Every time he bent over a freshly dead body, pathologist George Martin pondered the diversity before him. Although his cadavers almost always belonged to the elderly, they varied dramatically. One would have intestines pocked by polyps. Another’s arteries were plugged with plaque. Variety even existed within the same types of disease. For instance, the location of beta amyloid deposits in the brain of people who’d suffered from Alzheimer’s differed significantly. If each type of malady shared the identical underlying cause, bodies ravaged by that cause should look similar in death. But they didn’t. “I never saw two people who had aged in the same way,” Martin says.
Martin read everything he could on aging. He took particular interest in observations showing how organisms ranging from clonal yeast to human twins had wildly different lifespans. One of the more dramatic examples were reports of tiny worms, Caenorhabditis elegans, that varied in lifespan by up to five-fold even when the worms were genetically identical and lived in identical laboratory surroundings.
Biologists know how chance events in the environment (such as getting hit by a bus) impact lifespan. And they understand the role of chance in genetics (such as inheriting genes for Huntington’s disease and certain cancers). But it now seems a third realm of uncertainty emerges as animals grow older, causing them to age in different ways. Researchers are only beginning to figure out the basis of biological fluctuations that build up over time. Some result from mutations that slip into the genomes within cells as they replicate. Others occur because of changes in molecules that either shut off or activate genes.
The location of random variations within a cell’s nucleus matters, and that too may be determined by chance. It’s as if they were small tears forming on a blueprint that is sloppily folded, unfolded, and refolded over time. Depending on where the wrinkles occur, the building the blueprint encodes either remains intact or is rendered vulnerable to collapse.
Like a gust of wind into a glass shop, chance dictates the extent of the damage.
Why would evolution have allowed such instability to persist in our biological makeup over the eons? Martin, now an 89-year-old researcher who studies the genetic basis of aging at the University of Washington, and a handful of other researchers who study aging, speculate that a limited amount of internal uncertainty is beneficial because it helps animals adapt to changing environments.
This notion fits with a broader picture of the vital role of diversity in evolution: Variation among individuals in a population provides options for natural selection to choose from. After all, natural selection cannot weed out excessive variations if they occur after an animal has passed on its genes, since survival in the evolutionary sense only means survival of that lineage in the next generation.
As Martin puts it, “Nature doesn’t give a damn about us after we make our babies. Natural selection is pretty much gone around age 40—and that’s where aging begins.”
While mutations and fluctuations in gene expression within organisms can provide an adaptive boon, they become problematic late in life as mutations accumulate and the breadth of swings in gene expression grows wider. In turn, Martin says, these chance swings may lead to “geriatric disorders,” including cancer and degenerative brain diseases.
Throughout our lives, chance slips into our bodies through tiny, accidental mutations. Somatic mutations occur in cells as they divide over time. During every division, mistakes are made in the new strand of DNA. Genes involved with DNA repair normally fix those mutations—but if enough time passes, a mistake will inevitably occur within one of those repair genes, too. When, exactly, is a matter of chance—although the risk increases over time. But once that error occurs, additional mutations will be sustained rather than fixed. Again, chance determines how much more time passes until one of the mutations randomly affects a cancer-causing gene.
“Imagine that you are throwing darts at DNA,” Martin explains. “It might hit at some places where you’re lucky—and like me, you live until 89. Or you could be unlucky like my late wife who died of an aggressive form of cancer beginning in the brain. Even though she was in great physical shape, she got a hit in a dominant oncogene.”
Not all cancer is caused by chance. In January, 2015, a biostatistician and a cancer geneticist estimated that about a third of cancer cases can be attributed to inherited defects and assaults from environmental factors ranging from sun rays to cigarettes. The rest, they wrote in Science, may be due to random internal events like chance mutations.
Recent studies have indicated how random mutations with seemingly negligible effects may be a problem as they accumulate over time. Some of these mutations might cause changes in how genes are expressed. State-of-the-art technology has permitted cell biologists to analyze minute differences in the expression of genes within individual cells. It appears that genetically identical cells within a body gradually become more dissimilar in terms of the expression of their genes as an animal grows older.
“Nature doesn’t give a damn about us after we make our babies.”
In a 2006 Nature study, researchers monitored the expression of a dozen genes in heart cells extracted from young, 6-month-old mice and elderly, 27-month-old mice. Gene activation was interpreted through levels of the corresponding RNA because RNA is the middleman in the molecular pathway between a gene and a protein. Levels of RNA were relatively similar between the cells from young mice. However, heterogeneity elevated among cells from older mice. The researchers suggest this discordance provides an explanation for why the older mouse hearts functioned less well. “These results underscore the stochastic nature of the aging process,” concluded Jan Vijg, a molecular geneticist at Albert Einstein College of Medicine, and his co-authors on the manuscript.
Roger Brent, a molecular biologist at the Fred Hutchinson Cancer Research Center in Seattle, agrees. “If the higher functions in a vertebrate depend on a population of cells within an organ or tissue responding in some way, then mutations mean that the response will be less coherent, and that may contribute to a decline in function,” he explains. However, scientists have not yet proven that this noise causes diseases of aging, like heart disease. Vijg suspects it does, but says the evidence isn’t simple to come by. “It’s very hard to show precisely how variation among cells leads to, say, a loss of organ function,” he explains.
More than 60 years ago, C.H. Waddington predicted that too much fluctuation could be detrimental, and proposed the existence of a biological mechanism that kept fluctuation within a safe range. He coined the term “canalization” to describe the ability to remain stable. Currently, Martin’s lab is searching for genes that maintain homeostasis. His hypothesis is that something occurs to these cells later in life that increases the molecular “drift” of old age that results in cell-to-cell variation. “This drift might take on a life of its own after the reproductive age,” Martin says, “And if you get drift outside of a certain window of homeostasis, it’s possible you can’t come back.”
Martin says his research suggests that “drifts in gene expression become greater during aging,” and can lead to “quasi-stochastic” geriatric disorders. He hastens to add, however, that while there is evidence of increased variability in gene expression with age, the precise mechanisms behind it remain to be discovered.
Meanwhile, Alexander Mendenhall, a researcher who studies age-related disease at the University of Washington, is trying to figure out which cells and organs are most sensitive to increasing incoherence. “We want to figure out what breaks first by chance,” he says. To get there, the researchers in his lab observe individual cells of living C. elegans worms to understand when and how mitochondria, muscle cells, excretory cells, and other components fall apart. “Once there’s data on which cell types fail, perhaps we can predict the probability of particular problems,” he says.
As they research what goes wrong with cells, biologists are also learning how variability within individuals can be adaptive. Although cause-and-effect connections between non-heritable fluctuations and benefits have not been proven, analogous trade-offs occur within genetically identical populations of organisms in nature. Biologists call it bet-hedging—stealing a word from investors who want a cushion against monetary losses, and so put their money in opposing outcomes.
One example of bet-hedging occurs in single-celled, genetically identical populations of Escherichia coli bacteria. Many of the microbes die when exposed to antibiotics, but the slowest growing microbes among them seem to persist. Researchers don’t know what causes the variation in gene expression among the microbial clones, but they suspect it has lasted through the generations since it helps the lineage survive. There are some hints that cancerous cell populations employ bet-hedging too: Fast-dividing cells seem more vulnerable to chemotherapy.
“Chance is something that researchers have to deal with, whether they like it or not.”
Likewise, Mendenhall and others have shown genetically identical C. elegans worms vary in the amount of “heat shock” proteins they produce. These particular proteins protect other proteins, so they don’t become misshapen or otherwise faulty in response to triggers, ranging from cancer to a flash of high heat. Those who produce more of the protein have fewer offspring, but survive environmental stress better than those who make more babies but produce fewer heat shock proteins. Mendenhall says this variation likely persists because it diversifies the physiologies of the worms within a population. He has identified genes controlling signaling systems that he thinks may underlie the variation.
When many researchers consider how the expression of genes varies or drifts, epigenetics comes quickly to mind. Through epigenetics, expression changes without mutations to the nucleotides comprising genes. Specifically, epigenetics has to do with two key processes. One component involves histone proteins, which DNA coils around within the nucleus of cells. The tightness of those coils exposes or hides certain genes so that they are or are not expressed—and this condition can be altered by the binding of “histone factors” in a process called histone modification. A second epigenetic component involves methyl compounds that attach to the DNA, activating or repressing the underlying gene from expression—a process called methylation.
In the past decade, scientists have documented changes in patterns of methylation and histone modification that vary as organisms age. Because these alterations cause deviations in gene expression, some researchers call the phenomenon “epigenetic drift.”
In a 2012 report in the Proceedings of the National Academy of Sciences, researchers found that DNA from a 103-year-old was less methylated than DNA from a newborn. And not just a little less: The centenarian had almost 500,000 fewer points of methylation along their genome than the baby. In another study, researchers compared 3-year-old identical twins who had similar patterns of methylation and histone modification along their genomes to a pair of 50-year-old identical twins who differed dramatically in methylation, histone modification, and gene expression. No one knows what the effect of these different patterns in epigenetic modifications are—they simply indicate that epigenetic changes, known to alter gene expression, could be responsible for genetic drift.
Some external causes are known to underlie epigenetic changes, such as smoking. But the sheer extent of epigenetic drift observed between young and old organisms causes researchers to believe that other factors are at play—including those that are as random as a person being pelted by pigeon poop. “Chance is something that researchers have to deal with, whether they like it or not,” Mendenhall says. “It doesn’t make things so straightforward.”
As with the gradual accumulation of mutations, methylation and histone deacetylation could alter the activation of genes that don’t matter much—or those that dictate vital processes like DNA repair and cancer control. Like a gust of wind into a glass shop, chance dictates the extent of the damage. Researchers have found evidence the damage can lead to diseases we associate with aging. For example, some cancers begin when genes that normally repress tumors are silenced through methylation.
However, Tom Johnson, a molecular biologist at the University of Colorado, worries about how little evidence drives the current wave of enthusiasm over the role of epigenetics in aging. Most variation in gene expression still exists without a documented cause. What’s more, Johnson points out that C. elegans worms don’t display methylation at all, despite enormous, unexplained variation in the way they age. He doesn’t doubt that gene expression drifts over time—rather, he questions epigenetics as a main underlying cause. Another potential mechanism could be changes that occur to proteins during and after synthesis. “I’d rather just call [drift] stochastic, and not epigenetic,” he says, “because when you say epigenetics you pretend you know what’s going on.”
Johnson, who was a physics major for a year at the Massachusetts Institute of Technology, has never been shaken by the interference of chance in aging. “You can never measure where an atom is exactly because in measuring it you move it,” he says, referring to Heisenberg’s Uncertainty Principle, which asserts a limit to precision. “You cannot be completely specific about anything, not even an individual atom.”
Because Martin and his fellow researchers have yet to pinpoint the mechanisms of chance in aging, any mention of ways to address aging at a cellular level would be premature. (“If we really want people to live longer, we would work on our health system so that there is universal care,” Martin says sharply.) Still, the scientists have painted a fuller picture of chance in nature. Before and during reproductive age, as Darwin showed, variability can help organisms adapt to changing environments and pass on their genes. After reproductive age, though, Martin and other researchers are beginning to feel that variation becomes increasingly unkind. Nature is indeed a cruel mistress. Having set us on our way with elementary instructions, she abandons us to the fates.