Summary: Researchers have identified a precise molecular strategy to restore NMDA receptor (NMDAR) function, addressing a central pathology in autism spectrum disorder (ASD). The approach redirects attention from broadly acting, toxic systemic targets to the glycine transporter Slc6a20a / SLC6A20, which is concentrated in cognition-related brain regions such as the cortex and hippocampus.
Using antisense oligonucleotides (ASOs) in adult mouse models carrying SHANK2 and SHANK3 mutations, and in CRISPR-edited human cortical organoids, the research team restored NMDAR signaling, corrected synaptic phosphorylation imbalances, and reversed persistent behavioral deficits without inducing the respiratory or motor side effects seen with earlier therapies.
Key Facts
- The NMDAR activation challenge: NMDA receptors require both glutamate and glycine for full activation. NMDAR hypofunction is implicated in ASD, schizophrenia and intellectual disabilities. Prior clinical attempts to boost glycine by inhibiting GlyT1 failed because GlyT1 is abundant in brainstem centers, producing dangerous respiratory and motor side effects.
- Regional precision for safety: Slc6a20a (SLC6A20 in humans) is a glycine transporter highly expressed in higher-order cognitive regions—cortex and hippocampus—but sparsely present in brainstem motor and respiratory nuclei. Targeting Slc6a20a therefore offers an anatomically selective route to enhance NMDAR co-agonist availability where it matters for cognition and social behavior.
- Behavioral rescue in adults: In adult mice with mutations in autism-risk genes SHANK2 and SHANK3, a Slc6a20a-directed ASO increased NMDAR function and reversed established deficits in social interaction, communication, and repetitive behaviors, demonstrating efficacy beyond early developmental windows.
- Phospho-proteomic mechanism: Large-scale phospho-proteomic profiling showed the ASO did not broadly change protein abundance. Instead, it normalized abnormal phosphorylation patterns across synaptic signaling networks and NMDAR regulatory proteins, indicating functional restoration at the signaling level.
- Human cortical organoid validation: Human cortical organoids edited to carry SHANK2 or SHANK3 mutations exhibited reduced NMDAR activity. Treatment with a human-targeted SLC6A20-ASO restored NMDAR function toward normal levels, supporting translational relevance.
- Durable single-dose effect: A single ASO administration produced measurable rescue for at least eight weeks in vivo with no detectable adverse signals, suggesting a long therapeutic window from infrequent dosing.
- Broader neuropsychiatric potential: Because this approach modulates native signaling pathways rather than attempting gene re-expression, it could be adapted to treat other disorders characterized by NMDAR hypofunction, including some forms of schizophrenia and intellectual disability.
Source: Institute of Basic Science
Overview of the study
A team led by Director Eunjoon Kim at the IBS Center for Synaptic Brain Dysfunctions explored whether lowering Slc6a20a expression could safely increase local glycine availability and restore NMDA receptor activity. The strategy replaces prior, nonselective GlyT1 inhibition with a targeted ASO that selectively reduces Slc6a20a transcript levels in cognition-related regions while leaving brainstem circuits intact.
The NMDA receptor’s reliance on both glutamate and glycine makes glycine transporters attractive therapeutic targets. However, GlyT1 inhibitors broadly elevate glycine throughout the brainstem and forebrain, which led to adverse effects in prior human trials. The current study redirects the focus to Slc6a20a / SLC6A20, whose expression pattern favors cortex and hippocampus and avoids the brainstem centers that control breathing and basic motor functions.
In multiple mouse models with autism-associated Shank2 and Shank3 lesions, intracerebral administration of Slc6a20a-ASO restored NMDA receptor-mediated synaptic currents in the prefrontal cortex and hippocampus. Behavioral assays demonstrated improved social approach, normalized ultrasonic social communication, and reduced stereotyped repetitive actions. Crucially, adult animals with established phenotypes showed substantial recovery, indicating that NMDAR-targeted interventions can be effective after developmental periods.
Phospho-proteomic analysis of cortical tissue revealed that the ASO’s primary effect is to rebalance phosphorylation states in synaptic signaling proteins and regulators of NMDARs rather than to alter overall protein abundance. This suggests the therapy rescues functional signaling dynamics and synaptic responsiveness.
To test human relevance, the investigators generated cortical organoids with CRISPR-engineered mutations in SHANK2 or SHANK3. These organoids showed suppressed NMDAR responses similar to the mouse models. A human-directed SLC6A20-ASO reinstated NMDAR activity to near-normal levels, supporting the translational potential of selective glycine transporter modulation.
Director Eunjoon Kim emphasizes that SLC6A20 inhibition acts by modulating endogenous signaling pathways, offering a potentially practical and scalable therapeutic route compared to strategies that try to reintroduce or overexpress missing genes. The single-dose durability observed and the absence of detectable toxicity in treated animals strengthen the case for further development.
Key Questions Answered
A: Earlier therapies targeted the GlyT1 transporter, which is heavily expressed in brainstem areas that control breathing and motor function. Inhibiting GlyT1 raised glycine broadly across these regions and interfered with vital autonomic processes, producing severe respiratory and motor side effects that undermined clinical trials.
A: SLC6A20 is predominantly localized to higher-order cognitive regions such as cortex and hippocampus and is largely absent from brainstem respiratory centers. Using ASOs to reduce SLC6A20 expression increases glycine selectively in cognition-related areas, preserving breathing and motor circuits.
A: No. The study demonstrated effective rescue in fully mature adult mice with long-standing behavioral phenotypes. A single ASO treatment restored NMDAR function and corrected social and repetitive behavior deficits, indicating that intervention can be beneficial beyond developmental critical periods.
Editorial Notes
- This article was edited by a Neuroscience News editor.
- The journal paper was reviewed in full.
- Additional context was added by the editorial staff.
About this genetics and autism research news
Author: William Suh
Source: Institute for Basic Science
Contact: William Suh – Institute for Basic Science
Image: The image is credited to Neuroscience News
Original Research: Open access. “Glycine-modulating Slc6a20a-ASO restores NMDA receptor function in SHANK2 and SHANK3-mutant mice and cortical organoids” by Junyeop Daniel Roh, Mihyun Bae, Yusang Oh, Yeji Yang, Suho Lee, Woo-Chang Hwang, Esther Yang, Hyeonji Kim, Hyunjee Jang, Hyung-Wook Choi, Hyun Kim, Jin Young Kim & Eunjoon Kim. Nature Communications. DOI: 10.1038/s41467-026-73881-9
Abstract
Glycine-modulating Slc6a20a-ASO restores NMDA receptor function in SHANK2 and SHANK3-mutant mice and cortical organoids
Reduced NMDA receptor (NMDAR) function contributes to disorders such as schizophrenia, autism spectrum disorder, and NMDAR encephalitis. Attempts to boost ambient glycine by inhibiting GlyT1 have produced mixed clinical outcomes, largely because GlyT1 is abundant in brainstem regions. Slc6a20a, a glycine transporter enriched in cortex and hippocampus, represents a regionally selective alternative.
This study demonstrates that ASO-mediated suppression of Slc6a20a normalizes NMDAR function and behavioral phenotypes in male Shank2 and Shank3 mutant mice and corrects phospho-proteomic signaling profiles in the prefrontal cortex. A human-directed SLC6A20-ASO likewise rescues suppressed NMDAR function in CRISPR-edited cortical organoids with SHANK2 or SHANK3 mutations. These results highlight the therapeutic potential and boundaries of Slc6a20a/SLC6A20-targeted ASO interventions for disorders marked by NMDAR hypofunction.