A six-year-old girl in China died in March 2025 after receiving what researchers said they believed was the world’s first base-editing therapy directed at the brain, according to an investigation by Science and Retraction Watch.

The investigation has raised important questions about the oversight of an experimental personalised gene-editing trial and the way the researchers and their institutions communicated their choices.

The girl had Snijders Blok-Campeau syndrome, a rare neurodevelopmental disorder caused by a mutation in the CHD3 gene. Although the condition affects intellectual development and its severity varies widely, most people with the condition live full lives.

But that such a case could still clear the bar for an unprecedented, high-risk brain intervention says something about a society uneasy with disability and enamoured of techno-scientific interventions.

The girl’s parents sought the experimental treatment after learning of research led by neuroscientist Zilong Qiu at the Shanghai Jiao Tong University.

Base-editing is a more recent invention than CRISPR-Cas9. The American biochemist David Liu developed it at the Broad Institute in 2016 after finding a way to convert one DNA letter directly into another — say, C to T — without cutting both strands of the DNA double helix, as Cas9-based editing does.

As a result, base-editing is, in principle, more ‘gentle’ and can cause fewer unintended mutations. Its development thus renewed hopes that gene-editing could graduate from being a proof of concept and to treating rare disorders caused by defects in single genes.

That said, the memory of He Jiankui, the Chinese scientist who announced in 2018 that he had edited the genomes of twin girls at the embryonic stage without adequate ethical review, scientific justification, and even disclosure to his own university, has weighed heavily on the field.

He was later sentenced to three years in prison for his actions, and the episode forced China to tighten its rules on human gene-editing research. But as the new investigation by Science and Retraction Watch has found, the rifts between what the law says and how it is enforced have remained wide enough for a similar failure to recur.

Dr. Qiu’s team developed a personalised base-editing therapy designed to correct the genetic mutation in the child’s brain cells.

However, delivering the ‘editor’ to the brain required the researchers to inject large doses of adeno-associated viruses (AAVs), which are known to carry risks of severe immune reactions.

According to the investigation report, the child’s parents helped finance much of the therapy’s development, contributing around $860,000 (Rs 8.3 crore) from their personal savings and money borrowed from relatives.

The treatment proceeded as a clinical trial initiated by the investigator at the Xinhua Hospital in Shanghai. This process did not require the National Medical Products Administration, China’s national drug regulator, to review it first.

However, the report has quoted independent experts expressing concerns with several aspects of the process — including whether preclinical studies (in animals) had adequately proved the treatment’s safety for human use.

Toxicology studies in monkeys reportedly found that all the treated animals had developed liver injury and one monkey’s kidneys were damaged as well.

However, the researchers proceeded to inject the therapy into the girl’s cerebrospinal fluid on March 24, 2025. Within days, she developed a fever and kidney damage, and died a week later.

An internal review at the hospital concluded that the treatment was certainly linked to her death, and listed the cause of death as thrombotic microangiopathy, a complication that has previously been associated with genetic therapies involving high doses of AAVs.

According to Science and Retraction Watch, neither the researchers nor the hospital disclosed the girl’s death at the time. Several months later, local health authorities fined the hospital for shortcomings in trial oversight and registration, but did not sanction the lead researcher.

The family said they later asked that a research paper related to the trial be withdrawn, arguing that it did not reflect the appropriate outcomes.

The investigation also examined a paper published in Nature earlier this year describing the underlying preclinical work. Some independent experts questioned aspects of the reported data and argued that the journal should examine the underlying evidence and funding disclosures.

Nature said it had not been informed of the patient’s death or the regulatory issues before publication and that such information would have been considered during editorial review.

The case has prompted renewed debate about ‘first in human’ gene-editing trials, particularly for rare but serious disorders.

In June 2024, Nature had reported a similar story in India: researchers across the country raced against time to develop a treatment for Uditi Saraf, a 20-year old with familial encephalopathy with neuroserpin inclusion bodies, or FENIB — a rare and aggressive brain disorder with symptoms similar to dementia.

Sadly, Uditi passed away before the treatment was ready, although the Sarafs, who helped fund the tests, and the researchers expressed hope afterwards that their efforts could help other people with FENIB.

The researchers’ challenges included identifying the mutation, engineering the base-editor, manufacturing the AAV vectors, and improving institutional capacity to produce a gene therapy of direct clinical value.

Both instances — Uditi Saraf and the little girl in China — began with researchers committing themselves to a schedule they knew would be uncertain and high-pressure.

But in Uditi’s case, while the researchers worked 12-hour days and called in favours from colleagues abroad, their efforts also had to contend against the friction of collaborating across borders and dealing with the US’s more cautious regulatory system. In India as well, Arkasubhra Ghosh at Narayana Nethralaya Eye Hospital was waiting for approval from Indian regulators to manufacture AAVs when Uditi passed away.

In China, on the other hand, the researchers faced little resistance going from decision to injection.

Of course, this doesn’t mean India treats disability better than China. Stigma against disabled people here is widespread, sustained by misunderstanding, misinformation, and prejudice.

The Sarafs’ experience also opened a window to India’s opportunity to become a centre for affordable gene-editing as it wound through Ghosh’s ambition to manufacture AAVs within the country and Debojyoti Chakraborty’s efforts at the CSIR-Institute of Genomics and Integrative Biology to develop inexpensive CRISPR therapies.

That said, while India offered the possibility of acting before it was too late, the researchers might have faced similar questions over the preclinical evidence and toxicology reports and regulatory penalties against hospitals if the treatment had gone awry.

Finally, none of the regulatory philosophies at work in these episodes guarantees success. One emphasised exhaustive reviews before ‘first in human’ use. Another allowed investigator-initiated trials to proceed more rapidly. A third was willing to devolve trust based on reputation and clinical stature over external reviews, including of safety.

In fact, while the law in China forbids charging patients for unproven therapy, it is also written narrowly enough for informal payments to individual researchers — such as the gifts to Dr. Qiu — to fall outside its purview.

Communicating risk and the possibility of death also have thorny social dimensions, including the optimism surrounding frontier biotechnologies and the emotional entanglement that can develop between families and researchers.

This handout picture obtained on May 30, 2024 from the Institut Botanic de Barcelona (CSIC) shows a fern in New Caledonia.
| Photo Credit: AFP

New research shows that a fork fern species, called Tmesipteris oblanceolata, has a genome — all the genetic information of an organism — that is 7% larger than that of the previous record-holder, the Japanese flowering plant Paris japonica, and more than 50 times the size of the human genome.

The metric for genome size was the number of base pairs, the fundamental units of DNA, in an organism’s cellular nuclei. If stretched out like from a ball of yarn, the length of the DNA in each cell of this fern would extend nearly 350 feet (106 meters), taller than New York’s Statue of Liberty, the clock tower for London’s Big Ben or India’s Taj Mahal. The human genome would extend a mere 6-1/2 feet (2 meters).

The fern grows primarily on the ground or atop fallen tree trunks in New Caledonia, about 750 miles (1,200 km) east of Australia, and on neighboring islands such as Vanuatu.

“We can tell you that this species is not very showy. It is a small plant, typically 10-15 centimeters (4-6 inches) in height, which could easily go unnoticed to anyone who’s not specifically looking for it,” said evolutionary biologist Jaume Pellicer of the Botanical Institute of Barcelona (IBB), co-lead author of the study published on Friday in the journal iScience.

The specimens used in the study were collected last year on New Caledonia’s Grande Terre Island. Its leaf-like structures are not true leaves, but rather flattened stems. It belongs to a fern family whose evolutionary lineage split from other ferns around 350 million years ago, approximately 120 million years before dinosaurs appeared.

A large genome is not considered beneficial.

“We think that Tmesipteris oblanceolata has such a large genome not for any evolutionary advantage but simply because, for a reason that we do not know yet, it has failed over time to efficiently remove non-functional or redundant DNA sequences,” Pellicer said.

DNA, which carries an organism’s genetic information, is comprised of two linked strands that wind around each other in a shape called a double helix that looks like a twisted ladder. Base pairs make up the ladder’s rungs.

Multiple consequences

Genome size has multiple consequences. For instance, larger genomes require more resources for DNA replication, repair and transcription — a process a cell uses to make the proteins an organism needs to function.

“This increased demand can strain the plant’s energy and nutrient resources, which could otherwise be used for growth, reproduction and stress responses,” said IBB evolutionary biologist and study co-lead author Oriane Hidalgo.

Bigger genomes necessitate bigger cells to house them and take longer to replicate the DNA, slowing down cellular division underpinning growth. This could reduce the ability of a plant species to compete for light, water and nutrients, making it ecologically more restricted, Hidalgo said.

Scientists have wondered why some organisms have large genomes and others do not.

“This is indeed a puzzling question. But what is clear is that there is no relationship between genome size and organism complexity,” said plant geneticist and study co-author Ilia Leitch of the Royal Botanic Gardens Kew in London.

There also is no relationship between an organism’s physical size and its genome size. The little fern’s genome is about 6,000% bigger than that of a blue whale, Earth’s largest animal, and about 4,650% bigger than that of an African elephant, the largest land animal. It also is about 1,500% larger than that of the giant redwood, the tallest plant.

Africa’s marbled lung fish has the biggest-known animal genome. The fern’s is nearly 25% larger.

Scientists have gauged the genome size of about 20,000 organisms possessing clearly defined nuclei, with big genomes the exception rather than the rule.

“Genome size is an important biodiversity character as it has been shown to play a role in influencing how, where and when a plant is able to grow and compete, and respond to environmental challenges such as climate change and pollution,” Leitch said.