And almost immediately, genetic mistakes started to accrue.
“That process of accumulating errors across your genome goes on throughout life,” says Phil H. Jones, a cancer biologist at the Wellcome Sanger Institute in Hinxton, England.
Scientists have long known that DNA-copying systems make the occasional blunder — that’s how cancers often start — but only in recent years has technology been sensitive enough to catalog every genetic booboo. And it’s revealed we’re riddled with errors. Every human being is a vast mosaic of cells that are mostly identical, but different here or there, from one cell or group of cells to the next.
Cellular genomes might differ by a single genetic letter in one spot, by a larger lost chromosome chunk in another. By middle age, each body cell probably has about a thousand genetic typos, estimates Michael Lodato, a molecular biologist at the University of Massachusetts Chan Medical School in Worcester.
These mutations — whether in blood, skin or brain — rack up even though the cell’s DNA-copying machinery is exceptionally accurate, and even though cells possess excellent repair mechanisms. Since the adult body contains around 30 trillion cells, with some 4 million of them dividing every second, even rare mistakes build up over time. (Errors are far fewer in cells that give rise to eggs and sperm; the body appears to expend more effort and energy in keeping mutations out of reproductive tissues so that pristine DNA is passed to future generations.)
“The minor miracle is, we all keep going so well,” Jones says.
Scientists are still in the earliest stages of investigating the causes and consequences of these mutations. The National Institutes of Health is investing $140 million to catalog them, on top of tens of millions spent by the National Institute of Mental Health to study mutations in the brain. Though many changes are probably harmless, some have implications for cancers and for neurological diseases. More fundamentally, some researchers suspect that a lifetime’s worth of random genomic mistakes might underlie much of the aging process.
“We’ve known about this for less than a decade, and it’s like discovering a new continent,” says Jones. “We haven’t even scratched the surface of what this all means.”
Suspicious from the start
Scientists had suspected since the discovery of DNA’s structure in the 1950s that genetic misspellings and other mutations accruing in non-reproductive, or somatic, tissues could help explain disease and aging.
By the 1970s, researchers knew that growth-promoting mutations in a fraction of cells were the genesis of cancers.
“The assumption was that the frequency of this event was very, very low,” says Jan Vijg, a geneticist at the Albert Einstein College of Medicine in New York.
But it was extremely difficult to detect and study these mutations. Standard DNA sequencing could only analyze large quantities of genetic material, extracted from vast groups of cells, to reveal only the most common sequences. Rare mutations flew under the radar. That started to change around 2008 or so, says stem cell biologist Siddhartha Jaiswal of Stanford University in California. New techniques are so sensitive that mutations present in a tiny fraction of cells — even a single cell — can be uncovered.
In the early 2010s, Jaiswal was interested in how mutations might accumulate in people’s blood cells before they develop blood cancers. From the blood of more than 17,000 people, he and colleagues found what they’d predicted: Cancer-linked mutations were rare in people under 40, but occurred in higher amounts with age, making up about 10 percent or more of blood cells after the 70th birthday.
But the team also saw that the cells with mutations were often genetically identical to one another: They were clones. The cause, Jaiswal figures, is that one of the body’s thousands of blood cell-making stem cells picks up mutations that make it a little bit better at growing and dividing. Over decades, it begins to win out over normally growing stem cells, generating a large group of genetically matched cells.
Not surprisingly, these efficiently dividing mutated blood cell clones were linked to risk for blood cancer. But they were also associated with increased risk for heart disease, stroke and death by any cause, perhaps because they promote inflammation. And unexpectedly, they were associated with about a one-third lower risk of Alzheimer’s dementia. Jaiswal, who coauthored an article on the health impacts of blood cell clones in the 2023 Annual Review of Medicine, speculates that some clones might be better at populating brain tissue or clearing away toxic proteins.
As Jaiswal and colleagues were pursuing the blood clones they reported in 2014, researchers at the Wellcome Sanger Institute commenced investigations of body mutations in other tissues, starting with eyelid skin. With age, some people get droopy eyelids and have a bit of skin surgically removed to fix it. The researchers acquired these bits from four individuals and cut out circles 1 or 2 millimeters across for genetic sequencing. “It was full of surprises,” says Inigo Martincorena, a geneticist at Wellcome Sanger. Though the patients did not have skin cancer, their skin was riddled with thousands of clones, and one-fifth to one-third of the eyelid skin cells contained cancer-linked mutations.
The findings, that so many skin cells in people without skin cancer had mutations, made a splash. “I was blown away,” says James DeGregori, a cancer biologist at the University of Colorado Anschutz Medical Campus in Aurora, who was not involved in the study.
Wellcome Sanger researchers went on to identify clusters of identical, mutated cells in a variety of other tissues, including the esophagus, bladder and colon. For example, they examined colonic crypts, indentations in the intestinal wall; there are some 10 million of these per person, each inhabited by about 2,000 cells, all arising from a handful of stem cells confined to that crypt. In a study of more than 2,000 crypts from 42 people, the researchers found hundreds of genetic variations in crypts from people in their 50s.
About 1 percent of otherwise normal crypts in that age group contained cancer-linked mutations, some of which can suppress proliferation of nearby cells, allowing mutant cells to take over a crypt faster. This alone is not necessarily sufficient to create colorectal cancer, but on rare occasions, cells can acquire additional cancer-causing mutations, overflow crypt boundaries and cause malignancies.
“Everywhere people have looked for these somatic mutations, in every organ, we find them,” says Jones. He’s come to see the body as a kind of evolutionary battleground. As cells accumulate mutations, they can become more (or less) able to grow and divide. With time, some cells that reproduce more readily can overtake others and create large clones.
“And yet,” notes DeGregori, “we don’t turn lumpy.” Our tissues must have ways to stop clones from becoming cancer, he suggests. Indeed, overgrowing mutant clones in mice have been seen to revert to normal growth, as Jones and a coauthor describe in the 2023 Annual Review of Cancer Biology.
Jones and colleagues found one example of protection in the human oesophagus. By middle age, many oesophagus clones — often making up the bulk of oesophagus tissue — have mutations disrupting a gene called NOTCH1. This doesn’t affect the ability of the oesophagus to move food along, but cancers seem to need NOTCH1 to grow. Bad mutations may accumulate in oesophageal cells, but if NOTCH1 is absent, they appear less likely to become tumors.
In other words, some of the bodily mutations aren’t bad or neutral, but even beneficial. And, lucky for us, these good mutations prevail a lot of the time.
Mutant clones increase in size as people age. In this image, each panel represents one square centimetre of tissue from a subject’s esophagus. The youngest subject (top panel) was a moderate smoker; the other two were non-smokers. The size and color of each circle represents a clone with mutations in a particular gene (see key, at top). Sometimes, clones contain multiple mutations, represented by overlapping circles. Mutations in some genes, such as TP53 (orange) promote cancer, while mutations in other genes such as NOTCH1 (purple) suppress it.
Getting inside the brain
Our DNA-copying machinery has plenty of opportunity to make errors in cells of the esophagus, colon and blood because they divide constantly. But neurons in the brain stop dividing before or soon after birth, so scientists originally assumed they would remain genetically pristine, says Christopher Walsh, a neurogeneticist at Boston Children’s Hospital.
Yet there were hints that mutations accruing through life could cause problems in the brain. Back in 2004, researchers reported on a patient who had Alzheimer’s disease due to a mutation present in only some brain cells. The mutation was new — it had not been inherited from either parent.
And in 2012, Walsh’s group reported an analysis of brain tissue that had been removed during surgery to correct brain overgrowth that was causing seizures. Three out of eight samples had mutations affecting a gene that regulates brain size, but these mutations were not consistently present in the blood, suggesting they arose in only part of the body.
There are a couple of ways that brain cells could pick up mutations, says Lodato. A mutation could occur early in development, before the brain was completed and its cells had stopped dividing. Or, in a mature brain cell, DNA could be damaged and not repaired properly.
By 2012, interest in non-inherited brain mutations was heating up. Thomas Insel, director of the National Institute of Mental Health at the time, proposed that these kinds of mutations might underlie many psychiatric conditions. Non-inherited mutations in the brain could explain a longstanding puzzle in neurological diseases: why identical twins often don’t share psychiatric diagnoses (for example, if one twin develops schizophrenia, the other has only about a 50 percent chance of getting it).
Mosaicism provides “a very compelling answer,” says neuroscientist Mike McConnell, scientific director for the Lennox-Gastaut Syndrome Foundation in San Diego, a nonprofit that supports families and research into a severe type of epilepsy.
Starting in the early 2010s, McConnell, Walsh, Lodato and others began to catalog mutations large and small sprinkled across the brains of people who had died. They tallied deletions and duplications of individual genes, multiple genes or entire chromosomes; they spotted entire chromosome segments moved to new spots in the genome. And, eventually, Walsh, Lodato and colleagues found a thousand or more single-letter mutations in the genetic code within every nerve cell of people aged 50 or so. That last finding “seemed completely impossible to us,” recalls Walsh. “We doubted ourselves.”
In the face of such stunning results, the researchers investigated further. They looked at 159 neurons from 15 people who had died between four months and 82 years of age. They reported that the numbers of mutations increased with age, indicating that errors accumulated over time, just as in other body parts. “The brain is a mosaic, in a profound and deep way,” says Lodato.
To further explore that mosaicism, the National Institute of Mental Health funded a series of projects from 2015 to 2019 investigating brain tissue mosaicism in samples, mainly collected after death and deposited in tissue banks, from more than 1,000 people who were neurotypical or had conditions such as Tourette syndrome and autism spectrum disorder.
Single-letter mutations were most common, says McConnell, who co-led the project. Researchers accumulated more than 400 terabytes of DNA sequences and other data, and built analytical tools, creating a powerful platform on which to build the next round of brain mosaicism studies. From this work and other studies, scientists have linked brain mosaicism to neurological diseases including autism, epilepsy and schizophrenia.
In Lodato’s lab, graduate students Cesar Bautista Sotelo and Sushmita Nayak are now investigating how accumulated mutations might cause amyotrophic lateral sclerosis, a paralysing condition also known as Lou Gehrig’s disease. Geneticists can identify a known mutation in only about 10 percent of non-inherited cases. But the new data on mosaicism suggest that many more people may have mutations in ALS genes in their brains or spinal cords, even if they don’t have them in the rest of their body.
That matters, because scientists are working on therapies targeting some of the 40-plus genes that, when mutated, cause ALS. In 2023, the Food and Drug Administration approved the first such treatment, which shuts down a commonly mutated ALS gene. For patients to be eligible for such therapies, they will need to know their mutations.
Thus, says Nayak, “we strongly advocate for a change in the current practice of diagnosing ALS.” Instead of just looking at DNA in a blood sample, other tissues such as saliva, hair or skin could be examined too, in case an ALS mutation arose during development in cells that didn’t give rise to blood but did give rise to other tissues in the body.
Clues to how we age
For now, the health implications of our body’s mosaicism are mostly too fuzzy to warrant action, especially in cases like the blood clones where there is no relevant treatment to offer. “We don’t really advocate that people should worry about this,” says Jaiswal. “At this point in time, there’s no rationale to be testing people who are well.”
But many scientists do see the findings as evidence for a longstanding theory: that a lifetime’s worth of mutations leads to the inevitable condition we call aging.
Martincorena and colleagues tested an element of that theory in a 2022 study. If mutation buildup contributes to aging, they reasoned, then short-lived critters like mice should build up mutations fast, while longer-lived species like people should accumulate mutations more slowly, perhaps due to better repair mechanisms.
To investigate this idea, the researchers embarked on a five-year odyssey studying colon crypt samples from eight people plus a menagerie of creatures: 19 lab mice and rats; 15 domestic animals such as cats, dogs, cows and rabbits; and 14 more exotic creatures that included tigers, lemurs, a harbor porpoise and four naked mole rats, which are famed for their outsized rodent lifespan of 30-plus years. As predicted, the longer-lived the species, the slower its accumulation of mutations.
“This does not demonstrate that somatic mutations cause aging, but is consistent with the possibility that they play at least some role,” says Martincorena. There are two factors at play here: Accumulating mutations contribute to shorter lifespan, but then the shortened lifespan makes mutation protection less crucial, so short-lived species invest less in DNA repair.
The idea that mutations could contribute to aging is tantalizing, as it suggests vanquishing them would be a genetic fountain of youth. “If, tomorrow, I figure out a way to stop these mutations from accumulating, I think I would be a bajillionaire,” says Bautista Sotelo. Already, at least one biotech startup, Matter Bio in New York City, has raised funds with the aim of repairing the human genome. (Whether such a plan would ever be feasible across broad swaths of cells is another matter: “I don’t think you can get rid of the mutations,” says DeGregori.)
The story of body mutations is far from over. “Judging by the discoveries that we are making at the moment, the journey has only just started,” says Martincorena. “I expect many surprises in the next few years.”
