Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Cellular and Molecular Mechanisms of Biological Aging: The Roles of Nature, Nurture, and Chance in the Maintenance of Human Healthspan TASK GROUP DESCRIPTION Background The degree to which an individual organism maintains healthspan and lifespan is a function of complex interactions between genetic inheritance (nature), environment, including cultural inheritance (nurture), and sto- chastic events (luck or chance). This task group will focus upon the role of chance because it is so poorly understood and because it appears to be of major importance in the determination of individual variations in health- span and lifespan within species. The major factor determining variations in healthspan and lifespan between species is genetic inheritance. Broader aspects of cellular and molecular mechanisms of biological aging will also be considered, given their importance for understanding the cellular and molecular basis of successful aging. The task force will consider the cellular and molecular basis for nature, nurture, and chance in healthspan and lifespan determination. On the basis of comparisons between identical and nonidentical twins, geneticists have estimated that genes control no more than about a quarter of the interindividual differences in lifespan (Herskind et al., 1996). Twin studies of very old individuals, however, show substantially greater genetic contributions to healthspan (McClearn et al., 1997; Reed and Dick, 2003). The environment clearly plays an important role in the length and the qual- ity of life. Tobacco smoke, for example, has the potential to impact upon 73
74 the future of human healTHspan multiple body systems in ways that appear to accelerate the rates at which those systems age (Bernhard et al., 2007). To document the role of chance events on aging one must rigorously control both the genetic composition of an organism and its environment. This has been done to a remarkable degree in a species of nematodes, Cae- norhabditis elegans (Vanfleteren et al., 1998). The results confirm hundreds of previous studies with a wide range of species, especially those with inbred rodents housed under apparently identical but less well-controlled environ- ments. One observes wide variations in lifespan in all these studies. For the C. elegans experiments the distributions of lifespan fit best with two-param- eter or three-parameter logistic models and not with the classical Gompertz model or the Weibull model. Many mutations have been shown to substantially increase lifespan in C. elegans. It is of interest, however, that the ranges of the lifespan variations among such mutant strains overlap with those of wild type strains (Kirk- wood and Finch, 2002). Many of these long-lived mutant strains exhibit enhanced resistance to a variety of stressors, notably heat shock. It was there- fore predicted that variable degrees of response to heat shock stress might form a basis, or a partial basis, for individual variations in longevity. An initial set of experiments demonstrated that is indeed the case, at least for a transgenic construct that includes the promoter of a small heat shock gene (Rea et al., 2005). There was a very strong correlation between the response to heat stress and longevity, with good-responding worms living longer. Strikingly, this phenotype was not heritable. The progeny of a worm show- ing a strong heat stress reaction exhibited the broad distribution of lifespan shown by the starting population. The heat stress reaction was therefore sto- chastic. The nature of the chance events that determine the reaction remains unknown. They could be related to the intrinsic instability of the transgene, making it important to repeat such experiments utilizing endogenous genes as reporters of the response to heat shock and other stressors. It could be due to epigenetic drifts in gene expression, perhaps involving random changes in gene promoters or in the state of chemical modifications to histone proteins that coat chromosomes. Such changes have indeed been observed in aging human identical twins (Fraga et al., 2005). While those changes have been interpreted as being driven by the environment, one cannot at present rule out random variations unrelated to environmental influences. Variations in gene expression in genetically identical organisms exam- ined under environmentally identical conditions have also been attributable to intrinsic noise in fundamental molecular processes such as the transcrip- tion and translation of genes. Most such observations have been made using
CELLULAR AND MOLECULAR MECHANISMS OF BIOLOGICAL AGING 75 microorganisms (Elowitz et al., 2002), but stochastic bursts of transcrip- tion have also been noted in mammalian cells (Raj et al., 2006). Moreover, substantial variations in the levels at which genes are transcribed has been shown to occur in mouse tissues, and that variation was shown to increase with age (Bahar et al., 2006). Chance events are also of major significance in the determination of diseases of aging. For the case of cancer, mutations have been shown to be of major importance. A likely key to malignancy, however, is the chance event of suffering a mutation in a gene that when mutated, now greatly enhances the general frequency of mutation. Such genes are referred to as âmutator genesâ (Bielas et al., 2006). Chance events can make the difference between life and death of individuals with coronary artery atherosclerosis, as mortality often follows the rupture of an atherosclerotic plaque, an event that is likely to be due in part to a chance event (a trigger) leading to the rupture (Falk, 1992). Moreover, some genetic interventions that have been introduced into model organisms (nematodes, mice) increase mean but not maximum lifespan and appear to rectangularize the lifespan curve. A recent example is a mouse strain carrying extra copies of a tumor suppressor locus (Matheu et al., 2007). As expected, these mice are remarkably cancer free. Of particular interest, though, their mean but not maximum lifespan was extended. Does this locus and similar interventions rectangularize the lifespan curve by reducing random events? Initial Challenges to Consider â¢ What experiments might be designed in model organisms to probe the role of variations of endogenous gene expression at birth in the determi- nations of the remarkable stochastic variations in lifespan among genetically identical organisms? â¢ Which subset of endogenous genetic loci are major contributors to such stochastic variation? â¢ Are these variations in gene expression attributable to specific molecular events, such as chemical modifications to DNA CpG islands or histones? â¢ What are the molecular and biophysical mechanisms that lead to transcriptional bursts in gene expression? â¢ Do cells within an individual organism differ in their susceptibility to stochastic fluctuations in gene expression? For example, are postmitotic neurons more or less susceptible than cells that are destined to die or cells that turn over?
76 the future of human healTHspan â¢ Are there species-specific differences in the degree to which stochas- tic fluctuations in gene expression occur (in similar cell types)? â¢ Has evolution shaped the above stochastic variations in gene ex- pression, are they adaptive, and what are the selective pressures that led to such adaptations? How can one test the hypothesis that different degrees of stochastic variations in gene expression do in fact evolve and that they are adaptive? â¢ To what extent do early environmental influences in developing humans (fetal, neonatal, childhood, pubertal) determine patterns of gene expression and patterns of aging in human subjects? â¢ Do genetic interventions that increase mean but not maximum lifespan, and appear to rectangularize the lifespan curve, act by reducing random events? Can we learn about the cellular and molecular bases for stochastic variation by testing the hypothesis that some of these interven- tions act by this mechanism (reducing stochastic variation)? â¢ What are some candidate environmental agents and social influences responsible for such putative influences and how can their impacts upon public health be measured? Initial References Bahar, R., C. H. Hartmann, K. A. Rodriguez, A. D. Denny, R. A. Busuttil, M. E. Dolle, R. B. Calder, G. B. Chisholm, B. H. Pollock, C. A. Klein, and J. Vijg. 2006. Increased cell- to-cell variation in gene expression in ageing mouse heart. Nature 441:1011-1014. Bernhard, D., C. Moser, A. Backovic, and G. Wick. 2007. Cigarette smokeâan aging accelerator? Experimental Gerontology 42(3):160-165. Bielas, J. H., K. R. Loeb, B. P. Rubin, L. D. True, and L. A. Loeb. 2006. Human cancers express a mutator phenotype. Proceedings of the National Academy of Sciences U.S.A. 103:18238-18242. Elowitz, M. B., A. J. Levine, E. D. Siggia, and P. S. Swain. 2002. Stochastic gene expression in a single cell. Science 297:1183-1186. Falk, E. 1992. Why do plaques rupture? ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ Circulation 86:III-30-III-42. Fraga, M. F., E. Ballestar, M. F. Paz, S. Ropero, F. Setien, M. L. Ballestar, D. Heine-Suner, J. C. Cigudosa, M. Urioste, J. Benitez, M. Boix-Chornet, A. Sanchez-Aguilera, C. Ling, E. Carlsson, P. Poulsen, A. Vaag, Z. Stephan, T. D. Spector, Y. Z. Wu, C. Plass, and M. Esteller. 2005. Epigenetic differences arise during the lifetime of monozygotic twins. ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ Proceedings of the National Academy of Sciences U.S.A. 102:10604-10609. Herskind, A. M., M. McGue, N. V. Holm, T. I. Sorensen, B. Harvald, and J. W. Vaupel. 1996. The heritability of human longevity: A population-based study of 2872 Danish twin pairs born 1870-1900. Human Genetics 97:319-323.
CELLULAR AND MOLECULAR MECHANISMS OF BIOLOGICAL AGING 77 Kirkwood, T. B., and C. E. Finch. 2002. Ageing: The old worm turns more slowly. Nature 419:794-795. Matheu, A., A. Maraver, P. Klatt, I. Flores, I. Garcia-Cao, C. Borras, J. M. Flores, J. ViÃ±a, M. A. Blasco, and M. Serrano. 2007. Delayed ageing through damage protection by the Arf/p53 pathway. Nature 448:375-379. McClearn, G. E., B. Johansson, S. Berg, N. L. Pedersen, F. Ahern, S. A. Petrill, and R. Plomin. 1997. Substantial genetic influence on cognitive abilities in twins 80 or more years old. Science 276:1560-1563. Raj, A., C. S. Peskin, D. Tranchina, D. Y. Vargas, and S. Tyagi. 2006. Stochastic mRNA synthesis in mammalian cells. PLoS Biology 4:e309. Rea, S. L., D. Wu, J. R. Cypser, J. W. Vaupel, and T. E. Johnson. 2005. A stress-sensitive reporter predicts longevity in isogenic populations of Caenorhabditis elegans. Nature ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ Genetics 37:894-898. Reed, T., and D. M. Dick. 2003. Heritability and validity of healthy physical aging (wellness) ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ in elderly male twins. Twin Research 6:227-234. Vanfleteren, J. R., V. A. De, and B. P. Braeckman. 1998. Two-parameter logistic and Weibull equations provide better fits to survival data from isogenic populations of Caenorhabditis elegans in axenic culture than does the Gompertz model. Journals of Gerontology AâBiological and Medical Sciences 53:B393-B403. Due to the popularity of this topic, two groups explored this subject. Please be sure to explore the other write-up, which immediately follows this one. TASK GROUP MEMBERSâGroup A â¢ Diddahally Govindaraju, Boston University â¢ Stephanie Lederman, American Federation for Aging Research â¢ Kyongbum Lee, Tufts University â¢ Richard Mayeux, Columbia University â¢ Saira Mian, Lawrence Berkeley National Laboratory â¢ Chris Schaffer, Cornell University â¢ Nicholas Schork, Scripps Genomic Medicine, Scripps Health â¢ Rob Stephenson, Emory University â¢ David Stopak, The National Academies â¢ Richard Suzman, National Institute on Aging â¢ Woodring Wright, University of Texas Southwestern Medical Center â¢ Alissa Poh, University of California, Santa Cruz
78 the future of human healTHspan TASK GROUP SUMMARYâGroup A By Alissa Poh, Graduate Student, Science Writing Program, University of California, Santa Cruz Boundaries between the different sciences are gradually being torn down as interdisciplinary research and open laboratories become increas- ingly popular concepts. Similarly, we once regarded disease and aging as independent mechanisms, but have since realized that they are in fact closely intertwined; diseases with strong age-related incidences are likely to have a strong age-associated component. One key issue for those studying the biology of aging is that while in- terspecies variations in healthspan and lifespanâfor example, between mice and humansâcan be fully explained by genetics, the same is not true for variations within species. Here stochastic, or random, events are thought to be an important factor, but precious little is actually known about the role of chance in determining lifespan. This group was hence charged with examining stochastic variationâs po- tential influence on human lifespan. Very early in the discussion, however, it became clear that many group members were uncertain that investigations in this area should be a priority, and spent much of the first day debating its importance. Stochastic Events and LongevityâHow Much Should We Care? âItâs an enormously interesting and informative biological question,â said Woodring Wright of stochastic variation itself. âBut when talking about longevity, I think itâs an uninteresting question.â He then offered an analogy using restaurant glasses. âIf you look at the rate of breakage of restaurant glasses, you get curves that mimic those in lifespan studies,â he said. âIf you have a well-built glass with a thick wall versus a flimsy wall, youâll get differences in lifespan, but thereâs still going to be this variation. I donât think itâs an interesting ques- tion, whatâs responsible for the variation between when the first glass and the last one breaks, in terms of understanding the glassâs lifespan.â Nevertheless, the group decided that the consequences of stochastic factors on lifespan might be sufficiently pronounced to justify developing an experimental model addressing the issue. This would be no run-of-the-mill investigation either, since the group felt that most current human longevity
CELLULAR AND MOLECULAR MECHANISMS OF BIOLOGICAL AGING 79 studies have yielded insights into factors that increase or decrease lifespan only marginallyâsmoking cessation, dietary changes, and drug treatments, to name a few. âWeâre not looking to increase lifespan by a couple of years; rather, we need to ask what could be radically tweaked to dramatically extend lifes- pan,â said Nicholas Schork. A brief debate ensued around the definition of stochastic factors. The group eventually agreed that these involve subtle, random perturbations of an individualâs cellular and molecular physiological milieu that accumulate over time, gradually affecting function at the whole-organism level. The powwow then moved on to various ways to set up this experiment. Mathematical modeling was considered, where one could make statistical predictions about the impact on human lifespan of manipulating factors possibly associated with longevity. However, such modeling would only be as good as the empirical studies on which they were based and which have not proved particularly useful to date. Taking a different approach, each group member was asked the ques- tion: âWhat reasonable experiments would you pursue, given a blank check?â Genetic screens involving model organisms were suggested. The group acknowledged these to be incredibly valuable, although some members were skeptical about the potential for human genetic screens via genomewide as- sociation studies. However, such screens are unable to capture many factors implicated in the aging process. An argument for comparative physiology and genomic studies across species was made, but because there are many species, or lineage-specific factors at play in aging and senescence, the translatability of cross-species findings would always be in doubt. The group toyed with interventional studies, but these were also thought to be problematic. The intervention does not always work: they are only as good as the biological insights motivating the intervention. Where humans are concerned, the results are hardly radical, usually affecting lifes- pan in terms of only a few years. Another suggestion was to look at contrasts between long-lived and short-lived individuals within a species. Such studies could simultaneously investigate many factors, both genetic and nongenetic, but might produce species-specific insights. So each strategy had its drawbacks. However, by the end of the first day, the group settled on designing a study comparing individuals undergoing
80 the future of human healTHspan early and late senescence as a result of stochastic factors. This model would probably not reveal mechanisms leading, upon artificial manipulation, to dramatic lifespan differences. Still, the group felt that it would shed light on the potential role of stochastic factors in aging. Fine Tuning the Experimental Model Much of the second day was spent refining this proposed experiment. The group agreed that the nematode worm C. elegans would be an ideal test subject, since it has a very small number of cells (less than 1000), is easily manipulated, has a short lifespan, and much is already known about its biology. In particular, it has an interesting marker for senescence that can be exploited: before these worms die, they exhibit an identifiable change in swimming behavior. Researchers could thus readily access animals in the final stages of life while they are still relatively fit. âYou donât just want to pick up a dead worm and look at its gene ex- pression profile; itâs too late,â Chris Schaffer commented. Or, as another group member put it, âDeath is a lousy endpoint for measuring aging.â Among the group members, Wright remained openly skeptical about exploring stochastic variationâs role in lifespan. He did not take issue with the feasibility of any particular experiment; rather, he questioned whether time, money, and effort should be spent in this area, or if there were worthier aspects of aging biology that should be studied first. The group therefore decided to make him the official spokesman for delivering their conclusions. Why Genetic Screens Arenât the Last Word The group reiterated that genetic screens in model organisms are tre- mendously important, as they reveal evolutionarily conserved pathways, and thus important processes, in humans. These screens do, however, miss many categories of important mechanisms. For starters, aging is multifactorial, so the effects of individual pathways are minimal. While genetic screening has successfully identified pathways regulating multiple others downstream, single pathway effects are unde- tected and left by the wayside. Genetic screening also identifies processes ac- tive in postmitotic organisms that are shared by their mitotic counterparts, but additional processes could be involved in mitotic organisms. Finally,
CELLULAR AND MOLECULAR MECHANISMS OF BIOLOGICAL AGING 81 Figure 1â Lifespan distribution of C.elegans populations. it is not possible to identify potentially significant nongenetic effects with such screens. Among the examples of mechanisms these screens miss are the role of stem cells in aging, the importance of neoteny in the evolution of human lifespan, and most significantly, the key issue this group was asked to ex- plore, with longevity in mind: stochastic variation in gene expression. The Game Plan in a Nutshell As mentioned earlier, changes in swimming behavior can be detected in C. elegans several days prior to death, allowing one to predict longevity and select individuals soon destined to die. This fact was exploited in the proposed model, where 10 percent of the shortest-lived worms (early sink- ers) in an otherwise genetically homogeneous C. elegans population, as well as 10 percent of the longest-lived cohort (late sinkers), would be identified shortly before death. Healthy worms (swimmers) isolated at an early time point would serve as a control group. The lifespan distribution of these populations, marked by swimming ability, is illustrated in Figure 1. In analyzing these different groups of worms, three hypotheses would be considered, with the first two pointing to two different roles for nonge- netic factors in aging:
82 the future of human healTHspan 1. Stochastic epigenetic events generate phenotypic states associated with aging. 2. Increased variation among cells in a tissue induces an increased rate of aging due, for example, to loss of homeostatic capability. 3. Increased variation in gene expression at the single cell or organismal level does not explain the heterogeneity of lifespan in homogeneous geno- types (null hypothesis). Many Roads to Rome In the first hypothesis different phenotypes are produced by a change in state as a result of stochastic variation. So early and late sinkers would have identical global phenotypes due to both cohorts being ultimately channeled into senescence. This would be distinguishable from the early swimmers, as illustrated in Figure 2. âThe analogy here is that being healthy is like being in a little well at the top of a mountain, and a slight push knocks you out of the well,â Schaffer explained. âOnce youâre out, you run all the way to the bottom. And when youâve arrived there, it doesnât really matter how you were pushed; you got to the bottom of the hill, to a common state associated with senescence.â If this hypothesis is true, a potential secondary analysis would identify individual genes and pathways implicated in the fingerprint of senescence. Figure 2â HypothesisââMany roads to Rome.â
CELLULAR AND MOLECULAR MECHANISMS OF BIOLOGICAL AGING 83 Figure 3â HypothesisââVariation drives aging.â Cross-species conservation levels of genes and pathways, as well as DNA sequence polymorphisms in human orthologs, would also be assessed. Variation Drives Aging All cells within a tissue start off with prescribed levels of gene expres- sion, and any variation is small, controlled within a range where that tissue can maintain itself. In the second hypothesis, which Figure 3 illustrates, random variation in gene expression within or between specific cells drives lifespan differences. These cells become dysregulated, ultimately losing their ability to function as part of their tissue of residence. So there would be no common gene expression profile or other aspects of biology associated with senescence; rather, greater variation among cells within a tissue would result in loss of its capability. If results indicate that variation is important, individual genes vary- ing in expression from cell to cell would be identified, and vice versa (cells displaying cell-to-cell variation in gene expression), as secondary analyses. Cell-specific methylation patterns would be assessed, as well as cross-species conservation of the genes exhibiting greater variability. Finally, a variety of evolutionary and comparative biology experiments could be carried out on worms and flies possessing very different lifespans. The Bigger Picture Apart from designing this experimental model the group also pon- dered broader, scratch-the-surface research ideas not sufficiently refined to
84 the future of human healTHspan permit specific experiments, but nevertheless focused on developing tools to identify events critical to aging that would sail by a genetic screen. One particularly promising idea involved a series of transplant experiments: placing young tissue into an old animal and vice versa. This would help identify aging mechanisms associated with a particular cell or tissue versus organism-level processes. In the end as the group humorously illustrated with $$$ as a bullet point on a presentation slide, this new brand of aging-and-disease research will require a substantial financial investment to truly take wing. âThereâs a real case for individual RO1s versus programmatic policies,â Wright pointed out. âA lot of the things thatâre being found, we canât predict a priori. So thereâs still an important role for individual entrepreneurship to tackle isolated problems and find new handles to push, in terms of discover- ing new aging mechanisms.â TASK GROUP MEMBERSâGroup B â¢ Suresh Arya, National Cancer Institute â¢ Christine Grant, North Carolina State University â¢ Linda Miller, Nature â¢ Richard Miller, University of Michigan â¢ Santa Jeremy Ono, Emory University â¢ Chris Patil, Lawrence Berkeley National Laboratory â¢ Jerry Shay, University of Texas Southwestern Medical Center â¢ Eric Topol, Scripps Research Institute â¢ Michael Torry, Steadman-Hawkins Research Foundation â¢ Heinz-Ulrich G. Weier, Lawrence Berkeley National Laboratory â¢ Iris Tse, Boston University TASK GROUP SUMMARYâGroup B By Iris Tse, Graduate Writing Student, Boston University I have cataracts, lose uphill races, get really sick when I catch the flu, girls no longer whistle when I pass by, my joints ache, and if you looked closely youâd see preclinical signs of the cancer that will kill me.
CELLULAR AND MOLECULAR MECHANISMS OF BIOLOGICAL AGING 85 How Old Am I? Not all species of animals age at the same rate. Humans typically will have a longer lifespan than horses, which in turn will have a longer lifespan than most rodents. However, there is a synchrony in the way age-related decline in health and diseases appears across different species. At certain points in life, for example, health problems such as vision degeneration, cancer, diabetes, central nervous system degeneration, and organ failures, will appear in some members of each species. Even more surprising is that most of these age-related diseases will appear regardless of lifespan and size of the animals. It appears as if some yet unknown species- or breed-specific factors are tying together these degenerative functional changes. Therefore, the group found this to be an area worth pursuing. The Validity of the Stochastic Model The multidisciplinary group initially debated in earnest the validity of the stochastic model in the context of healthspan and lifespan. The stochas- tic model was initially assigned to the group by the conference organizers as a springboard for discussion. Previous research has found that there is a wide range of lifespan within a genetically homogenous group. Therefore, there must be some unspecified nongenetic factor that can influence the aging trajectory and lifespan. Lifespan is not entirely controlled by genes and controllable environmental factors, and undefined gaps in current knowledge still need to be examined and studied. However, some members of the group felt it would be premature to attribute the entirety of this unknown area to the stochastic model. The stochastic model proposed that factors extending average lifespan, but not the maximum lifespan, played a major role in extending healthspan. The group felt that the basis of this view was somewhat biased and would exclude many healthspan factors that also extend maximum lifespan. The stochastic model, while important, is not the complete answer. After an hour of vigorous discussion and scrutiny, the group leader, Jerry Shay, a professor of cell biology at the University of Texas Southwest- ern Medical Center, articulated the groupâs thinking. âTo think that aging is stochastic, and therefore not within our control, greatly diminishes the impact and window of ability to understand, change, and manipulate the process of aging,â said Shay.
86 the future of human healTHspan Important Topics Surrounding Healthspan and Lifespan Once the group set aside the stochastic model, the group members were free to explore other facets of aging. Linda Miller, the U.S. executive editor of Nature and the Nature journals, sparked the next round of discussion by pointing out neurodegeneration as an area worth exploring. Topics sug- gested by other members of the group included cell regenerative capacity, cellular replication and control, broad-spectrum genomic analysis of aging, and biomaterials used in current research. However, the topic that seized the groupâs attention was the synchronic- ity of age-associated decline. Richard Miller, a professor of pathology at the University of Michigan, pointed out that many organs and cells fail at more or less the same time within a species. The sequence of illnesses are often synchronized across various species of animals, albeit at a different rate, but seemingly related to the lifespan of specific species. These co-morbid events, such as the onset of diabetes or cataracts, are not necessarily terminal illness- es and may not directly affect lifespan. However, they do affect healthspan. While the scientific basis of these assertions was unclear to some members of the group, the idea was thought to be meritorious. If we can understand why humans get cataracts at 60 years old and why mice get cataracts at 2 years old, we might be able to delay the onset of these symptoms, prolong healthspan, and understand aging a little better. Miller insisted that it will be useful to find themes, or common fami- lies, of underlying cellular or molecular factors that time aging sequentially. Because of the lack of critical knowledge of the field by some members of the group, the initial goal was created to develop a new set of hypotheses for future experiments. Experimental Approach Miller already had a rough idea of the experimental approach and the group spent the remaining day and a half deliberating those points, includ- ing lengthy discussions regarding the potential benefits and pitfalls of the many different types of study designs that may best accomplish the groupâs target. An important first step is to generate a list of late-life dysfunctions plausibly related to the timing of aging. These age-related dysfunctions can be diseases, such as cancer, or they can be symptoms, such as cataracts or cognitive failure, that affect many animals. Experts knowledgeable in aging or pathogenesis, or both, will need to
CELLULAR AND MOLECULAR MECHANISMS OF BIOLOGICAL AGING 87 be solicited to provide hypotheses. They also will be called upon to brain- storm a body of known factors, such as proteins, macromolecules, enzy- matic activities, and genetic expressions that might influence healthspan. By understanding the synchrony of these factors, perhaps we will understand more about the synchrony of aging. Since the group hoped to get the experiments started as quickly as pos- sible, an important criterion is that these factors must be easy and practical to measure using established methodology and reagents. Assays that can be done on multiple species using existing methods and technology will be more useful for the project than those that depend on the development of new, species-specific reagents. âBasically, you would want to be able use preexisting reagents or kits straight from a Sigma catalogue. Otherwise, youâd get mired over things like the trial and error of a new experimental design,â said Chris Patil, a postdoc- toral scholar of life sciences from Lawrence Berkeley National Laboratory. The next step will be to select the proper animal models for interspe- cies comparison. The experiments will use healthy young adults with no age-related diseases because it is thought that the selective factors that mold maturation are not the same as aging. These animals will be evaluated to uncover patterns of protein expressions, or possibly some other cellular events, that determine the rate of aging of that particular species. With some justification, four types of animals were chosen for this large-scale study: 1. PrimatesâFor their similarity to humans. 2. BatsâIn most animals, lifespan is directly correlated to body weight. However, bats have a long lifespan for their small body size, therefore allow- ing researchers to observe trends that are actually related to lifespan and not weight. 3. RodentsâLifespan varies drastically across different species of ro- dents, ranging from 2 years for mice to 30 years for naked mole rats. The variety will allow experimenters to adjust for confounders. 4. BirdsâThe nonmammalian out-group to anchor the phylogenetic tree during statistical analysis. Additional funding will be necessary to trap and collect these animals. Around 10 species for each clade of animals were thought to be adequate to start. Animals from the wild will be ideal, since they provide a more realistic snapshot of the aging process. But the group is also open to using captive
88 the future of human healTHspan animals, such as those from zoos, since they are likely to show species-spe- cific cellular and biochemical traits that correlate with lifespan. This broad-spectrum analysis is meant to generate useful hypotheses. Itâs not meant to be a hypothesis-testing exercise. The end product from these experiments will produce information necessary to create mecha- nistic hypotheses for further testing. It is important to provoke further questions and examinations. Interspecies contrast will be the first step. Once something comes up as a hit, then the next step will be intraspecies comparison. âFocusing on the basic biology of aging, instead of approaching dis- eases one at a time, may actually speed up the process of learning how to postpone individual diseases.â said Miller. âThis umbrella approach is a better way to probe the age-diseases nexus by exploiting the power of comparison.â TASK GROUP MEMBERSâGroup C â¢ Craig Atwood, University of Wisconsin-Madison â¢ Miles Axton, Nature Genetics â¢ Rita Effros, University of California, Los Angeles â¢ Nan Jokerst, Duke University â¢ Jay B. Labov, National Academy of Sciences â¢ Valter Longo, University of California, Los Angeles â¢ Joao Magalhaes, Harvard University â¢ Ken Turteltaub, Lawrence Livermore National Laboratory â¢ C atherine Wolkow, National Institute on Aging, Intramural Research Program, National Institutes of Health â¢ Natalia Mackenzie, Boston University TASK GROUP SUMMARYâGroup C By Natalia Mackenzie, Graduate Writing Student, Boston University For many people, aging is taken for granted. Like an old machine, the human body starts to fail in its functions and eventually it stops working. Although aging seems to be an obvious and expected state of life, it has been difficult for scientists to achieve consensus on the characteristics of aging and why some organisms live longer than others. One thing though is clear:
CELLULAR AND MOLECULAR MECHANISMS OF BIOLOGICAL AGING 89 Genes, the environment, and their interactions throughout life are key to understanding of aging. As the nine scientists of task group C met for the first time, an engineer, an adviser in education, a science editor, and a chemist surrounded by five biologists got ready to discuss the roles that genes and the environment play in aging. As the topic was quite broad, the first day felt like a warm-up. Intense brainstorming included a discussion about one of the main problems that scientists face while studying the biology of aging: the search for repre- sentative animal models. Because aging is a complex process that involves multiple body-part failures, it is difficult to find animal models that have all the different aspects associated with human aging. âAll models simulate only one part of aging, so I am not sure we do have a model for studying aging,â said one of the participants. They concluded that in addition to the most widely used animal models in researchâworms, flies, and miceâanimals, such as tortoises, that live much longer than humans also should be consid- ered. This statement turned the conversation to the understanding of why some animals live longer than others. Following this idea, they discussed the concept that animals may live as long as they need to in order to optimize reproduction. In addition, they proposed that as a consequence of food scarcity, some animals may have to live longer in order to wait for better conditions to reproduce. Another possibility was that animals exposed to less predation could reproduce later and therefore live longer. But why do animals have to age at all? âTo avoid competing with our kids,â said one member of the group. Another participant thought that âwe have to recycle matter, we donât have infinite mass.â During the second round of discussion in the first day, the group fo- cused on the characteristics of aging and the biological and environmental events that cause it. Aging can be seen as the progressive deterioration of the organism by the accumulation of mutations and epigenetic alterations in the DNA over time. In other words, the longer we live, the more time our DNA diverges from what it was at the beginning of life. Such DNA alterations are generated by environmental factors such as sunlight (UV), ionizing radiation, pathogens (bacteria, viruses), carcinogens, replication errors, among others. Another crucial environmental factor that affects aging is socioeconom- ic status (SES). SES depends on family income, parental education level, and social status in the community. It may be also understood as quality of life, considering nutrition, exposure to violence, self-esteem, and stress.
90 the future of human healTHspan One key idea proposed as one of the main causes of animal aging was the progressive decline in the ability of cells to communicate with one an- other. Organs are made of different kinds of cells. For organs to properly function, the cells have to be able to converse with one another in order to synchronize growth and function. While we age, accumulation of DNA mutations and epigenetic changes generated by environmental factors affect the ability of cells to communicate with one another and therefore organs begin to fail. One good example is brain cells. The brain is composed of bil- lions of cells that electrochemically communicate with one another. If cells are damaged and are not able to communicate, the chain of transmission is interrupted and information is lost. The result may be a decrease of memory in older people, or slower muscle reactions. According to one member of the group, epigenetic DNA changes, such as histone methylation and DNA CpG methylation, accumulate and respond faster to environmental changes than mutational damage to DNA. Fortunately, there are intrinsic cellular mechanisms that patrol the integrity of the DNA. If the repairing machinery detects a change, it can repair the mistake from within the cell. Therefore, how long an animal lives also depends on the efficiency of its repairing machinery. However, at some point in life, not even the repairing machinery can save the body from deteriorating, because mutation accumulation can also affect the integrity of the repairing machinery itself. As a consequence, mistakes are no longer repaired and an overwhelming accumulation of altered DNA damages or kills the cells. Another flaw of the repairing machinery is that it cannot com- pare genome sequence between individual cells in order to âhomogenizeâ their genetic information after they start accumulating differences. It was proposed that the way life is designed to reset all this damaged information is by eliminating the old organism, generating a whole new one by fertiliza- tion of the egg by the sperm. During the second day, the group focused on building a tangible pro- posal that would represent their ideas and conclusions. They first decided that their working hypothesis would be that aging results from increasing cellular damage that compromises communication pathways, leading to impaired cellular functioning and organ failure. The group felt that there was currently a great deal of research and information about aging at the cellular and organismal levels, but that there remain large gaps of knowledge in between (e.g., at the tissue, organ, and organ system levels). According to the group, one reason for that was that scientists studying different aspects of aging do not communicate enough with one another. Agreeing with one
CELLULAR AND MOLECULAR MECHANISMS OF BIOLOGICAL AGING 91 memberâs comment that âin the intersection among different points of view is where creativity lies,â the group decided to propose an interdisciplinary research initiative on aging that would begin to break down miscommu- nication among disciplines and mine data from research on aging at the cellular and organismal levels to offer possible new insights into the process at additional levels of biological organization. Their final proposal was a five-year program that would focus on creat- ing biologically guided tools that help extend life- and healthspan. The idea emerged when the only engineer in the room said that engineers only need to know âwhat is brokenâ in aging so they can fix it. âWe donât know what you biologists want,â said the engineer. The group immediately understood how powerful the fusion of both disciplines could be for contributing not only to aging research but also to science in general. The main goal of the proposed program would be to use biology to guide machines that would allow early detection and prevention of genetic and epigenetic modifications that influence aging. The idea is to use these new tools to intervene in the aging process by repairing, reprogramming, removing, or replacing damaged biological components of cells that con- tribute to the deterioration of the human body. Specifically, they proposed to focus on organ systems that are mostly affected in aging like vasculature, the nervous system, musculoskeletal, immune, vision, and hearing. The group decided that the first year should be oriented to data mining, the integration of all the existing knowledge of aging at a cellular, tissue, organ, and organism level. The next two to five years, experiments and tech- nology development would take place where the most promising research would be prioritized and engineering prototype devices would be made. One example of a tool that the group proposes relates to hormonal imbalances that normally inhibit physical, sexual, and cognitive functions as a result of aging. The idea is to first identify those hormones that are altered and create devices to monitor and redirect hormones and behavior to what is observed in younger organisms. But the conclusion this task group presented the last day of the con- ference was not only a very practical one. They were also able to call the scientific communityâs attention to breaking down the silos and to moving to a multidisciplinary approach in aging research.