Sci. Aging Knowl. Environ., 3 May 2006
Vol. 2006, Issue 8, p. nf12
[DOI: 10.1126/sageke.2006.8.nf12]


What Good Is Growing Old?

Renegade researchers argue that aging is adaptive for the community

Mitch Leslie

Don't mourn the loss of your wrinkle-free complexion, full head of hair, and wide-open, limber arteries. Don't fret if your knees now bend as easily as rusty hinges, if your endurance is flagging, your eyesight dimming, and your memory faltering. Your deterioration and demise are for the best. By conveniently dying, you will clear space for your children and possibly prevent the population from crashing to extinction.

Ascribing a function to aging is a sure way to tick off an evolutionary biologist. But a small group of researchers asserts that aging is an adaptation just like an eagle's keen eyesight or a leopard's camouflage coat. Citing theoretical studies and experimental results, these scientists contend that genes drive organisms to an early death to benefit relatives, neighbors, or the population in general. The notion that aging is programmed clashes with the conventional wisdom that creatures fall apart because natural selection becomes feeble as they age. Backers of the controversial idea, such as population biologist Joshua Mitteldorf of the University of Arizona, Tucson, call their views revolutionary. A century of research supports the conventional theory that explains how populations evolve, he says, and "you have to throw all that out to believe what I'm suggesting." But the extreme course is the only way to explain a wealth of data, he claims. Earlier this year, Mitteldorf published a new mathematical model that provides a possible explanation for how programmed aging could evolve. Most evolutionary biologists remain skeptical, but some say that reconsidering the idea might help us gain a clearer understanding of why we get old.

The Virtue of Selfishness

When you ask beginning students why we age, they usually respond that physical decay culls the old to make way for the young, says evolutionary biologist Ophélie Ronce of the University of Montpelier in France. That explanation carries a long pedigree--it dates back to Alfred Russel Wallace, the co-discoverer of natural selection--but most modern evolutionists spurn it.

They object to the assertion that organisms are programmed to die because it relies on a disputed form of natural selection called group selection. As the name suggests, traits evolve under this type of selection because they help a group of organisms survive, even if they harm individuals. Fifty years ago, some ecologists would invoke group selection to, say, explain why birds don't lay as many eggs as possible. Their restraint prevents overpopulation, the argument went. Similarly, group selection provided a rationale for why organisms grow decrepit and die: They are sacrificing themselves for the common good.

But as many researchers have pointed out, evolutionary scenarios based on self-sacrifice usually fail because cheaters prosper. A selfish creature that refuses to die on schedule, for example, gains more time to pump out offspring. Because egocentric organisms propel more copies of their genes into the next generation, eventually they swamp the altruists. Today, researchers generally accept that traits evolve because they help individuals persist and reproduce--what's known as individual selection. Group selection can work only under extremely rare conditions, says evolutionary biologist Steven Austad of the University of Texas Health Sciences Center in San Antonio, and therefore it can't explain a ubiquitous process such as aging.

At first glance, however, individual selection also seems incompatible with aging, which causes organisms to forfeit chances to reproduce. The leading explanation that reconciled the two phenomena coalesced in the mid-20th century and drew on insights from immunologist Peter Medawar, evolutionary biologists George C. Williams and W. D. Hamilton, and other researchers (see "Aging Research Grows Up" and Williams Classic Paper). Aging is not an adaptation or the result of a "death program," according to this theory, but a side effect of how natural selection changes over a lifetime. Even without senescence, organisms would die from disease, predators, and accidents. As a result, natural selection loses its power over organisms as they get older. To understand why, consider a harmful gene that sickens an animal in its youth. Because the gene curtails reproduction, natural selection can weed it out. But selection can't eliminate injurious genes that exert their effects later in life because the organism has almost completed its reproduction. The gene has little impact on the lifetime output of offspring. So genes that act up in old age can persist and trigger deterioration. Researchers are debating whether these genes start working at advanced ages or whether they aid organisms during youth and turn against them later on, a situation called antagonistic pleiotropy. Field and lab studies back the theory that aging results from weakening natural selection. For example, in a 1993 paper, Austad showed that opossums shuffling around an island that lacked predators aged more slowly than did animals from the neighboring mainland, where the opossum's enemies abounded. That finding meshes with the theory's prediction that organisms living in hazardous habitats should divert limited resources to reproduction--and thus break down faster. Although evolutionists need to sort out a few complications, the theory remains solid, says Austad.

Get With the Program

As solid as mush, counter some researchers. According to these heretics, experimental results offer clear evidence that aging is programmed, and scientists can't explain why deterioration occurs by invoking individual selection alone (1, 2, 3, 4). One line of support, they assert, comes from lab studies of long-lived mutant animals. For example, nematodes can survive twice as long as normal without daf-2 , which is a part of the insulin/insulin-like growth factor 1 pathway. Tampering with genes in the mouse and fly versions of the pathway also adds time (see "One for All"). Although some long-lived mutants are invalids, others are spry and appear to reproduce normally (see "Guinness-Bound"). Their vigor suggests that, contrary to what the standard theory predicts, organisms can live longer without incurring a cost such as reduced reproduction, says Mitteldorf. So dying sooner must provide some benefit, he claims, and genes such as daf-2 might promote aging. That the same pathways govern senescence in such distantly related species also supports the notion that falling apart is part of a genetic plan, says Mitteldorf.

Suicidal yeast supply further backing, say other researchers (see "I Regret That I Have But One Life to Give for My Colony" and "Death Comes for the Fungus"). No, the fungi aren't depressed about having to crank out yet another bland Merlot. But when researchers put the cells on a harsh diet, large numbers of them expire. They deploy proteins that spur cell suicide, which "suggests that this is an organized form of death," says geneticist Valter Longo of the University of Southern California in Los Angeles.

Work by Longo's group indicates that some cells choose death to help others ("Culture Clash"). After most of the yeast in a hungry colony have perished, a few hardy survivors resume reproducing. The liquid from colonies of suicidal yeast triggers faster regrowth than does broth from selfish cultures, which suggests that colonies recover by feeding on the remains of their selfless comrades.

A 2004 study of guppies also contradicts the standard theory, says Mitteldorf. In Trinidad, some of the gaudy, stream-dwelling fish share their habitat with predators. The evolutionary theory of aging predicts that threatened guppies should reproduce and age faster, as do the mainland opossums Austad studied. But in the lab, evolutionary biologist David Reznick of the University of California, Riverside, and colleagues found that guppies from the predator-prowled waters aged more slowly than did fish from benign environments (see "Prey for Long Life"). The rapid demise of the safe guppies might check their population growth, Mitteldorf speculates. Most of the guppies from the dangerous streams get snarfed, so they don't need such a control mechanism and live longer when predators disappear, his reasoning goes.

Some mathematical models also suggest population-level upsides to curtailing life span, such as slowing the spread of diseases (5). To Mitteldorf, the case is convincing: "There's strong experimental and observational evidence for aging as an adaptation."

Such talk riles most evolutionary biologists for two reasons (6). First, the idea that group selection controls populations "was thoroughly discredited by people like George Williams and John Maynard Smith in the 1960s," says Austad. "It would be a colossal waste of time and energy to refight those old, decisive battles today." Second, many researchers denounce the mere suggestion of programmed aging (see Olshansky Perspective). To understand why aging can't be programmed, Austad says, compare it to development, a process that biologists agree follows a script (7). Development involves intricate choreography of cell movements and scheduling of gene activity. Aging, by contrast, is messy. A cage full of inbred mice, which are nearly uniform genetically, will die at a variety of ages and from multiple causes, Austad notes. "If aging is programmed, it should happen in a more stereotyped way," he says.

The argument for programmed aging and group selection rests on the erroneous conclusion that long-lived mutants are healthy, says Austad. Data from the original papers describing these organisms reveal small but significant handicaps, he says, and careful experiments have accentuated the defects. When molecular geneticist Gordon Lithgow of the Buck Institute for Age Research in Novato, California, forced worms with daf-2 defects to compete for food against normal worms, the altered animals died out quickly (see "Paying the Price"). And a study by Wayne Van Voorhies of New Mexico State University in Las Cruces found that the genetically altered critters bit the dust when reared in rich Wisconsin soil, supposedly their native habitat (see "Death in the Dirt"). To programmed-aging doubters, these observations suggest that unless mutant nematodes are pampered, they don't measure up to normal worms.

The Selection That Dare Not Speak Its Name

Although most researchers keep their distance from group selection, Mitteldorf has been an unabashed groupie for more than a decade. But he admits that previous mathematical models of how group selection could promote aging don't work. They share a flaw, he says: The cost of shortening life span--fewer offspring--is immediate, whereas the benefits--less competition between old and young--accrue slowly. He's devised a new model that he says provides a quicker payoff for reduced longevity by damping fluctuations in population size (8).

The key factor in the model is the disparity between how fast a population grows and how fast its habitat renews. By favoring creatures that squeeze out more offspring, individual selection pushes up the population's growth rate and increases the disparity between the two parameters. As the gap grows, the changes in population size switch from gentle fluctuations to what mathematicians call chaos, in which the number of organisms can swing wildly. Once chaos sets in, the population is doomed; it will inevitably fluctuate to zero and become extinct. Aging can save the day, Mitteldorf's simulations indicate, because it prevents this erratic behavior. Instead of dying out, a population of aging organisms lingers at the edge of chaos, with individual selection continually nudging the growth rate higher and group selection tamping it down. The model shows how aging can be adaptive by helping to stabilize population dynamics, says Mitteldorf.

Figure 1
View larger version (28K):
[in this window]
[in a new window]
Chaos theory. As individual selection drives growth rates higher and higher, gentle oscillations in population size (top) give way to chaotic growth (bottom). Chaos eventually leads to extinction. [Source: Joshua Mitteldorf]

Family Ties

Other models suggest that senescence can be beneficial if it helps kin survive. The work comes from evolutionary biologist Calvin Dytham of York University in the U.K. and evolutionary ecologist Justin Travis of the Center for Ecology and Hydrology in Banchory, U.K. Both researchers are careful to say that they aren't modeling group selection. But they do stray from orthodoxy by incorporating programmed aging into their simulations, which probe whether organisms die before their time in order to reduce competition with their offspring. This suggestion involves another type of selection called kin selection, which W. D. Hamilton spelled out more than 40 years ago. Kin selection involves sacrifice for relatives, not for the group as a whole; it isn't controversial because it's an extension of individual selection. The measure of a gene's evolutionary success is how many copies of it wind up in the next generation. Hamilton noted that a creature can pass on its genes directly by having offspring. It can also pass on its genes indirectly by helping relatives--which carry many of the same genes--survive and reproduce. He showed that genes that promote sacrifice for relatives could evolve even if they lead to an individual's death. For example, kin selection explains why worker bees kill themselves to protect the hive; they die to save their mother and sisters.

To gauge whether kin selection adjusts life span, Travis and Dytham have crafted computer simulations of populations in which aging occurs. Under these conditions, dying might be the best thing an old, feeble organism can do to aid its younger, healthier relatives. Travis's 2004 model (9) backed that suggestion, revealing that kin selection could speed senescence. Long life evolved if the silicon creatures moved around freely. But early death evolved if organisms remained near their birthplace, where their demise would likely benefit kin. Using a more complicated model (10), Dytham and Travis showed similarly that the closer organisms remained to their home, the shorter their life span.

The researchers plan to test the models' predictions by comparing nematodes with different life spans. If kin selection is driving the evolution of aging in the animals, long-lived strains should be wanderers, whereas animals that die young should be homebodies. Human life span might even be evolving because of kin selection, Travis says. Because fewer and fewer people live where they were born, kin selection might be weakening in our species, and its decline could gradually stretch human longevity.

Kin selection could also explain kamikaze yeast. In the wild--or at least in the vineyard--a colony of the fungi spreading across a grape will probably consist of a single cell's descendants. But Longo notes that the situation doesn't meet Hamilton's criterion, which requires that beneficiaries outnumber altruists. The opposite occurs in yeast--most of the colony perishes to save a few cells. Group selection might offer a better explanation, Longo says.

But dying might not be the best way to help your relatives, at least in social species. If offspring need lots of nurturing to survive, living longer might be a better strategy for parents and other relatives. That's the conclusion from a model by demographer Ronald Lee of the University of California, Berkeley (11), which gauged the effect on mortality of care by parents and other relatives. The same idea might explain why women and females in some other species live beyond the age when they stop reproducing. Some researchers argue that this extra time evolved because oldsters can then help raise their relatives' offspring, the so-called grandmother hypothesis (see Holmes Perspective). Whether life after menopause is an evolved trait or a consequence of modern medical care remains controversial, however.

Figure 2
View larger version (176K):
[in this window]
[in a new window]
Born to baby-sit? Women survive past their childbearing years so they can help care for relatives' offspring, according to the grandmother hypothesis. [Credit: James Ross/Taxi/Getty Images]

Rehashing or Revamping?

Whether the researchers investigating programmed aging and group selection are bravely challenging orthodoxy or belaboring ancient arguments depends on whom you talk to. But some middle-of-the-road scientists such as Ronce see new questions opening up. Researchers should consider the possibility of natural selection operating at many levels of biological organization, even down to organelles within cells, she says. Mitochondria provide one example. The organelles reproduce independently of the cell, and mutations that speed mitochondrial duplication are harmful to yeast cells (12). So scientists might have to modify the theory of aging to include more than individual selection, she says. Future work should determine how factors such as kin selection and population growth patterns affect life span, she says. If kin selection does prove to be a major force in our deterioration, older folks will have something else to make their kids feel guilty about.

May 3, 2006

A writer in sunless Portland, Oregon, Mitchell Leslie hopes that gray hair is adaptive.

  1. V. P. Skulachev, Programmed death phenomena: From organelle to organism. Ann. N.Y. Acad. Sci. 959, 214-237 (2002). [Abstract] [Full Text] [Medline]
  2. D. E. Bredesen, The non-existent aging program: How does it work? Aging Cell 3, 255-259 (2004). doi:10.1111/j.1474-9728.2004.00121.x [CrossRef][Medline]
  3. V. D. Longo, J. Mitteldorf, V. P. Skulachev, Programmed and altruistic ageing. Nat. Rev. Genet. 6, 866-872 (2005). doi:10.1038/nrg1706 [Medline]
  4. J. Mitteldorf, Ageing selected for its own sake. Evol. Ecol. Res. 6, 1-17 (2004).
  5. J. W. Kirchner and B. A. Roy, The evolutionary advantages of dying young: Epidemiological implications of longevity in metapopulations. Am. Nat. 154, 140-159 (1999). [CrossRef]
  6. T. B. L. Kirkwood, Understanding the odd science of aging. Cell 120, 437-447 (2005). doi:10.1016/j.cell.2005.01.027 [CrossRef][Medline]
  7. S. N. Austad, Is aging programmed? Aging Cell 3, 249-251 (2004). doi:10.1111/j.1474-9728.2004.00112.x [CrossRef][Medline]
  8. J. Mitteldorf, Chaotic population dynamics and the evolution of aging. Evol. Ecol. Res. 8, 1-14 (2006).
  9. J. M. J. Travis, The evolution of programmed death in a spatially structured population. J. Gerontol. A Biol. Sci. Med. 59, B301-B305 (2004). [Abstract] [Full Text] [Abstract/Free Full Text]
  10. C. Dytham and J. M. J. Travis, Evolving dispersal and age at death. Oikos, 23 February 2006 [e-pub ahead of print]. doi:10.1111/j.2006.0030-1299.14395.x
  11. R. D. Lee, Rethinking the evolutionary theory of aging: Transfers, not births, shape senescence in social species. Proc. Natl. Acad. Sci. U.S.A. 100, 9637-9642 (2003). doi:10.1073/pnas.1530303100 [Abstract/Free Full Text]
  12. D. R. Taylor, C. Zeyl, E. Cooke, Conflicting levels of selection in the accumulation of mitochondrial defects in Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. U.S.A. 99, 3690-3694 (2002). doi:10.1073/pnas.072660299 [Abstract/Free Full Text]
Citation: M. Leslie, What Good Is Growing Old? Sci. Aging Knowl. Environ. 2006 (8), nf12 (2006).

Science of Aging Knowledge Environment. ISSN 1539-6150