After an exposure to ionising radiation, cells can quickly repair damage to their genomes; however, a few unrepairable DNA double-strand breaks (DSBs) emerge in the nucleus in a prolonged culture and perpetuate as long as the culture continues. These DSBs may be retained forever in cells such as non-dividing ageing tissues, which are resistant to apoptosis. We show that such unrepairable DSBs, which had been advocated by the classical target theory as the ‘radiation hit’, could account for permanent growth arrest and premature senescence. The unrepairable DSBs build up with repeated irradiation, which accounts for an accumulated dose. Because these DSBs tend to be paired, we propose that the untethered and ‘torn-off’ molecular structures at the broken ends of the DNA result in an alteration of chromatin structure, which protects the ends of the DNA from genomic catastrophe. Such biochemical responses are important for cell survival but may cause gradual tissue malfunction, which could lead to the late effects of radiation exposure. Thus, understanding the biology of unrepairable damage will provide new insights into the long-term effects of radiation.
Many repair and signalling proteins are recruited to the site of double-strand breaks (DSBs) where they form ionising radiation induced foci (IRIF) (Rogakou et al., 1999; van Gent et al., 2001; Mailand et al., 2007; Shao et al., 2009). The formation and resolution of IRIF represent the repair process of DSBs. Although unrepairable DSBs are the most important type of damage for elucidating the effects of radiation on living organisms, little is known about the dynamics of how these damages are generated and affect cell fate. We believe that an understanding of the long-term, biological effects of unrepairable DSBs is crucial because the majority of the in vivo functions of cells in tissues are under non-dividing and terminally differentiated conditions and these cells exhibit radioresistance, i.e. apoptosis resistance. This evidence indicates that damaged cells can survive for a long time and may eventually affect tissue function. We used irradiated, human diploid fibroblasts as a model to define long-lasting, large repair foci as a ‘mark’ of unrepairable DSBs. Then, we investigated the functional consequences and molecular mechanisms of these foci. Our results show that the damages induce two important cellular responses: permanent growth arrest and premature senescence. Our results also demonstrate that unrepairable DSBs are the fundamental factors that underlie the classical target theory in radiation biology (Crowther, 1926; Atwood and Norman, 1949).
We further illustrate that the accumulated dose in cells and tissues can be estimated by measuring the number of persisting, unrepairable foci. Microscopic observation revealed that unrepaired foci tended to exist in pairs, which suggested that the two broken ends from each break had separated from each other and that the chromatin had torn off or become untethered. This process occurred at the unrepaired DSB sites and resulted in the alteration of heterochromatin structure and gene silencing at the broken ends. This novel characteristic of unrepairable DSBs may have important implications in the late effects of radiation on cells.
For better reproducibility of the experiments, we used automated systems to score IRIF (supplementary material Fig. S1 shows the macro program and actual foci scanning data). To avoid the ambiguous, small foci that are frequently observed in non-irradiated, cycling cells (Ichijima et al., 2005), we used young, quiescent (G0) normal human diploid fibroblasts (NHDFs). For the majority of experiments, the quiescent NHDFs were prepared by culturing the cells with MEM that was supplemented with 0.1% FCS for more than 1 week. Under such conditions, cells could be maintained for longer than 1 year without affecting their growth potential.
Persistent repair foci form after exposure to ionising radiation
Ionising radiation causes the rapid accumulation of phosphorylated H2AX (γH2AX) at the sites of DSBs (Rogakou et al., 1999; Shao et al., 2009), and the formation of γH2AX foci can be detected within 5 min after irradiation. The number of IRIF per nucleus reaches a plateau ∼0.5–1 hr after exposure, at which point the 53BP1 and phosphorylated ataxia-telangiectasia mutated (p-ATM) proteins are also colocalised to the foci (Fig. 1A). Subsequently, the foci begin to disappear with a half-life of 0.5–1 hr during the first few hours (Fig. 1D, upper panel); however, a small fraction of these foci persist, even after an overnight incubation (Fig. 1B), and perpetuate for prolonged periods (Fig. 1D, upper panel). In a 1-month culture of 6-Gy-irradiated cells (6-Gy/1-month), each nucleus carried one or two large foci, which persisted as long as the culture was continued (at least up to 6 months). The foci contained all of the components that were previously reported to colocalise in IRIF, e.g. γH2AX, p-ATM, MDC1, 53BP1, Ser15-p-p53, Rap80 and polyubiquitin. Thus, when we stained the foci with immunocytochemistry-compatible antibodies, it was found that the persistent IRIF had no differences from those that formed rapidly (Fig. 1C). Because quiescent cells lack replication activities, the repair of DSBs solely depends on the non-homologous end-joining (NHEJ) pathway. Indeed, no BRCA1 foci were observed in quiescent cultures, whereas in growing cell populations (i.e. with 10% FCS), we detected BRCA1-containing IRIF in ∼30% of the nuclei, which correlated to the labelling index of exponentially growing cells (data not shown). Therefore, BRCA1 colocalises and fulfils its function only in replication-mediated homologous recombination repair (HR) in S phase.
To examine the roles of HR in the formation of persistent IRIF, we examined the repair kinetics of irradiated cells when under growing conditions (i.e. MEM that contained 10% FCS). We confirmed that the decay of IRIF was comparable to that of quiescent cells (Fig. 1D, lower panel). Specifically, we found that irradiation with 6 Gy had completely suppressed cell growth via the activation of the cell-cycle checkpoint, which consequently forced the young NHDF cells into a state that was similar to quiescence. The persistent IRIF may represent unrepairable DSBs, which activate growth arrest signals regardless of the cellular conditions, and if so, that result could indicate that one or two persistent IRIF, i.e. unrepairable DSBs, are sufficient for the cells to undergo permanent growth arrest. In our experiments, because the formation of one IRIF after one hour of irradiation required 28 mGy, and 6 Gy of irradiation resulted in one or two persistent foci in each nucleus after 1 month, we concluded that ∼1% of the IRIF became unrepairable.
The role for unrepairable IRIF in cell survival
Microscopic observations showed that persistent IRIF were larger and more brilliant than repairable foci. A size distribution of 53BP1 foci in the nuclei indicates that cells that were irradiated with 6 Gy and cultured for 30 days (1 month) had a 1.5-times larger diameter and a 2-fold larger area than 1-Gy/1-hr cells (Fig. 2A; supplementary material Fig. S2A–C, we show the changes in the size distribution of the foci for up to 70 days after 6-Gy irradiation). The size distribution of the repairable foci (1-Gy/1-hr) showed typical, normal distribution (Fig. 2A, red bars, and their numbers were, on average, 35.7/nucleus at 1-Gy/1-hr). The unirradiated background, naturally occurring γH2AX/53BP1 foci (0.34 foci/nucleus) exhibited the same size and they were also included in the 1-Gy/1-hr histogram. By assuming that the normal distribution was also applicable to the persistent IRIF [Fig. 2A, green bars (6-Gy/30-day)], we extracted the fraction and calculated the number of unrepairable DSB foci. The result from this analysis indicates a linear correlation between radiation doses and the number of unrepairable DSBs (Fig. 2B). If we assume that one unrepairable DSB is sufficient to cause cell death (i.e. replicative death and permanent growth arrest), then the mean dose that causes one such IRIF corresponds to a mean lethal dose (D0) in the formula that is used in the Target Theory. Fig. 2B shows that the mean dose to cause one unrepairable focus per cell was estimated to be 3.2 Gy, which led us to predict that the exposure of cells to 3.2 Gy caused 37% cell survival (i.e. e−1). The dose that caused two unrepairable foci was 5.5 Gy, which led us to predict a survival level of 13.5%. Similarly, a hypothetical cell survival curve was drawn from Fig. 2B and was superimposed onto actual human cell clonogenic survival curves, which are in general agreement with the survival curves that were observed for several human cell strains (Fig. 2C). These results suggest that the unrepairable DSBs, which are observed here as perpetuating large IRIF, play a crucial role in human cell death, which is defined as the loss of mitotic activity to form a colony. Therefore, the old theory of radiation biology (Crowther, 1926; Atwood and Norman, 1949) now comes back to life.
Unrepairable IRIF account for an accumulated dose of radiation and result in paired structures
To better understand the repair competency of cells that contain unrepairable DSBs, 6-Gy/31-day cells were challenged with 1 Gy of irradiation and analysed for the clearance of the newly arisen IRIF. The results show that the cells were competent in repairing newly generated IRIF. Approximately 35 new foci were formed per cell; however, these foci quickly disappeared with kinetics that were similar to those of the cells that received only an initial dose of 1 Gy (Fig. 3A,B). The cells appeared to return to the 6-Gy/31-day status within a few days. However, we hypothesised that a fraction of the newly arisen IRIF might eventually form unrepairable DSBs, which might result in the accumulation of unrepairable foci. To test this hypothesis, 4-Gy/14-day cells were irradiated with 4 Gy and cultured for another 7 or 15 days. As expected, the cells accumulated unrepairable IRIF that were comparable with cells that were irradiated with a single dose of 8 Gy (Fig. 3C). Therefore, our results indicate that unrepairable IRIF accumulate in the nuclei of non-dividing cells, which implies that unrepairable damage accumulates in tissue in vivo, especially in cells with long life spans that are resistant to apoptosis, such as cells in the ageing brain, pancreas and kidney.
Although the biochemical characteristics of unrepairable IRIF remain elusive, these foci often appear paired, as if they had originated from a single DSB (i.e. in even numbers per nucleus): they are often paired, similar in their respective sizes and are separated. For example, Fig. 3D shows the distribution of unrepaired foci per nucleus of 6-Gy/30-day cells (in supplementary material Fig. S5, we show serial photo images from 6-Gy/30-day nuclei). Cells with two and four foci are overrepresented and those with 1 and 3 foci are underrepresented when compared to the expected foci from a Poisson distribution. Because the two partner foci appear to be distributed more than several micometers apart, which is a great distance at the molecular scale, one interpretation of the nature of unrepairable IRIF is that the broken chromosome ends are disconnected and physically separated. A recent report showed that acentric chromosome fragments are tethered by protein linkers and can be delivered normally into two daughter cells upon cell division (Royou et al., 2010; Lammens et al., 2011). Therefore, the physical disconnection of the broken ends of DNA might be the true nature of unrepairable IRIF.
Persistent accumulation of foci components at unrepairable DSBs
Equilibrium signalling that involves ubiquitin metabolism at DSB sites is a critical determinant for IRIF formation and resolution (Stewart et al., 2009; Messick and Greenberg, 2009). In that context, the continued binding of phosphorylated Rap80 (p-Rap80), which is under the control of p-ATM, to K63-polyubiquitylation substrates at unrepaired DSBs may prevent the access of deubiquitylating enzymes (Shao et al., 2009). If unknown protein(s) participate in this regulation, one of these proteins might be a key factor that could distinctly characterise unrepairable DSBs. Polyubiquitylation is maintained at high levels in unrepairable IRIF (Fig. 1C), and pretreatment of cells with the polyubiquitylation inhibitor MG132 (Mailand et al., 2007; Murakawa et al., 2007; Shao et al., 2009) results in the formation of IRIF that lacks 53BP1 immediately after irradiation (Fig. 4A). However, MG132 has no inhibitory effect on the formation of γH2AX foci; therefore, the drug has no inhibitory effect on the formation of γH2AX foci in newly generated DSBs. Surprisingly, we observed a rapid disappearance of 53BP1 immunofluorescence, which had a half-life of 1 hr, at unrepairable IRIF after treatment with MG132 (Fig. 4B). These results indicate that even after a significant period (e.g. 1 month) subsequent to irradiation, 53BP1 accumulation at IRIF requires the continuous polyubiquitylation of substrates. Moreover, inhibition of the feedback signals also caused an impairment of γH2AX accumulation at unrepairable IRIF (Fig. 4B). Consequently, the inhibition of IRIF maintenance caused the cells to become more sensitive to radiation (Fig. 4C).
Several key observations support the idea that DSBs are present at persistent IRIF. (1) Phosphorylation of Ser2056 of DNA-PKcs (DNA-dependent protein kinase catalytic subunit) is essential for the onset of NHEJ at DSBs, which is a hallmark of DSB repair (Brugat et al., 2010). We found that the phosphorylated form of DNA-PKcs accumulated at the persistent IRIF (Fig. 4D). (2) Treatment of cells with the ATM inhibitor KU55933 completely abolished the unrepairable IRIFs that were formed in 12-Gy/6-month cells; however, the foci recovered and paired again after the inhibitors were removed (Fig. 5; supplementary material Fig. S3A). (3) L189 is a recently developed metabolic inhibitor of DNA ligase 4, which sensitises cells following radiation exposure (Chen et al., 2008). If persistent IRIF consist of unrepairable DSBs, then the treatment of cells with this drug should inhibit radiation-induced NHEJ, and therefore, increase the frequency of unrepairable DSBs. In fact, treatment of cells with 400 µM L189 increased the number of unrepairable IRIF by 30% (supplementary material Fig. S3B). (4) Transfection of cells with a siRNA that targets the DNA ligase 4 mRNA increased the number of unrepairable IRIF by 40% (supplementary material Fig. S3C). The above results indirectly indicate that the persistent, broken DNA ends are localised at the IRIF. However, the status of all of the persistent IRIF could not be precisely substantiated. Therefore, cells are forced to continue consuming energy to maintain the architecture of the persistent IRIF. Because such IRIF will not be ameliorated in any way, the biological response to process might be an underlying cause of the early decline in tissue function.
The role of unrepairable DSBs in the induction of premature senescence
Unrepairable damage continuously activates checkpoint mechanisms that result in the growth arrest of cells. Even in the presence of 10% FCS in the culture medium (i.e. normal culture), irradiation with 6 Gy resulted in the permanent growth arrest of young NHDFs. Apart from serum starvation-induced quiescence, this growth-arrested state is reminiscent of the cellular conditions that occur under replicative senescence in the Hayflick model. In this case, even cells at an early passage may display a phenotype of premature senescence if they carry unrepairable DSBs. Indeed, 6-Gy/1-month cells at 20 population doublings (PDs 20) exhibited a senescent phenotype (i.e. SA β-gal-positive; Fig. 6A). Conversely, we found that nearly all senescent cells (PDs 70) carried large γH2AX/53BP1 foci (Fig. 6B). In particular, senescent cells that carried large foci were all SA β-gal positive (Fig. 6C). Furthermore, a close association has been indicated between radiation-inducible senescence and the formation of unrepairable DSBs (Sedelnikova et al., 2004). The results from our study do not necessarily indicate that unrepairable foci are the cause of senescence because the senescent cells are permanently growth arrested and could have accumulated rare, spontaneous DNA damage during long-term culture.
The localisation of DSB foci on an eroded telomere has been reported in cells that are undergoing replicative senescence. In M-phase analyses of pre-senescent replicating cells, two-thirds of the γH2AX foci are telomeric (Nakamura et al., 2008), although there are some contradictory observations (Ksiazek et al., 2007). Since the 6-Gy/1-month cells at PDs 20 never start to divide even after the serum stimulation, analyses of the unrepairable foci in the M-phase chromosomes are not applicable. We therefore performed G1 cell telomere fluorescence in situ hybridisation (FISH) along with 53BP1 foci staining. Unrepairable IRIFs, which we observed in non-replicating cells, do not appear to be colocalised with telomeric signals (Fig. 6D; supplementary material Fig. S4B; note that strong 53BP1 signals leak into the green filter, but these signals do not match telomeric signals). Our results might not prove the non-colocalisation of the two signals, since erosion of telomere, if it occurs, may hamper the hybridisation of telomere probes. However, our observation seems reasonable because the NHDFs used in this study are at low PDs (young cells) and have long telomeres, and no cell division occurred after the irradiation. Because radiation causes DSBs to occur randomly throughout the genome, the formation of repair foci and eroded telomeres should be less associated with our experiments, unless radiation specifically induces telomere erosion in non-dividing young cells. Radiation does not specifically hit the chromosome ends, therefore the quick erosion of telomere sequences following the irradiation without cell division will not happen. Indeed, radiation could not reduce both the numbers of telomere FISH signals as well as their fluorescence intensities (supplementary material Fig. S4C). Moreover, colocalisation of TRF1 and 53BP1 signals, which has been demonstrated as dysfunctional telomeres in senescent cells (Benson et al., 2010), did not seen in our young quiescent cells bearing unrepairable DSBs (supplementary material Fig. S4D). Enforced expression of telomerase ameliorates the dysfunctional telomeres thereby reduces telomere–DSB crisis signals (Benson et al., 2010); however, the telomerase expression did not reduce the radiation-induced unrepairable DSBs (supplementary material Fig. S4E). Unrepairable DSBs might have a specific molecular structure, i.e. cluster damages, that prevent the access of repair enzymes, or broken ends that are distally separated, as previously mentioned (see Fig. 3D).
Another aspect of cellular senescence is the emergence of genome-wide epigenetic changes. The formation of facultative heterochromatin foci (Narita et al., 2003; Bandyopadhyay et al., 2007; Sedivy et al., 2008), which are known as the senescence-associated heterochromatin foci (SAHF), occurs on each chromosome (Zhang et al., 2007). SAHF contain HP1 proteins, which modulate histone modifications and epigenetic silencing. Although the SAHF and unrepairable IRIF are seemingly disparate in their numbers and nuclear localisation (supplementary material Fig. S4A), we found that HP1β proteins (Fig. 6E) accumulated in unrepairable IRIF, which indicates that facultative heterochromatin is formed anew at the unrepairable IRIF. Moreover, we observed mono-ubiquitylation of H2A, which is a hallmark of histone silencing that depends on ATM (Shanbhag et al., 2010), at most of the unrepairable IRIF (supplementary material Fig. S3D). Therefore, the chromatin structure at unrepairable DSBs may be inhibitory to the processes of DNA repair. There are two waves of γH2AX protein accumulation in meiotic X-Y chromosome pairing: DSB-dependent accumulation of γH2AX for HR and then, damage-independent accumulation of γH2AX with mono-ubiquitylated H2A for gene silencing at unsynapsed chromosomes (Turner, 2007). Thus, a dual role for γH2AX at DSBs could be postulated. The accumulation of γH2AX with mono-ubiquitylated H2A at unrepairable DSBs may imply that gene silencing occurs at the broken ends of DNA. However, we cannot formally exclude the possibility that unrepairable DSBs exclusively consist of those that initially occurred at the heterochromatic region of the chromosome, and therefore, HP1β remains in the foci until the damage is repaired. DSB repair mechanisms differ between euchromatin and heterochromatin in ATM function and HP1 isoforms (Goodarzi et al., 2008; Dinant and Luijsterburg, 2009). However, the fact that almost all the unrepairable IRIF contained HP1β suggests their common roles in causing unrepairable damage. As in the case of the telomere, end-masking of the DNA double strand is critically important for the cells to avoid genomic catastrophe. Therefore, it is reasonable to assume that the formation of facultative heterochromatin contributes to the silencing of the region around unrepairable DSBs, which could be a reasonable requirement for cells to survive and live long.
In this study, we measured and characterised radiation-induced, perpetuating IRIF as unrepairable DSBs. In addition, we presented the relevance of these foci to cell fate. One of the key findings from our study is the formation of separated and stable pairs of unrepairable IRIF with heterochromatic structures at each broken end. Terminally differentiated cells and tissues are radiation-resistant because they are non-dividing (i.e. no occurrence of replicative death) and have a very long half-life (non-apoptotic). Under those circumstances, cells will survive as long as the unrepairable broken ends are stably maintained. We have confirmed this finding in irradiated NHDFs. The damage accounts for the target of the ‘radiation hit’ that was referred to in the classical theory.
Our findings also provide a way of using unrepairable DSBs to measure damage in tissues that have accumulated a radiation dose, such as previously irradiated, archived specimens, and for evaluating the involvement of radiation in tissue aging. We identified persistent, large foci in 6-Gy/1-month tissues such as the mouse pancreas (supplementary material Fig. S2D,E). In mice, pancreatic cells have a very long life span (i.e. the turnover time is estimated to be ∼500 days, and thus, the cells are nearly quiescent) (Wiktor-Brown et al., 2008). Therefore, the late effects of radiation exposure on the function of this tissue may relate to unrepairable damage. However, our strategy has some caveats that are partly due to the autofluorescence of tissue components, which causes higher background levels. In this regard, the measurement of only large foci does not seem sufficient; however, the development of specific antibodies against unrepairable IRIF will aid in alleviating this problem.
The association of DSBs with senescence is also suggested by the overproduction of ROS. MTH1 knockdown results in premature aging due to the accumulation of ROS-induced damage (8-oxoG) that eventually produces multiple DSBs during S phase (Rai et al., 2009). With regards to oncogene-induced senescence, the expression of an activated RAS or BRAF in non-tumour cells results in enforced multiple rounds of replicon firing during S phase, and thereby, produces excess DSBs, which eventually causes cells to undergo premature senescence (DiMicco et al., 2006; Bartkova et al., 2006). However, because the function of the senescence marker SA β-gal is still unknown, it remains to be determined whether the observations in this study are primary events that are related to senescence. Nevertheless, we have learned, even in recent years, that senescent cells in vitro maintain their ability to repair DNA damage (Mayer et al., 1986). Our results in Figs 3 and 4 are in line with the notion that cells are functionally active but cannot continue to divide in the presence of unrepairable DSBs. Although residual, unrepaired DSB/foci have also been reported in cells that were treated with low doses of radiation (Grudzenski et al., 2010) or mild replication stress (Harrigan et al., 2011; Lukas et al., 2011), these foci are distinct from our observations with regards to their dependence on radiation doses and growth conditions. We have observed the formation of unrepairable DSB foci and the induction of premature senescence in non-replicating, irradiated cells.
Studies on DSB repair have received considerable attention from various fields of the biomedical sciences because the components of the IRIF have been assigned as tumour suppressor proteins and checkpoint regulators, and new factors are still in the process of being identified. Currently, more emphasis has been placed on the analyses of the repair mechanisms of repairable DSBs, but these types of unrepairable damages are still elusive. We know that radiation effects result from unrepairable damages. Thus, studying the biology of unrepairable DSBs is opportune.
Materials and Methods
Cells, culture conditions and X-ray irradiation
HCA2 cells (normal human skin diploid fibroblasts; NHDF) (Noda et al., 1994) were used for the majority of this study. The cells were made quiescent by incubating them with MEM+0.1% FCS for at least one week (quiescent culture). Such quiescent cells can be maintained for at least one year with weekly changes of the medium (A.N., unpublished results). After a week of induced quiescence, cells were X-irradiated (at 1 Gy/min), and the quiescent culture was maintained until fixation and immunostaining. For the irradiation of exponentially growing cells, semi-confluent cultures were incubated with MEM+10% FCS (normal culture), and these cells were maintained in the same media after irradiation. Senescent cells were made by culturing HCA2 and Jas 3 cells (Noda et al., 1994) until the completion of their in vitro replicative life spans (ca. PDs 70), and then the culture continued for 1–2 months (less than a 0.1% labelling index). Clonogenic cell survival for X-rays (i.e. survivals determined by the colony formation of irradiated cells) was examined using various human cell lines in normal culture: HCA2, HCA2 TERT-neo (telomerase integrated HCA2), WI26VA4 (SV40 immortalised WI26), HeLa S3, and HT1080 (human sarcoma). AT2KY (ataxia telangiectasia) fibroblasts were also examined to show the survival index of typical, repair-deficient cells.
Immunohistochemistry and IRIF analyses
Cells were fixed and stained using standard protocols. Briefly, cells were plated on coverslips in 35-mm dishes and made quiescent or allowed to grow exponentially. The cells were irradiated using X-rays followed by post-irradiation culturing for the indicated periods. The cells were then fixed with 4% formaldehyde/PBS, treated with 0.5% Triton X-100/PBS, and blocked with 10% non-immune goat serum (Invitrogen). The primary antibodies that were used were as follows: anti-γH2AX [mouse monoclonal JBW301 (Millipore-Upstate) or rabbit monoclonal 20E3 (Cell Signaling)], anti-phospho-ATM [mouse monoclonal 10H11.E12 (Cell Signaling)], anti-53BP1 [rabbit A300-272A (Bethyl)], anti-MDC1 [rabbit A300-051A (Bethyl)], anti-Rap80 [rabbit A300-763A (Bethyl)], anti-poly-ubiquitin [mouse monoclonal FK2 (Nippon Biotest)], anti-HP1 [mouse monoclonal 1MOD-1A9 (Millipore)], anti-DNA-PKcs [phospho-S2056, rabbit ab18192 (Abcam)], anti-ubiquityl-histoneH2A [mouse monoclonal E6C5, (Millipore)]. The secondary antibodies that were used included the Alexa Fluor 488 goat anti-mouse IgG (Invitrogen), TRITC goat anti-rabbit IgG (Jackson), or Alexa Fluor 488 goat anti-rabbit IgG (Molecular Probes). Images were taken and incorporated into the Image-Pro software to automatically select and count cell nuclei and nuclear IRIF. The macro-program for detecting and counting the IRIF under the specified foci diameter and fluorescence is shown in supplementary material Fig. S1.
Proteosome, ATM and ligase 4 inhibitors
MG132, a polyubiquitylation inhibitor (Boston Biochem), was added 2 hr before irradiation to a final concentration of 5 µM. The drug concentration was sustainable for short-term culture (up to 2–3 days); however, a lower concentration was required for prolonged culture to examine radiosensitisation of the cells. KU99533, an ATM inhibitor (Tocris) was used at a concentration of 10 µM. L189, a DNA ligase 4 inhibitor (Chen et al., 2008) (Tocris), was used at a concentration of 400 µM and was added 2 hr before irradiation. The siRNA sequence that targets human DNA ligase 4 was as follows: 5′-GGAGATGTATATAAATACT-3′ and induced an ∼75% knockdown of human ligase 4 mRNA (data not shown).
SA β-gal assay
The senescent phenotype was detected using a commercial kit (Sigma CS0030). After X-gal staining, the cells were treated with 0.5% Triton X-100/PBS for 5 min on ice and then processed for immunostaining, as described above.
Immunofluorescence and telomere FISH of quiescent cells
Quiescent, 6-Gy/1-month cells were treated with trypsin/EDTA, the cell suspension was attached onto a glass slide by Cytospin. This attachment was followed by a short fixation with 4% formaldehyde/PBS for 10 min, and then the slide was stained with anti-53BP1. Shortly after staining, the slide was subjected to telomere FISH with Alexa488-OO-(CCCTAA)3 PNA probes, as described by Cesare et al. (Cesare et al., 2009).
We thank H. Oomine for technical assistance and Dr Evan Douple for kind comments and discussion.
This publication was supported by the Radiation Effects Research Foundation (RERF) [Research Protocol RP A4-09 to all the authors]; and Grant-in-Aid for Scientific Research from the Ministry of Education, Sports, Science and Technology (MEXT) of Japan [s-21221003 to A.N. and H.M]. The Radiation Effects Research Foundation (RERF), Hiroshima and Nagasaki, Japan is a private, non-profit foundation funded by the Japanese Ministry of Health, Labour and Welfare (MHLW) and the U.S. Department of Energy (DOE), the latter in part through DOE Award DE-HS0000031 to the National Academy of Sciences.
Note added in proof
After the submission of this manuscript, a subset of our cultures survived longer than 1 year after irradiation. We found that 6-Gy/1-year quiescent cells still retained unrepairable, persistent foci, as shown in supplementary material Fig. S6.
Supplementary material available online at http://jcs.biologists.org/lookup/suppl/doi:10.1242/jcs.101006/-/DC1
- Accepted June 6, 2012.
- © 2012. Published by The Company of Biologists Ltd