ARTICLE IN BRIEF
Whole-genome sequencing of neurons from postmortem tissue from neurologically normal individuals and those with neurodegenerative diseases — ages 4 months to 82 years old — shows that somatic mutations may be present in infancy and increase with age, putting individuals at risk for neurodegenerative diseases.
A team of scientists at Boston Children's Hospital and Harvard Medical School have discovered that single neurons have hundreds of somatic mutations in infancy and over a lifespan can accumulate a few thousand more, reflecting immense genetic diversity, even in the same brain. The scientists believe that these increasing genetic mutations may put people at risk for neurodegenerative disease.
The proof comes in a study, published February 2 in Science. Christopher A. Walsh, MD, PhD, the Bullard professor of pediatrics and neurology at Harvard Medical School and Boston Children's Hospital, and colleagues developed a technique to carry out whole-genome sequencing in single neurons. It's been clear that somatic mutations occur in many cell types outside of the brain, and can trigger cancer, but until now it has been impossible to know whether non-dividing cells have the same vulnerability to mutations.
It turns out they do. Dr. Walsh said that the findings could have implications in understanding human vulnerability to age-related neurodegenerative diseases.
“When genes turn on, there is always a risk of damage, and that implies that somatic genetic mutations would increase with age,” explained Dr. Walsh. “We wanted to see whether this was true in the brain and understand the mechanism that drives these mutations. Our expectation is that every cell type has its unique set of mutational signatures and every neuron is living its own life,” genetically speaking.
STUDY METHODS, FINDINGS
Dr. Walsh and his colleagues collected 116 single neurons culled from two different brain regions — the prefrontal cortex and the dentate gyrus of the hippocampus — in autopsy tissue from 15 neurologically normal individuals from four months to 82 years old. They also collected 42 single neurons from the brains of nine people who died of two early-onset neurodegenerative diseases, Cockayne syndrome (CS) and xeroderma pigmentosum (XP). These conditions are caused by genetic disorders of DNA repair that trigger a range of developmental disabilities and accelerated aging in childhood. Patients generally live only a decade or two.
They removed neurons from each brain — newborn through old age — so that they could compare the similarities and differences of single cells from the same person. They isolated the single neurons using flow cytometry and amplified their genomes and then used the amplified genomes for whole-genome sequencing. They used a bioinformatics strategy to separate out true double-stranded single nucleotide variants from single-stranded variants and artifacts.
The results were surprising, said Dr. Walsh. The researchers found that the somatic mutation count increased with age. This was a widespread phenomenon from the normal brain to the brains of patients with CS and XP, who showed more than a two-fold increase in single cell mutations. There were regional-specific and age-specific variations in the neurons, even those that came from the same brain.
Newborn neurons had an average of 300 to 900 mutations. By comparison, single cells from an 80-year-old brain had an average of 2,500 mutations. “Do the math,” said Dr. Walsh. “That is about 25 mutations a year.”
Some brains had a lot of mutations, and others very little. For instance, one teenager had 3,000 mutations in the single cells studied while a 42-year-old woman who died in an accident had the neuronal mutational profile of a teenager. “Every individual neuron has its own life history,” Dr. Walsh added.
The dentate gyrus is the only area of the brain known to undergo neurogenesis throughout the lifespan, Dr. Walsh noted, but the researchers found even more mutations — about 4,000 — in the dentate gyrus in the oldest normal brains studied.
The brains from the children and teenagers who died with CS or XP had 2.5 times the number of mutations, which makes sense because the DNA repair mechanisms are damaged. The teenage diseased brain has about 2,500 mutations in the individual neurons compared to 1,500 in a normal young adult, Dr. Walsh said.
With these data in hand, the researchers wanted to determine where the DNA damage was coming from. They used technology applied in cancer research — mutational signature analysis, which basically sorts the mutations into specific signatures that, in some cases, implicate specific mutations' mechanisms like UV light damage or cigarette smoke. They identified three signatures in the brain and found that these signatures resemble some of those seen in cancers. These signatures may help to explain what may be driving the brain mutations. It also may suggest possible ways to target specific disease prevention strategies at the level of DNA repair, Dr. Walsh said.
The first signature A is also found in cancer cells and in normal stem cells. It is like a genetic clock, ticking off time over the lifespan. These mutations are not seen in newborns but commonly seen in middle age and onward. It is a mutational signature tied to age. They accumulate at the same rate in both the prefrontal cortex and the dentate gyrus, Dr. Walsh explained.
The second signature B involves genetic mutations that are present at birth, and they don't change much over time. These mutations bear the hallmark of mutations that occur during cell division, and therefore represent a “genomic barcode,” that marks an individual brain cell's history of cell divisions. There are hints that this signature slopes up a bit in older people in the dentate gyrus, which is a rare part of the brain that continues to show cell division after birth.
The third and last basket of mutation — signature C — appears to reflect oxidative damage. These signatures also resemble those seen in certain cancers and have been implicated in neurodegenerative conditions. They increase with age in the prefrontal cortex but not in the dentate gyrus.
The scientists suspect that changes in the slope of the number of mutations may put people at risk for dementia, amyotrophic lateral sclerosis, and other neurodegenerative diseases. The Harvard team is now collecting autopsy tissue samples from patients with amyotrophic lateral sclerosis and Alzheimer's disease to see if they have higher numbers of somatic mutations.
“The hope is that the accumulations of mutations may be part of a unifying model for neurodegenerative disease,” Dr. Walsh explained.
DNA repair mechanisms are always at work in the brain and probably fix more than 99 percent of the damage to the cell. But that means about 1 percent of the damage is not fixed and can lead to a permanent change in the neuron. “Mutations accumulate linearly,” Dr. Walsh added. “We can probably get by with damage to one copy of the gene, but maybe not two.”
In another study in the same issue of Science, Flora Maria Vaccarino, MD, Harris professor in the Child Study Center and professor in the department of neuroscience at Yale University School of Medicine; Alexej Abyzov, PhD, assistant professor of biomedical informatics at Mayo Clinic; and their colleagues found a similar number of mutations when they looked at fetal brain tissue. The team analyzed the genomes of single cells extracted from the forebrains of three fetuses, ranging in age from 15 to 21 weeks in fetal development. They used a different technique to expand the single cells: They clonally expanded them. This gave them enough DNA for sequencing, and it bypassed the artifacts commonly induced by amplification of single cells.
Still, they reached similar conclusions to those observed by the Harvard group that began their mutational search postnatally at four months. They found 200 to 400 single-nucleotide variations per cell at mid-fetal stage. There was also a linear correlation with age. Interestingly, they found that the more common mutations (present in more than 2 percent of the cells) were also present outside of the central nervous system, in the spleen.
“These mutations probably occurred before the differentiation of mesodermal and neuroectodermal tissues, early in embryogenesis, and before gastrulation,” Dr. Vaccarino explained.
“This is a process that is happening normally throughout the lifespan,” she added. “More work needs to be done to understand the functional consequences of this. The fact that both of our numbers added up using different techniques is amazing. The two studies validate one another.”
Both laboratories are part of the Brain Somatic Mosaicism network, sponsored by the National Institute of Mental Health.
“I find this study by Walsh surprising but convincing, especially in light of the accompanying paper in the same issue by Vaccarino's group at Yale,” said Fred H. Gage, PhD, head of the laboratory of genetics at The Salk Institute for Biological Studies. “It is not clear why these mutations accumulate with age, or what the consequences are, at present, but the fact seems clear. Now the search for cause and consequences can begin.”
“I was blown away by the sophistication of the technology and the possible implications,” added Gregory L. Holmes, MD, chair of the department of neurological sciences at the University of Vermont College of Medicine. “This is a tour de force, but it is just the beginning. To show that humans have a genetic clock that could potentially contribute to disease is exciting. If we could understand the genetics of these somatic mutations, it would open up the door to possible therapeutic interventions and treatments.”
THE SCIENCE EXPLAINED: WHOLE GENOME SEQUENCING
WHAT IT IS: Whole genome sequencing (WGS) is a comprehensive method that can rapidly sequence large amounts of DNA and provide genetic analysis of an individual's entire genome, including coding and non-coding regions of nuclear DNA. In the current study, the researchers used this technique to study the genome of single cells.
THE TECHNIQUE: The scientists start with neuronal nuclei from postmortem brain. They individually sort single nuclei into wells on a microtiter plate. Then, then they amplify the entire genome from a single DNA copy using a highly efficient DNA polymerase that makes millions of copies of the original genome. They generate a tube of DNA as if it came from the blood of an individual, but it is obtained from a single cell's genomes. Then, they sequence that DNA sample from the single cell. The sequences are then mapped to the consensus genome and the variants that pop up in the sequencer that are different from the genome reference are annotated.
HOW IT IS USED: WGS has been used mostly as a research tool to identify inherited disorders, mutations, and large and small variants that put people at risk for disease. More recently, costs for WGS have dropped substantially: It cost $350,000 in 2005 and today it can cost a few thousand dollars. It is being used clinically to diagnose disorders.