies; BD Pharmingen, San Diego, CA, USA) and incubated in the dark for 15 minutes at room temperature. Following incubation, cells were washed with PBS containing 1% FCS, centrifuged for five minutes at 2,500 rpm. After aspiration, approximately 50 μl of the cells were resuspended in PBS containing 1% FCS and underwent fluorescence activated cell sorter analysis immediately.
For cell cycle analysis, approximately 1 × 106 cells in an aliquot of 1 ml were fixed by the addition of 3 ml cold absolute ethanol. After one hour, cells were pelleted, washed twice in PBS and treated with propidium iodide/RNAse solution for 15 minutes at 37°C. Samples were placed in 12 × 75 mm Falcon tubes and analysed by a Coulter Epics flow cytometer.
Clonogenic cell survival assay
Stock cultures of MCF-7 monolayer and mammosphere cells were trypsinized, resuspended in DMEM/F12 medium supplemented with 10% FCS, and plated in 60 mm tissue culture dishes at densities of 200 to 600 cells per dish. The cells, as a single-cell suspension, were then subjected to γ-radiation. After radiation, the cells were allowed to adhere to the plates and incubated at 37°C for 18 days, after which they were fixed and stained with crystal violet. Colonies consisting of 30 or more cells were counted and the surviving fraction (in comparison to unirradiated controls) was plotted as a function of dose using PRISM software (version 5.0 GraphPad, San Diego, CA, USA)
Cell proliferation
Cellular survival of monolayer and mammospheres were measured by cell proliferation assay using CellTiter 96 AQueous One Solution Cell Proliferation Assay (Promega, Madison, WI, USA). For this assay, both monolayer and mammospheres were trypsinized and plated in a 96-well plate at a density of 2 × 103 cells per well. Monolayer cells were cultured in DMEM/F12 supplemented with 10% FCS and mammosphere media consisted of DMEM/F12 supplemented with 0.4% fetal serum albumin, 20 ng/ml basic fibroblast growth factor, 10 ng/ml epidermal growth factor, 5 μg/ml bovine insulin, and 4 μg/ml heparin. The cells were then subjected to increasing doses of γ- radiation (2 to 8 Gy) and incubated for five days at 37°C. Following incubation, 20 μl of CellTiter reagent (Promega, Madison, WI, USA) was added per well and after four hours of incubation at 37°C the absorbance was recorded at 490 nm and plotted verses radiation doses. All measurements were carried out in triplicate.
Detection of reactive oxygen species
For the detection of highly ROS we employed an aminophenyl fluorescein dye-based kit (Cell Technology Inc, Mountain View, CA, USA) according to the manufacturer's protocol. Briefly, cells were rinsed in modified Hank's balanced salt solution (10 mM 4-(2-hydroxyethyl)piperazine -N'-(2-ethanesulfonic acid), 1 mM MgCl2, 2 mM CaCl2, 2.7 mM glucose) and aminophenyl fluorescein was added (to a final concentration of 10 μM) to approximately 1.5 × 106 cells. Cells were incubated for 30 minutes in the dark and then exposed to 1, 4, or 10 Gy γ-radiation (60Co Gammacell; Atomic Energy of Canada Limited, Ottawa, Canada) and plated at 3 × 105 cells per 96-well plate. The fluorescence was immediately measured by excitation at 485 nm and reading emission at 515 nm using a fluorescence plate reader (Cell Technology Inc., Columbia, MD, USA).
Single-cell gel electrophoresis (Comet assay)
At various times after irradiation of the cells, approximately 1 × 10
5 cells were trypsinized and mixed with 1.0 % molten low melting point agarose at 42°C at a 1:10 ratio (10 μl cells per 100 μl of agarose). The mixture was then spread on a glass slide (Trevigen, Gaithersburg, MD, USA). Following solidification of the agarose, slides were immersed in prechilled lysis solution (2.5 M NaCl, 100 mM Na
2EDTA, pH 10, 10 mM Tris base, 1% SDS, 1% Triton X-100), followed by incubation on ice for 40 minutes and immersion in an alkaline solution (300 mM NaOH, 1 mM Na
2EDTA) for 30 minutes at room temperature. Slides were then placed in an electrophoresis apparatus, filled with fresh alkaline solution, and run at 1 V/cm and approximately 300 mA at 4°C for 40 minutes. The slides were washed in 70% ethanol for five minutes and the DNA was stained with SYBR Green I (Molecular Probes-Invitrogen, Carlsbad, CA, USA) and viewed with an AxioScope 2 fluorescence microscope (Carl Zeiss, Toronto, ON, Canada). For each data point we visualized and analyzed a minimum of 100 random cell images based on the 'comet' category shown by Collins [
24]. (We have changed the numbering 0 to 4 to 1 to 5).
For the neutral non-denaturing comet assay the cells were lysed as described above followed by rinsing in 50 ml 1 × TBE (89 mM Tris base, 89 mM boric acid, 2 mM EDTA) buffer. The slides were then subjected to electrophoresis at 1 volt/cm for 40 minutes in a horizontal electrophoresis apparatus, followed by fixation in 70% ethanol for five minutes and air dried. The slides were then stained and scored as for alkaline comet.
For preparation of cell lysate we followed a previously published method [
25]. For western blotting and immunohistochemistry standard protocols currently applied to breast stem cells were used [
26,
27]. Antibodies to the following proteins were obtained commercially: PCNA (monoclonal, sc-25280), XRCC1 (polyclonal, sc-11429), Chk1 (polyclonal, sc-7898), Chk2 (polyclonal, sc-9064), p21 (monoclonal, sc-6246), pRb (monoclonal, sc-103), phospho-Chk1 (polyclonal, sc-17922R), Cdc25A (polyclonal, sc-7157), MDM2 (monoclonal, sc-5304), p53 (monoclonal, sc-126), E2F1 (monoclonal, sc-251), and actin (polyclonal, sc-1616) from Santa Cruz Biotech (Santa Cruz, CA, USA); ataxia telangiectasia mutated (ATM; polyclonal, NB 100-104), apurinic/apyrimidinic endonuclease 1 (APE1; monoclonal, NB 100-116) from Novus Biologicals (Littleton, CO, USA); phospho-Chk2 (polyclonal, 2661), and phospho-p53 (sampler kit, #9919) antibodies from Cell Signaling (Boston, MA, USA); phospho-ATM (polyclonal, ab2888) and phospho-DNA-dependent protein kinase catalytic subunit (DNA-PKcs; polyclonal, ab18192) from Abcam (Cambridge, MA, USA); PNK (monoclonal, MAB005) from CytoStore (Calgary, AB, Canada); epithelial specific antigen (FITC-labeled monoclonal, MM-1014) from Immunobiosciences (Raleigh, NC, USA). Polyclonal anti-Ku80 antibody was a gift from Dr. Susan Lees-Miller (University of Calgary, Calgary, AB, Canada), antibody to polymerase-β was a gift from Dr. Sam Wilson (National Institute of Environmental Health Sciences, Research Triangle Park, NC, USA), and monoclonal antibody to ING1 was kindly provided by Dr. Karl Riabowol (University of Calgary, Calgary, AB, Canada). Proteins on western blots were visualized by enhanced chemiluminescence (Amersham Pharmacia Biotech, Arlington Heights, IL, USA) and X-Omat K film (Eastman Kodak Co., Rochester, NY, USA). Films were scanned as image files and the optical densities of the bands were quantified using ImageJ software version 1.33u (NIH, Bethesda, MA, USA).
Statistical analysis by two-tailed Student's t-test was performed with Microsoft Excel software (Microsoft Corp, Redmond, WA, USA).
For mammosphere immunostaining, we used a Cytospin centrifuge (Thermo Shandon, Pittsburgh, PA, USA) to prepare thin-layer mammospheres slides. Cells were dried and fixed by paraformaldehyde. After blocking with 5% milk powder, the Cytospin smears were stained with appropriate primary and secondary antibodies.
Imaging of γH2AX, 53BP1 and Rad51 foci
We monitored the level of H2AX phosphorylation before and after ionizing radiation by the modified protocol of Furuta and colleagues [
28]. Briefly, 1 × 10
5 MCF-7 cells were seeded onto each glass coverslip inside 35 mm tissue culture dishes. For mammospheres, however, we used a Cytospin to prepare thin-layer mammosphere slides. After incubation overnight at 37°C in a carbon dioxide incubator, the cells were exposed to 1 or 10 Gy γ-radiation and then incubated for different times at 37°C. The cells were then fixed in 2% paraformaldehyde in PBS for five minutes, followed by washes with PBS and 50:50 methanol/PBS solution, and permeabilization with 100% methanol for 20 minutes at -20°C. Cells were then blocked with 5% milk powder at room temperature and then exposed to a monoclonal antibody to anti-phospho-histone H2AX (Ser-139) (Upstate, Charlottesville, VA, USA) for one hour in the dark. After washing the coverslips with PBS, Alexa Fluor 488-labelled goat anti-mouse secondary antibody (Molecular Probes-Invitrogen, Carlsbad, CA, USA) was added to the coverslips and incubated at room temperature for one hour in the dark. The coverslips were again washed twice with PBS and PBS/Tween 20 and finally rinsed with water. The coverslips were mounted on a microscope slide using 95% glycerol in PBS containing 3 µg/ml 4',6-diamidino-2-phenylindole. Slides were stored at 4°C in the dark. Phosphorylated H2AX foci were viewed with a LSM510 laser-scanning confocal microscope (Carl Zeiss, Toronto, ON, Canada) mounted on a Axiovert 100 M microscope (Carl Zeiss, Toronto, ON, Canada). Images were taken with a ×40 (NA 1.3) objective, following the Nyquist sampling requirement, with the same instrument settings for different slides. The average integrated intensity per nucleus was determined by using METAMORPH OFFLINE 6.1 software (Universal Imaging, Downingtown, PA, USA), as described previously [
29].
The same procedure was followed to monitor 53BP1 and Rad51 foci. For the former we used an anti-53BP1 rabbit polyclonal antibody (ab36823) from Abcam (Cambridge, MA, USA), and for the latter an anti-Rad51mouse monoclonal antibody (ab213) from Abcam (Cambridge, MA, USA).
Real-time quantitative RT-PCR
Total RNA was extracted from 2 × 106 cells using TRIZOL Reagent (Invitrogen, Carlsbad, CA, USA) as per the manufacturer's instructions. RNA concentration was determined spectrophotometrically with a Life Science UV/Vis Spectrophotometer (Beckman Coulter Du 730, Brea, CA, USA).
The quantitative RT-PCR assay was performed on a 7900HT Real-Time PCR system (AB Applied Biosystem, Carlsbad, CA, USA). Total RNA (0.5 μg) from each sample and standard was amplified using Platinum SYBR Green quantitative PCR SuperMix-UDG (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's protocol. The primer sequences for APE1 gene were 5'-GCTGCCTGGACTCTCTCATC-3' (sense), 5'-GCTGTTACCAGCACAAACGA-3' (antisense) to generate a 180 bp product. The primer sequences for ING1b were 5'-CAACAACGAGAACCGTGAGA-3' (sense) and 5'-GAGACCTGGTTGCACAGACA-3' (antisense) to give a 195 bp product. For GAPDH, as control, the primer sequences were 5'-GTCTCCTCTGACTTCAACAGCG-3' (sense) and 5'-ACCACCCTGTTGCTGTAGCCAA-3' (antisense) to yield a 130 bp product. PCR conditions were as follows: 50°C for 6 minutes, 95°C for 10 minutes and 40 cycles of 95°C for 15 seconds and 60°C for 30 seconds followed by 2 minutes incubation at 60°C. The cycle threshold value for each gene was normalized by subtracting the cycle threshold value of the GAPDH control gene, and then the relative fold increase in transcript level was calculated.
Telomerase assay
The telomerase assay was performed with a TRAPeze Telomerase Detection kit (Chemicon, Temecula, CA, USA) using non-radioactive detection. Whole cell extracts were prepared from 106 unirradiated and irradiated MCF-7 monolayer cells and 106 unirradiated and irradiated mammosphere cells in CHAPS lysis buffer provided with the kit according to the manufacturer's instruction. Samples of extract equivalent to 104 cells were used in each assay. Following incubation with cell extract and PCR amplification, the samples were analyzed by polyacrylamide gel electrophoresis and ethidium bromide staining. The optical densities of the scanned bands were quantified using ImageJ software version 1.33u (NIH, Bethesda, MD, USA). Statistical analysis by one-tailed Student's t-test was performed with Microsoft Excel software (Microsoft Corp, Redmond, WA, USA).
Apoptosis and senescence-associated β-Gal assays
For the detection of apoptosis, unirradiated cells, or cells exposed to 4 Gy γ-radiation (60Co Gammacell; Atomic Energy of Canada Limited, Ottawa, Canada) three days prior to collection, were harvested and analyzed by flow cytometry for the induction of apoptosis using the BD Pharmingen-Annexin V: FITC Apoptosis Detection Kit (San Jose, CA, USA) according to the manufacturer's instructions.
Induction of senescence was observed by morphological changes detectable under light microscopy, and by looking at cellular β-galactosidase activity [
30], a known marker of senescent cells, using a Senescence Galactosidase Staining Kit (Cell Signaling, Beverly, MA, USA) according to the manufacturer's protocol. As mammospheres are non-adherent cells we used a Shandon Cytospin centrifuge (Waltham, MA, USA) to spread cells over the glass slides prior to fixation and staining.
Discussion
The data presented here demonstrate that mammospheres, as a representative of epithelial mammary cancer progenitors, are distinct from their differentiated MCF-7 cells. It has been previously shown by others that MCF-7 mammospheres are more resistant to ionizing radiation than the monolayer cells and have lower ROS content in unirradiated and irradiated cells [
21,
22], and we have corroborated their findings. Phillips and colleagues [
21] also ascertained that the observed effects were not the result of differences in the growth factor content of the media in which the two cell types were cultured. In our laboratory, the clonogenic survival assays of the two cell types was carried out under identical growth conditions post-irradiation while the cell proliferation assays were carried out under the respective optimal growth conditions for the two cell types. That both assays showed the radio-resistance of the mammosphere cells further supports the idea that this reflects the intrinsic radiosensitivity of the cells rather than an artifact of the culture conditions.
We have now established several other facets of the response of mammospheres to endogenously and exogenously induced DNA damage. It is clear that mammospheres possess efficient base excision repair/SSBR, probably mediated by an elevated level of APE1. Although base excision and SSBR are largely responsible for cellular protection from endogenous oxidative damage, they also play a role in response to radiation and antineoplastic agents. For example, in breast cancer patients receiving the radiomimetic drug, anthracycline, lower APE1 expression was associated with a better pathologic response [
56]. We also observed that mammospheres exhibit a similar DSBR response after irradiation as the monolayer cells in spite of the lack of H2AX phosphorylation, implying that mammospheres use a γH2AX-independent pathway for DSBR. The formation of 53BP1 and Rad51 foci in the irradiated mammospheres strongly suggests that both the NHEJ and HR pathways can be initiated, although further study will be required to fully elucidate the DSBR pathway utilization by these cells.
In terms of radiation-induced cell death, the most important pathway for MCF-7 cells appears to be replicative senescence but the mammospheres displayed a markedly reduced tendency to undergo senescence. The immediate molecular basis for the senescent behaviour can be attributed to the robust telomerase activity and negligible p21 expression. However, p21 expression is regulated by ATM through Chk2 and p53 [
57], and thus implicit in our findings is that reduced Chk2 phosphorylation/activation may play a significant role in the senescence phenotype of MCF-7 mammospheres.
Finally, it is to be anticipated that cellular pathways utilized by cancer stem cells to enhance survival will depend on the genetic background of the stem cells. For example, MCF-7 cells are representative of cancer cells that do not readily undergo apoptosis, and thus enhanced survival of MCF-7 cancer-initiating cells is primarily dependent on down-regulation of the senescence pathway. A similar response may be expected in other cancer stem cells, such as those with inactivating mutations in the proline-rich domain of p53, which is required for p53-dependent transactivation of key apoptotic genes but not p21 [
58,
59]. On the other hand, avoidance of apoptosis is likely to play a more critical role in the survival response exhibited by cancer stem cells that possess potential functional apoptotic activity. This may explain the differences in DNA repair/cell cycle protein expression and post-translational modification seen between the current study and Bao and colleagues [
20].
Acknowledgements
We thank Ms Ann Berg (Flow Cytometry Facility, Cross Cancer Institute) and Dr. Xuejun Sun (Cell Imaging Facility, Cross Cancer Institute) for assistance with flow cytometry and cell imaging, respectively. We also thank Jane Lee and Jodi Wilkinson for their technical assistance, Drs. Susan Lees-Miller (University of Calgary), Karl Riabowol (University of Calgary) and Sam Wilson (NIEHS) for generously providing antibodies, and Drs. Gordon Chan and Razmik Mirzayans (Dept of Oncology, University of Alberta) for useful discussions. This work was supported by grants from the Canadian Breast Cancer Foundation and Alberta Cancer Foundation to MW, FKB, and JRM.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
FKB and ARN carried out all the experiments. FKB, JRM and MW designed the experiments and prepared the manuscript.