Researchers suppressing a glycine transporter called Slc6a20a in mice carrying autism-linked genetic mutations reported changes in social interaction, social communication and repetitive grooming behaviors, alongside restored function in a brain receptor that had been underperforming.
The work, published in Nature Communications by a team led by Eunjoon Kim, director of the Institute for Basic Science Center for Synaptic Brain Dysfunctions in South Korea, used antisense oligonucleotides to reduce the transporter's expression in mice with mutations in the SHANK2 and SHANK3 genes. The first author is Junyeop Daniel Roh.
This is laboratory research in animals and in cultured human brain tissue. It does not establish that the target is effective or safe in people, and no human trial has tested it.
The Receptor Problem the Approach Targets
The NMDA receptor is a central component of learning and memory, and reduced NMDA receptor function has been implicated in several brain conditions including autism spectrum disorder, schizophrenia, and NMDA receptor encephalitis.
Activating it requires two molecules rather than one. Glutamate binds at one site, and glycine, a critical co-agonist, must occupy a second site before the receptor opens. Insufficient glycine at the synapse leaves the receptor underactive even when glutamate is present.
Glycine transporters remove glycine from the space between neurons, and blocking a transporter raises the local concentration. That logic is not new, and it is why the specific transporter chosen matters more than the general strategy.
Earlier efforts targeted GlyT1, a transporter distributed broadly across the brain. Attempts to restore NMDA receptor activity through GlyT1 inhibition yielded mixed outcomes, which researchers attribute partly to GlyT1's extensive expression in essential brainstem regions.
Slc6a20a is widely expressed in cognition-relevant regions including the cortex and hippocampus. The premise of this work is that targeting it offers a more targeted alternative.
The Experiments and What They Showed
The researchers used antisense oligonucleotides, short synthetic strands of genetic material that bind to messenger RNA and reduce production of a specific protein.
They tested this in male mice carrying Shank2 and Shank3 mutations. Both genes are strongly associated with autism and with Phelan-McDermid syndrome, a rare genetic condition involving developmental delay, intellectual disability, absent or severely delayed speech and often seizures.
The reported results included restored social interaction, improved social communication, reduced excessive self-grooming, and normalized NMDA receptor function and synaptic phospho-proteomic profiles in the prefrontal cortex. The paper describes model-dependent rescue profiles, meaning effects differed between the Shank2 and Shank3 models rather than being uniform. Large-scale phospho-proteomic analysis indicated the treatment caused relatively little change in total protein amounts.
The team also used CRISPR gene editing to generate human cortical organoids, small clusters of brain cells grown in culture, carrying SHANK2 or SHANK3 mutations. These showed reduced NMDA receptor activity, and an antisense oligonucleotide targeting human SLC6A20 restored function to near-normal levels.
Two results drew particular attention. Effects were observed in adult mice, not only during early development, which the researchers suggest means correcting NMDA receptor dysfunction may remain possible after key developmental stages have passed. And a single administration remained effective for at least eight weeks, with no detectable adverse effects in treated mice during that period. Both findings are from mice.
"Unlike gene re-expression strategies, SLC6A20 inhibition works by modulating endogenous signaling pathways and may offer a more practical therapeutic route," Kim said.
The Distance Between This and Any Human Treatment
Several boundaries deserve explicit statement.
Mouse behavior is not autism. Researchers measure proxies including time spent near another mouse and duration of self-grooming, and label them social interaction and repetitive behavior. Those measures are useful for testing a molecular hypothesis. They do not capture the human experience of autism, which involves communication, sensory processing, executive function, and social understanding in ways no rodent assay approximates.
The models are genetically specific. Shank2 and Shank3 mutations account for a small fraction of autism diagnoses. Most autism has no single identified genetic cause, and a mechanism operating in these models may not generalize.
Only male mice were used, a persistent limitation in autism research given documented sex differences in presentation and diagnosis.
Organoids are not brains. They lack vasculature, immune cells, sensory input and the organized architecture of a developing brain, and results in them do not predict clinical outcomes.
Delivery is a practical obstacle. Antisense oligonucleotides do not cross the blood-brain barrier and generally require injection into spinal fluid, a procedure carrying its own risks. Approved drugs in this class exist for other neurological conditions, so the route is feasible, but it is a substantial intervention rather than a pill.
The Framing This Research Deserves
Coverage of autism research frequently uses language about reversing or curing, and that framing is both scientifically inaccurate here and contested by many autistic people and advocacy organizations who describe autism as a difference in neurological development rather than a disease requiring elimination.
The more precise description of what this work targets is a specific molecular deficit associated with particular genetic syndromes in which affected individuals often experience significant disability, including absent speech, seizures, and intellectual disability. Families affected by Phelan-McDermid syndrome and related conditions are the population for whom this line of research is most directly relevant.
That distinction is not semantic. It determines who a future therapy would be for and what outcome would count as success.
Nothing here changes anything available today. Families should be cautious about clinics or products claiming to act on glycine, NMDA receptors or related pathways, since no such treatment has been shown safe or effective for autism and supplements marketed on this basis are not evaluated for either.
Evidence-based supports, including speech and language therapy, occupational therapy, and applied behavioral supports, remain what is actually available, and access questions rather than molecular ones determine most families' outcomes. MedicalDaily has covered federal investment in community mental health infrastructure that serves developmental disability populations, and research complicating another long-assumed brain target.
Key Questions Answered
What did the researchers do? They used antisense oligonucleotides to suppress a glycine transporter called Slc6a20a in mice carrying SHANK2 and SHANK3 mutations, and reported restored NMDA receptor function along with changes in social interaction, social communication, and repetitive grooming.
Why does glycine matter for this receptor? The NMDA receptor needs both glutamate and glycine to activate fully. Transporters remove glycine from the synapse, so blocking one raises local concentration and can restore receptor function.
How is this different from earlier attempts? Earlier work targeted GlyT1, which is extensively expressed, including in essential brainstem regions, and yielded mixed outcomes. Slc6a20a is concentrated in cognition-relevant regions such as the cortex and hippocampus.
Does this work in humans? Unknown. No human trial has been conducted. The research was done in mice and in CRISPR-edited human cortical organoids, neither of which predicts clinical outcomes.
Which forms of autism does this concern? The models carry SHANK2 and SHANK3 mutations, associated with Phelan-McDermid syndrome and related conditions. These account for a small fraction of autism diagnoses, and most autism has no single identified genetic cause.
What are the main limitations? Male mice only, model-dependent results, organoids that lack brain architecture, and antisense oligonucleotides that cannot cross the blood-brain barrier and would require spinal injection.
Should families act on this? No treatment exists. Families should be wary of clinics or supplements claiming to act on glycine or NMDA receptor pathways, as none has been shown safe or effective for autism.
