Sci. Aging Knowl. Environ., 23 April 2003
Vol. 2003, Issue 16, p. pe8
[DOI: 10.1126/sageke.2003.16.pe8]

PERSPECTIVES

Is DNA Cut Out for a Long Life?

David Sinclair

The author is in the Department of Pathology, Harvard Medical School, Boston, MA 02115, USA. E-mail: David_Sinclair{at}hms.harvard.edu

http://sageke.sciencemag.org/cgi/content/full/sageke;2003/16/pe8

Key Words: DNA repair • yeast • DNA damage • chronological aging

Fortunately for us, our cells manage to survive hundreds of potentially lethal hits each day by using an impressive array of defenses and repair mechanisms. The amount of damage inflicted is hard to appreciate because it is repaired so efficiently. For example, it is estimated that close to 10,000 purine bases in DNA are hydrolyzed per cell per day (1). Unfortunately, because of the competing priorities of growth and reproduction, no organism can afford to invest all its energy in defending itself against such assaults (2). As a result, over the years, many of our cellular constituents--DNA, proteins, and lipids-- suffer irreparable damage. This accumulating damage is now one of the prime suspects in the global search for the major causes of human aging.

Reactive Molecules Wreak Havoc

Of all the damaging agents our cells must deal with, the most ubiquitous and relentless are the molecules known as reactive oxygen species (ROS) [see "The Two Faces of Oxygen" and the Nicholls Perspective). Unlike damaging agents that originate from the environment, there is no escaping ROS because they are natural byproducts of aerobic metabolism. There is now convincing evidence that the damage caused by ROS is one of the primary causes of aging in many species. Increased life span in nematodes, flies, and mice positively correlates with an increased ability to defend against ROS (3, 4). Moreover, administration of small-molecule compounds in the worm Caenorhabditis elegans (see EUK-8 and EUK-134, but also see "Casting Doubt") and over-expression of genes in Drosophila (see SOD1 and SOD2) that help neutralize ROS can extend the life spans of these species (5, 6).

Although ROS are almost certainly relevant to aging, it is not yet clear which types of ROS damage actually cause aging. Among the three main constituents damaged by ROS, DNA is the most likely candidate. The life span of various mammalian species correlates with their capacity to repair DNA damage (7, 8). In addition, human genome instability diseases, such as Werner syndrome (see Fry Review), Rothmund-Thomson syndrome, trichothiodystrophy, and Cockayne syndrome, exhibit symptoms that resemble accelerated aging (see "Of Hyperaging and Methuselah Genes") (9). Findings such as these have lent credence to the so-called DNA damage hypothesis of aging, which has been explored in a variety of organisms. In this Perspective, I discuss these findings in light of results presented in this month's issue of Aging Cell that address the role of a specific type of DNA damage in yeast aging (10) (see McClean et al.).

Break Excision Repair Fights Back

DNA is constantly exposed to ROS, especially in mitochondria, and a single damaging event can be lethal to the cell. Over 100 different DNA lesions are caused by ROS, including base modifications, interstrand cross-links, and DNA breaks (see Shcherbakova Review). Of all the many DNA repair processes, base excision repair (BER) is considered to be one of the most important defenses against aging (11-14). Although BER mends the majority of oxidatively damaged bases and apurinic/apyrimidinic (AP) sites in nuclear DNA, its ability to repair mitochondrial DNA is likely the most relevant to aging. In mammals, there are two BER subpathways. The long-patch BER pathway results in a two- to six- nucleotide repair patch (15, 16). The enzymes involved include a DNA glycosylase; the AP endonuclease (APE-1), which cleaves the lesion to produce a 3' OH and a 5' deoxyribose phosphate (dRP); DNA polymerase {beta} (see Shcherbakova Review); the flap endonuclease (FEN-1); and DNA ligase I, which joins DNA ends. In short-patch BER, the pathway used also involves glycosylase action, but only one base is removed and the final step is catalyzed by DNA ligases I or III (Fig. 1).



View larger version (14K):
[in this window]
[in a new window]
 
Fig. 1. BER pathways in S. cerevisiae and mammals. DNA bases that become damaged by ROS can be removed enzymatically by an array of glycosylases (for example, yeast Ntg1/2, Ogg1, and Mag1) to create an AP lesion (yellow). AP lesions may also be induced chemically by oxidation or alkylation and are particularly cytotoxic. AP lesions are processed by AP lyases/endonucleases. Apn1 and APE1 are the major AP endonucleases in yeast and mammals, but they show little structural homology. In mammals, depending on the nature of the 5' moiety and the DNA substrate, either short-patch or long-patch BER is initiated, resulting in the resynthesis of one or two to six nucleotides, respectively. In short-patch repair, DNA polymerase {beta} typically inserts a single nucleotide, and DNA ligase I or a complex of DNA ligase III/XRCC1 acts to seal the final nick. Yeast apparently does not possess a ligase III, and the major function of yeast DNA polymerase {beta} is in DNA end-joining, not BER (42).

 
DNA Repair Knockouts Present Their Evidence

The DNA damage hypothesis of aging makes the following prediction: In genetically engineered animals, gene knockouts that decrease the repair of oxidative DNA damage might produce signs of premature aging, whereas animals engineered to repair oxidized DNA more efficiently might live longer. In practice, however, it has not been simple to test these predictions. Of the 120 or so DNA repair genes that have been knocked out in mice, approximately half are embryonic lethal. Another confounding problem is the considerable redundancy in mammalian DNA repair. For the most part, only single knockouts have been examined thus far. But possibly the greatest obstacle to rigorously testing the DNA damage hypothesis of aging is the predisposition of animals that carry mutations in the DNA repair machinery for cancer (17, 18). Despite these obstacles, there is mounting evidence that DNA repair defects can accelerate particular aspects of aging (9). Examples include the Ku80 knockout mouse (19), which is defective in DNA break repair and the rearrangement of immunoglobulin genes, and the TTD mouse model (20), which is defective in transcription and repairing DNA lesions by nucleotide excision repair (NER). These mice exhibit a reduced life span and accelerated bone loss and skin atrophy, among other phenotypes. Knockout mice deficient in many key BER genes, including Udg (encoding the uracil DNA glycosylase), Aag (3-methladenine glycosylase), Apex (AP endonuclease), Pol{beta} (DNA polymerase {beta}), Myh (adenine DNA glycosylase), and Ogg1 (8-oxoguanine glycosylase), have been created (18, 21). Of these, the Ogg1, Udg, ApnG, and Myh knockouts are viable. Curiously, all of these mice develop (or are currently developing) normally (21), which argues either that there is considerable redundancy in mammalian BER pathways or that the mice can tolerate a high degree of DNA damage. The Ogg1 mouse knockout is reported to accumulate 8-oxoguanine (8-oxoG) damage despite having a normal lifespan, so perhaps mice can tolerate more DNA damage than was first thought (see "Trash Cache").

Yeast: An Organism with Two Lives

Although the jury is in final deliberations on the DNA repair hypothesis of aging, we may be able to predict the verdict by studying simpler organisms such as the budding yeast Saccharomyces cerevisiae. Compared to mammals, the study of DNA repair and aging in yeast has considerably fewer complicating factors. Yeast mutants lacking key DNA repair processes are viable, and life span experiments take weeks, not years. In this month's issue of Aging Cell, Peter Piper and colleagues present experiments in yeast that test whether the types of DNA damage repaired by the BER pathway are a cause of aging in this organism (10).

The life span of S. cerevisiae can be measured in two ways (Fig. 2) (see Kaeberlein Perspective). One measure, known as replicative life span, is simply the number of divisions that an individual yeast cell undergoes before dying (22). This type of aging is caused by the inherent instability of repeated DNA, which in yeast can recombine to produce replicating circular DNAs (23-27). These circles accumulate to such high levels in old cells that they account for more DNA than the rest of the genome (22) and likely cause cell death by titrating transcription and/or DNA replication factors. Studies of yeast replicative aging have identified components of a conserved calorie restriction pathway and provided insights into the relation between environment, metabolism, silencing, and genome stability (4, 28-30) (see "High-Octane Endurance--Yeast in the Metabolic Fast Lane Live Longer".



View larger version (22K):
[in this window]
[in a new window]
 
Fig. 2. The two measures of life span in S. cerevisiae. Budding yeast divides asymmetrically, so the daughter cell emerging from its mother can be distinguished by its smaller size. The number of daughter cells produced by a mother cell is known as its replicative life span. A major cause of replicative aging is genetic instability at the repetitive ribosomal DNA locus, which is suppressed by the formation of a more condensed form of the DNA and histones (known as heterochromatin). The other measure, known as chronological life span, is the length of time that a cell survives when deprived of nutrients in a nondividing, often quiescent, state. A major cause of chronological aging is thought to be mitochondrial DNA damage by ROS. Both measures of yeast life span have proven useful in identifying functionally conserved longevity genes.

 
The other measure of yeast life span, utilized by Piper and colleagues, is the length of time that a population of cells survives when deprived of nutrients (31). This number, called known chronological life span, is ~10 to 15 days for an average strain, but it can vary considerably, depending on the growth medium and incubation temperature used (31). Some studies suggest that mitochondrial ROS damage is a major cause of chronological aging. SCH9 is a homolog of the AKT kinases that regulate life span in C. elegans as part of the insulin/insulin-like growth factor 1 pathway (32). Mutations in SCH9 greatly extend chronological life span (33), and this extension depends on Sod2, a mitochondrial superoxide dismutase that detoxifies ROS (34).

Yeast Presents Its Evidence

The S. cerevisiae BER pathway analyzed in the Maclean et al. paper (10) is well characterized and relies on a variety of DNA glycosylases. Three DNA glycosylases, Ogg1, Ntg1, and Ntg2, catalyze the majority of BER. Ogg1 primarily removes 8-oxoG lesions (16), whereas Ntg1 and Ntg2 remove a variety of oxidatively damaged bases (15, 35, 36). Mag1 is a 3-methyladenine DNA glycosylase that initiates the repair of DNA damage due to alkylation (37, 38). The action of each of these glycosylases generates an AP site, which is particularly cytotoxic (39). AP lesions can also be generated spontaneously or by ROS action. The first step in the repair of AP sites is the excision of 3' phosphate/phosphoglycolate groups by one of two AP endonucleases, Apn1 and Apn2 (39-41). Apn1, a member of the exonuclease IV family, catalyzes 95% of all AP incisions in yeast (42). In contrast, there is no apparent Apn1 homolog in humans, and almost all of all the AP incisions are catalyzed by the exoIII/Apn2 homolog APE2. Excellent reviews of BER are available [see Shcherbakova Review and (15, 42)].

The overall experimental design of Piper, Mclean, and colleagues was straightforward: determine the chronological life span of strains lacking or overexpressing critical BER genes. The absence of any one of the DNA glycosylase genes was found to have little or no effect on chronological life span. The double and triple mutants, however, had greatly reduced life spans. A similar result was obtained for the two AP endonuclease genes. The main conclusions to be drawn from these findings are that (i) related BER enzymes in yeast show considerable redundancy, and (ii) BER is necessary for cells to realize a full chronological life span. A secondary conclusion is that NER, which can also remove oxidized bases, is unable to compensate for a serious defect in BER in chronologically aging cells. Surprisingly, Piper and colleagues found no increase in the spontaneous generation of respiratory-deficient cells in the BER mutant strains, implying that BER mutants die from damage to nuclear, not mitochondrial, DNA. This finding casts doubt on the relevance of BER in yeast aging, which is almost certainly caused by mitochondrial damage (43).

Simple Yeast Has Its Drawbacks, Too

In the present study, Piper and colleagues run into the potential drawbacks of using yeast as a model for aging. The authors concede that the direct analysis of damaged DNA in chronologically aged cells can be technically challenging, in part because it is difficult to separate the few live from the many dead cells. An alternative approach adopted in this paper was to analyze mutation rates indirectly in cells carrying a conditionally lethal reporter gene. Chronologically aged cells were plated back onto rich medium, and the number of survivors indicated the frequency of loss-of-function mutations in the reporter gene. One caveat of this approach is that oxidatively damaged bases may not be lethal in diploid cells as long as the cells remain quiescent, but they can become problematic and potentially lethal if struck by a DNA replication fork after cells reenter the cell cycle. Therefore, a large number of potential survivors may be missed by the replating assay. Perhaps this explains the surprising finding that the decreased life span of the mag1 mutant does not correlate with an increase in mutation frequency. Given that differentiated mammalian cells rarely reenter the cell cycle, conclusions drawn from damaged cells that have reentered the cell cycle should be made with caution.

Piper and colleagues typically grow their yeast strains under conditions that optimize life span, in order to increase the likelihood that they are really studying aging. This is a subtle but crucial point. Aging is a process that occurs in the majority of individuals in a species (44). Precautions have to be taken to avoid the risk of studying life spans that are shortened because of cell sickness (which could be strain-specific) rather than bona fide aging (22). For studies of S. cerevisiae, this is particularly important, because it is currently impossible to distinguish a sick yeast cell from one experiencing accelerated chronological aging.

Have We Reached a Verdict?

A true test of any hypothetical cause of aging is that its suppression should extend life span. For example, if it could be shown that up-regulating a specific DNA repair pathway extends life span, then DNA damage would be exposed as a culprit. When Maclean et al. (10) applied this test to the BER pathway in yeast by overexpressing the APN1 gene, they found that it did not extend chronological life span. This is good evidence that DNA damage repaired by BER does not cause chronological aging. Of course, it is possible that overexpressing other DNA repair genes will extend life span, either singly or in combination. But clearly the jury will be deliberating a while longer.


April 23, 2003
  1. T. Lindahl, Instability and decay of the primary structure of DNA. Nature 362, 709-715 (1993).[CrossRef][Medline]
  2. T. B. Kirkwood, R. Holliday, The evolution of ageing and longevity. Proc. R. Soc. Lond. B. Biol. Sci. 205, 531-546 (1979).[Medline]
  3. E. Migliaccio, M. Giorgio, S. Mele, G. Pelicci, P. Reboldi, P. P. Pandolfi, L. Lanfrancone, P. G. Pelicci, The p66shc adaptor protein controls oxidative stress response and life span in mammals. Nature 402, 309-313 (1999).[CrossRef][Medline]
  4. S. Hekimi, L. Guarente, Genetics and the specificity of the aging process. Science 299, 1351-1354 (2003).[Abstract/Free Full Text]
  5. S. Melov, J. Ravenscroft, S. Malik, M. S. Gill, D. W. Walker, P. E. Clayton, D. C. Wallace, B. Malfroy, S. R. Doctrow, G. J. Lithgow, Extension of life-span with superoxide dismutase/catalase mimetics. Science 289, 1567-1569 (2000).[Abstract/Free Full Text]
  6. H. Ruan, X. D. Tang, M. L. Chen, M. L. Joiner, G. Sun, N. Brot, H. Weissbach, S. H. Heinemann, L. Iverson, C. F. Wu, T. Hoshi, M. A. Joiner, High-quality life extension by the enzyme peptide methionine sulfoxide reductase. Proc. Natl. Acad. Sci. U.S.A. 99, 2748-2753 (2002).[Abstract/Free Full Text]
  7. G. A. Cortopassi, E. Wang, There is substantial agreement among interspecies estimates of DNA repair activity. Mech. Ageing Dev. 91, 211-218 (1996).[CrossRef][Medline]
  8. E. Wang, A. Wong, G. Cortopassi, The rate of mitochondrial mutagenesis is faster in mice than humans. Mutat. Res. 377, 157-166 (1997).[Medline]
  9. V. A. Bohr, Human premature aging syndromes and genomic instability. Mech. Ageing Dev. 123, 987-993 (2002).[CrossRef][Medline]
  10. M. J. Maclean, R. Aamodt, N. Harris, I. Alseth, Erling Seeberg, M. Bjørås, P. Piper, Base excision repair activities required for yeast to attain a full chronological life span. Aging Cell 2, 93-104 (2003). http://www.blackwellpublishing.com/abstract.asp?ref=1474-9718&vid=2&iid=2&aid=3&s[CrossRef][Medline]
  11. D. C. Cabelof, S. Yanamadala, J. J. Raffoul, Z. Guo, A. Soofi, A. R. Heydari, Caloric restriction promotes genomic stability by induction of base excision repair and reversal of its age-related decline. DNA Repair (Amsterdam) 2, 295-307 (2003).[CrossRef][Medline]
  12. T. Stevnsner, T. Thorslund, N. C. de Souza-Pinto, V. A. Bohr, Mitochondrial repair of 8-oxoguanine and changes with aging. Exp. Gerontol. 37, 1189-1196 (2002).[CrossRef][Medline]
  13. V. Bohr, R. M. Anson, S. Mazur, G. Dianov, Oxidative DNA damage processing and changes with aging. Toxicol. Lett. 102-103, 47-52 (1998).
  14. N. C. de Souza-Pinto, B. A. Hogue, V. A. Bohr, DNA repair and aging in mouse liver: 8-oxodG glycosylase activity increase in mitochondrial but not in nuclear extracts. Free. Radic. Biol. Med. 30, 916-923 (2001).[CrossRef][Medline]
  15. P. M. Girard, S. Boiteux, Repair of oxidized DNA bases in the yeast Saccharomyces cerevisiae. Biochimie 79, 559-566 (1997).[Medline]
  16. P. M. Girard, N. Guibourt, S. Boiteux, The Ogg1 protein of Saccharomyces cerevisiae: a 7,8-dihydro-8-oxoguanine DNA glycosylase/AP lyase whose lysine 241 is a critical residue for catalytic activity. Nucleic Acids Res. 25, 3204-3211 (1997).[Abstract/Free Full Text]
  17. L. B. Meira, A. M. Reis, D. L. Cheo, D. Nahari, D. K. Burns, E. C. Friedberg, Cancer predisposition in mutant mice defective in multiple genetic pathways: Uncovering important genetic interactions. Mutat. Res. 477, 51-58 (2001).[Medline]
  18. P. Hasty, J. Campisi, J. Hoeijmakers, H. van Steeg, J. Vijg, Aging and genome maintenance: Lessons from the mouse? Science 299, 1355-1359 (2003).[Abstract/Free Full Text]
  19. H. Vogel, D. S. Lim, G. Karsenty, M. Finegold, P. Hasty, Deletion of Ku86 causes early onset of senescence in mice. Proc. Natl. Acad. Sci. U.S.A. 96, 10770-10775 (1999).[Abstract/Free Full Text]
  20. J. de Boer, J. O. Andressoo, J. de Wit, J. Huijmans, R. B. Beems, H. van Steeg, G. Weeda, G. T. van der Horst, W. van Leeuwen, A. P. Themmen, M. Meradji, J. H. Hoeijmakers, Premature aging in mice deficient in DNA repair and transcription. Science 296, 1276-1279 (2002).[Abstract/Free Full Text]
  21. E. C. Friedberg, L. B. Meira, Database of mouse strains carrying targeted mutations in genes affecting cellular responses to DNA damage. Version 4. Mutat. Res. 459, 243-274 (2000).[Medline]
  22. D. A. Sinclair, Paradigms and pitfalls of yeast longevity research. Mech. Ageing Dev. 123, 857-867 (2002).[CrossRef][Medline]
  23. D. A. Sinclair, L. Guarente, Extrachromosomal rDNA circles--a cause of aging in yeast. Cell 91, 1033-1042 (1997).[CrossRef][Medline]
  24. K. Ashrafi, S. S. Lin, J. K. Manchester, J. I. Gordon, Sip2p and its partner Snf1p kinase affect aging in S. cerevisiae. Genes Dev. 14, 1872-1885 (2000).[Abstract/Free Full Text]
  25. R. M. Anderson, K. J. Bitterman, J. G. Wood, O. Medvedik, H. Cohen, S. S. Lin, J. K. Manchester, J. I. Gordon, D. A. Sinclair, Manipulation of a nuclear NAD+ salvage pathway delays aging without altering steady-state NAD+ levels. J. Biol. Chem. 277, 18881-18890. (2002).[Abstract/Free Full Text]
  26. P. A. Defossez, R. Prusty, M. Kaeberlein, S. J. Lin, P. Ferrigno, P. A. Silver, R. L. Keil, L. Guarente, Elimination of replication block protein Fob1 extends the life span of yeast mother cells. Mol. Cell 3, 447-455 (1999).[CrossRef][Medline]
  27. M. Kaeberlein, M. McVey, L. Guarente, The SIR2/3/4 complex and SIR2 alone promote longevity in Saccharomyces cerevisiae by two different mechanisms. Genes Dev. 13, 2570-2580 (1999).[Abstract/Free Full Text]
  28. R. M. Anderson, K. J. Bitterman, J. G. Wood, O. Medvedik, D. A. Sinclair, Nicotinamide and Pnc1 govern lifespan extension by calorie restriction in S. cerevisiae. Nature , in press.
  29. S. J. Lin, P. A. Defossez, L. Guarente, Requirement of NAD and SIR2 for life-span extension by calorie restriction in Saccharomyces cerevisiae. Science 289, 2126-2128 (2000).[Abstract/Free Full Text]
  30. H. A. Tissenbaum, L. Guarente, Increased dosage of a sir-2 gene extends lifespan in Caenorhabditis elegans. Nature 410, 227-230 (2001).[CrossRef][Medline]
  31. V. D. Longo, P. Fabrizio, Regulation of longevity and stress resistance: A molecular strategy conserved from yeast to humans? Cell. Mol. Life Sci. 59, 903-908 (2002).[CrossRef][Medline]
  32. C. Kenyon, A conserved regulatory mechanism for aging. Cell 105, 165-168 (2001).[CrossRef][Medline]
  33. P. Fabrizio, F. Pozza, S. D. Pletcher, C. M. Gendron, V. D. Longo, Regulation of longevity and stress resistance by Sch9 in yeast. Science 292, 288-290 (2001).[Abstract/Free Full Text]
  34. P. Fabrizio, L. L. Liou, V. N. Moy, A. Diaspro, J. SelverstoneValentine, E. B. Gralla, V. D. Longo, SOD2 functions downstream of Sch9 to extend longevity in yeast. Genetics 163, 35-46 (2003).[Abstract/Free Full Text]
  35. L. Eide, M. Bjoras, M. Pirovano, I. Alseth, K. G. Berdal, E. Seeberg, Base excision of oxidative purine and pyrimidine DNA damage in Saccharomyces cerevisiae by a DNA glycosylase with sequence similarity to endonuclease III from Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 93, 10735-10740 (1996).[Abstract/Free Full Text]
  36. S. Senturker, P. Auffret van der Kemp, H. J. You, P. W. Doetsch, M. Dizdaroglu, S. Boiteux, Substrate specificities of the Ntg1 and Ntg2 proteins of Saccharomyces cerevisiae for oxidized DNA bases are not identical. Nucleic Acids Res. 26, 5270-5276 (1998).[Abstract/Free Full Text]
  37. W. Xiao, B. L. Chow, L. Rathgeber, The repair of DNA methylation damage in Saccharomyces cerevisiae. Curr. Genet. 30, 461-468 (1996).[CrossRef][Medline]
  38. W. Xiao, V. Penugonde, G. H. Rank, The MAG1* 3-methyladenine DNA glycosylase gene is closely linked to the SPT15 TATA-binding TFIID gene on chromosome V-R in Saccharomyces cerevisiae. Yeast 10, 687-691 (1994).[CrossRef][Medline]
  39. J. Y. Masson, D. Ramotar, Normal processing of AP sites in Apn1-deficient Saccharomyces cerevisiae is restored by Escherichia coli genes expressing either exonuclease III or endonuclease III. Mol. Microbiol. 24, 711-721 (1997).[CrossRef][Medline]
  40. S. C. Popoff, A. I. Spira, A. W. Johnson, B. Demple, Yeast structural gene (APN1) for the major apurinic endonuclease: homology to Escherichia coli endonuclease IV. Proc. Natl. Acad. Sci. U.S.A. 87, 4193-4197 (1990).[Abstract/Free Full Text]
  41. R. E. Johnson, C. A. Torres-Ramos, T. Izumi, S. Mitra, S. Prakash, L. Prakash, Identification of APN2, the Saccharomyces cerevisiae homolog of the major human AP endonuclease HAP1, and its role in the repair of abasic sites. Genes Dev. 12, 3137-3143 (1998).[Abstract/Free Full Text]
  42. M. R. Kelley, Y. W. Kow, D. M. Wilson III, Disparity between DNA base excision repair in yeast and mammals: Translational implications. Cancer Res. 63, 549-554 (2003).[Abstract/Free Full Text]
  43. V. D. Longo, C. E. Finch, Evolutionary medicine: From dwarf model systems to healthy centenarians? Science 299, 1342-1326 (2003).[Abstract/Free Full Text]
  44. D. A. Sinclair, in The Encyclopedia of Aging , R. Abruzzi, Ed. (Macmillan, New York, 2002).
Citation: D. Sinclair, Is DNA Cut Out for a Long Life? Sci. SAGE KE 2003, pe8 (23 April 2003)
http://sageke.sciencemag.org/cgi/content/full/sageke;2003/16/pe8








Science of Aging Knowledge Environment. ISSN 1539-6150