But they’re not the social gaffes from seventh grade that keep you lying awake at night. No, these are tiny typos in your DNA—the code that produces your proteins, cells, the brain itself. 

Every time a cell divides, it must copy its entire genome, a massive tome of 3.1 billion DNA base pairs. While there are mechanisms at work to check each strand for its spelling, no system is perfect. For those of us who cannot write a single sentence without a typo, it comes as no surprise that in every replication, some mistakes fall through the cracks. The result is a mosaic of cells throughout the body, no two cells quite alike, each with its own unique set of spelling mistakes. 

Some of these changes have difficult—even deadly—consequences. Specific forms of this somatic mosaicism has been linked to pediatric epilepsy, autism, ALS, schizophrenia, and dementia. It has been linked to cancer. 

But not all changes need to be bad ones. Some mistakes could, as Bob Ross famously said, be happy little accidents. Small DNA typos in cells in the brain—typos that will never be passed down to the next generation—could be a way for DNA to try new things, says Christopher Walsh, a neurologist and chief of the division of genetics and genomics at Boston Children’s Hospital in Massachusetts. These tiny changes could merely tweak the functions of individual neurons—and result in changes to our behaviors and personalities that make us, fundamentally, distinct individuals.

With funding from the John Templeton Foundation, neurologist Christopher Walsh and geneticist Alice Eunjung Lee, both at Boston Children’s Hospital, are studying the neuronal lineages of identical twins. They hope to find out how somatic mosaicisms—the tiny changes that result in a distinct genetic code in every cell—occur in the brain and what patterns they form. The result could change how we understand brain development—and what makes each person unique.

We are all mutants

All of us come from a single original cell—the zygote, a union of one ovum and one sperm. That cell has our complete genome, which then gets copied out, over and over—forming cells which become bone, muscle, gut, brain. The result is the roughly 30 trillion cells that make up an adult human body. Many of those cells continue to divide at different rates throughout our lives—blood cells renew every few days, gut cells live less than a week, fat cells 12 years, and muscle cells decades.  

Whole systems exist inside the cell to prevent errors during the copying process, running along each strand to ensure its integrity. But a few mutations always slip through the net.

Walsh explains. “We think of these mutations occurring as like a photocopying machine where the photocopy is never good as the original.” Unlike photocopies, however, when cells divide and replenish, they are not going back to that original zygote for their next copy. Instead, they are forming photocopies from photocopies, accumulating tiny mutations as they go. 

In the zygote, the number of mutations is very low, explains Alexej Abyzov, a computational biologist at the Mayo Clinic in Rochester, NY. “So the exact number is being debated,” he says, but for the sake of argument, assume 100. Now assume, he says, that cell gets multiplied trillions of times—each time accumulating another mutation. “If we take the length of our genome, which is 9 billion [base pairs]…it will mean that every position in our genome has mutated roughly 10,000 times before we’re born.” That’s just at birth. “And then we’ll live another 80 years, let’s say. So the number of mutations in our body to probably every position has been mutated to about…a million times.” Every cell has their own pattern of these somatic mutations, rendering us beautiful genetic mosaics. 

There is one part of the body, however, where cell division ceases. Neurons in the adult human brain do not generally get replaced. They live the full 80 years or so that the person does. This means that they won’t accumulate mistakes from a full genetic photocopy. Scientists used to assume that neurons, then, would not have as many mutations in their DNA. 

“Neurons are on the lower end in terms of rates of accumulation of mutations,” Walsh says. And yet, mistakes are still happening. How? “Even cells that don’t replicate their whole genome are actually replicating little parts of it,” he explains. Every cell suffers DNA damage over time—from ultraviolet light to the reactive chemicals our own cells produce as they run. Repair pathways come in and cut out the damaged DNA base, and pop in a new one, using the opposite DNA strand as a template. “That fill-in process is a form of replication. And so that has a certain error associated with it,” Walsh notes.

The millions of mutations might seem terrifying, but we are one of the longest-lived mammal species on earth, and our mutation rate is actually much lower than it is in other species. 

“My own favored hypothesis is that this accumulation of mutations is how nature sets our lifespan,” Walsh says. A 2022 study compared somatic mutation rates across 16 mammal species. The researchers showed that the shorter the lifespan, the faster mutations slip past quality control. Mice live a 30 times shorter lifespan than humans, and they accumulate mutations 30 times faster. “Dogs accumulate mutations in dog years, seven times faster than humans.” 

While a mouse may die in two years and a human in 60, they will die with a roughly similar number of cellular mistakes. “So this rate is under the control of evolution in a global way,” Walsh says. “That’s a remarkable thing that nature chooses a lifespan somehow and adjusts the fidelity of DNA repair to get there.”

Malignant mosaics

If every cell in the body—neurons included—has its own unique collection of mistakes—some of those mutations will matter. They could matter especially if they are on genes that control cell proliferation—tumor-suppressor genes. “We’ve long understood that cancers are driven by mosaic mutations that drive the cells to proliferate,” Walsh says. “It’s only somewhat more recently been recognized that this mosaicism is present in all dividing cells in the body, but that it doesn’t always cause cancer.”

Knowing how and where the genetic typos occur could help scientists hunt down causes of disease, says Flora Vaccarino, a developmental neuroscientist and functional genomicist at Yale University. Other typos could hold clues to neutral or positive traits—like disease resistance or personality.  

One way to study our mosaicisms is to track the mutations themselves. An early typo that is preserved “can be used to mark cells during development, during each division, [to] build a sort of a genealogical history of each side,” Vaccarino says. “And what I mean by that is understanding what’s the mother cell? What are the daughter cells, and what are their daughters, and so forth and so on.” Tracking the development of these mutations across cells as they divide can produce, in the end, “the cellular history of a person.” 

Those lineages can also provide important clinical clues. In fact, Walsh’s entry into somatic mosaicism was through pediatric epilepsy. In the most common versions of this condition, “you could see an abnormal spot on the brain. And it’s usually just on one side of the brain or one patch.” 

To find out more about these abnormal spots, “we started collecting these things that the surgeons would take out and hypothesized that, ‘oh, you know, maybe there’s a mutation that’s just in that patch, and not in the whole brain,’” Walsh says. He, Lee and their colleagues found that these cell clumps were part of the same lineage of somatic mutations—a single family all descended from one cell with a particular typo. In every clump, the researchers found “particular kinds of mutations in a particular list of about a dozen genes,” Walsh says. “And those genes all do the same thing. They’re one of the most familiar pathways involved in cancer.” The pathway is called mTOR and regulates cell proliferation. It’s also involved in many cell-signaling pathways. 

These clumps of cells were not cancers. Instead, “it makes those cells bad neighbors, basically, it makes the neurons very, very active,” Walsh says. “They’re firing all the time, they’re partying late at night, they’re making a lot of noise, and they create this focal epilepsy and you cut them out, you make them go away, and the epilepsy goes away.”

While this condition can often be addressed with surgery, Walsh’s discovery pointed to another potential approach. “There are drugs that regulate the mTOR pathway developed to treat cancer,” he explains. “And so they have been immediately repurposed to treat those kids that have focal epilepsy that does not respond to surgery.” 

Effects of different mutations are probably not limited to epilepsy alone, Vaccarino notes. “That’s one possibility in terms of autism and more common disorders, schizophrenia, other neuropsychiatric disorders,” she says. 

Indeed, Walsh, Abyzov, and Vaccarino conducted a 2022 analysis of mutations in 131 human brains, 59 of which had autism. While the brains of people with autism did not have more mutations, they tended to have mutations that affected how often genes got expressed in particular cells. Some of those mutations could be inherited, but others could be the result of somatic mosaicism, coming together to produce someone with autism, Vaccarino explains. Walsh and Lee have also shown somatic mosaicisms connected with frontotemporal degeneration, Alzheimer’s, amyotrophic lateral sclerosis (ALS), and schizophrenia. 

Mapping out genetic playgrounds

It might seem as though somatic mosacism is only creating harmful DNA gibberish with its typos. But typos aren’t always negative—they can also produce new words, jokes, slang terms and poetry. And Walsh and Lee argue that mosaicism might play a similar role in our brains—but instead of new slang or puns, it could produce personality, and even genius. 

The positive effects of somatic mutations could lie in the outer layers of our brains, explains Walsh. “Those upper layers of the cortex are particularly elaborated in humans, compared to non-human primates,” he notes. The neurons there are constantly busy, integrating systems like touch, scent and vision together into the unified way we see the world. 

Importantly, the neurons making up these layers are also the last to develop. “Because they’re formed last, they will have the largest number of somatic mutations compared to any of the other neurons in the brain,” Walsh says. “Those neurons are nature’s playground, because those are the neurons that are most subject to these somatic mutations.” The many somatic mutations these cells have could result in in different neuron networks, new sequences of development. All sources of potential human diversity, he explains. 

Of course, these cells in the cortex do not and cannot pass on their unique DNA to the next generation. Only cells in the testes or ovary will. But what does get passed on, Walsh notes, is the ability to let some somatic mutation typos through. “In other words, evolution might favor a certain leakiness in the repair of DNA as a way of generating a certain diversity in the standing population,” he says. 

Creating lineages by tracking mutations worked to understand pediatric epilepsy. Now, Walsh and Lee want to apply it at a larger scale, tracking the genetics of single cells through to adulthood in the brain. 

To do it, Lee is working with Walsh to develop and implement new ways to reliably sequence the DNA from single cells—to capture each cell’s unique genome. “Typically, when we study genomes, we extract DNA from millions of cells,” Lee explains. “But if a mutation is present in only a small fraction of cells, or even in single cells, the classical…DNA-based method wouldn’t work. So we need to have much more sensitive methods.” 

One of the challenges is ensuring the DNA results are accurate, Lee notes. Just as when cells divide, creating photocopies of photocopies, most current methods for amplifying DNA create copies of copies of copies, until there are enough to establish the DNA sequence. “If errors happen in the middle, it can be pulled through and propagated,” Lee says. So she is combining sequencing of single neurons with primary template amplification—a method that amplified DNA by always going back to the original strand, no copies of copies allowed. 

With accurate sequencing of single cells, Lee and Walsh want to add barcoding—a way to identify each cell’s lineage. “One of the dream situations, which is actually impossible to do in in humans, is to trace the cellular origins from the zygote to all the cells in the human body, how they were actually made,” Lee says. But of course, once the human is an adult, that zygote is long past. But with sequencing and barcoding, Lee says, “we can tackle it to some degree.”

There are around 86 billion neurons in the brain, and so to help distinguish between the mutations that someone might have from their parents, and the newly acquired somatic typos, Lee and Walsh are focusing on two sets of identical twins. 

Walsh says. “Identical twins will not develop exactly the same way just in terms of the genetic constitution of their brain. They’ll develop mostly the same way. But there will be little variations.” Creating detailed lineages of neurons could help the scientists find out which of those little variations lead to disease or disorder, and which to unexpected success. 

Somatic mosaicism may hold clues to human disease, but also to how such different cells come together in the end to produce a person. Because tracking down genetic spelling mistakes, Lee notes, highlights not just what can go wrong, but how incredibly often things go right. “What amazes us is even with that high chance of getting mutation here and there, our brains have very consistent structure and function,” Lee says. “It’s really amazingly controlled.”