Of Worms, Mice & Men: Altering Rates of Aging
By Partridge, Linda
The word ‘aging’ has very different implications for wine and women. Matters can improve or deteriorate over time, and the everyday usage of the term can cover either situation.
In the natural world things tend to get worse over the course of adulthood. In mammals and birds, for instance, many long-term field studies have revealed that both likelihood of survival and production of offspring decline later in life. There are exceptions, to be sure. Especially where animals start to procreate before they are fully grown, as is the case in many fish, chances of survival and fecundity can improve over at least part of adulthood.
But in the organism in which most people are really interested, humans, there is a clear deterioration in function over time, after puberty. This is the sense in which I will use the word ‘aging’ in this essay, as the intrinsic decay in function that sets the ultimate limit to life span.
At first sight, aging does not require any special explanation. Machines wear out and fail, so why not living things too? Aging does indeed involve accumulation of damage. The molecules that make up bodies acquire lesions, as do whole cells and tissues. Decrements in the ability to run or to solve problems quickly, for example, reflect this accumulation of damage.
If this were the whole story, then we would expect similar organisms to age at roughly similar rates, but they do not. For instance, bats and monkeys are peculiarly long-lived mammals for their body size; birds, in general, also live much longer than comparably sized mammals. These differences persist in captivity and are thus properties of the species rather than of the individual lifestyle. The rate at which organisms age can, therefore, evolve. By changing the rate at which unrepaired damage accumulates with time, certain biological processes are able to influence the rate of intrinsic decline. It is the activity of these processes that can change during evolution.
Considering its genetic basis and evolutionary change, aging has some very odd characteristics. It is unconditionally deleterious ; yet, as far as we know, no genes have evolved to directly cause accumulation of damage, lowered fertility, and death. Earlier ideas that the aging of individuals could be beneficial to the species, by removing the old to make way for the young, are now very largely discredited. If aging itself is a disadvantage, then it cannot evolve by natural selection in its favor and must instead evolve as a side effect of something else.
Several of the intellectual giants of theoretical populations genetics and a large amount of empirical research in the laboratory and the field have produced quite a clear picture of the two routes through which aging can evolve.
First, it can appear as a side effect of mutation pressure, in a process known as mutation accumulation. Sporadic alterations to the genetic material transmitted from one generation to the next can often cause genetic diseases. Some of these diseases, however, become manifest only as the bearer of the mutation gets older, as in the case of Huntington’s disease. The later in life that the mutation’s bad effects appear, the greater the mutation’s chance of reproduction because external hazards such as infections, predators, and accidents will cause attrition of the bearers of the mutation and not all of them will survive long enough to express its effects. Natural selection therefore acts more weakly to remove these later- acting mutations from the population. The balance between the occurrence of mutations and their removal from the population by natural selection is hence shifted ; later-acting mutations can achieve a higher frequency in the population than can equivalent mutations that also affect the young. These late-acting, deleterious mutations can therefore cause aging.
In addition, aging can evolve as a side effect of earlier success. If a mutation benefits the young, perhaps by making them more fecund, natural selection can act in its favor, even at the price of a higher subsequent rate of aging. Again, death by natural hazards means that more of the mutation’s bearers will survive to gain the early benefits than to show the elevated rate of aging.
In either case, whether as a side effect of mutation pressure or of earlier reproductive benefits, the intrinsic rate of aging seems to evolve according to the level of external hazard. In an environment where external risks minimize the likelihood of survival beyond age 10, natural selection will not eliminate a mutation that causes problems for n-year-olds from the population. However, if the external hazards ameliorate so that survival to age 14 becomes routine, then natural selection against the mutation will become stronger.
We therefore expect to find slow-aging creatures in less hazardous environments. This is roughly what we observe in nature. Species that are well protected such as tortoises and turtles, species that can fly such as birds and bats, and mammals that live in trees instead of on the ground are all capable of longer lives.
Aging involves change in the organism with time, which has often led people to think that it is a process like development or growth. Our evolutionary understanding of aging says otherwise. Aging evolves in response to extrinsic hazard, with more rapid aging arising in more dangerous environments, as a side effect of accumulating more mutations that either affect an organism later in life or cause more intense early reproduction. Aging is therefore not a programmed process like development or growth, with a well- orchestrated hierarchy of genetic control ensuring that the right things happen in the right place and at the right time to make a well-formed organism. The genes that affect aging did not evolve to control it therefore it is a much more haphazard and variable kind of process. In humans, the sorts of damage that accumulate are very distinct in different parts of the body, and the rate at which these forms of damage accumulate can vary greatly between individuals. People develop different problems from each other as they age, and they die of diverse causes. Because aging is such a complex process, it probably involves a large variety of parallel processes. Therefore, many rather than few genes likely influence aging.
Aging’s complexity has long colored scientific and medical attitudes toward it. Scientists have tended to assume that aging is too difficult and intractable a trait for experimental analysis. Although aging is the major risk factor for multiple diseases, including major killers such as cancer, circulatory disease, and neurodegeneration, the medical community treats these diseases separately rather than as different manifestations of a single, underlying aging process. Aging is seen as inevitable, too complicated to do anything about, and best treated by piecemeal intervention into its undesirable manifestations.
Tempering pessimism about the prospects for intervening in the aging process are the dramatic improvements in health during aging that have occurred in industrialized human societies worldwide. The rate of aging certainly has a genetic and evolutionary basis, but life span can vary depending upon the environment encountered. Beginning in the mid-nineteenth century, survival rates have risen steadily, probably mainly because of various public-health measures such as improved sanitation and hygiene and a great reduction in the impact of infectious diseases. These increases in survival have affected all age groups. Thus not only average life expectancy but also the longevity of the oldest segment of the population has increased. Furthermore, there is no sign that the upward swings in survival rates are slowing down for any age group, which suggests that there is no impending limit to maximum human life span. Such a limit may exist, but at the moment we cannot see what it is.
People of all ages are certainly healthier than they used to be, although the approaching wave of obesity in the young, if unchecked, may counter this trend. But is the improvement in health the result of the slowing of the aging process itself? The figures suggest not. If we define a population’s death rate as deaths per thousand in a year, then this rate increases with chronological age – aging. Were aging checked, we would see a decline in this latter rate, say over ten-year blocks of time, but we do not. Rather, survival at all ages has increased, with no evidence of a slow down in the rate of aging itself. Human death rates tend to be somewhat elevated at birth, fall to a minimum around the age of puberty, and start to rise steadily thereafter. This pattern has remained unchanged over the one and a half centuries during which life expectancy has lengthened, and the rate of increase in death rates after puberty has not declined. Health during aging is better, but the underlying aging process seems to have eluded modification. There is good news and bad.
As a result of longer life expectancy, industrialized societies face many challenges, particularly for health-care systems. Although older people are now healthier, health-care demand is steadily growing as more people reach the older ages at which aging-related health problems occur. This provides strong motivation for the biomdical community to undertake scientific research that will ameliorate the i\mpact of aging-related disease and disability. In recent years, the discovery that mutations in single genes can greatly lengthen healthy life span in experimental animals has galvanized research into the mechanisms of aging. This has come as a surprise to many, and it has opened up new vistas in our understanding of how healthy life span is controlled.
The process of discovering the single gene mutations that extend life span started with a tiny roundworm called Caenorhabditis elegans. It was in the context of studies of these worms’ development that the first mutation came to light. Before becoming reproductive adults, the worms can take two different developmental routes. If conditions are good, they grow straight through to adulthood and start reproducing. If, on the other hand, the worms are crowded or short of food, they arrest their development and form a dauer larva. Dauer larvae stop feeding, store fat, resist various environmental stresses, and are very long-lived. Because of these characteristics, they can sit out hard times and resume normal development when conditions ease. Initially, work on dauer development produced mutations in single genes that caused the developing worms to form dauers even under good conditions. Different, weaker mutations in these same genes, however, made the adult worm long-lived. And one of these mutations turned out to be in a gene that encoded a part of a signaling pathway that was clearly similar in its evolution to the insulin and insulin-like growth factor signaling pathways of mammals.
These findings were fascinating. They revealed that adult life span could be under simple genetic control, and that a signaling pathway more familiar for its effects on the regulation of blood sugar and growth in mammals could also affect life span in a lowly invertebrate. However, the full implications of these findings took some time to become apparent. For starters, the long life of the mutant adult worms was probably simply a reexpression in the adult of the traits that make for long life in dauer larvae. If this were true, then the findings would unlikely to be of any relevance to mammals, which do not have dauer larvae or their equivalent. Furthermore, it was quite unclear how this insulin-like pathway could have any bearing on life span.
There matters rested for some years until an insulin-like signaling pathway came to light in another invertebrate inhabitant of research laboratories, the fruit fly Drosophila. Again, the initial discovery occurred during studies of development rather than of aging. This insulin-like signaling pathway controls growth in the pre-adult period. If the activity of the pathway is elevated, growth rate and the size of the adult are increased. Conversely, lesions in components of the pathway result in dwarf adult flies. Because of this work, scientists were able to make mutations in several genes that encode components of the pathway and measure the effect on adult life span. Some of these mutations did indeed increase life span. The results for Drosophila were also interesting, for instance, in showing that lesions in the pathway seemed to increase life span much more in females than in males and to impair the fecundity of females.
But the findings from the fly had broader implications. The worm and the fly are very different kinds of organisms ; they are only distantly related to each other and have diverged over millions of years of evolutionary time. If the genes of the insulin-like pathway could affect life span in both, then it was distinctly possible that the pathway could control life span in mammals as well.
Some straws in the wind were already suggesting that insulin- like growth factor signaling could play a role in the control of mammalian life span. It was known, for instance, that several mutations that cause lesions in the development of the pituitary gland and in the signaling by the growth hormone produce long-lived dwarf adult mice. Then, in 2003, two key papers implicated both the insulin-like growth factor signaling pathway and the insulin signaling pathway in the control of life span in the mouse. In one, scientists manipulated the gene that encodes a receptor on the surface of cells that responds to signals from the insulin-like growth factor. They reduced the number of copies of this gene from two copies to one throughout the mouse. The resulting female, but not male, mice lived longer. In the other, scientists removed the receptor on cell surfaces that reacts to insulin from the mice’s fat cells, their white adipose tissue. The engineered mice were leaner and longer-lived.
These results had huge implications. The insulin or insulin-like growth factor signaling pathway, which is present in all multicellular animals, appears to have conserved one of its functions, the control of life span, over the very large evolutionary distances between the invertebrate worm, the fly, and the mouse. The pathway is therefore a strong candidate for the control of human life span. Furthermore, the evolutionary conservation of the pathway means that we can use powerful analytical methods with fewer ethical implications on the invertebrates, with all their advantages of relatively short life spans (about three weeks in the worm and three months in the fly, as opposed to three years in the mouse), simplicity, and low maintenance costs to understand how this pathway might control mammalian life span. This has long been the Holy Grail of aging research. But it opens up a major paradox.
How can a mutation in a single gene produce such a large increase in life span if many genes influence the multiple, parallel processes that control the rate of aging?
A clue may come from the effects of diet. Dietary restriction is an environmental intervention that, along with mutations in single genes, has long been known to extend life span in laboratory rodents. Its effects have been conserved over evolution ; first discovered to extend life span in laboratory rats in 1935, dietary restriction has since been shown to have a similar effect in organisms as diverse as yeast, worms, flies, and mice as well as other less intensively studied species. During dietary restriction the amount of food an animal consumes is reduced dramatically. In rats and mice a reduction to about 50 percent of voluntary levels can produce a substantial increase in life span. Dietary restriction does not merely reverse the effects of a sedentary existence and overeating. In worms, flies, and mice dietary restriction also reduces fecundity; female mice subjected to strong dietary restriction become completely infertile.
This correlation between the longer life span that results from dietary restriction and lowered fertility has led some to suggest that reduced fertility is an evolved, adaptive mechanism for sitting out hard times. Reproduction is expensive in nutrients and can compromise the survival of the parents. If food is scarce, parents cannot produce many offspring anyway and their offspring’s likelihood of survival is low. Under these circumstances, it may pay for a parent to lower its reproductive rate and thereby increase its own chance of survival until the food supply improves.
Numerous species show greater longevity and reduced fecundity in response to lowered food intake. However, we do not know precisely what it is about lower intake that influences longevity, for instance, whether the overall intake of energy or particular dietary components are critical. Nor do we fully understand how the animal senses the change in nutrition and how it uses this information to change its internal state and extend its life span. For these reasons, it is not yet certain if different animals achieve these responses to dietary restriction in the same way, as we would expect if the processes at work have been conserved over large evolutionary distances. Thus, we cannot exclude the alternative possibility – that these responses have evolved independently in different lineages. Still, there is intense interest in the mechanisms by which dietary restriction lengthens life span in laboratory organisms.
Dietary restriction also keeps laboratory rats and mice healthy for longer, delaying the impact of aging-related disability and disease. Dietary restriction enables animals to remain active and able to reproduce for longer, maintains better immune function, slows down the changes in the musculoskeletal system that alter body shape as the animal ages, preserves the structure and function of the nervous and endocrine systems, and reduces the frequency of cancers and other diseases. In the fly, too, several lines of evidence, including the ability to reproduce for longer, suggest that dietary restriction helps animals remain youthful. For these reasons, it seems that dietary restriction really might slow down the aging process itself. We might, then, have two interventions – conserved by evolution that can increase life span by slowing down the aging process. But if aging occurs through multiple, parallel pathways of accumulation of damage, how could dietary restriction slow down all of them? As for the effects of the insulin or insulin- like growth factor pathway, the evolutionary and mechanistic findings seem to be at variance with each other.
We are making steady progress toward understanding how lesions in the insulin or insulin-like growth factor signaling pathway and dietary restriction extend life span. We have already made some important findings. One is that these interventions do not slow down the rate of living; the metabolic rate of these long-lived animals is normal. Nor does the reduction in reproductive rate we often see in conjunction with the increase in life span appear to play any simple direct role. We have also sequenced all the genomes of the worm, fly, and mouse, opening up powerful avenues to understanding how genes control vital processes. \While we have a long way to go to this goal, it is becoming clear that when life span is extended, the genes that control processes such as detoxification and turnover of damaged molecules, resistance to stresses, immunity and inflammation, and the metabolic pathways by which nutrients are stored and used can all show changes in expression. Some of these may be irrelevant to the extension of life span; only experimental work will eventually reveal which are the critical changes, where in the organism they occur, and how exactly they lead to increased survival.
The discovery of interventions – be they diet or single gene mutations – that have the potential to improve health during aging is a massive step forward. The current intense interest in their mode of action will undoubtedly continue amidst high hopes that the results will translate into medical practice. But do these interventions really slow aging, or could they be prolonging life in some other way?
Since aging is the accumulation of unrepaired, irreversible damage, we would expect an organism that has aged more slowly to have less cumulative damage. Flies that have lived at different temperatures demonstrate this phenomenon. Flies – reared to adulthood at a single intermediate temperature and then kept at different temperatures as adults live longer the lower the temperature at which they are kept. Flies are much too small to regulate their own body temperature and immediately adopt the temperature of the environment in which they live. Lower temperatures slow down the rates of all chemical processes, including, presumably, those leading to death.
Low temperatures extend the life span of flies by slowing down the rate of aging. If flies that have been in the cold for varying lengths of time are moved to higher temperatures, their death rates are lower than those of flies that have been previously kept at the higher temperatures. Also, the degree of protection that they enjoy is greater the longer that they are left in the cold before being transferred to the warmer environment. Similarly, the death rates of flies switched to a cooler environment after living in high temperatures are permanently elevated above those of flies with a history of life in a cooler environment and the longer that they have stayed in the warmer conditions the higher their death rates. These findings demonstrate that, in flies, warmer temperatures induce more irreversible damage that leads to death ; in other words, they elevate the rate of aging.
This kind of experimental approach, where animals are switched between regimes partway through adult life, has revealed some surprising findings about dietary restriction in flies. When flies that have been on a normal diet are subjected to dietary restriction, they adopt within forty-eight hours the lower death rates of flies that have been on restricted diets. Similarly, when flies that have been on restricted diets are fully fed for the first time, they show within fortyeight hours the elevated death rates characteristic of flies that have been on normal diets throughout adulthood. It does not seem to matter how late in life they switch ; after a short lag their life spans show no memory of nutritional history, and the death rates of flies that have switched diets converge with those of the flies that have been kept permanently in that nutritional regime. At least in flies, then, dietary restriction does not slow down aging. Rather, dietary restriction somehow acts acutely to make the flies less likely to die from the damage that they have accumulated.
We do not know if dietary restriction has similarly acute effects on death rate in mammals because we have not carried out the appropriate experiments yet. If the effects were acute, then we would expect the switch in death rates after a change in nutritional regimes to take longer in mammals than in flies, perhaps on the order of some weeks. At first sight, the idea that dietary restriction acts acutely to reduce death rates seems incompatible with the finding that it also delays the impact of agingrelated disability and disease. However, these two findings can be reconciled. Dietary restriction could act acutely to lower the likelihood that aging-related damage will lead to the appearance of a lethal, aging-related pathology. The lag period during which death rates switch over to those characteristic of the new nutritional regime would then represent the period during which individuals that have acquired a lethal pathology are lost from the high-risk population, with a switch to dietary restriction, or gained by it, in a switch to full feeding.
We have not yet determined, even in flies, if the reduced activity of the insulin or insulin-like growth factor signaling pathway extends life span by slowing the rate of aging or through an acute effect. Answering that question will require an experiment switching the activity of the pathway partway through adulthood, which is now coming within the realms of technical feasibility. These are important questions for the future. The acute effects of dietary restriction in the fly have unexpectedly revealed the existence of a type of intrinsic risk factor whose mechanistic basis requires elucidation. If similar acutely acting interventions exist for humans, we could eliminate the adverse long-term consequences of some risky habits.
So far, we have not proven that dietary restriction or single gene mutations extend life span by reducing the rate of accumulation of aging-related damage in any organism. The information that we have about their effects is compatible with the alternative possibility that they increase life span by making animals less likely to die of damage already accumulated, leaving the rate of aging itself unaffected. Perhaps the pervasive effects of altered temperature for ectothermic animals and the slow, cumulative effects of evolution really are the only means by which the rate of aging can be altered.
The fact that we may not yet be able to alter its rate in humans is disappointing. But the discovery that there are interventions that can act acutely to delay the impact of multiple forms of aging and pathology is in some ways even better. For those with unhealthy habits, it may mean it is never too late to adopt a lower-risk lifestyle.
Linda Partridge is Weldon Professor of Biometry and BBSRC Professorial Fellow at University College London. The author of numerous articles and the editor of several books, Partridge has researched the genetics of fitness-related traits, with a particular interest in life histories and aging. Currently, she is working on understanding how dietary restriction and mutations in the insulin or insulin-like growth factor signaling pathway extend life span.
2006 by the American Academy of Arts & Sciences
Copyright MIT Press Winter 2006