Sci. Aging Knowl. Environ., 20 October 2004
Vol. 2004, Issue 42, p. pe39
[DOI: 10.1126/sageke.2004.42.pe39]


Aging-Related Research in the "-omics" Age

Matt Kaeberlein

The author is in the Department of Genome Sciences at the University of Washington, Seattle, WA 98195, USA. E-mail: kaeber{at}

Key Words: microarray • gene expression profiling • biomarker • high-throughput screen • RNAi • calorie restriction


Welcome to the "-omics" age, where we can do science bigger, faster, better; or at least we think we can. For the past few years, biological research has been undergoing a revolution of scale; instead of studying one gene or one protein at a time, thousands, or even tens of thousands, of variables are being examined in one experiment. The application of genome-scale technologies to the study of aging has been slower than in many other fields; however, things are beginning to change. This Perspective will review some of the ways in which high-throughput technologies are being successfully developed for, and applied to, the study of aging in humans and model systems, with particular emphasis on microarrays and methods for genome-wide life-span analysis.

Microarrays and Aging

Global gene expression profiling ("transcriptomics") has matured from infancy to the point where microarrays are used to study nearly every area of molecular biology. The use of microarrays to study aging, however, has not been without its pitfalls and controversies. Recent reviews discuss several important considerations in experimental design and the myriad technical aspects of microarray applications (1-3) (see Melov Review, Lee Perspective, and Becker Perspective). Issues such as cross-platform reproducibility, tissue and cell-type heterogeneity, insufficient sample size, lack of rigorous statistical analysis, and failure to independently validate reported changes in mRNA expression have raised serious questions regarding many of the gene expression studies related to aging (3). A number of speakers at the recent Expression Array Workshop held at the University of Washington Nathan Shock Center of Excellence in the Basic Biology of Aging pointed out the need for improvement in these areas; however, there was also a high degree of optimism that the technical and design hurdles can be overcome. In fact, several recent reports, discussed in the following paragraphs, demonstrate that microarrays can be used successfully to identify both biomarkers of aging and longevity, as well as genes that determine life span.

Age-associated changes in gene expression

The most common application of microarrays to aging-related research has been in studies designed to identify gene expression changes that correlate with age, by comparing mRNA obtained from "young" individuals to mRNA from "aged" individuals (Fig. 1A). These types of studies are particularly useful for understanding the molecular changes associated with aging. It is often difficult, however, to distinguish gene expression changes that are causally related to aging and the onset of age-associated phenotypes from gene expression changes that occur as a response to age-associated pathologies. In general, it has been observed that stress response genes, particularly those involved in the oxidative stress response, show increased expression with age in both invertebrates (4-6) and mammals (7, 8). This finding is intriguing because some long-lived mutant worms also show increased expression of oxidative stress response genes as young animals (9, 10), suggesting that oxidative stress-related damage increases during normal aging, and enhanced protection against this type of stress can confer longer life.

View larger version (7K):
[in this window]
[in a new window]
Fig. 1. Strategies for using microarrays to study aging. (A) In young-versus-old designs, gene expression from young control (YC) animals is compared to expression from old control (OC) animals, and biomarkers of age are obtained. Designs in which multiple time points correspond to different chronological ages are superior to the simple two-component young-versus-old design shown here. (B) In control-versus-long-lived (or -short-lived) designs, gene expression from YC animals is compared to expression from young mutant (YM) animals with increased (or decreased) longevity. This type of design offers the opportunity to obtain biomarkers of longevity. (C) More complicated designs involve combining a comparison of age-associated changes with expression changes in models of altered longevity. In addition to YC-versus-OC and YC-versus-YM comparisons, a YM-versus-old mutant (OM) comparison is added. As well as offering the ability to detect biomarkers of age and longevity, this type of design allows the identification of gene expression changes associated with age that are attenuated in the mutant longevity model.

This type of old-versus-young comparison has been carried out in several different model systems, providing an opportunity to ask how similar the age-associated changes in gene expression are between evolutionarily divergent organisms. McCaroll et al. took a pioneering step in this direction by comparing age-associated changes in gene expression in worms and flies (11). In this way, it was discovered that these highly diverged animals share several dozen functionally related orthologous gene pairs that show similar changes in expression during early adulthood. It will be of great interest to see how well this adult-onset expression program is conserved in other organisms, particularly mammals. Similar types of analyses regarding gene expression changes associated with models of increased longevity, such as calorie restriction (CR), would be of value and are likely to become a high priority.

Longevity-associated changes in gene expression

Rather than examine gene expression changes that correlate with age, an alternative approach is to examine gene expression changes that correlate with longevity (or rate of aging) (Fig. 1B). In model systems where life-span-enhancing mutations and interventions are known (see the SAGE KE Genes/Interventions Database), this can been done by comparing mRNA from young long-lived (or possibly short-lived) organisms to mRNA from young wild-type organisms. In this way, gene expression biomarkers that correlate with longevity (but are present before age-associated changes occur) can be obtained, some of which are potentially causative for the longevity phenotype. The value of such gene expression biomarkers of longevity should not be underestimated. By comparing gene expression changes correlated with enhanced longevity in multiple models, it should be feasible to identify several highly reliable biomarkers of longevity. In principle, these biomarkers could then be used to screen for drugs that cause a similar gene expression profile and, perhaps, increased life span.

One model of enhanced longevity that has been studied in detail is provided by CR (see Masoro Subfield History). CR is the only intervention known to increase life span in yeast, worms, flies, and mammals (for example, mice and rats), and there is much interest in understanding the mechanism(s) by which CR promotes longevity. As such, many of the microarray studies related to aging have examined the gene expression changes associated with CR. Several studies in mice have combined microarray analysis of CR versus control fed animals with a young-versus-old comparison (Fig. 1C), resulting in the discovery that a subset of age-associated gene expression changes are attenuated or reversed by CR (8). Other studies have compared gene expression changes associated with CR to gene expression changes associated with genetic mutations that increase life span (12-15) (see "Common Ingredients"). A smaller-scale analysis of CR-induced gene expression changes was recently carried out in long-lived Ames dwarf mice, with the observation that (i) CR and dwarfism result in different gene expression changes, and (ii) dwarf mice respond differently to CR than do normal mice (16). Although perhaps somewhat surprising on the surface, this result fits well with genetic data from worms and mice indicating that CR can further increase the life span of long-lived insulin/insulin-like growth factor-1 (IGF-1) pathway mutants (17-19) (see "Dieting Dwarves Live It Up"). Whether CR increases life span by a different mechanism in dwarf animals than in control animals will require additional study.

Microarrays have also been used to study the temporal response of animals to CR. In old mice, gene expression changes in the liver that correlate with long-term CR are observed within 2 to 8 weeks of initiating a CR-fed regimen (20), which is consistent with the observations that adult-onset CR results in a rapid decrease in mortality in both flies (21) and mice (20). Although it is still unclear which, if any, of the specific gene expression changes detected in response to CR are responsible for increased longevity, the ability to detect short-term gene expression changes associated with CR suggests that it might be feasible to use gene expression biomarkers as a screening method for compounds that mimic the physiological state associated with CR. The identification of drugs that increase life span by mimicking CR but do not require reduced caloric intake would clearly be of immense potential value.

Forging the link between microarrays and life span

Microarray analysis can provide useful information regarding gene expression changes that correlate with age or longevity, but tell us nothing about whether these changes are functionally related to the aging process. In two important studies carried out in worms, microarrays were used to identify genes that act downstream of the DAF-16 transcription factor (which functions in the insulin/IGF-1 pathway) to determine life span (9, 10) (see Larsen Perspective). What makes these studies so impressive is that they represent the first examples in which gene expression changes have been used as a discovery method to identify putative longevity-determining genes, followed by validation of the findings by determination of the life span of animals in which expression of the identified genes was down-regulated. Using clever experimental designs to take advantage of the genetic relationship between the insulin-like/IGF-1 tyrosine kinase receptor DAF-2 and its downstream target DAF-16 (which is under the negative control of DAF-2 signaling), these groups were able to identify genes that showed opposite expression profiles in animals that were long-lived because of down-regulation of daf-2 expression by RNA-mediated interference (RNAi) relative to animals that were short-lived because of down-regulation of both daf-2 and daf-16 by RNAi. Subsequent life-span analysis was then used to demonstrate a causal role for several of these genes in determining longevity, including genes involved in oxidative stress response, antimicrobrial response, and protein turnover.

The Search for Long Life Goes Genome-Wide

High-throughput longevity screens

In addition to gene expression microarrays, several other high-throughput technologies are making an impact on aging-related research. Although yet to come of age, the use of proteomic and metabolomic technologies to measure changes in protein and metabolite concentrations associated with age and longevity will undoubtedly become important in the future (see Gibson Perspective). Other methods such as genome-wide single-nucleotide polymorphism typing and genome-wide linkage mapping are already being used successfully to identify genes important for longevity in humans (22-24). The remainder of this Perspective will focus on new technologies with the potential to greatly accelerate the identification of environmental and genetic interventions that increase life span in model systems.

Genetic screens to identify mutations that increase life span are nothing new in the field of aging-related research. Technological advances, however, have made it possible to assay secondary phenotypes associated with longevity more directly in a high-throughput manner. Recent reports describing studies of three different organisms illustrate novel applications of high-throughput methodologies to identify mutations and small molecules that increase life span. In yeast, a screening method for cells with increased replicative life span has been developed, in which magnetic sorting of old mother cells is coupled with the use of a fluorescence-activated cell sorter to estimate the aging potential of a cell population (25). This method relies on the binding of a fluorescent molecule to the scars left on a mother cell after budding, which represent the number of daughters produced by that cell. The sensitivity and ability of this method to reflect replicative life span accurately remains to be demonstrated; however, if successful, it will be feasible to screen thousands of mutations or chemical compounds rapidly for those that increase life span. In worms, automated fluorescence-based sorting has also been used to develop a high-throughput survival assay that relies on the uptake of a fluorescent dye by dead animals (26). In flies, a different approach has been taken to accelerate the life-span assay by coupling expression of a toxin protein to a known biomarker of age, so that short-lived flies (which express the toxin relatively early in life) die earlier than usual, allowing long-lived flies to be identified comparatively quickly (27) (see "Hastening Death to Delay Aging"). As with the yeast method, the general utility of each these latter two methods remains to be determined. Nonetheless, these reports represent exciting progress, and the continued development of these and other promising new technologies should greatly accelerate our ability to identify new genes and compounds that slow aging and increase life span.

Genome-wide longevity phenotyping

Global gene expression studies and other high-throughput screening methods for longevity-associated secondary phenotypes have the potential to greatly aid and accelerate our understanding of the molecular causes of aging. It is important to keep in mind, however, that in the absence of phenotypic validation, these studies can only show us correlative, not causative, effects. Too often, the inclination is to assume a causal relation with longevity or age-associated disease. It is worth stating explicitly: The only method for demonstrating that an observed change in gene expression (or change in protein concentration, metabolite concentration, or stress response) is causing the longevity phenotype is to undertake independent phenotypic validation of altered life span. Given that life span is the ultimate phenotype for aging-related studies, the ability to carry out life-span assays in a high-throughput manner would be of immense value. Clearly, in mammals that have life spans measured in years or decades, high-throughput longevity phenotyping is not likely to be feasible. In simple eukaryotes such as yeast, flies, and worms, however, recent technological advances will allow true genomic-scale exploration of life-span determinants. In fact, with the advent of RNAi, genome-wide longevity phenotyping in worms has already become a reality.

RNAi is a technology used to decrease expression of a specific gene by selective degradation of target mRNA. A library of Escherichia coli strains carrying RNAi sequences corresponding to nearly 17,000 unique genes (~85% of the predicted open reading frames in the Caenorhabditis elegans genome) was created (28) and has been partially screened to identify genes that increase life span when expression is reduced (29) (see Melov Perspective). Based on this analysis, it was discovered that reduced expression of several genes important for mitochondrial function can increase life span (29, 30); many genes that limit life span were also identified. Thus far, a genome-wide approach to life-span analysis has not been described for flies. The recent release of a Drosophila RNAi library by Ambion and Cenex, however, is likely to make this type of analysis a reality in the near future.

In yeast, no counterpart to RNAi is known; however, the Saccharomyces Genome Deletion Project (31) has generated a collection of over 6000 strains, each of which carries a single gene-deletion mutation. This collection covers more than 96% of the yeast genome. Taking advantage of this resource, we have recently developed new technologies that will allow the determination of both replicative and chronological life-span phenotypes for every strain contained in this collection (32). With the discovery that Sir2 and CR act in parallel to promote longevity in yeast (33) (see "Calorie Restriction Un-SIR-tainty"), we anticipate the identification of genes that function in both of these pathways, and perhaps in as yet unidentified pathways, to regulate aging in this simple eukaryote. This study will provide a genome-wide longevity data set for yeast that is directly comparable to the genome-wide longevity data obtained from RNAi screens in worms and flies, and will allow for the large-scale identification of conserved regulators of eukaryotic aging (Fig. 2). A few such genes are already known (Fig. 3), and it seems certain that many others are waiting to be discovered. Orthologous gene families that regulate aging in yeast, worms, and flies are likely to share a similar function in vertebrates and will make excellent candidates for further study in mammals.

View larger version (19K):
[in this window]
[in a new window]
Fig. 2. Identification of evolutionarily conserved aging-related genes through genome-wide screens in simple eukaryotes. Genome-wide RNAi screens in flies and worms and analysis of the yeast deletion collection for genes that increase life span will yield comparable sets of genes from each organism that increase life span when function is decreased. Orthologous genes present in more than one set (and perhaps all three sets) are potential conserved regulators of longevity. The role of these genes in mammalian aging can then be studied by knocking out the orthologous gene in mice.


View larger version (5K):
[in this window]
[in a new window]
Fig. 3. Potential conserved aging-related genes. Orthologous genes known to increase life span in more than one organism when expression is altered are indicated. (See Genes/Database entries for yeast genes SIR2, SCH9, RPD3, SOD1, and SOD2; worm genes sir-2.1, sgk-1, daf-2, and let-363; fly genes RPD3, InR, SOD1, and SOD2; and mouse genes Ig1r and Insr.)


One of the most exciting aspects of aging-related research over the past few years has been the growing recognition that many of the same processes determine aging and longevity in different organisms. Microarray studies are beginning to identify conserved transcriptional changes associated with aging, as well as molecular signatures associated with life-span extension by CR and longevity-enhancing mutations. The recent trend toward technology development directed at high-throughput screens for increased life span signals a growing recognition that a molecular understanding of the factors that regulate aging is achievable. The application of these new technologies may ultimately provide a path toward intervention into the human aging process.

October 20, 2004
  1. P. N. Nair, T. Golden, S. Melov, Microarray workshop on aging. Mech. Ageing Dev. 124, 133-138 (2003).[CrossRef][Medline]
  2. E. S. Han, Y. Wu, R. Mccarter, J. F. Nelson, A. Richardson, S. G. Hilsenbeck, Reproducibility, sources of variability, pooling, and sample size: Important considerations for the design of high-density oligonucleotide array experiments. J. Gerontol. A Biol. Sci. Med. Sci. 59, 306-315 (2004).
  3. F. N. Hudson, M. Kaeberlein, N. Linford, D. Pritchard, R. P. Beyer, P. S. Rabinovitch, in Handbook of the Biology of Aging, E. J. Masoro, S. N. Austad, Eds. (Academic Press, San Diego, CA, ed. 6, in press), chap. 12.
  4. S. Zou, S. Meadows, L. Sharp, L. Y. Jan, Y. N. Jan, Genome-wide study of aging and oxidative stress response in Drosophila melanogaster. Proc. Natl. Acad. Sci. U.S.A. 97, 13726-13731 (2000).[Abstract/Free Full Text]
  5. G. N. Landis, D. Abdueva, D. Skvortsov, J. Yang, B. E. Rabin, J. Carrick, S. Tavare, J. Tower, Similar gene expression patterns characterize aging and oxidative stress in Drosophila melanogaster. Proc. Natl. Acad. Sci. U.S.A. 101, 7663-7668 (2004).[Abstract/Free Full Text]
  6. S. D. Pletcher, S. J. Macdonald, R. Marguerie, U. Certa, S. C. Stearns, D. B. Goldstein, L. Partridge, Genome-wide transcript profiles in aging and calorically restricted Drosophila melanogaster. Curr. Biol. 12, 712-723 (2002).[CrossRef][Medline]
  7. T. Kayo, D. B. Allison, R. Weindruch, T. A. Prolla, Influences of aging and caloric restriction on the transcriptional profile of skeletal muscle from rhesus monkeys. Proc. Natl. Acad. Sci. U.S.A. 98, 5093-5098 (2001).[Abstract/Free Full Text]
  8. R. Weindruch, T. Kayo, C. K. Lee, T. A. Prolla, Microarray profiling of gene expression in aging and its alteration by caloric restriction in mice. J. Nutr. 131, 918S-923S (2001).[Abstract/Free Full Text]
  9. J. Mcelwee, K. Bubb, J. H. Thomas, Transcriptional outputs of the Caenorhabditis elegans forkhead protein DAF-16. Aging Cell 2, 111-121 (2003).[CrossRef][Medline]
  10. C. T. Murphy, S. A. Mccarroll, C. I. Bargmann, A. Fraser, R. S. Kamath, J. Ahringer, H. Li, C. Kenyon, Genes that act downstream of DAF-16 to influence the lifespan of Caenorhabditis elegans. Nature 424, 277-283 (2003).[CrossRef][Medline]
  11. S. A. Mccarroll, C. T. Murphy, S. Zou, S. D. Pletcher, C. S. Chin, Y. N. Jan, C. Kenyon, C. I. Bargmann, H. Li, Comparing genomic expression patterns across species identifies shared transcriptional profile in aging. Nat. Genet. 36, 197-204 (2004).[CrossRef][Medline]
  12. M. Kaeberlein, A. A. Andalis, G. R. Fink, L. Guarente, High osmolarity extends life span in Saccharomyces cerevisiae by a mechanism related to calorie restriction. Mol. Cell. Biol. 22, 8056-8066 (2002).[Abstract/Free Full Text]
  13. M. Kaeberlein, A. A. Andalis, G. Liszt, G. R. Fink, L. Guarente, Saccharomyces cerevisiae SSD1-V confers longevity by a Sir2p-independent mechanism. Genetics 166, 1661-1672 (2004).[Abstract/Free Full Text]
  14. S. J. Lin, M. Kaeberlein, A. A. Andalis, L. A. Sturtz, D. P., V. C. Culotta, G. Fink, L. Guarente, Calorie restriction extends Saccharomyces cerevisiae life span by increasing respiration. Nature 418, 344-348 (2002).[CrossRef][Medline]
  15. R. A. Miller, Y. Chang, A. T. Galecki, K. Al-Regaiey, J. J. Kopchick, A. Bartke, Gene expression patterns in calorically restricted mice: Partial overlap with long-lived mutant mice. Mol. Endocrinol. 16, 2657-2666 (2002).[CrossRef][Medline]
  16. M. M. Masternak, K. Al-Regaiey, M. S. Bonkowski, J. Panici, L. Sun, J. Wang, G. K. Przybylski, A. Bartke, Divergent effects of caloric restriction on gene expression in normal and long-lived mice. J. Gerontol. A Biol. Sci. Med. Sci. 59, B784-788 (2004).[Abstract/Free Full Text]
  17. K. Houthoofd, B. P. Braeckman, T. E. Johnson, J. R. Vanfleteren, Life extension via dietary restriction is independent of the Ins/IGF-1 signalling pathway in Caenorhabditis elegans. Exp. Gerontol. 38, 947-954 (2003).[CrossRef][Medline]
  18. S. Hekimi, L. Guarente, Genetics and the specificity of the aging process. Science 299, 1351-1354 (2003).[Abstract/Free Full Text]
  19. A. Bartke, J. C. Wright, J. A. Mattison, D. K. Ingram, R. A. Miller, G. S. Roth, Extending the lifespan of long-lived mice. Nature 414, 412 (2001).[CrossRef][Medline]
  20. J. M. Dhahbi, H. J. Kim, P. L. Mote, R. J. Beaver, S. R. Spindler, Temporal linkage between the phenotypic and genomic responses to caloric restriction. Proc. Natl. Acad. Sci. U.S.A. 101, 5524-5529 (2004).[Abstract/Free Full Text]
  21. W. Mair, P. Goymer, S. D. Pletcher, L. Partridge, Demography of dietary restriction and death in Drosophila. Science 301, 1731-1733 (2003).[Abstract/Free Full Text]
  22. S. Kammerer, L. L. Burns-Hamuro, Y. Ma, S. C. Hamon, J. M. Canaves, M. M. Shi, M. R. Nelson, C. F. Sing, C. R. Cantor, S. S. Taylor, et al., Amino acid variant in the kinase binding domain of dual-specific A kinase-anchoring protein 2: A disease susceptibility polymorphism. Proc. Natl. Acad. Sci. U.S.A. 100, 4066-4071 (2003).[Abstract/Free Full Text]
  23. A. A. Puca, M. J. Daly, S. J. Brewster, T. C. Matise, J. Barrett, M. Shea-Drinkwater, S. Kang, E. Joyce, J. Nicoli, E. Benson, et al., A genome-wide scan for linkage to human exceptional longevity identifies a locus on chromosome 4. Proc. Natl. Acad. Sci. U.S.A. 98, 10505-10508 (2001).[Abstract/Free Full Text]
  24. B. J. Geesaman, E. Benson, S. J. Brewster, L. M. Kunkel, H. Blanche, G. Thomas, T. T. Perls, M. J. Daly, A. A. Puca, Haplotype-based identification of a microsomal transfer protein marker associated with the human lifespan. Proc. Natl. Acad. Sci. U.S.A. 100, 14115-14120 (2003).[Abstract/Free Full Text]
  25. C. Chen, S. Dewaele, B. Braeckman, L. Desmyter, J. Verstraelen, G. Borgonie, J. Vanfleteren, R. Contreras, A high-throughput screening system for genes extending life-span. Exp. Gerontol. 38, 1051-1063 (2003).[CrossRef][Medline]
  26. M. S. Gill, A. Olsen, J. N. Sampayo, G. J. Lithgow, An automated high-throughput assay for survival of the nematode Caenorhabditis elegans. Free Radic. Biol. Med. 35(6), 558-565 (2003).[CrossRef][Medline]
  27. J. H. Bauer, S. Goupil, G. B. Garber, S. L. Helfand, An accelerated assay for the identification of lifespan-extending interventions in Drosophila melanogaster. Proc. Natl. Acad. Sci. U.S.A. 101, 12980-12985 (2004).[Abstract/Free Full Text]
  28. R. S. Kamath, A. G. Fraser, Y. Dong, G. Poulin, R. Durbin, M. Gotta, A. Kanapin, N. Le Bot, S. Moreno, M. Sohrmann, et al., Systematic functional analysis of the Caenorhabditis elegans genome using RNAi. Nature 421, 231-237 (2003).[CrossRef][Medline]
  29. S. S. Lee, R. Y. Lee, A. G. Fraser, R. S. Kamath, J. Ahringer, G. Ruvkun, A systematic RNAi screen identifies a critical role for mitochondria in C. elegans longevity. Nat. Genet. 33, 40-48 (2003).[CrossRef][Medline]
  30. A. Dillin, A. L. Hsu, N. Arantes-Oliveira, J. Lehrer-Graiwer, H. Hsin, A. G. Fraser, R. S. Kamath, J. Ahringer, C. Kenyon, Rates of behavior and aging specified by mitochondrial function during development. Science 298, 2398-2401 (2002).[Abstract/Free Full Text]
  31. E. A. Winzeler, D. D. Shoemaker, A. Astromoff, H. Liang, K. Anderson, B. Andre, R. Bangham, R. Benito, J. D. Boeke, H. Bussey, et al., Functional characterization of the S. cerevisiae genome by gene deletion and parallel analysis. Science 285, 901-906 (1999).[Abstract/Free Full Text]
  32. M. Kaeberlein, B. K. Kennedy, Large-scale identification of conserved aging genes in yeast. Mech. Ageing Dev., in press.
  33. M. Kaeberlein, K. T. Kirkland, S. Fields, B. K. Kennedy, Sir2-independent life span extension by calorie restriction in yeast. PLoS Biol. 2, E296 (2004).[CrossRef][Medline]
Citation: M. Kaeberlein, Aging-Related Research in the "-omics" Age. Sci. Aging Knowl. Environ. 2004 (42), pe39 (2004).

Extension of chronological life span in yeast by decreased TOR pathway signaling.
R. W. Powers III, M. Kaeberlein, S. D. Caldwell, B. K. Kennedy, and S. Fields (2006)
Genes & Dev. 20, 174-184
   Abstract »    Full Text »    PDF »

Science of Aging Knowledge Environment. ISSN 1539-6150