Mini-brains in the lab: what one gene reveals about how the autistic brain is built

Sooner or later, almost every family asks the same question: why is my child autistic? Sometimes it comes with guilt, as if something done during pregnancy or in the first months of parenting could explain it. Science has been answering clearly for years that this is not the case, and studies like this one help us understand where the story really begins: in the way the brain builds itself, cell by cell, long before the first day of life.

What they studied

Over the past decade, large genetic studies have identified hundreds of genes linked to autism. Many of them take part in two tasks: forming the connections between neurons and regulating which instructions are switched on while the cerebral cortex takes shape. There is, however, a third, less explored group: the genes of the so-called ubiquitin-proteasome system, a kind of recycling service that tags the proteins a cell should discard and removes those tags when they are worth keeping. Large databases list 125 notable variants in these genes in autistic people, compared with 28 in their siblings who are not on the spectrum.

One of them is USP15, an enzyme whose job is precisely to remove those tags. Earlier work by the same group had found, in an autistic person from a Korean cohort, a change in the very first «letter» of this gene that reduces the amount of protein produced. As a single case, it was not enough to know whether that change influenced development.

To find out, the team from Seoul National University, Korea University and Ewha Womans University turned to brain organoids: small spheres of tissue grown from human stem cells that reproduce, on a small scale, the first stages of cortex formation. They created three versions identical except for that gene: one unchanged, one with a single altered copy, like the person studied, and a third with both copies altered. They grew them for 107 days and analyzed 62,897 cells one by one, grouped into 14 different types.

What they found

The central finding concerns the progenitor cells, the seedbed from which neurons are born. In the organoids with one altered copy, that seedbed emptied ahead of time. At 60 days, the three versions had a similar proportion; at 80 and 100 days, the version carrying the change found in the person studied had about 21 to 22 % fewer progenitors. In the cell-by-cell analysis, they went from 47.32 % of all cells in the unchanged organoids to 29.21 % in the altered ones.

The cells also moved faster along the path that turns them into neurons: they stopped dividing earlier and matured ahead of schedule. The organoids with one altered copy were smaller and had fewer neurons from the deep layers of the cortex, the ones that send messages to distant regions of the brain.

What happened with the synapses, the contact points between neurons, was striking. The genes that build them were more active than usual, yet when the researchers counted the contacts actually assembled under the microscope, there were fewer. It is as if the blueprints were ready but there were not enough hands to raise the building on time. The authors describe it as a mismatch between the instruction and the construction.

Finally, the genes altered in these organoids overlapped significantly with genes already associated with autism and with the changes described in mice with alterations in Fezf2 and Foxp1, two genes that define the identity of those deep-layer neurons. In short, the USP15 variant touched the same developmental gears as other known genetic pathways.

The version with both copies altered, which does not match the person studied, showed a different picture, not just a stronger one: larger organoids, slowed cell division and signs of stress. The team sees this as evidence that the gene dose matters and that losing it completely is not the same as doubling the effect of one copy.

What is still unknown

The article itself spells out its limits, and they are worth keeping in mind:

• An organoid is not a brain. It has no body, senses or relationships, and it reproduces only the earliest stages, equivalent to life before birth. Nothing observed describes how a person thinks, feels or learns.
• The cell-by-cell analysis used one pooled sample per version, made of six organoids from a single culture. Independent repetitions are needed to confirm the patterns.
• The areas where synapses were counted were chosen with fixed criteria, but without hiding which version each sample belonged to.
• A single variant, found in a single person, was studied. The exact mechanism by which USP15 produces these effects is still unknown; the authors propose new experiments to find out.

In other words: it is a valuable step toward understanding the early biology of autism, and its results do not yet have practical applications. Nor does it justify genetic testing on your own.

What we can take home

Although the study takes place in a laboratory, it leaves ideas that do help day to day:

1. Let go of guilt. This work adds to a long list of evidence: the roots of autism lie in the biology of early brain development, often before birth. They do not depend on how you raise your child. What you do at home does not explain why your child is autistic; it does shape how confidently they grow with your support.
2. Compare your child only with themselves. The clearest finding of the study is about timing: when each cell matures and when each connection is assembled. Every brain follows its own calendar. Instead of measuring your child against cousins or classmates, write down their achievements month by month and celebrate how far they have come.
3. Give connections chances to settle. This study does not measure what happens after birth, but decades of research on neuroplasticity show that the brain’s connections grow stronger with repeated experience. A song for washing hands, the same order when getting dressed or a game that comes back every afternoon are brief, predictable chances to practice.

The next step: meet your child at their developmental stage with TAEO | EdTech

This work illustrates something that guides our approach: autism has neither a single cause nor a single form. Hundreds of genes, each with its own variant in each person, shape brain development in different ways. That is why two autistic children may need very different support, and why at TAEO | EdTech we do not start from a label but from what your child already does. The Skills and Competencies Profile places their developmental level across the eight aspects of neurodevelopment and, from there, the platform orders daily activities tailored to them.

The study also reminds us that the brain is built in stages, at its own pace. After birth that work continues, and everyday life, with its routines, games and conversations, becomes the place where each connection can take hold. No app changes genes; what we can offer you is an ordered path to support your child exactly where they are.

At the TAEO Psychoeducational Organization we have poured more than sixteen years of experience into TAEO | EdTech so that every family has clear activities and a human team that guides them step by step.

For our official launch in December 2026, you can request an invitation for the free trial using the form below.