In a laboratory at UCLA, scientists grew clusters of human brain-like tissue from stem cells carrying autism-linked mutations, then quietly strengthened the one healthy gene copy those cells still possessed. The cells drifted back toward typical development, a modest but meaningful sign that a faulty gene's effects may be softened from the other side rather than only managed afterward. Much of autism's genetic risk arises in a child anew rather than being inherited, and a single working copy often cannot carry the load alone. This work, published in Nature Neuroscience, offers a proof of conce
Boosting the Healthy Gene Copy Offsets Autism-Linked Mutations in Lab-Grown Brain Cells
Turn up the one healthy copy the cell still has
Why does it matter that they boosted the healthy copy instead of fixing the broken one?
Many of these mutations leave a person with one working copy instead of two. Turning up that remaining copy works with what the cell already has, and using enhancers aims to keep activity matched to natural development.
What exactly did they test it on?
Human stem cell-derived excitatory neurons and cortical organoids, the small brain-like clusters. Two genes, CHD8 and SCN2A.
Before anyone gets excited, all of this is in dishes. What did it actually show?
In CHD8 models, less abnormal organoid overgrowth, better cell maturation, and gene activity closer to typical. In SCN2A models, much of the normal electrical activity came back.
Much of it. Not all.
Right. Some finer electrical properties normalized only in certain cell lines, and the shift from mutant to normal levels varied between lines.
What about the specificity?
Healthy cells showed no increase in gene activity and neighboring genes weren't affected. Chen said that matters for any future therapy.
That's in the tested cell lines, though. Nothing here says how it behaves in a living brain.
Correct. Delivery to the brain is unsolved, and nobody yet knows if it works after early development. Timing mattered in the experiments.
So what's the honest takeaway?
A proof of concept that could extend to other conditions caused by one faulty gene copy, with a long road ahead.
Der Puls
- Lab-grown human brain tissue carrying autism-linked mutations in CHD8 and SCN2A showed abnormal growth and electrical activity that the team set out to correct.
- The team turned up the activity of the remaining healthy gene copy using a modified CRISPR tool, aiming at DNA dials called enhancers rather than the faulty gene itself.
- Overgrowth eased, cells matured, and gene activity moved closer to typical levels, though the degree of change varied across cell lines.
- The effect appeared only in mutant cells, leaving healthy cells and neighboring genes untouched, a specificity researchers call essential for any future therapy.
- The results remain confined to laboratory models, with delivery to the living brain, treatment timing, and effectiveness after early development still unresolved.
In a laboratory at UCLA, scientists grew clusters of human brain-like tissue from stem cells carrying autism-linked mutations, then quietly strengthened the one healthy gene copy those cells still possessed. The cells drifted back toward typical development, a modest but meaningful sign that a faulty gene's effects may be softened from the other side rather than only managed afterward. Much of autism's genetic risk arises in a child anew rather than being inherited, and a single working copy often cannot carry the load alone. This work, published in Nature Neuroscience, offers a proof of concept that remains far from any patient, yet it reframes the biology in a way that invites patience and careful steps.
At UCLA Health, researchers grew small clusters of brain-like tissue from human stem cells, each carrying a mutation linked to autism. Rather than trying to replace the faulty gene, they turned up the activity of the one healthy copy those cells still had. The cells moved back toward typical development, and the findings were published in Nature Neuroscience.
The starting point is a genetic reality. Autism is highly heritable, and roughly 15% of its genetic risk comes from new mutations that appear for the first time in a child. Many of these disable one of a gene's two copies, and a single working copy often cannot produce enough of the gene's product. Most current treatments address the symptoms that follow, not the underlying biology.
Led by Dr. Daniel Geschwind, the team focused on two genes: CHD8, which helps regulate other genes in brain development, and SCN2A, which neurons need to fire electrical signals. They used a modified CRISPR tool aimed at enhancers, the DNA segments that act like dials controlling how strongly a gene is expressed. Adjusting the dial rather than the gene was meant to keep activity in step with natural development and avoid harmful overexpression.
In CHD8 models, boosting the healthy copy reduced abnormal overgrowth in organoids and helped cells mature. In SCN2A models, much of normal electrical activity returned, though some finer properties normalized only in certain cell lines. Two findings stood out: the effect appeared only in mutant cells, and timing seemed to matter.
First author George Chen cautioned that the work remains far from the clinic, and the authors stress its limits. This is proof of concept in lab models. Safe delivery to the brain, the right developmental window, and whether the approach works after early development must all be solved. Because the strategy relies on a healthy gene copy, it may extend to other conditions caused by one faulty copy, but delivery and timing are the questions to watch.
In a lab at UCLA Health, researchers grew small clusters of brain-like tissue from human stem cells, each carrying a mutation tied to autism. They then turned up the activity of the one healthy gene copy those cells still had. The result, published in Nature Neuroscience, was that the cells moved back toward typical development.
The work starts from a genetic fact about autism. The condition is highly heritable, yet roughly 15% of its genetic risk comes from new mutations that appear for the first time in a child rather than being passed down from a parent. Many of these mutations disable one of a gene's two copies. A single working copy often cannot make enough of the gene's product on its own, and neurodevelopment can change as a result. Most current autism treatments address symptoms that follow from these mutations, not the underlying biology.
The team, led by Dr. Daniel Geschwind, focused on two genes strongly linked to autism. CHD8 helps regulate other genes involved in brain development. SCN2A is essential for neurons to fire electrical signals.
The researchers did not try to replace or edit the faulty gene. They used a modified form of CRISPR to raise the activity of the remaining healthy copy, aiming it at enhancers. These are stretches of DNA that work like dials, controlling when and how strongly a gene is expressed, often in a cell-type-specific way. Targeting the dial rather than the gene itself was meant to keep activity in step with natural brain development and avoid overshooting into harmful overexpression.
They tested the approach in two kinds of human stem cell-derived models: cultures of excitatory neurons, and cortical organoids, which mimic aspects of early brain development and key features of neurodevelopmental disorders seen in affected people. In CHD8 models, boosting the healthy copy reduced abnormal organoid overgrowth, helped cells mature, and shifted gene activity closer to typical levels. In SCN2A models, the treatment restored much of the neurons' normal electrical activity, though some finer electrical properties normalized only in certain cell lines.
Two other findings stood out. The effect appeared only in cells carrying the mutation; healthy cells showed no increase in gene activity, and neighboring genes were unaffected. And timing mattered, which suggests the approach may work best at particular stages of development. First author George Chen, a former postdoctoral fellow in the Geschwind lab and now an assistant professor of neurology at Children's Hospital of Los Angeles, said that specificity matters for any future therapy, while cautioning that the work remains far from the clinic.
Geschwind, the Gordon and Virginia MacDonald distinguished professor of Human Genetics, Neurology and Psychiatry at UCLA, described the result as a key first step that offers a new way to think about the biology underlying these conditions.
The authors stress the limits. Everything was done in lab models, and the degree of shift from mutant to normal expression varied between cell lines, so this is proof of concept, not a ready-to-use treatment. Safe delivery to the brain will need better methods, and researchers still have to test whether the approach works after the early developmental window has passed. Because the strategy relies on a healthy gene copy, it is expected to extend to other neurodevelopmental conditions caused by one faulty copy of a gene. Delivery and timing are the questions to watch.
Bemerkenswerte Zitate
We are still a long way from the clinic, and there is a lot of work ahead on delivery and timing, but these findings provide hope.— George Chen, first author, Children's Hospital of Los Angeles
In human cell models, it's possible to turn up the remaining copy and move cells back toward typical development.— Dr. Daniel Geschwind, UCLA (paraphrased)