This view unites several features of aging. Importantly, the model applies equally well to cells that are mitotic, such as skin, intestine, blood, etc., as to cells that are postmitotic, such as neurons and muscle. In the case of dividing cells, much of the DNA damage is likely generated because of errors that occur during DNA replication. Mutations in factors of the replication of repair machinery in diseases such as Werner syndrome (Yu et al. 1996), Cockayne syndrome, and ataxia telangiectasia (Savitsky et al. 1995) may exaggerate the generation of DNA damage and give rise to the apparent acceleration of aging, especially in organs with mitotic cells. In postmitotic cells, DNA damage may be generated by agents such as ROS. In the case of muscle or brain cells, the rate of respiration is especially high, which may favor the production of ROS, thereby triggering DNA damage.

Can anything be done to slow the aging process? The model outlined above suggests that interventions resulting in an increase in the capacity of Sir2 proteins to sustain silencing ought to slow aging. In yeast, an extra copy of SIR2 does slow aging and increase life span. It should be possible to carry out similar tests in experimental animals, such as Caenorhabditis elegans and mice. In humans the task is more formidable. Any novel Sir2 agonists, for example, molecules that would enter cells and mimic the effect of NAD on Sir2, might offer the promise of long-term intervention to slow the aging process. Although it seems unlikely that any drastic extension in life span is imminent, a therapeutic that compresses the period of morbidity may be more than a flight of fancy.

I thank S. Imai, B. Jegalian, B. Johnson, and H. Tissenbaum for comments on the manuscript. Work in my lab was supported by grants from the NIH, The Ellison Medical Foundation, The Seaver Foundation, and The Howard and Linda Stern Fund.