Research: Genetics and Evolution of Longevity

Evolutionary Causes of Aging

Two evolutionary models can explain the phenomenon of senescence. The antagonistic-pleiotropy theory of senescence postulates that genetically-caused correlations between fitness traits are the underlying cause of senescence. These genetic correlations arise because individual genes affect more than one trait (pleiotropy). It is these pleiotropic genes that cause health and performance to decrease with advancing age. To understand the theory, first assume that some alleles (variant forms of a gene) have pleiotropic effects that are limited to certain ages (are age-specific in expression). If an allele has beneficial effects at one age, but deleterious effects at a different age, it is subject to conflicting selection pressures. An example of such an allele would be one that boosts production of a hormone that increases fertility during early adulthood, but also increases risk of disease later in life. The strength of selection acting on this allele is inversely proportional to the age of expression. This is because late-life increases or decreases in reproduction and survival have little effect on Darwinian fitness compared to similar effects early in life (see figure above, based on data from human males). Natural selection will thus tend to favor alleles with beneficial early-age effects and deleterious late-age effects. Conversely, alleles with the reverse pattern of action (early deleterious effects and late beneficial effects) will be selected against. As the frequency of the first type of allele increases in the population and the frequency of the second type decreases, the population will evolve a senescent life history. Individual differences in longevity can be caused by polymorphism maintained by these genetic 'trade-offs'.

A competing theory of the evolution of senescence is based on a model of mutation-selection balance. This model assumes that some alleles will have unconditionally deleterious effects, but the effects will be confined to certain ages. One potential example of this kind of allele is the one causing Huntington's Disease (HD) in humans: people with the HD allele usually have no symptoms until late middle age. The premise of the mutation-accumulation model is that natural selection will be efficient at eliminating deleterious mutations if their effects are expressed early in life, but much less efficient if the effects are only expressed late in life. Again, this is because the strength of selection declines with age. Thus, the population-wide frequency of a mutant allele (determined by the balance between mutation and selection) will be higher the later the age of expression. High frequency of deleterious alleles that are only expressed at late ages will cause performance to decrease with advancing age, and will contribute to individual longevity differences.

Both the antagonistic-pleiotropy and mutation-accumulation models have been tested several times (most extensively in D. melanogaster). The general conclusion emerging from these reviews is that both mechanisms probably contribute to senescence in fruit flies.

Causes of Mortality Plateaus

Common experience and data on humans suggest that mortality rates in a cohort should increase with age, (delete by some function), after those individuals reach adulthood. Contrary to this expectation, Carey et al. and Curtsinger et al. independently reported experimental results showing that rates of aging after reaching adulthood can plateau, or even become negative, at very late ages (1992). This has since been found in both experimental organisms and in humans (Vaupel et al. 1998).

Not only is this an interesting phenomenon, but understanding the causes of these patterns is critical for demographers and public health specialists who interpret the patterns as they relate to humans, and to biologists concerned with the genetic and cellular processes causing aging. The underlying biological causes of the leveling-off and/or decline of mortality rates have been discussed by many scientists (Curtsinger et al. 1995; Partridge & Barton 1996; Charlesworth & Partridge 1997; Vaupel et al. 1998; others) and at least four different hypotheses have been proposed. Although mortality plateaus are characteristics of populations and not of individuals, both population-level and individual-level factors can contribute to the overall phenomenon. Population-level factors hypothesized to contribute to mortality plateaus are genetic and environmental heterogeneity, while individual-level factors may be changes in the physiology or behavior of the organism.