Science word of the day: Allele

Pronunciation: Ah-LEEL'

What it means: Genes contain the instructions for the various proteins that determine traits, such as eye color or blood type. An allele is a specific variant of a gene. Alleles come in pairs, one from each parent. Some are dominant, meaning that copy will override any instruction from the other allele in the pair, and some are recessive, meaning there have to be two copies of that particular allele for those instructions to translate into an observable feature. Of course, not all inheritance is as simple as dominant and recessive patterns, particularly for traits that arise from combinations of hundreds of genes.

How to use it in a sentence: We learned in grade school that alleles for brown eyes are dominant over alleles for blue eyes, leaving the kids with hazel eyes feeling lost and confused.

Can you crack our science word of the day puzzle, Chain Word?

A 6-year-old girl was given a first-of-its-kind gene therapy in a hospital in China for a nonlethal condition. Within a week, she was dead.

Her death was never publicly disclosed by the scientists who developed and administered her experimental gene therapy. But now, more than a year after her death, her parents are telling her story in the hope that others will be spared the same fate.

A new joint investigation from Science and Retraction Watch, published July 23, details the case of "Mei," a young girl given gene therapy in March 2025 at Xinhua Hospital, which is affiliated with the Shanghai Jiao Tong University School of Medicine. (Mei's parents asked that the family's names be changed for the report. Her name was changed to Mei, which is Chinese for "beautiful," and they are referred to as "Jason" and "Linda.")

Mei's trial was led by Zilong Qiu, a neuroscientist at the university's brain center. Live Science has contacted Qiu for comment.

Mei had a rare genetic disorder called Snijders Blok-Campeau syndrome, which is caused by mutations in the gene CHD3. The gene carries instructions for a protein that helps control how DNA is packaged in cells, which in turn, affects how genes can be switched on and off. The gene plays an important role in early brain development. In people with the syndrome, this process is disrupted, leading to brain changes that can cause speech problems and intellectual disability.

Mei's parents had sought further tests after their daughter was diagnosed with global developmental delay in 2023; this broad diagnosis is given to young children who show delays in several areas, such as speech and motor skills. Genetic sequencing following her diagnosis revealed she had the syndrome.

People with the condition typically have normal life expectancy, although symptoms can range in severity, with severe cases coming with seizures and heart problems. Mei's case was mild, and her parents secured her speech and occupational therapy alongside special education support to help manage her symptoms.

Jason worried about her future, though, anticipating that she'd continue to trail behind in school and may never live independently. On WeChat (the instant messaging, social media and mobile payment app), Jason and Linda joined a group of parents of kids with autism and similar disorders. There, they learned about Qiu, the neuroscientist, who has a gene-therapy company called Lanqi Xintu Gene Technology. Qiu had been developing a gene therapy for Rett syndrome, another genetic disorder that disrupts brain development.

Jason contacted Qiu about possibly creating a gene therapy for Mei, and Qiu immediately arranged a meeting. He explained that such a therapy would require delivering viruses into Mei's spinal fluid; modified viruses are often used as vehicles to deliver gene therapies into the body. The viruses would then travel to her brain, where they'd make the desired edits. To get a big enough dose into the brain, hundreds of trillions of viruses would be needed.

Qiu planned to use a base editor, a CRISPR-based approach that swaps out just one letter in DNA's code for another. (The technique was famously used in the case of a U.S.-based baby, called KJ, whose life-threatening genetic disease was successfully treated.)

Because Mei's condition was not life-threatening, the approach would need to clear a high bar to pass the university's ethics board, Qi said. It would need to be trialed in mouse and monkey experiments first. Qiu insisted, though, that the university could ensure Mei's safety, although it couldn't guarantee the treatment would work. The possibility of death wasn't discussed — and it would not be in the years to come.

Jason and Linda agreed to the plan. They would ultimately put the equivalent of $860,000 toward the cost of developing, testing and deploying the therapy; those costs included informal payments made directly to various members of the research team. Charging for an unproven therapy is forbidden under Chinese law, but Qiu explained this away by saying the parents were paying into his gene-therapy company, which then paid for the procedure.

In January 2025, Qiu submitted a manuscript describing the team's lab-animal experiments to the journal Nature. (A revised version of the manuscript was published in February 2026 and contains no mention of the family nor its financial contributions.) Science had several experts review the manuscript, who expressed concerns about how well the monkey study was conducted.

Live Science contacted Nature for comment, and a spokesperson said "We are looking into the matter carefully following an established process. We are unable to provide further information at this time."

The next month, a toxicology study conducted by the team found that four monkeys given the treatment developed moderate to severe liver damage, and one also had kidney damage. This should have prompted additional safety studies, experts told Science, and hinted that the therapy was triggering a dangerous inflammatory response — a known risk of virus-based gene therapies.

The hospital's ethics committee did not review this data before giving the OK for Mei's procedure. Her parents were told about the results of the animal studies, but they said the scientists seemed unconcerned about them. Under China's regulatory system, the trial did not need to be cleared by the country's equivalent of the Food and Drug Administration before proceeding.

On March 24, 2025, after giving her a steroid to dampen potential immune reactions, Dr. Yongguo Yu injected gene therapy-laden viruses into Mei's spinal fluid. Although some virus-based gene therapies in the U.S. can be given with steroids, as of late, doctors often opt for more aggressive immune-suppressing drugs for safety.

Within days, Mei developed a range of symptoms; she got a fever, wasn't urinating, and her platelet levels plummeted. She was taken into an intensive care unit for treatment, but she died seven days after receiving the injection.

The hospital determined her death was directly related to the treatment and caused by thrombotic microangiopathy. This is an immune reaction that causes blood clots and can arise from virus-based gene therapies.

The hospital was later fined the equivalent of about $3,600 for failing to properly oversee the trial and not registering it as commercially sponsored research in China's national database. The hospital required that the doctor involved receive verbal counseling.

The family asked Qiu to withdraw his Nature paper, worried that other families might be tempted to try the same therapy, and he initially agreed, although it was later published. The paper was covered positively in the press at the time, and parents in the WeChat group began expressing interest. Seeing that, Jason and Linda submitted a complaint to Shanghai Jiao Tong University School of Medicine, calling for a retraction and investigation. The university took no action.

They also sent a letter to Nature after the publication; at the time, the journal said that the ethical issues raised "fall outside our purview in terms of data integrity" and that the matter should be handled by the university.

Regarding the investigation, neither Qiu, Lu, the university or the hospital responded to Science's requests for comment. Nature told Science that it was not aware of the issues surrounding the clinical trial before it published the group's paper.

You can read the full investigation by Brendan Borrell at Science, along with a breakdown of the top four takeaways from the case. In particular, Mei's death raises questions about China's regulations around experimental therapies and human clinical trials, especially in the wake of the infamous "CRISPR baby" cases that purportedly prompted the country to tighten its restrictions around gene editing.

Editor's note: This story was updated at 3:00 p.m. on July 24, 2026, with new comments from Nature. It was first published at 2:00 p.m. the same day.

Researchers have discovered that children's brains can contain cells with their mother's DNA and that these cells can persist for decades.

The findings, which were posted to the preprint database bioRxiv June 10 but have not been peer-reviewed yet, are part of a growing body of work showing that a mother and fetus exchange cells during pregnancy ‪—‬ a phenomenon known as "microchimerism." Previously, scientists had found that a mother's brain harbors cells with her children's DNA.

The findings are important for several reasons, said Amy Boddy, co-director of the Microchimerism, Human Health and Evolution Project at the University of California, Santa Barbara, who was not involved in the study. Past work mostly found evidence of maternal microchimerism in infancy, and in blood samples, she told Live Science in an email. "What's exciting here is that it's tissue, not blood; it's real human data, not an animal model; and the methods are cutting-edge."

More broadly, the work reinforces the idea that microchimerism is "a normal process of mammalian biology," Boddy said.

Hunting down maternal cells in the brain

Before this study, there was sparse evidence for maternal microchimeric cells in brains, mostly because it is hard for researchers to get samples of human brain tissue and DNA from both parents and their children.

To overcome this challenge, a team led by Sami Kanaan, a staff scientist at the Fred Hutchinson Cancer Center in Seattle, analyzed brain tissue that had been surgically removed from dozens of children with severe epilepsy as part of their treatment. The patients ranged in age from 28 days to 19 years at the time of their surgery, and their mothers provided DNA samples through cheek swabs.

Kanaan's team used a tool called quantitative PCR to identify and count maternal cells hiding among millions of cells in the children's brains.

Out of 37 mother-child pairs, 26 children ‪—‬ 70% ‪—‬ had their mother's cells in their brains. These maternal cells were distributed across multiple regions of the brain, including the frontal, temporal and parietal lobes, which sit on the brain's outer surface, and the hippocampus, which is buried deep inside. Each sample had, on average, about 2.2 maternal cells per 100,000, though one sample from the hippocampus had 459 maternal cells per 100,000 and 11 children had no evidence of maternal DNA in their brains.

That prevalence is probably underestimated, Boddy said, "due to the limits of detecting rare cells at low frequency with this method."

Being a firstborn child seemed to increase the odds of having maternal cells in the brain. Of the children who carried their mother's cells, 14 were firstborns and 12 were later-born. In contrast, among the children who didn't have these cells, only one was a firstborn and 10 were later-born.

Using a technique called single nucleus RNA sequencing, which reveals what a cell is doing by showing which genes within the cells are switched on, the researchers found that the maternal cells had transformed into several types of brain cells. Likely, these cells were originally leukocytes and stem cells and had been transferred to the fetus via the placenta or during pregnancy or breastfeeding.

The transformed cells included neurons; oligodendrocytes, which produce the protective sheath around neurons; astrocytes, which support many brain functions and help fuel neurons; microglia, the brain's immune cells; and endothelial cells, which line blood vessels.

"Being able to use single-nucleus RNA sequencing to identify what type of cells the maternal microchimeric cells actually are is amazing," Boddy said. "We've been so limited in understanding the function of these cells, and papers like this, with these methods, are getting us closer."

Checking healthy brains

To see whether these findings also applied to people without epilepsy, the researchers examined brain autopsy data from 29 individuals with no known neurodevelopmental conditions, ranging in age from 22 weeks of gestation to 40 years. They also analyzed brain tissue from three men in their late 80s and early 90s without any known brain conditions originally collected as part of an Alzheimer's disease study.

In total, the researchers found foreign cells in 25 of 32 people — about 78% — including in the brain of a man in his 90s.

The researchers suspect these foreign cells are maternal, but they couldn't confirm it because they didn't have DNA from the mothers. It is possible that the foreign cells came from a twin; an older sibling; a past pregnancy, miscarriage or abortion (in females); or, rarely, from a maternal grandmother, the team wrote in their paper.

As in the epilepsy patients, these cells had matured into several types of working brain cells. In younger brains, they most often became a specific kind of neuron, a layer 2/3 neuron, while in older brains, maternal cells were more likely to be microglia.

Given that these microchimeric cells take on very different roles, it would be interesting to know whether that diversity reflects their origins ‪—‬ for example, whether they came from a mother, an older biological sibling or a maternal grandmother, said Dr. Sing Sing Way, a microchimerism researcher at Cincinnati Children's Hospital Medical Center who was not involved in the study.

The number of maternal cells declined with age, but they never completely disappeared.

Boddy said the findings mostly raise new questions. "Do we maybe need microchimeric cells to 'help out'?" she asked. "Is diversity of cells in the brain important, or is it just a byproduct of being a placental mammal?”

Because microchimeric cells appear to be common, Boddy suspects they "are doing an important job in the brain, so understanding their function could be very important for understanding healthy brain development."

By using new analytical tools to pinpoint microchimeric cells, the study "pushes the boundaries" of previous research, but future studies would benefit from larger and more uniform datasets, Way said. That would mean obtaining more brain biopsies, analyzing more cells from each sample, collecting specimens at different ages, and sampling similar brain regions across individuals to better compare the results.

This article is for informational purposes only and is not meant to offer medical advice.

Scientists long thought that when RNA kick-started life on Earth 4 billion years ago, it could form only small, simple structures. But new research shows that naturally occurring RNA molecules can also adopt large, sophisticated geometries, like filaments and cages. Now, scientists wonder whether the structures were present at life's beginning.

According to an idea known as the RNA world hypothesis, RNA-based life-forms preceded modern ones that use DNA and protein. RNA, a molecular cousin of DNA, still plays roles in modern cells but does not serve as the primary genetic material. By comparison, primordial species used RNA to store genetic information and to catalyze reactions as stand-in enzymes.

Proteins eventually dominated as enzymes, perhaps because they can fold into more diverse figures than RNA can. That's because proteins are composed of 20 types of subunits, called amino acids, each with a unique structure, whereas RNA is composed of only four subunits, called nucleotides, that all adopt similar shapes.

Scientists originally thought that only proteins were varied enough to assemble into large structures, but a new paper has demonstrated that RNA — though more limited in its variety — also has the capacity to form these large configurations. The research was posted to the preprint server bioRxiv July 1 and has not been peer-reviewed yet.

"We show RNA can do things which we have never seen before," said study co-author Lin Huang, an RNA biologist at Sun Yat-Sen University in China. "It suggests that at the origin of life RNA could assemble into all kinds of shapes," he told Live Science.

Huang and his colleagues had hypothesized that RNA molecules could link together if they possessed sequences that fold into "kissing stem loops." This occurs when an RNA strand folds over on itself, forming a structure that resembles a loop in a shoelace. If loops from different RNAs bond together, or "kiss," the molecules could link up and form larger complexes, the researchers proposed.

After sifting through a bevy of RNA sequences, the researchers found a family of RNA molecules encoded by bacteriophages ‪—‬ viruses that infect bacteria ‪—‬ that form these loops. They purified several of these RNA molecules in the lab, allowed them to assemble in a dish, and then captured their structures using cryo-electron microscopy.

They discovered that some of the RNA molecules formed long filaments. These resembled protein-based filaments such as the cellular cytoskeleton, a scaffold that participates in many functions, including shaping and moving the cell.

Other RNA molecules assembled into cages as large as common viruses. Some of these took the form of an icosahedron — a 3D shape that is built from 20 equilateral triangles and resembles a soccer ball. Many viruses, including herpesviruses, package their genome into protein-based icosahedra called capsids. This raises a question: Could RNA-based capsids have packaged genomes in the RNA world?

This work demonstrates that RNA had the capacity to assemble into these elaborate structures during the RNA world, Huang said, but that doesn't prove it actually happened.

"I definitely think that environmental parameters are a question," Anna Medvegy, an evolutionary biologist at Eötvös Loránd University in Hungary, told Live Science in an email. "Can these structures form in the environment in which the hypothetical RNA World existed?" said Medvegy, who was not involved in the new work.

If scientists could recreate these environmental conditions at the dawn of life, such as high temperatures and low pH, and still observe that these structures take shape, that would strengthen the theory that they could have been present in the RNA world, she said.

Although the RNA cages and filaments were large, Huang's team generated them using only short RNA strands, each no longer than 200 subunits. Medvegy said long RNAs are susceptible to breaking, so if short strands can assemble into these structures, that provides more promise that these multi-tiered molecules could have formed in the RNA world.

Another question is whether these elaborate RNA complexes currently assemble inside the bacteriophage-infected bacteria from which they were derived. So far, Huang's team has only seen these structures form in a lab dish, so they need to determine if factors inside bacteria, such as proteins, would either disrupt or enable their formation inside cells.

Beyond providing insight into life's beginnings, these RNA cages could have potential applications in biotechnology, Huang thinks. Efforts are underway to use DNA folded into "DNA origami" to deliver drugs into cells, and Huang thinks DNA's older cousin, RNA, could one day play a similar role in medicine.