DNA Double-strand Breaks and Aging
umans and animals have a limited lifespan, and cell cultures derived from them cease proliferating and enter senescence after a finite number of population doublings. Oxidative stress, DNA damage, and in humans, telomere shortening are all suggested as major factors in aging at cellular levels. However, how well cellular senescence in in vitro models corresponds to aging in vivo is unclear. Our study aims at understanding the relationship between cellular and organismal aging in mammals.
Several findings have supported a role for DNA damage/repair in aging. Mice deficient in DNA repair proteins exhibit aging phenotypes earlier than wild-type mice (Celeste A et al. Science 296: 922–7, 2002). In addition, several DNA damaging agents induce a cellular senescence-like state in culture without extensive cell division (premature senescence) (Horikawa I et al. J Anti Aging Med 3: 373–82, 2000). To obtain more direct evidence for a direct link between DNA damage and mammalian aging, we immuno-stained cultures of normal human cells at different population doublings and touch prints of tissues from mice of different ages for foci of γ-H2AX, the phosphorylated form of histone H2AX that specifically marks DNA double-strand breaks (DSBs) (Rogakou EP. J Cell Biol 146: 905–16, 1999; Sedelnikova OA et al. Radiat Res 158: 486–92, 2002).
DNA DSBs were revealed in both experimental systems as γ-H2AX foci (γ-foci). In addition, the incidences of γ-foci were found to increase with senescence in vitro and age in vivo as measured by counting γ-foci in individual nuclei. In three normal human cell strains, the number of γ-foci increased with replicative aging from 0.1 to 0.3 foci per young cell to 2.2 to 4.1 foci per senescent cell. Fibroblastic and epithelial cells exhibited similar increases. Cultures subjected to premature senescence with the DSB inducer bleomycin and the oxidative stress inducer hydrogen-peroxide also exhibited increased incidences of γ-foci characteristic of more senescent phenotypes.
Confocal microscopy image showing accumulation of γ-H2AX signal in nuclei of testes cells from an old mouse (A) compared with a young mouse(B). The tissues were touch printed and processed for immunocytochemistry. Large green spots are pachytene spermatides with ongoing homologous recombination.
Studies with mice yielded similar findings (Figure 1). Touch-print immunostaining of cells from the liver, kidney, lung, brain, and testes taken from mice of different ages revealed increased incidences of γ-foci during in vivo aging similar to those found in vitro. Thus, the accumulation of γ-foci is a common process in mammalian aging in vivo and in culture.
These foci of unknown origin are named cryptogenic foci to differentiate them from those of known causes such as ionizing radiation. When cell cultures or mice were irradiated with 0.6 Gy, we observed the expected increases in the incidences of γ-foci 30 minutes post-irradiation, followed by the expected slower decreases, until at 24 hours post-irradiation, the incidences of γ-foci returned to values near the pre-irradiation values characteristic of the stage of senescence or age of the treated cells. Importantly, this finding indicates that the cryptogenic foci are stable before, during, and after the induction and disappearance of those induced by ionizing radiation. Thus, it is likely that the cryptogenic foci mark persistent and unrepairable DNA lesions.
Cryptogenic foci accumulate the DNA repair proteins, 53BP1, Mre11, Rad50, and Nbs1, indicating that they are sites of ongoing or attempted DNA repair. In addition, since repair proteins accumulate at γ-foci with different kinetics, it may be possible to differentiate cryptogenic foci and nascent foci induced by ionizing radiation by their content of DNA repair factors soon after irradiation. We observed that the accumulation of DNA repair proteins is incomplete on a subset of foci at 30 minutes post-irradiation but complete on all at 60 minutes, indicating that DNA repair factors were already present at the cryptogenic foci but in the process of accumulation at the nascent foci. Thus, this observation supports a model of two types of γ-H2AX foci: (1) transient, where successful DSB rejoining occurs, and (2) persistent, containing unrepairable DSBs. On the other hand, cryptogenic foci do not colocalize with telomeres to a significant extent, suggesting that the senescence/aging-associated DSBs in cryptogenic foci are primarily at non-telomeric sites.
This study demonstrates that during cellular senescence or organismal aging, mammalian cells accumulate persistent DNA lesions that contain unrepairable DSBs. The similar incidence of persistent γ-H2AX foci associated with cellular senescence, either prematurely induced by the exogenous agents or after cell divisions, supports a model in which the accumulation of unrepairable DSB-containing lesions may play a causal role in aging. Hence, this study establishes the physiological importance of unrepairable DSB-containing lesions in cellular and organismal aging in mammals, and raises the possibility that diverse factors that affect aging may all act ultimately through the accumulation of persistent DNA lesions containing unrepairable DSBs. One area of future interest is to determine whether γ-foci accumulate at specific chromosomal sites or structures. Our study also suggests that γ-foci may be useful markers for detecting individual senescent cells in aged mammals. We believe that our findings will lead to a better understanding of how cellular senescence contributes to organismal aging and how aging can be a major risk factor in human carcinogenesis.