**2. Radiation-induced genomic instability**

It has been well established that ionizing radiation induces delayed effects in the progeny of surviving cells (Little, 2003, Lorimore et al., 2003, Morgan et al., 1996, Suzuki et al., 2003). This phenomenon is now called radiation-induced genomic instability, which is manifested as the delayed expression of various radiation effects, such as delayed reproductive death, delayed chromosomal instability, and delayed mutagenesis (Figure 1). Radiation-induced genomic instability has been commonly observed in many cell culture systems as well as in various animals (Lorimore et al., 2003, Morgan, 2003). In addition, there are a series of studies showing that radiation-induced genomic instability is attributed to transgenerational effects in mice using hypervariable minisatellite sequences, which have been renamed as expanded simple tandem repeats (Niwa, 2006). Radiation-induced genomic instability results in accumulating gene mutations and chromosomal rearrangements in addition to the direct genome damage caused by the primary radiation exposure. Therefore, it has been thought to play a pivotal role in accelerating the process of radiation-induced carcinogenesis (Huang et al., 2003, Kadhim et al., 1992, Niwa, 2003, Suzuki, 1997).

Fig. 1. Radiation-induced genomic instability.

Because radiation-induced genomic instability is induced in a certain fraction of the progenies originated from a single survived cell, not a single gene mutation but some epigenetic changes could be involved in the initiation of radiation-induced genomic instability (Wright, 2010). Persistently elevated levels of oxidative stress was found is association with radiation-induced genomic instability (Azzam et al., 2003, Kim et al., 2006a, Limoli et al., 2003). Further association was confirmed in studies where chronic hydrogen peroxide treatment initiated instability. Administration of free-radical scavengers, antioxidant treatment, and reducing oxygen tension each reduced delayed chromosomal instability, suggesting a role of oxidative stress in perpetuating genomic instability (Roy et al., 2000). Moreover, unstable clones showed mitochondrial dysfunction, which might explain elevated levels of oxidative stress (Kim et al., 2006b). An inflammatory response has also been proposed to be involved in radiation-induced genomic instability particularly in bone marrow (Wright, 2010). Furthermore, possible involvement of bystander effects has also been discussed. Bystander effects are responses observed in unirradiated cells as a

manifestation of radiation effects and the integrity of the genome in the cells surviving

It has been well established that ionizing radiation induces delayed effects in the progeny of surviving cells (Little, 2003, Lorimore et al., 2003, Morgan et al., 1996, Suzuki et al., 2003). This phenomenon is now called radiation-induced genomic instability, which is manifested as the delayed expression of various radiation effects, such as delayed reproductive death, delayed chromosomal instability, and delayed mutagenesis (Figure 1). Radiation-induced genomic instability has been commonly observed in many cell culture systems as well as in various animals (Lorimore et al., 2003, Morgan, 2003). In addition, there are a series of studies showing that radiation-induced genomic instability is attributed to transgenerational effects in mice using hypervariable minisatellite sequences, which have been renamed as expanded simple tandem repeats (Niwa, 2006). Radiation-induced genomic instability results in accumulating gene mutations and chromosomal rearrangements in addition to the direct genome damage caused by the primary radiation exposure. Therefore, it has been thought to play a pivotal role in accelerating the process of radiation-induced carcinogenesis

Because radiation-induced genomic instability is induced in a certain fraction of the progenies originated from a single survived cell, not a single gene mutation but some epigenetic changes could be involved in the initiation of radiation-induced genomic instability (Wright, 2010). Persistently elevated levels of oxidative stress was found is association with radiation-induced genomic instability (Azzam et al., 2003, Kim et al., 2006a, Limoli et al., 2003). Further association was confirmed in studies where chronic hydrogen peroxide treatment initiated instability. Administration of free-radical scavengers, antioxidant treatment, and reducing oxygen tension each reduced delayed chromosomal instability, suggesting a role of oxidative stress in perpetuating genomic instability (Roy et al., 2000). Moreover, unstable clones showed mitochondrial dysfunction, which might explain elevated levels of oxidative stress (Kim et al., 2006b). An inflammatory response has also been proposed to be involved in radiation-induced genomic instability particularly in bone marrow (Wright, 2010). Furthermore, possible involvement of bystander effects has also been discussed. Bystander effects are responses observed in unirradiated cells as a

radiation exposure will be discussed.

**2. Radiation-induced genomic instability** 

Fig. 1. Radiation-induced genomic instability.

(Huang et al., 2003, Kadhim et al., 1992, Niwa, 2003, Suzuki, 1997).

result of receiving signals from irradiated cells (Mothersill and Seymour, 2004, Prise and O'Sullivan, 2009). A variety of responses have been described including DNA damage induction, chromosomal instability, and cell death. As bystander effects have been observed in coculture of irradiated and unirradiated cells, and after the transfer of medium from irradiated cells to unirradiated cells, secreted factor(s) may be involved in transducing the bystander signals (Sowa and Morgan, 2004). It has been hypothesized that increased secretion of transforming growth factor beta results in stimulation of production of reactive oxygen species through a membrane NADPH oxidase. In fact, previous study demonstrated that transforming growth factor beta increased oxidative stress through decreased activity of mitochondrial complex IV (Kim et al., 2006b).

Although oxidative stress is surely involved in perpetuation of radiation-induced genomic instability (Azzam et al., 2003, Coates et al., 2008, Miller et al., 2008, Kim et al., 2006a, Limoli et al., 2003, Wright, 2007), alternative mechanisms could be associated with manifestation of radiation-induced genomic instability in non-hematopoietic cells. We have shown that delayed unscheduled induction of DNA double strand breaks is involved in the manifestation of delayed phenotypes (Suzuki et al., 2003). In fact, our study indicated that increased phosphorylated histone H2AX foci, which correspond to DNA double-strand breaks, were frequently detected in the progeny of normal human diploid cells surviving Xrays. Moreover, delayed reactivation of p53 in response to DNA damage was manifested in the surviving clones. Delayed induction of DNA double strand breaks was also confirmed by delayed induction of chromosomal aberrations (Toyokuni et al., 2009). Thus, it is evidenced that induction of DNA double strand breaks is induced indirectly in surviving cells from exposure to radiation, indicating that DNA repair pathways could play roles in amending delayed DNA double-strand breaks in surviving cells.

Previously, Chang and Little reported that radiation-induced genomic instability was absent in xrs5 cells, which are NHEJ-deficient Chinese hamster cells defective in Ku80 protein (Chang and Little, 1992a). Interestingly, delayed reproductive death was not observed in these cells, even though they harbor sufficient amount of DNA double-strand breaks. Since the mechanism of delayed reproductive death has not been fully described yet, we have hypothesized that defective NHEJ in xrs5 cells decreases the chance of mis-rejoining of the broken ends, which result in the formation of dicentric chromosomes involved in division halt. To test this possibility we examined delayed chromosomal instability in two NHEJdefective cells, xrs-5 and xrs-6 cells, and compared the frequency with the wild-type CHO cells. Furthermore, delayed induction of dicentric chromosomes was examined in cells defective in the catalytic subunit of DNA-dependent protein kinase (DNA-PKcs).

### **3. Non-homologous end-joining and its mutants**

Non-homologous end-joining is one of the two major pathways involved in amending DNA double-strand breaks in multicellular eukaryotes. It primarily plays a critical role during G0-, G1- and early S-phase of cell cycle (Lieber 2010, Polo and Jackson, 2011). Non-homologous end-joining is initiated by binding a heterodimeric protein complex consists of Ku80 and Ku70 to both ends of DNA double-strand breaks (Jackson and Bartek, 2009, Mahaney et al., 2009, O'Driscoll and Jeggo, 2006, Weterings and Chen, 2007, Wyman and Kanaar, 2006). Then, DNA-PKcs, a catalytic subunit of DNA-dependent protein kinase, is recruited to Ku-DNA complex, which results in activation of the protein kinase activity of DNA-PKcs and tethering two broken DNA ends. DNA-PKcs phosphorylates a number of proteins,

Involvement of Non-Homologous End-Joining in Radiation-Induced Genomic Instability 161

These cells were irradiated with X-rays from an X-ray generator at 150 kVp and 5 mA with a 0.1-mm copper. The dose rate was 0.44 Gy/min. Dose rates were determined with an

Since radiation-induced genomic instability is manifested as the expression of delayed effects in the progeny of surviving cells, most of the study isolated colonies formed by the cells surviving X-irradiation. The primary colonies were cloned 10 days after irradiation. The cells obtained from each colony have already passed 15 to 20 population doublings. Then, the primary clones were subjected to the secondary colony formation. After 10 day, the secondary clones were isolated, and the cells were at 30 to 35 population doubling levels after irradiation. Delayed reproductive death was examined by colony-forming ability. Colonies derived from the surviving cells often contain giant cells, which occupied an area in the colony several times greater than the rest of the cells. Colonies containing at least one giant cell were judged as giant cell-positive colony. Delayed chromosomal bridges were detected between two dividing daughter nuclei in the anaphase cells in the surviving colonies. Delayed chromosomal aberrations were examined in metaphase cells derived from the primary and the secondary clones. Both chromatid- and chromosome-type aberrations

Delayed induction of DNA double strand breaks was determined by 53BP1 foci by immunofluorescence. 53BP1 is a protein, which was originally discovered as a p53-binding protein. Lately, it turns to be clear that 53BP1 is a critical component of DNA damage signalling pathway (Polo and Jackson, 2011). ATM-dependent DNA damage checkpoint plays a central role in protecting integrity of the genome (Kastan and Bartek, 2004, Lavin, 2008, Shiloh and Kastan, 2001, Shiloh, 2003). ATM, which forms dimer or oligomer in the control cells, dissociates into monomer through autophosphorylation at serine 1981 in response to ionizing radiation (Bakkenist and Kastan, 2003). Activated ATM transduces DNA damage signal through phosphorylation of down-stream factors, including histone H2AX, MDC1 and 53BP1 (Ciccia and Elledge, 2010, Kastan and Lim, 2000)(Figure 3), and DNA damage signal is amplified through the formation of ionizing radiation-induced foci

**4. Induction of delayed phenotypes in DNA repair-deficient cells** 

were analyzed, and total 200 metaphases were counted per each sample.

Fig. 2. Survival curves of CHO cells and Ku80-deficient cells.

**4.1 Methods for detection of delayed phenotypes** 

ionization chamber.

including Ku80, Ku70, XRCC4, Artemis, DNA ligase IV and DNA-PKcs itself (Mahaney et al., 2009, Weterings and Chen, 2007). Autophosphorylation alters the conformation of DNA-PKcs, allowing the recruitment of DNA end-processing enzymes, such as MRE11 and Artemis (Huertas, 2010). MRE11 exhibits a 3' to 5' exonuclease activity and plays a critical role in homologous recombination, but its function in non-homologous end-joining is still to be determined. On the other hand, Artemis shows a 5' to 3' exonuclease activity, and it is suggested to be involved in a subset of DNA double-strand breaks. In fact, cells defective in both nucleases do not show significant radiation sensitivity. Processing of broken ends might create DNA single-strand breaks, which are amended by the DNA polymerase X family, such as polymerase mu and polymerase lambda, and terminal deoxyribonucleotidyltransferase. DNA double-strand breaks are finally rejoined by a complex composed of DNA ligase IV, XRCC4, and XLF.

Until now, several mutants defective in DNA double-strand break repair have been cloned (Jeggo, 1998). They are highly sensitive to radiation and show chromosome instability in response to DNA damaging agents. As shown in Table 1, at least four independent groups were identified. XR-1 mutant belong to X-ray repair complementing defective repair in Chinese hamster cells 4 (XRCC4) group holds mutations in the XRCC4 gene. Both xrs-5 and xrs-6 cells are the members of the XRCC5 group, and they show profound defect in the Ku80 function (Singleton et al., 1997). In xrs-5 cells, the expression of the Ku80 gene is significantly compromised. In xrs-6 cells, a 13-base pair insertion causes a truncation of KU80 protein, which accelerates degradation of KU80 protein. The responsible gene for the XRCC7 group is the DNA-PKcs (Zdzienicka, 1999). Both scid and V-3 mutants harbor premature termination of the protein in the C-terminus. In irs-20 a mutation that causes substitution of the amino acid located in the C-terminal region.



Radiation sensitivity of these mutants was determined by colony formation assay. A single cell is able to form a cluster of progenitor cells, namely, a colony, when it is incubated for 10 days or more. After X-irradiation, numbers of colonies are decreased in a dose dependent manner. Consequently, cell survival can be calculated by dividing the number of colonies after X-irradiation by the number of colonies formed by the control cells. As shown in Figure 2, both xrs-5 (closed diamond), xrs-6 (closed triangle), and scid (closed square) cells show significant reduction of cell survival as compared to CHO cells (open circle). These cells lose the shoulder of the survival curve observed in CHO cells, indicating that they lose DNA repair capacity. To compliment the defect in xrs5 cells, the human Ku86 gene was introduced by electroporation, and as a result, radiation sensitivity was significantly recovered (open diamond), confirming that a single gene mutation causes profound DNA repair defect.

including Ku80, Ku70, XRCC4, Artemis, DNA ligase IV and DNA-PKcs itself (Mahaney et al., 2009, Weterings and Chen, 2007). Autophosphorylation alters the conformation of DNA-PKcs, allowing the recruitment of DNA end-processing enzymes, such as MRE11 and Artemis (Huertas, 2010). MRE11 exhibits a 3' to 5' exonuclease activity and plays a critical role in homologous recombination, but its function in non-homologous end-joining is still to be determined. On the other hand, Artemis shows a 5' to 3' exonuclease activity, and it is suggested to be involved in a subset of DNA double-strand breaks. In fact, cells defective in both nucleases do not show significant radiation sensitivity. Processing of broken ends might create DNA single-strand breaks, which are amended by the DNA polymerase X family, such as polymerase mu and polymerase lambda, and terminal deoxyribonucleotidyltransferase. DNA double-strand breaks are finally rejoined by a

Until now, several mutants defective in DNA double-strand break repair have been cloned (Jeggo, 1998). They are highly sensitive to radiation and show chromosome instability in response to DNA damaging agents. As shown in Table 1, at least four independent groups were identified. XR-1 mutant belong to X-ray repair complementing defective repair in Chinese hamster cells 4 (XRCC4) group holds mutations in the XRCC4 gene. Both xrs-5 and xrs-6 cells are the members of the XRCC5 group, and they show profound defect in the Ku80 function (Singleton et al., 1997). In xrs-5 cells, the expression of the Ku80 gene is significantly compromised. In xrs-6 cells, a 13-base pair insertion causes a truncation of KU80 protein, which accelerates degradation of KU80 protein. The responsible gene for the XRCC7 group is the DNA-PKcs (Zdzienicka, 1999). Both scid and V-3 mutants harbor premature termination of the protein in the C-terminus. In irs-20 a mutation that causes substitution of

Table 1. Rodent cell mutants with defective non-homologous end-joining repair.

Radiation sensitivity of these mutants was determined by colony formation assay. A single cell is able to form a cluster of progenitor cells, namely, a colony, when it is incubated for 10 days or more. After X-irradiation, numbers of colonies are decreased in a dose dependent manner. Consequently, cell survival can be calculated by dividing the number of colonies after X-irradiation by the number of colonies formed by the control cells. As shown in Figure 2, both xrs-5 (closed diamond), xrs-6 (closed triangle), and scid (closed square) cells show significant reduction of cell survival as compared to CHO cells (open circle). These cells lose the shoulder of the survival curve observed in CHO cells, indicating that they lose DNA repair capacity. To compliment the defect in xrs5 cells, the human Ku86 gene was introduced by electroporation, and as a result, radiation sensitivity was significantly recovered (open diamond), confirming that a single gene mutation causes profound DNA

complex composed of DNA ligase IV, XRCC4, and XLF.

the amino acid located in the C-terminal region.

repair defect.

Fig. 2. Survival curves of CHO cells and Ku80-deficient cells.

These cells were irradiated with X-rays from an X-ray generator at 150 kVp and 5 mA with a 0.1-mm copper. The dose rate was 0.44 Gy/min. Dose rates were determined with an ionization chamber.
