he human body is composed not of perishable materials like wood or metal, but of living cells that can grow and replenish themselves. So if the body's individual units are renewable, why do we not live forever?

Theologians explain that Adam and Eve were denied immortality and expelled from Eden because they dissed the Demiurge. Evolutionary biologists hold that natural selection favors genes that promote having many offspring over those that might ensure longer life. In their view, life span is a trade-off between fertility and longevity, and infinite life would be allied with childlessness and rapid extinction.

A chance discovery by a Texas biologist has provided a different kind of explanation, although one on a very practical level. His answer, appropriately paradoxical for so profound a riddle, is that we age in order to live longer.

The biologist, Dr. Lawrence A. Donehower of the Baylor College of Medicine in Houston, has found that aging seems to be a necessary cost of suppressing cancer. A well-known mechanism that forces cells to commit suicide rather than become cancerous turns out to have the unfortunate side effect of making the body's tissues age faster.

Dr. Donehower's finding, reported in the current issue of Nature, concerns an intensively studied cellular agent that is the body's first line of defense against tumors. The agent, a protein present in every cell, is a stern moral guardian of the cell's behavior; at the slightest appearance of abnormality, especially damage to the cell's DNA, the agent will force the cell either to stop growing or to destroy itself in a suicidal spasm of self- liquefaction.

The agent is known to biologists by the deceivingly humble name of p53 — p for protein, 53 a measure of its weight. In fact, p53 is one of the most important hubs in the living cell's vast network of interacting components. Many circuits that monitor different aspects of the cell's integrity report back to p53. The reporting takes the physical form of adding or removing small chemical elements that alter p53's function. Placed at the center of this intelligence network, p53 can switch on different sets of genes to perform any of several major actions.

It seems that cancer can only occur when the p53 system is somehow damaged. In half of all human cancers the DNA of the p53 gene itself is directly damaged by a mutation, or change, that renders the protein useless. In many other tumors where the p53 itself seems intact, nearby components in its network are sabotaged, thereby inactivating p53 indirectly.

Dr. Donehower works with mice, a standard organism for geneticists because their cells are very similar to those of people. In 1992 he engineered a strain of mouse that lacked a working gene for p53. This is a common technique for understanding what a gene does — knock it out of a mouse and figure how the mouse has been made deficient. Sure enough, the mice without p53 died of cancer at an early age.

Dr. Donehower next tried to make another strain of mice in which the p53 gene had a single inactivating change, one that is commonly found in human cancers. But the alteration did not go as planned, a disappointment because engineering a new strain of mouse is hard work. "We made the mice and kind of forgot about them for a year," he said.

He and a colleague, Dr. Stuart D. Tyner, and others expected the mice to develop cancer early, just as the mice without p53 had done. "But first we noticed they were just not getting cancer when they should have gotten it. We also noted that they looked kind of decrepit; they just looked like old mice," Dr. Donehower said.

It turned out that the Baylor team had inadvertently created mice that possessed unusually large amounts of p53 in their cells. The extra p53 was vigorously suppressing cancers. But it was doing something else that looked suspiciously like aging. The mice did not live long, suffered from osteoporosis, had shriveled organs and did not bear up well under stress.

Dr. Donehower said that aging was hard to characterize and that critics might argue the mice were suffering from some subtle pathology different from aging. But he has convinced some biologists that extra p53 may indeed hasten senescence.

The Baylor experiments "raise the shocking possibility that aging may be a side effect of the natural safeguards that protect us from cancer," wrote Dr. Scott Lowe of the Cold Spring Harbor Laboratory in a commentary in Nature.

Mice and people have several tissues, like blood, skin and stomach lining, that suffer constant wear and tear and need a stream of new cells throughout the organism's life. These are generated by a pool of special cells known as stem cells. While an animal is young, the stem cells generate plenty of new cells, and it does not much matter if the tumor-screening p53 mechanism forces many suspect cells to commit suicide.

But it does matter, Dr. Donehower's experiment suggests, as the animal ages and the stem cells themselves may become less prolific. The Baylor mice with extra p53 may be aging prematurely because too many cells are being forced to destroy themselves, and the tissues can no longer function properly.

If so, what does the finding mean for the hopes of increasing human longevity?

Biologists are just beginning to come to grips with the new finding, but some are already hoping it may prove possible to cheat the p53 system — to somehow keep p53's cancer-fighting potency but to curb its promotion of aging. Dr. Donehower said that among his 200 mice with extra p53 there were two that neither got cancer nor grew old prematurely — a hint that some fortunate individuals may naturally possess a way around the cancer-or-aging impasse.

Dr. Leonard Guarente, an expert on the genetics of aging at the Massachusetts Institute of Technology, thinks that the activity of p53 could be manipulated through a gene known as sir2. The gene responds to signals in the cell's metabolism, and extra doses of it can lengthen life in both yeast and roundworms. It could be part of the mechanism underlying the well known finding that calorie- deprived mice live longer.

Dr. Guarente reported in October that sir2 inhibited p53 in mouse cells grown in the laboratory. He hopes there can be a range of doses, or a window, within which "you could increase sir2 activity and downregulate p53 and slow down the unwanted cell death related to aging but still have adequate cancer surveillance." He has formed a company, Elixir Pharmaceuticals, that is screening for drugs that affect sir2.

One drawback of Dr. Donehower's study is that his experiment may be hard to repeat because his mice with extra p53 were an accident, not rationally designed. Another, as he has pointed out, is the still lingering question of whether his mice showed true aging or some pathology that mimicked it. But those caveats aside, his finding has provocatively brought two vast and separate areas of study, p53 and aging, into sudden conjunction.

"I think it's a very significant paper because it opens your mind to thinking about p53 and aging and allows one to speculate about a lot of things," said Dr. Lowe of Cold Spring Harbor. Why does the skin age so fast in people who are exposed too long to the sun? The answer is not known, Dr. Lowe said, but ultraviolet radiation is one of the damage signals that switches on p53 in a cell, and the extra p53 could be the reason why sunburned skin ages faster.

"With the current study we realize the double-edged sword of p53," Dr. Lowe said. "Without it we'd probably all die of cancer before the age of 30, which is what happens to Li-Fraumeni patients," he said, referring to a syndrome in which many patients have an impaired p53 gene. "With it, only one in three of us get cancer late in life."

"The fountain-of-youth pill will be a challenge if this holds true," Dr. Lowe said, but he did not rule out the possibility of drugs that would deal with the effects of aging directly, leaving p53's role intact.

The new role of p53 is likely to provide just a piece in the puzzle of aging, not the whole picture. Nature can make a mouse live three years, and an elephant 70 years, but probably by some deeper genetic mechanism that has yet to be discovered. The p53 system, Dr. Guarente said, is probably just a way of modulating this deeper mechanism so as to couple the aging rate with the metabolic rate. The reason for this coupling is the enormous evolutionary advantage of being able to bank down metabolism when food gets scarce, outlive the bad times and reproduce later.

Many animals seem to possess this genetic reflex, and it is the reason a low-calorie diet prolongs life in mice and rats. But this mechanism increases life span by only 50 percent or so, Dr. Guarente said, suggesting that the natural life span of a species must be controlled by a separate system.