WITHSTANDING FEAST AND FAMINE:Courtesy of Leonard P. GuarenteYeast form dormant spores during times of starvation in order to survive for future times of plenty. The cycle between diploid a/α cells and haploid spores is regulated by SIR2 and the related gene HST1.Since the dawn of consciousness, humans have been in a unique position to contemplate their own mortality. While this exercise has been a boon to philosophical musings, it has not led to any real scientific progress in understandin

Leonard Guarente
Apr 25, 2004

Courtesy of Leonard P. Guarente

Yeast form dormant spores during times of starvation in order to survive for future times of plenty. The cycle between diploid a/α cells and haploid spores is regulated by SIR2 and the related gene HST1.

Since the dawn of consciousness, humans have been in a unique position to contemplate their own mortality. While this exercise has been a boon to philosophical musings, it has not led to any real scientific progress in understanding aging. Yet studies over the past several years provide evidence that the aging process is regulated under certain environmental conditions, such as caloric restriction (CR). Therapies based on this knowledge may not be far behind.

In contradiction to the classical views about aging in evolution, simple organisms such as yeast or worms have mechanisms regulating aging that are easily altered through external stimuli, such as CR, or...


Current findings have led to a reevaluation of aging in the context of evolutionary theory. Classically, evolution places aging as a default mechanism; cellular and organismal processes across the board decline in an unpredictable way. This generates deterioration. Haphazard decline occurs in the postreproductive phase of life, and Darwinian selection does not effectively prevent such regression if an individual's genes have already been passed on to the next generation.

Thus, our genes have a limited shelf life, sufficient to get us through reproductive age, but in functional decline thereafter. By this reckoning, aging would have many causes, which could vary from organism to organism. Moreover, the postreproductive shortcomings of many genes would have to be remedied to slow aging.

Recent genetic studies in the budding yeast, Saccharomyces cerevisiae, and the roundworm, Caenorhabditis elegans, appear to contradict this theory. Single gene mutations can lead to a substantial life-span extension. The molecular examination of the genes in which mutations extend life span has been revealing. To wit, we found that the SIR2 gene determines the life span of both yeast cells and roundworms. In both organisms, if we add an extra copy of this gene, the life span is extended. Conversely, if we delete SIR2, life span is shortened. SIR2 genes are universally conserved from yeast to mammals. And though there's not yet evidence to support a direct role, this conservation suggests that SIR2 genes are key regulators of lifespan in a wide spectrum of eukaryotic organisms, including mammals.

The critical function of Sir2 proteins is an enzymatic one. Sir2 is an NAD-dependent protein deacetylase. Such activity is revealing, because it couples gene silencing, via local histone deacetylation, to the cell's metabolic state. Sir2 proteins cannot bind DNA themselves, but must be directed to genes by partner proteins. The genes upon which Sir2 proteins act determine the pathways affected. In different organisms, Sir2 proteins have evolved to bind different partner proteins and thereby target different genes. In yeast, these include the ribosomal DNA, which is known to determine the life span for the organism. In C. elegans, Sir2 regulates an insulin-like signaling pathway, found to be a key regulator of worm life span. In this way, SIR2 genes can regulate life span for evolutionarily divergent organisms, even if they age by different mechanisms.

The Sir2-NAD link may allow organisms to sense food availability and stall the aging process in times of deprivation. CR prolongs life span for a wide variety of organisms, including mammals, but it was previously supposed that it altered metabolism and slowed damage independently of any specific regulator. But in yeast, the ability of CR to extend the life span requires SIR2, implying that the protein mediates CR regulation.

Evolutionarily adaptive responses to CR come in many flavors. In times of starvation, both yeast and worms form specialized body types. Yeast spores and worm dauers can survive for extraordinarily long periods during famine. When conditions improve, these dormant life forms become rejuvenated and give rise to progeny. Once again, SIR2 genes control yeast spore and worm dauer development. It will be of great interest to learn whether SIR2 genes regulate dormant states in mammals, such as hibernation.

These considerations lead to a new view of how a single gene can determine the life span of an organism. SIR2 slows aging and reproduction during scarcity, so that the organism will survive to reproduce when conditions improve. This survival function would be evolutionary adaptive and therefore pervasive in nature.

So the classical evolutionary theory of aging must be modified. In times of plenty, aging plays out in accord with evolutionary theory, with reproduction followed by wholesale decline. The failure of many genes evidently leads to this decline. But in times of scarcity, the survival program kicks in to decelerate aging. And a single gene can promote this survival mechanism across a wide swath of nature's creatures.


Courtesy of Leonard P. Guarente

Sir2p's dependence on the metabolic coenzyme nicotinamide adenine dinucleotide (NAD) to deacetylate and silence genes suggests a provocative connection to observed life extension by caloric restriction.

This means the question "What causes aging?" may be a futile one with many answers (oxidative damage, short telomeres, unfolded or glycated proteins, etc.). Rather, the important question may be "What regulates aging?" because a regulator would have to slow any and all causes, under the right conditions. The answer to this latter question may well be SIR2.


Of course, for the purposes of human health, SIR2 would have to promote survival in mammals. In cultured mouse or human cells, the mammalian SIR2 ortholog, SIRT1 (one of seven known mammalian SIR2 homologs) determines cellular survival in response to DNA damage. SIRT1 dampens the tendency of mammalian cells to apoptose. The protein binds to and deacetylates the tumor-suppressor protein, p53. It likely also deacetylates the histones in the vicinity of p53 proapoptotic target genes to downregulate them. Thus, as in yeast and worms, mammalian SIRT1 promotes survival, in this case of cultured cells.

This is a "good news/bad news" finding. The good news is that we may have identified a contributor to mammalian aging: the gradual death of cells by apoptosis and the accompanying failure of organs, which can be mitigated by SIRT1. But the bad news is that by downregulating p53, SIRT1 may actually impede tumor surveillance and favor cancer. But, this scenario strikes me as unlikely for two reasons: First, if nature has evolutionarily anointed SIR2 to promote survival, it is not sensible that the gene favors cancer; second, recent experiments in genetically altered mice show that the cell-death response may be slowed without causing cancer.

I think it likely that any procancer effect from SIRT1 on p53 will be overridden by another regulatory activity of SIRT1. CR does trigger a reduction in some hormones, including growth factors. This makes it tempting to speculate that SIRT1 regulates CR by lowering levels of circulating growth factors. The concomitant downregulation of p53 in hormone-responsive cells may compensate for the altered hormonal landscape by raising the threshold for apoptosis. The net effect on the animal would then not be cancer predisposition, but stress resistance and long life.


The practical benefit of these new findings may be enormous. If single genes can determine the life span of mammals, it should be possible to develop drugs that bind to the protein products of these genes to alter their activities. In this regard, I, along with Cindy Bayley and Cynthia Kenyon, founded Elixir Pharmaceuticals (Cambridge, Mass.). The goal of Elixir is to take the recent advances in aging research and identify new drug targets in aging. Thus, drugs that slow the aging process may be on the horizon. But simply extending life may not be enough. Whether such drugs prolong youth and vitality or simply increase our stay in nursing homes is an important matter.

We can consider the effects of CR on animal health. Calorie-restricted rodents not only live longer, but are vigorous and healthy for the duration of their long lives. Moreover, diseases of aging, such as tumors in mice and kidney disease in rats, are forestalled or prevented altogether. We believe that targets like SIRT1 offer a new opportunity to pharmacologically act on a wide swath of diseases associated with aging.

The vagaries of drug development defy an accurate prediction as to when we might see such advances, but I believe that I will be taking an anti-disease-of-aging pill in my lifetime. Any life extension is likely to be modest. Based on the effects of CR in mice, and the likelihood that such a pill would be taken later in life, I would expect a benefit on the order of a decade. This is nothing to sneeze at, but far less than we have already achieved with medical intervention in the past century. More important than longevity may be the health benefits. Maintaining vigor at advanced ages would hold considerable benefits not just in quality of life but in societal costs. The future is approaching and it looks longer and better than ever before.


Leonard P. Guarente is the Novartis Professor of Biology at The Massachusetts Institute of Technology and a director at Elixir Pharmaceuticals. His recent book, Ageless Quest, chronicles the path of aging research in his lab over the past 12 years.

He can be contacted at