Although we do not know the targets of Sir2 silencing in the genomes of multicellular organisms, Sir2 proteins may help demarcate active and inactive regions that determine cell type. This surmise is strengthened by the finding that a murine Sir2 also can function as an NAD-dependent histone deacetylase (Imai et al. 2000). Any gradual loss in silencing would lead to an erosion of the required chromatin landscape, perhaps resulting in an alteration of phenotype such as dedifferentiation (Ono and Cutler 1978) or even cell death. These changes may play a causative role in the progressive deterioration of vitality during aging.
Is there any evidence that changes in chromatin structure may lie at the heart of aging in mammals? The cloning of animals by the “reprogramming” of adult cell nuclei in oocytes (Wilmut et al. 1997; Wakayama et al. 1998) may be relevant in this regard. Studies in sheep and mice, although still rather preliminary, give no evidence that the clones will show accelerated aging or a reduced life span. Assuming that the cloning process does not strongly select for “young” cells in the adult soma, any aging-related changes that have occurred in adult cells are reversed by the reprogramming in oocytes. This reprogramming may well be the resetting of the chromatin structure to a zygotic landscape, which, in turn, resets the aging clock. Any irreversible changes in the DNA of adult soma, for example, deletions or other mutations, could not be reset by cloning and are therefore not likely causes of aging. That said, it has been repeatedly observed that DNA mutations do occur over time in adult soma (Melov et al. 1995; Vijg 2000). These mutations may not contribute to any aging phenotype, but it is possible that they will cause other anomalies in cloned animals, such as higher cancer rates.

How might silencing be lost over time?

If, in fact, a loss of silencing can cause aging, it begs the question of how this loss might occur. Here, I present a speculative model that takes into account the second function of Sir2 proteins—as mediators of DNA repair. I suggest that Sir2 is chronically recruited from silenced chromatin to sites of DNA damage to aid their repair (Fig. 5). After the DNA has been repaired, Sir2 would return to its prior residence. However, if this mobilization and resetting of Sir2 proteins were <100% efficient, there would be a gradual erosion in the integrity of silenced chromatin over time. The image that comes to mind is the futility of trying to repack what once was a perfectly packed suitcase at the end of a trip. As discussed above, any erosion in silencing could lead to inappropriate gene expression and phenotypes of aging.