Aging is an inevitable and progressive degeneration of biological functions that is universal among all living organisms. While some scientists believe that it results from the manifestation of a pre-programmed internal clock that determines life span, many others argue that aging stems from a slow and progressive loss of biological functions caused by cumulative metabolic damage. One widely held hypothesis is the "free radical theory" of aging, which states that the generation of free oxygen radicals inside cells accelerates aging through random and sequential damage to cell components (Harman, 1956). These chemically diverse oxygen species generated as byproducts of aerobic metabolism are usually grouped together as reactive oxygen species (ROS) and include molecules such as superoxide anions, hydroxyl radicals and hydrogen peroxide (Figure 1). Although traditionally ROS have been depicted as damaging to cells, evidence from recent years indicates that ROS are also essential components of normal cellular processes and may be indispensable to survival.

Direct comparison of the relative life spans of different animal species has shown that, in general, animals with short life spans have high metabolic rates, while longer-lived animals exhibit lower metabolic rates. This correlation led to the proposal of the "rate of living" hypothesis, which states that the metabolic rate of each species determines its life span. Because there is usually a good correlation between the metabolic rate of an animal and the level of ROS generated by their mitochondria, the rate of living hypothesis is generally associated with the free radical hypothesis.

Although the rate of living hypothesis fits well for most animal species, there are several animal groups that do not fit into it easily. In particular, the life spans of birds and primates appear to be much longer than predicted based on their observed metabolic rates. One explanation for this apparent contradiction is the finding that mitochondria isolated from birds and primates appear to generate fewer oxidants than mitochondria isolated from other animals with similar metabolic rates (Ku et al., 1993). More striking is the comparison of oxidant production by mitochondria isolated from two closely related mouse species that exhibit significantly different life spans: M. musculus (3.5 years) and P. leucopus (8 years). Even though P. leucopus has a higher metabolic rate, M. musculus produces significantly more oxidants than its longer-lived counterpart (Sohal et al., 1993). This demonstrates that, although ROS production and metabolic rate are often related, ROS production is perhaps the best predictor of longevity.

It is generally acknowledged that the majority of intracellular ROS production is generated in the mitochondria. The production of ROS is mostly dependent on the intrinsic metabolic rate and oxygen consumption. Since we live in an oxygenated environment, our cells have developed an impressive repertoire of strategies to detect and detoxify metabolites of molecular oxygen that could impair the organism's survival. Antioxidant enzymes are universally expressed across all species to protect cells from oxidative stress. Enzymatic ROS scavenger proteins such as superoxide dismutase (SOD), catalase, glutathione peroxidase and peroxiredoxins have been extensively studied (Dröge, 2002). Superoxide dismutase is an enzyme that speeds the conversion of superoxide to hydrogen peroxide, while catalase and glutathione peroxidase convert hydrogen peroxide to water. In addition, small non-enzymatic antioxidant molecules -- some derived from dietary sources -- are important in trapping ROS. These include ascorbate, pyruvate, flavonoids, carotenoids and glutathione (Beckman and Ames, 1998).

A delicate equilibrium between ROS production and antioxidant defenses determines the degree of intracellular oxidative stress. The consequences of long-term exposure to ROS include extensive alterations to DNA, lipids and proteins. Some studies have demonstrated that aging cells and organisms progressively accumulate increased levels of ROS-damaged nuclear and mitochondrial DNA. Mitochondrial DNA is actually more sensitive than nuclear DNA to oxidant insult since this organelle is less able to repair its damaged DNA (Dröge, 2002). Mice with defects in mitochondrial superoxide scavenging capacity show a significant increase in the frequency of mitochondrial DNA mutations. Their flawed mitochondria are thought to release additional ROS, which in turn induces further mitochondrial DNA damage (Esposito et al., 1999). This vicious ROS regulated cycle represents a potential positive feedback loop that may fuel the aging process. This concept is supported experimentally by the observation that older animals produce more mitochondrial oxidants than their younger counterparts. Besides damage to DNA, ROS can also induce damage to proteins, with levels of oxidatively modified proteins significantly increasing with age (Stadtman, 2001). Free radical attack also decreases membrane fluidity by modifying lipids via a process termed lipid peroxidation, which can significantly alter membrane properties and possibly disrupt the function of membrane-associated proteins (Beckman and Ames, 1998).

Even though an organism is clearly more than the sum of its individual cells, in vitro cell models of aging have yielded additional support linking oxidative stress and aging. Studies performed in human fibroblasts grown in low oxygen tension showed a significantly prolonged life span compared to cells grown under normal room air oxygen (Packer and Fuehr, 1977). In contrast, cells cultured in the presence of high oxygen concentrations have a reduced life span accompanied by increased ROS production. Exogenous oxidants are also known to contribute to development of a senescent phenotype. Exposure of human fibroblasts to moderate, nonlethal doses of hydrogen peroxide induces a rapid, senescence-like growth arrest. Recent experiments have also demonstrated that overexpression of an activated Ras oncogene can trigger a senescence-like state in human fibroblasts (Serrano et al., 1997). The Ras gene product is a protein involved in a number of critical signal transduction pathways in cells and is frequently mutated in human cancers. The ability of Ras to induce senescence is consistent with the notion that senescence may be part of the body's defense against cancer formation. Interestingly, human diploid fibroblasts expressing the Ras oncogene have higher levels of ROS, and these cells can be rescued from senescence by lowering oxygen concentration or treatment with an antioxidant (Lee et al., 1999). Together, these studies suggest an intimate link between intracellular ROS levels and activation of cellular senescence.

Recent technological advances have allowed scientists to analyze integrated cellular responses to oxidative stress at the genetic level using DNA microarrays. This technique provides a snapshot of all the genes that are expressed at a given time point and reveals the molecular fingerprint of a given normal or pathological state. Analysis of gene expression profiles in aging mice revealed genetic changes that suggested activation of cellular stress responses (Lee et al., 2000). Increased expression of heat shock factors and stress-induced proteins was observed in a variety of tissues. Microarray analysis of the skeletal muscle of aging rats also showed significant increases in the hydrogen peroxide scavenging enzymes catalase and glutathione peroxidase. Age-associated elevations in the expression of heat shock proteins have also been observed in model organisms such as flies and worms (Finkel and Holbrook, 2000). Although these results allow a glimpse of what aging is on a molecular level, it is hard to know if such genetic reprogramming is a cause or a consequence of the aging process.

The rate of ROS production in cells is largely determined by the availability of mitochondrial energy substrates. It is therefore not totally unexpected that caloric restriction has been found to prolong life span. Animals fed restricted amounts of a well-balanced diet also show diminished accumulation of oxidative damage that presumably stems from a lower rate of ROS production in mitochondria. In addition, there may be a concomitant increase in enzymatic antioxidant defenses and repair processes in these animals. Caloric restriction has also been shown to delay the onset of age-associated diseases, such as cancer, and to increase the ability of rodents to endure a number of physiological stresses (Sohal and Weindruch, 1996).

The evidence to support the free radical theory of aging has been further strengthened by the recent identification of specific genes influencing longevity. Links have been established in a variety of animal models that clearly correlate longevity with stress resistance. In the worm C. elegans a variety of mutants have been isolated that extend life (Guarente and Kenyon, 2000). These mutants have increased resistance to oxidative stress and, in some cases, increased antioxidant scavenging capacity. Pharmacological studies using synthetic compounds that can mimic either SOD or catalase have been shown to lengthen the life span of the worm, proving that, at least for this simple organism, reducing ROS levels prolongs life span (Melov et al., 2000). Similarly, other mutations in the worm that increase ROS levels have been shown to shorten life span (Taub et al., 1999). In flies, strains selected for prolonged life span also display increased resistance to oxidative stress, which in turn correlates with enhanced activity of antioxidant enzymes (Dröge, 2002). Taken together, these studies in lower organisms imply that ROS are critical determinants of life span.

The first evidence in mammals showing a link between resistance to oxidative stress and extended life span was described in mice carrying a mutation in the p66^shc protein. Mice that have been genetically engineered not to express p66^shc appear to live approximately 30% longer than littermates expressing this protein. Interestingly, cells without p66shc appear to be able to withstand exposure to hydrogen peroxide better than normal cells (Migliaccio et al., 1999). We have also recently demonstrated that p66^shc can actually regulate the level of ROS in cells (Nemoto and Finkel, 2002). This suggests that a connection between ROS and longevity might also be present in mammals.

In addition to the harmful effects of oxidants, a host of normal signaling pathways in cells appear to be regulated by ROS. Studies in a variety of cells have suggested that growth factors or cytokines can trigger the production of ROS. Blocking these ROS actually interferes with the normal ability of the cell to respond correctly to these signals. Indeed, the normal growth, death and differentiation signals present in cells might critically depend on the proper oxidation-reduction balance within cells (Finkel and Holbrook, 2000). These observations raise important questions as to how oxidants participate in the aging process (Figure 2). In particular, do ROS trigger aging by damaging random components of the cell such as DNA, proteins and lipids, or do oxidants function as regulators of specific pathways that are important to aging? Undoubtedly, further research is needed before such questions are answered.

To summarize, intracellular ROS are both harmful and necessary for cellular functions. Maintaining a delicate homeostatic balance between oxidant generation, antioxidant protection and repair of ROS-induced damage appears to be intimately connected to the regulation of life span. Developing pharmacological approaches that would influence the production of ROS or the pathways they regulate may assist in the prevention of aging as well as many age-related diseases. Caution is appropriate, however, since initial studies with broadly acting antioxidant vitamin supplementation have not proven to be particularly effective in preventing the onset of many age-related maladies (Yu, 1999). A wealth of laboratory studies have revealed that the response to oxidative stress is, in fact, significantly more complex than initially envisioned. Nonetheless, further studies in the area may give additional insight into how we have adapted to live under the dual edge sword of an oxygenated environment. In the end, such studies might also provide important clues as to what regulates our mortality.

Dr. Finkel is chief of the cardiovascular branch at the National Heart, Lung, and Blood Institutes of Health in Bethesda, Md.

Ms. Rovira is a biologist in his laboratory.

Beckman KB, Ames BN (1998), The free radical theory of aging matures. Physiol Rev 78(2):547-581.

Dröge W (2002), Free radicals in the physiological control of cell function. Physiol Rev 82(1):47-95.

Esposito LA, Melov S, Panov A et al. (1999), Mitochondrial disease in mouse results in increased oxidative stress. Proc Natl Acad Sci U S A 96(9):4820-4825.

Finkel T, Holbrook NJ (2000), Oxidants, oxidative stress and the biology of ageing. Nature 408(6809):239-247.

Guarente L, Kenyon C (2000), Genetic pathways that regulate ageing in model organisms. Nature 408(6809):255-262.

Harman D (1956), Aging: a theory based on free radical and radiation chemistry. J Gerontol 2:298-300.

Ku HH, Brunk UT, Sohal RS (1993), Relationship between mitochondrial superoxide and hydrogen peroxide production and longevity of mammalian species. Free Radic Biol Med 15(6):621-627.

Lee AC, Fenster BE, Ito H et al. (1999), Ras proteins induce senescence by altering the intracellular levels of reactive oxygen species. J Biol Chem 274(12):7936-7940.

Lee CK, Weindruch R, Prolla TA (2000), Gene-expression profile of the ageing brain in mice. Nat Genet 25(3):294-297.

Melov S, Ravenscroft J, Malik S et al. (2000), Extension of life-span with superoxide dismutase/catalase mimetics. Science 289(5484):1567-1569.

Migliaccio E, Giorgio M, Mele S et al. (1999), The p66^shc adaptor protein controls oxidative stress response and life span in mammals. Nature 402(6759):309-313 [see comment].

Nemoto S, Finkel T (2002), Redox regulation of forkhead proteins through a p66shc-dependent signaling pathway. Science 295(5564):2450-2452.

Packer L, Fuehr K (1977), Low oxygen concentration extends the lifespan of cultured human diploid cells. Nature 267(5610):423-425.

Serrano M, Lin AW, McCurrach ME et al. (1997), Oncogenic ras provokes premature cell senescence associated with accumulation of p53 and p16^INK4a. Cell 88(5):593-602.

Sohal RS, Ku HH, Agarwal S (1993), Biochemical correlates of longevity in two closely related rodent species. Biochem Biophys Res Commun 196(1):7-11.

Sohal RS, Weindruch R (1996), Oxidative stress, caloric restriction, and aging. Science 273(5271):59-63.

Stadtman ER (2001), Protein oxidation in aging and age-related diseases. Ann N Y Acad Sci 928:22-38.

Taub J, Lau JF, Ma C et al. (1999), A cytosolic catalase is needed to extend adult lifespan in C. elegans daf-C and clk-1 mutants. Nature 399(6732):162-166.

Yu BP (1999), Approaches to anti-aging interventions: the promises and the uncertainties. Mech Ageing Dev 111(2-3):73-87.

Surviving an Aerobic Environment: Aging Under Oxidative Stress