Sci. Aging Knowl. Environ., 3 October 2001
HOT TOPIC ORIENTATIONS
More Than a Sum of Our Cells
After decades of research, the changes in cellular function that underlie aging remain a major source of debate--and an area of intense investigation
Abstract: Cells in the body grow and die, cells in lab dishes grow and die, and individual organisms grow and die. The parallels seem maddeningly obvious, but scores of scientists still labor to draw the correct connections, to uncover the mechanisms that underlie aging in cell culture flasks and in whole animals. Do our cells stop growing, quit working, cease dividing, or start dying as we age? Do we die when our cells do, or are we somehow more than the sum of our cells? For decades, scientists have searched for evidence that links changes in cell growth, cell function, cell division, and cell death to the phenomenon we call aging. Although definitive proof eludes them, researchers continue to conduct experiments in tissue culture and in animal models, amassing information that points us toward a greater understanding of what aging is--and is not.
Complex Questions, Ambiguous Answers Back to TopHerrick's inspirational words remind us that aging, in many ways, is in the mind of the beholder. We are old when we no longer seek new experiences, challenge our intellectual foundations, indulge our curiosities, or surrender to our whimsies.
But the maxim may also resonate in a strictly biological realm. Perhaps we age because our cells stop growing or stop dividing. For some cells it wouldn't matter--cells in the brain and the heart generally don't divide in an adult. But maybe other types of cells begin to dodder or to die long before we switch to bifocals or experience that first "senior moment." A question then arises: How might such changes in cellular activity or fecundity render us more susceptible to the diseases and dysfunction we associate with growing old?
Although researchers have been investigating this question for decades, their studies have yet to produce incontestable conclusions. Part of the problem lies in the difficulty of experimental design. How, for example, would one determine whether some cells cease to divide in a living organism, and whether that failure to replicate drives tissue deterioration, and indeed, aging itself? "If the definitive experiment were easily doable, it would have been done," says Judith Campisi, a cell biologist at Lawrence Berkeley National Laboratory in Berkeley, California. Instead, scientists interested in aging must fashion tests in cell culture or animal models, indirect studies whose results are open to different interpretations. Although the ultimate answers regarding the cellular basis of aging remain elusive, the ongoing research efforts allow biologists to accrue evidence that points the way toward a better understanding of how cells, tissues, and organisms change with age.
Technically speaking, cells do not age. "Aging is something that turns a healthy young animal into an old one that's more vulnerable to disease," says Richard Miller, an immunologist at the University of Michigan, Ann Arbor. Because cells are not animals--with the exception of, say, yeast, which is a single-celled organism (see Kaeberlein Perspective)--they don't really fit the definition of "things that age."
But a cell's survival and function clearly play into the well-recognized phenotypes associated with aging (1). When cells in the fatty layer under the dermis disappear, our skin wrinkles; loss of bone-forming osteoblasts leads to osteoporosis; depletion of skeletal muscle fibers weakens muscles; and diminished T cell performance contributes to an enfeebled immune system.
Although we don't all experience identical patterns of decline--some octogenarians are crippled by arthritis but have strong hearts, others sport healthy joints but suffer from cardiovascular problems and cognitive deterioration--tissue aging appears to be synchronized, albeit imperfectly. "In mouse it takes 2 years for all the tissues to go bad; in a dog it takes 10 years; in a human, 80," says Miller. "So there's a fundamental mechanism that we have to pay attention to." But what might that mechanism (or those mechanisms) be?
For a long time researchers hung their hopes on replicative senescence, a state in which cells no longer proliferate (2). Perhaps, they reasoned, senescence in a whole organism--a product of pandemic tissue malfunction--hinges on senescence at the level of the cell. Cells that would normally continue to proliferate throughout the lifetime of the organism--such as those in the skin, the intestine, and the immune system--in older individuals might simply throw in the towel, exit the cell cycle for good, and never divide again.
Pushing the Hayflick Limit Back to TopLeonard Hayflick was working at the Wistar Institute in Philadelphia when he discovered that human cells in culture divide a finite number of times and no more (3, 4). He was trying to isolate cancer-causing viruses by exposing normal human cells to cancer-cell extracts and then identifying the carcinogenic culprit. But Hayflick never got past the first step because his cell cultures kept crashing--dividing only 50 or 60 times before they went kaput. Thus Hayflick concluded that normal cells in culture divide a limited number of times--a number eventually dubbed the Hayflick limit--before they senesce.
Perhaps, Hayflick and others then reasoned, cells in the body have similar limitations, depleting their replicative potential after dividing a set number of times (4, 5). And maybe, over a lifetime, such cellular exhaustion can account for the changes of function that accompany aging. An attractive hypothesis, but how to test it?
In the early 1960s, George Martin of the University of Washington, Seattle, removed cells from old animals and people, and tested whether they could divide in culture. Cells from younger individuals continued to divide after cells from older donors retired, suggesting that cells' ability to replicate may decline with age (6, 7). Similar studies revealed that cells from patients with Werner syndrome, a form of premature aging that takes hold after puberty and causes those affected to appear old by age 30 or 40 (see "Of Hyperaging and Methuselah Genes"), also show diminished replicative endurance.
Then in 1998, Vincent Cristofalo, now at the Lankenau Institute for Medical Research in Wynnewood, Pennsylvania, and his colleagues published a report stating that the donor age does not dictate the cells' replicative ability--or inability. Cell lines established from skin biopsies of young people were just as likely to proliferate poorly as cell lines prepared from their more senior comrades (8).
To resolve the discrepancy, Martin scrutinized his own data--and discovered that the correlation he was seeing related to whether the skin samples came from donors who were living or dead. Remove the data obtained from autopsy material, and the relation between donor age and replicative ability evaporates (7).
Culture Wars Back to TopRegardless of the results, most researchers now agree that tallying the number of times cells can divide in culture may not reliably measure their replicative vigor in a tissue. By the time a culture has gone through 50 generations, only the offspring of the most prolific cell remain. In a typical skin biopsy, fibroblasts that have already divided dozens of times live side by side with cells that have not yet duplicated once. Culture conditions will select for the most robust cells in this mosaic: the ones that, by chance, can multiply more rapidly. These types of experiments, therefore, do not necessarily reflect the proportion of cells in a tissue explant that can proliferate, but rather, how many times the ones most dedicated to propagation can spawn descendants (9) (Fig. 1).
The bottom line: Whether cells survive in culture is pretty much "a matter of luck," says Cristofalo. For example, allow a single cell to divide, and then separate the sister clones into two cultures. These cultures don't divide the same number of times--a result that highlights the limitations of using such culture studies to draw inferences about replicative capacity, much less the aging process.
"It's a shame that people trapped themselves into thinking that aging is something that happens to cells growing in a plastic dish," says Miller. Of course no one advocates discarding the data with the depleted media. Culture experiments have proven invaluable for probing cellular processes and modeling cell behavior. But conclusions must be drawn with care--and confirmed in vivo. "I'm not dismissing cell culture observations," says Harry Rubin, a cancer biologist at the University of California, Berkeley, who has worked with cell culture systems for 50 years. "But you have to be prudent about your interpretation."
Tweaking the experimental design can also render cell culture studies easier to evaluate. For example, to skirt the problem of measuring the relative health of only a subset of cells--the ones that are strong enough or lucky enough to survive in culture--some researchers have turned to so-called clone size distribution studies (10-13). In these experiments, investigators separate the initial tissue sample into individual cells and then follow the fate of each clone in culture, counting how many times both the lackadaisical cells and their more fruitful relatives double (Fig. 2). These experiments allow researchers to assess whether, as a collection, the cells taken from a younger donor tend to divide more extensively than cells from an elder.
Here the results seem more definitive. Although some cells surrender quickly while others thrive--no matter what the age of the donor--younger organisms have a higher proportion of cells with greater replicative potential than do older ones, according to studies performed by Martin and by Norman Wolf and his colleagues at the University of Washington, Seattle (11-13).
Still, a culture dish is not an animal. So Wolf and his colleagues have attempted to extend their findings to living organisms. In fact, their studies in intact animals reinforce the observation that cells in aging animals show a waning enthusiasm for replication. The researchers implant in living animals a tiny pump that releases BrdU, a nucleotide analog that gets incorporated into the newly synthesized DNA in replicating cells and can be detected with an appropriate antibody. Using this labeling technique in mice and other mammals, Wolf finds that in a variety of tissues--skin, liver, kidney, and pancreas--cells in the younger animals divide more frequently than do cells in the older animals (13, 14). He also finds that calorie restriction, the only intervention known to extend life-span in mammals, seems to restore cells in older animals to a more youthful state, replicatively speaking.
Although these animal studies suggest that some cells in an aging organism lose their drive to divide, they don't explain why. Maybe in older animals, cells no longer receive the proper growth or proliferation signals--the correct hormones and cytokines--or sufficient nutrients. For that reason, many researchers prefer to study cells in culture, even though such experiments have their limitations. In vitro, the concentrations of nutrients and growth factors can be carefully controlled and maintained at the same level for all cells. But plopping cells into a culture dish is not the only way to level the physiological playing field. Peter Hornsby, a cell biologist at Baylor College of Medicine in Houston, finds that young animals can also provide a stimulating environment for cells of all ages.
Young Mice: A Different Dish Back to TopHornsby and his colleagues are moving away from simply counting the number of times aging cells can divide, instead assessing whether older cells are still capable of acting young, physiologically speaking, when given the opportunity. The researchers remove steroid-producing cells from the adrenal cortex of a cow or a human and transplant them into young immune-deficient mice--young so that the circulating growth factors will promote the cells' survival, and immune-deficient to circumvent rejection. They are then able to evaluate whether the cells form functional tissue that produces hormones--and whether that ability differs depending on the age of the donor or the number of times the donor cells are forced to divide in culture before transplantation (14-17).
The results: Bovine adrenocortical cells transplanted into mice can form cortisol-producing tissue that promotes the survival of mice whose own adrenal glands have been removed (17). And it works with human donors too--the cells produce plenty of hormone and the animals survive--regardless of the age of the donor, in this case ranging from 6 to 50 (16). However, when the adrenocortical cells are aged in culture--allowed to undergo 35 or more population doublings--the transplants fail and the mice die.
Hornsby concludes that something happens to cells dividing in culture that does not happen when they age in vivo. Perhaps some change in their patterns of DNA methylation or chromatin structure alters their gene expression profiles and affects their ability to function and produce hormone, he suggests (see "Dangerous Liaisons").
But even old cultured cells--cells that have doubled more than 100 times--can be coaxed to form functional tissue following transplant, Hornsby finds. All they need is a boost from an enzyme that restores their telomeres (15).
Telomere Tales Back to TopTelomeres are the structures, made of repetitive bits of DNA sequence and associated proteins, that cap the ends of our chromosomes. Thanks to a quirk in the design of the DNA replication machinery, chromosomes lose a bit off their ends when a cell copies its DNA. Each time a cell divides, its telomeres shrink. At some point they appear to reach a critical length--so withered that chromosomes start fusing at the ends, which precipitates disaster for the cell.
Thus, scientists reasoned that shortened telomeres were a sure sign of aging--or at least a good indicator of how many times a cell has divided. The connection between telomeres and aging appears to hold true in cell culture. Numerous studies have confirmed that as human cells divide in culture, their telomeres shorten, until the cells senesce and cease to divide (14, 18-21). Furthermore, many cells can escape this replicative senescence by the addition of telomerase--the enzyme that synthesizes and maintains telomeres (22, 23) (see the Genes/Interventions Database).
But what happens to telomeres in vivo? In mice, the situation is complex. Unlike humans, mice express telomerase in all of their non-sex cells, which keeps their telomeres luxuriant (24, 25). To examine the relationship between telomere length and aging in mice, Ron DePinho and his colleagues at the Dana-Farber Cancer Institute in Boston generated mice that lack telomerase (26). As these animals age, their telomeres indeed grow short. And after six generations, the mice show a decreased life-span and a reduced ability to heal wounds or handle stress, such as radiation (26, 27). The animals do not, however, suffer many of the other maladies and malfunctions normally considered hallmarks of human aging, such as cataracts, osteoporosis, or heart disease--a finding that leads DePinho to conclude that telomere erosion is not the mechanism responsible for aging. "It would be na�ve to think that in telomere shortening we found the magic pathway to aging," says Jerry Shay of the University of Texas Southwestern Medical Center (UT Southwestern) in Dallas. "It may be part of the story, but [it's] not the whole story."
Gauging how big a part telomere wasting might play in human aging is made even more difficult by the fact that, in people, the relation between age and telomere length differs from tissue to tissue (14). "Some show shortened telomeres, some don't," says Hornsby, who has measured telomere shortening with age in human adrenocortical cells. Some of the strongest evidence that telomeres dwindle through the years comes from cells in the blood cell lineage. Telomeres are shorter in peripheral white cells from older individuals than in those from youthful cohorts (19, 28, 29). Perhaps, some argue, this observation explains why the immune system weakens and fails to fight infections as we age, or why older people are more sensitive than young ones to cancer chemotherapy, which depletes immune cells that must then be replaced by rigorous proliferation of surviving cells.
Not all studies, however, support a steady diminution of immune-cell telomeres with age (30, 31). Elizabeth Blackburn and her colleagues at the University of California, San Francisco, examined blood cells from people from infancy to 80 and found that their telomeres do shrivel with age, but they diminish most dramatically between birth and the age of 4--a time of life that does not spring to mind when we think of aging (Fig. 3). In this study, telomere lengths appeared to hold steady from age 4 to about 24--a period that boasts abundant cell proliferation. Telomere lengths then begin their gradual decline into advancing age (30).
In the long run, immune system function may have more to do with which types of T cells are present than how many survive with ample telomeres, says Miller. A healthy individual may have a million different T cell clones, each trained to recognize a different antigen. In a young person, these million clones will be represented in roughly equal numbers. But in older people, a small handful of clones predominate. This imbalance may leave the body susceptible to infection by organisms whose antigens are recognized by one of the underrepresented T cell clones.
Less Division: Does It Add Up? Back to TopEven if telomeres do atrophy with age, what does that mean for an organism? Do cells with truncated telomeres stop dividing in vivo? "It's an almost impossible question to answer," admits Hornsby. One would need to measure the lengths of the telomeres in individual cells from an intact animal and then determine whether those cells are capable of dividing. "We don't have the technology to test that at the moment," says Hornsby.
In fact, researchers are not even 100% clear on what stubby telomeres mean for cells in vitro. For example, cells that are allowed to divide a number of times in culture and are then rescued by the addition of telomerase may continue to divide even with shrunken telomeres (31). Telomerase apparently does more than build up telomeres; it might also safeguard chromosome ends, possibly by facilitating the formation of protective protein cap structures.
"This focus on length may be a red herring," concludes Cristofalo of the Lankenau Institute. When it comes to a telomere's ability to preserve chromosomal integrity and keep cells youthful, he says, "it may be quality rather than quantity." Mouse telomeres are 10 times longer than human telomeres, but mice don't live 10 times longer than us. And, Cristofalo adds, "mice get old and die with long telomeres."
As to whether a decrease in cell proliferation can account for the changes we attribute to aging, no one has been able to show that cells in an organism run out of replicative oomph. And certain cells appear to harbor limitless reproductive stamina: Some cells that line the gut divide thousands of times during the life of an individual. In sum: "There is still no convincing evidence that replicative cell senescence has anything to do with organismal aging," says Martin Raff of the Medical Research Council, University College London.
Even if senescence did occur in the intact animal, most healthy tissues should be able to handle the loss. "We know that people can live 122 years and they don't run out of hair or skin or gut cells," says UT Southwestern's Shay. Many tissues, including the gut, skin, and blood, contain stem cells that are capable of dividing many times and giving rise to a range of cell types. "So senescence could go on without dooming an organ," notes the University of Washington's Wolf. In some situations, however, serious chronic injury to an organism may exceed the capacity of its constituent cells to repair an impaired organ. For example, cirrhosis of the liver may occur when hepatocytes become exhausted after having to proliferate continuously to replace dying or damaged cells (14).
"In general, however, it's hard to pin down how a limited replicative life-span could relate to general decline," says the University of Washington's Martin, particularly when not all cells in an organism reproduce. "That's the problem with senescence," says Leonard Guarente of the Massachusetts Institute of Technology. "How can it explain aging in nondividing cells: muscle cells in the heart, neurons in the brain?" Although these cells don't replicate, their physiology changes over time, and a separate theory would need to be devised to account for these events. "That would be an awkward situation," says Guarente.
Essence of Senescence Back to TopBut senescence encompasses more than changes in replicative capacity: It may also drive some of the physiological dysfunction associated with aging. Senescent cells stop dividing, but they don't die. And they don't sit idly in the culture dish thinking back on their glory days. In culture, senescent cells secrete a number of proteins that their proliferating counterparts do not. Senescent human fibroblasts, for example, churn out growth factors and collagenase (2, 32) (see "Dangerous Liaisons"). Production of such proteases by even a few senescent cells could contribute to the tissue degeneration seen in aging (Fig. 4). "It's like having one rotten apple that spoils the barrel," says Martin. And the secretion of growth factors might explain some of the heterogeneity seen in aging tissue, in which regions of atrophy often abut patches of excessive cellular expansion (10).
Cells that are allowed to senesce in culture also harbor changes in proteins that interpret growth signals, says Cristofalo. He finds that old fibroblasts in culture, for example, are less adept than their younger counterparts at ushering activated signaling molecules into the cell nucleus. Such a failure could explain why cells from older individuals no longer respond robustly to growth factors and cease to divide in culture.
Miller finds similar changes in the signaling pathways that control T cell activation (33). To catalog such cellular changes, Miller advocates using microarray-style screening of the large-scale gene expression patterns of cells from aging animals (34). Data generated by such experiments would allow researchers to monitor changes in gene expression that accompany aging--changes that might provide clues about the underlying mechanisms. At the very least, such arrays might pinpoint genes whose corresponding proteins could serve as markers for aging cells, something that researchers have been hoping to find for decades (see Miller Perspective).
Proteins that are produced preferentially during senescence, for example, would help scientists determine whether these cells do, in fact, accumulate with age. LBNL's Campisi has used one such protein, a particular form of galactosidase, to track senescent cells in vivo. By staining tissue slices from skin biopsies, Campisi finds that older individuals harbor more cells that contain this enzyme than do their younger counterparts (35). Although Cristofalo and others have been unable to replicate her findings (36), some researchers have found evidence that cells ripe with galactosidase accumulate in aging organisms.
In the meantime, the search for senescent cells continues. And as they are detected, Campisi says, the question then becomes "So what?" Researchers will need to show that these cells exert some detrimental effect. If senescent cells damage their surroundings, perhaps the youthful physiology of certain tissues could be preserved by killing these senescent cells--or by getting them to kill themselves. That might not be easy: In culture, senescent cells are somewhat resistant to suicide by apoptosis (37).
Death Before Aging Back to TopMaybe spurring miscreant cells to commit suicide would promote youthfulness in certain tissues such as skin, but generally speaking, rampant cell death would probably not benefit an adult organism. Indeed, perhaps our tissues degenerate--and we age--not because cells cease to divide or fail to function, but because they start dying in droves. As tissues lose their physical integrity, we begin to show the signs of age: fragile bones, enfeebled immune response, weakened heart, wrinkly skin.
Certainly bone cells, immune cells, muscle cells, and fat cells die. But are they deliberately eliminating themselves en masse by a directed process of programmed cell death? Could apoptosis have something to do with aging? "The general thinking is that it does not," says Guarente. Why not? "Because there's no evidence that it does" (38).
Take worms, for example. In Caenorhabditis elegans, scientists have painstakingly mapped out the pathways that lead to apoptosis of select cells during development (39). Other researchers have carefully characterized a pathway that controls life-span (40). Defects in the latter pathway allow mutant worms to live up to four times longer than their wild-type counterparts.
But mutations that affect one process don't influence the other. "There's just no clear evidence in worms that cell death is causally involved in aging," concludes Michael Hengartner, an apoptosis expert now at the University of Zurich, Switzerland. What's more, the connection between apoptosis and aging is so weak that not even biotech companies are looking to exploit it. "And entrepreneurs are generally more optimistic than us regular academic guys," notes Hengartner. "So if biotech companies aren't trying to sell it, there's not much evidence for it."
The situation may be different in mammals. Pier Giuseppe Pelicci and his colleagues at the European Institute of Oncology in Milan have shown that cells from mice lacking p66--a protein involved in the cellular response to oxidative stress--are resistant to apoptosis when they are exposed to hydrogen peroxide or UV light (see Genes/Interventions Database). And these mice live 30% longer than wild-type animals (41). These results suggest that in some organisms apoptosis might be associated with aging, and researchers--including Guarente--are pursuing this lead.
Perhaps the one place where cell death might promote an aging phenotype is in the dramatic loss of neurons seen in the brains of patients with certain neurodegenerative disorders, such as Alzheimer's disease (see "Detangling Alzheimer's Disease" and Honig case study). But what controls the death of these neurons? Cells that harbor mutations in genes involved in neurodegenerative disease--presenilin, for example--readily undergo apoptosis in culture. "But it's no-go in vivo," says Dale Bredesen of the Buck Institute for Age Research in Novato, California. "You don't see classical apoptosis in these cells in vivo or in mouse models of the disease." Maybe a different kind of programmed cell death is operating in aging tissue--one that Bredesen and his colleagues have characterized in cultured fibroblasts (42).
Apoptosis is characterized by specific morphological changes: The cell nucleus implodes, leaving behind a telltale heap of molecular debris--DNA fragments and membrane shards. But the self-destructing fibroblasts and neurons that Bredesen studies leave no such trace of a violent end (Fig. 5). And proteins that traditionally inhibit apoptosis are unable to stop these suicidal cells from offing themselves. The cells are, however, dying by a programmed process that requires synthesis of new proteins, including, most likely, those that form the machinery that takes the cell down. Further, Bredesen finds that this form of cell death, which he calls paraptosis, appears to occur during some neurodegenerative diseases, such as amyotrophic lateral sclerosis. Perhaps paraptosis could account for the loss of neurons in Alzheimer's disease or other neurodegenerative diseases of aging.
These observations, however, may have limited implications for aging in general, because not much neuronal death occurs in the normal aging brain--no matter what the mechanism. The cognitive decline that sometimes accompanies aging might be due more to a loss of synaptic connectivity, suggests Martin.
To address the many unanswered questions about the cellular basis of aging, researchers will no doubt continue to explore the process from all angles, using genetics, cell culture, and powerful new technologies that may someday allow the determination of gene expression profiles for single cells. "We'll need a lot more basic research to tie the molecular findings to cellular findings to the animal models," says Bredesen. "And then we'll need to figure out the best way to relate human aging to these models."
Such studies should reveal what happens to our cells as we age, knowledge that will be rewarding and will lead, if not to longer lives, perhaps to healthier ones. "There's been a lot of progress made in the last decade," says Guarente. Undoubtedly, our curiosity and insight will continue to thrive even as our cells cease to grow.
October 3, 2001
Karen Hopkin is freelance writer living in Somerville, Massachusetts. She ain't what she used to be--but on a cellular level, none of us is.
Abbreviations: Adrenal cortex. In mammals, the outer portion of the adrenal gland (which sits atop the kidney); the adrenal cortex produces and secretes various steroid hormones, including glucocorticoids (which regulate carbohydrate metabolism) and mineralocorticoids (which control water and salt balance). Apoptosis. A genetically regulated process in which a damaged or unneeded cell commits suicide; also called programmed cell death. galactosidase. An enzyme that in cells breaks down the sugar lactose; one form of the protein, active at pH 6, is produced selectively by senescent cells in culture and may serve as a marker for senescence in cells in the organism. Fecundity. A measure of an organism's penchant for propagation, or its ability to produce offspring. Presenilin. Serpentine, membrane-spanning protein (and the gene that encodes it) that's necessary for -secretase cleavage of -amyloid precursor protein to produce amyloid. Replicative senescence. The state in which a cell exhausts its ability to divide and no longer reproduces; such cells permanently withdraw from the cell cycle that normally regulates the process of proliferation. The phenomenon is clearly seen in mammalian cells grown in culture, although it is less clear whether cells in the organism similarly cease to divide. Senescence. Another term used to refer to aging in an organism. Telomerase. An enzyme that maintains the structure of telomeres, the repetitive DNA sequences and their associated proteins that cap the ends of chromosomes. Telomerase synthesizes telomeric DNA sequences and promotes the formation of the protein complexes that protect the chromosome's ends. Telomeres. Structures made of repetitive bits of DNA sequence and associated proteins that cap the ends of chromosomes.
Suggested ReadingBack to Top
Science of Aging Knowledge Environment. ISSN 1539-6150