Scientists have found a possible new way to improve a type of brain signaling that is disrupted in some forms of autism.
The approach worked in genetically altered mice and in small human brain-like models grown in the laboratory.
The research was led by Director Eunjoon Kim at the Institute for Basic Science Center for Synaptic Brain Dysfunctions in South Korea. The study was published in Nature Communications on May 29, 2026.
The work focuses on NMDA receptors, which sit on the surface of brain cells and help the cells communicate. These receptors are important for learning, memory and the ability of connections between brain cells to change with experience.
Reduced NMDA receptor activity has been linked with several brain conditions, including some forms of autism, schizophrenia and intellectual disability. However, autism is highly varied, and there is no single biological cause or treatment that applies to everyone.
NMDA receptors need more than one chemical signal to work properly. Along with glutamate, they require glycine, a small molecule that helps the receptor become active.
Scientists have previously tried to increase glycine around brain cells by blocking a transporter called GlyT1. Transporters act somewhat like cleanup systems, moving chemicals from one place to another and helping control their levels.
But GlyT1 is found widely in the brain, including important areas of the brainstem. This has made it difficult to increase NMDA receptor activity in selected brain regions without affecting other functions.
The new study examined a different glycine transporter called Slc6a20a in mice and SLC6A20 in humans. This transporter is found in brain regions involved in thinking and memory, including the cortex and hippocampus.
The researchers reasoned that reducing this transporter might leave more glycine available near NMDA receptors. In theory, this could improve receptor activity in relevant brain areas in a more targeted way.
To test the idea, the team used short pieces of genetic material called antisense oligonucleotides, or ASOs. These molecules can be designed to reduce the production of a specific protein by interfering with the instructions cells use to make it.
The scientists tested the treatment in male mice carrying changes in Shank2 or Shank3. Human versions of these genes, SHANK2 and SHANK3, are important for communication points between brain cells and are associated with some neurodevelopmental conditions.
Changes in SHANK3 can also cause Phelan-McDermid syndrome, a rare genetic condition that can involve developmental delay, intellectual disability and autistic features. Studying these genes gives researchers a way to examine specific biological pathways rather than treating autism as one uniform condition.
In the mouse models, reducing Slc6a20a restored weakened NMDA receptor activity. The treatment also improved some behaviors related to social interaction, social communication and repetitive actions, although the pattern of improvement differed between models.
One especially interesting result was that the treatment produced benefits in adult mice. This suggests that at least some brain signaling problems linked with these genetic changes may remain changeable after early brain development.
The team also examined proteins in the brain to understand what had changed. They found relatively small changes in the total amounts of proteins but clearer changes in chemical tags that control how proteins behave.
These tags, known as phosphorylation, can act like switches that change protein activity. The treatment corrected abnormal patterns involving proteins important for communication between brain cells and NMDA receptor signaling.
The researchers then moved beyond mice and tested the idea in human cortical organoids. These are tiny laboratory-grown collections of human cells that develop some features of the outer part of the brain, although they are far simpler than a real human brain.
Using gene editing, the scientists created organoids carrying SHANK2 or SHANK3 changes. These models showed reduced NMDA receptor function, similar to the problem seen in the mice.
An ASO designed against human SLC6A20 brought NMDA receptor activity closer to normal in the organoids. Reproducing the basic effect in human cells strengthens the biological case for studying the treatment further.
The researchers also reported that a single ASO treatment remained effective in mice for at least eight weeks. They did not detect clear harmful effects during the period examined.
These results are promising, but they are still early-stage research. Improving autism-related behaviors in mice is not the same as treating autism in people, and brain organoids cannot reproduce the full complexity of a human brain, behavior or daily life.
There are also important questions about safety, dose, delivery and which patients might benefit. Because autism has many different genetic and biological causes, a treatment aimed at low NMDA receptor activity would probably be relevant only to a subset of people if it eventually proves effective.
The study is nevertheless important because it offers a more focused way to influence NMDA receptor signaling. Rather than replacing a missing autism-related gene, the treatment changes an existing pathway that may partly compensate for disrupted signaling.
Overall, the strongest finding is the consistency across several experimental systems: the approach improved NMDA receptor function in multiple mouse models and in human organoids. The adult-mouse results are also encouraging because they suggest some signaling problems may remain open to treatment later in life.
At the same time, the study should not be described as a cure for autism. It identifies SLC6A20 as a promising drug target and provides a strong reason for further research, but carefully designed animal safety studies and eventually human clinical trials would be needed before anyone could know whether the approach is safe or useful for patients.
If you care about autism, please read studies that cats may help decrease anxiety for kids with autism and new study may develop better treatment for autism
For more information about health, please see recent studies about how to eat your way to a healthy brain, and results showing this type of food may contribute to autism.
Source: Institute for Basic Science
