Calming one small brain region was enough to switch off seizures, repeated movements, hyperactivity, and social withdrawal in mice. Turning it back up brought every one of those behaviors back. That happened even in mice that had never shown them before.
That finding, published in Science Advances in August 2025, comes from Stanford Medicine researchers. It offers a new clue about the brain circuits involved in autism spectrum disorder, also called ASD. To be clear up front: this research was conducted entirely in mice. The drug involved has not been tested in autistic people for this purpose. No cure or treatment for humans exists from this study. And the mouse model used reflects only one specific genetic form of autism. It does not reflect the full range of human experience. These are early clues about the brain, not clinical instructions.
Editor’s note: The headline says scientists “successfully reversed autism symptoms.” The study was done in mice, not people. No human treatment comes from this research.
Autism spectrum disorder is a condition of brain development, as defined by the American Speech-Language-Hearing Association. It involves differences in social communication, or how people connect, and restricted or repeated patterns of behavior. For many autistic people, seizures, sensory overwhelm, and epilepsy that occurs alongside autism can be a serious daily burden. Researchers now believe that overlap may trace back to one overactive circuit in the brain.
What brain region were researchers watching?
The thalamus sits near the center of the brain and acts as a relay station. It sorts and sends sensory signals from the outside world to the cerebral cortex. That is the brain’s outer layer, where perception and thought take shape. Wrapped around part of the thalamus is a thin shell of neurons. It is called the reticular thalamic nucleus, or RT. Its job is to act as a gatekeeper. It regulates which sensory signals get passed on and which get filtered out before they reach awareness.
The RT has been studied in relation to sleep, attention, and epilepsy. Its role in autism had not been clearly mapped. Senior author John R. Huguenard is a neurology professor at Stanford University School of Medicine. Lead author Sung-Soo Jang, PhD, is a neurology researcher at Stanford. They suspected the region might matter because of its known connection to seizures. Many autistic people also have a seizure disorder. Huguenard’s lab had spent years studying how the thalamus misfires in those conditions. That overlap prompted a closer look.
The team used mice genetically changed to lack the Cntnap2 gene, a well-established animal model for autism-related traits. These Cntnap2 knockout mice display a cluster of behaviors that parallel autism traits in humans. They’re overly active, prone to repeated movements, and less likely to seek out other mice socially. They’re also more reactive to sensory stimuli. They are much more likely to have seizures. Recordings of their brain activity showed the reason for the last trait clearly. The RT neurons were firing in abnormal bursts. That drove waves across connected circuits and triggered seizures without any outside trigger.
Read More: Researchers Identify Four Distinct Autism Subtypes in Study of 5,000 Kids
What did the data show?
When researchers recorded the RT activity of these mice during everyday experiences, the pattern stood out. The reticular thalamic nucleus showed higher activity when animals were exposed to light, an air puff, or other mice. In healthy mice, the RT filters these inputs. It passes along only the most relevant signals. In the Cntnap2 mice, the region over-responded to nearly everything. It essentially jammed the brain’s sensory filter.
This finding matters because of the known overlap between autism and seizure disorders in humans. Epilepsy is much more common among autistic people than in the general public. That gap has been recognized for decades. The process linking the two conditions has never been fully explained. The Stanford findings suggest that the RT’s tendency to become overactive may be a shared root. It may help explain why seizures are so much more common among autistic people than in the broader public.
The researchers then tested whether calming the RT would reverse the behaviors they had documented. It did. Suppressing RT activity reduced seizure risk and dialed down reactions to sensory stimulation. It also reduced extra activity and repeated behaviors. It improved social interaction in the autism-model mice. The effect was specific enough to suggest a clear cause. Overactivity in this one region was driving the traits seen in these mice. It did not appear to be wide whole-brain dysfunction.
What drug quieted the circuit?
The first tool the team used to suppress RT activity was a drug. Z944 is a test drug that blocks T-type calcium channels. Those channels are the electrical mechanism that drives the RT’s abnormal burst firing. It was already being studied as a possible seizure drug. When given to the Cntnap2 mice, Z944 calmed the RT when it was too active. It restored more typical social behavior and reduced repeated movements. The mice also had fewer seizures.
Z944 has not been tested in autistic people for this use. It remains a test compound in the epilepsy pipeline. It is not an approved or available therapy. The mouse data, according to researchers cited by The Transmitter, suggest that drugs already being developed for seizures could possibly serve two purposes. That would depend on the RT circuit idea holding up in future human research. That is a hypothesis for future study. It is not a current treatment option.
The second approach used a technique called DREADDs, a lab method used to control specific brain cells. DREADDs, or Designer Receptors Exclusively Activated by Designer Drugs, are engineered proteins inserted into specific neurons. They allow researchers to turn those neurons on or off. The chemicals have no effect on other brain cells. The team used a calming DREADD approach to quiet RT neurons in the autism-model mice. They repeated what Z944 had achieved with a drug. In this case, the tool was used to calm the cells. They saw the same behavioral reversal, through a more targeted method.
Then researchers reversed the experiment. Using the same DREADD technology, they ramped up RT activity in normal, healthy mice. Those animals had no gene change and no prior history of autism-like behavior. Those mice began displaying extra activity, repeated behaviors, and reduced social interaction. They also became more sensitive to stimulation. Normal mice developed autism-like behaviors simply because one brain region was made to fire too much.
That two-way finding is the strongest scientific part of the study. Reversing a behavior in a disease model could reflect off-target effects. Those are effects the researchers were not aiming for. Producing that same behavior in a healthy animal by targeting a single circuit is a much stronger argument. It suggests the circuit plays a causal role, rather than only lining up with other changes.
What are the limits of this research?
Understanding what this study does not show is as important as understanding what it does show. The Cntnap2 knockout mouse model represents one specific genetic path to autism-related traits. Autism in humans is a broad spectrum of brain development. It is shaped by hundreds of different gene variants, environmental factors, and developmental paths. No single mouse model captures that breadth. Findings from Cntnap2 mice may apply most directly to autistic people who carry changes in the CNTNAP2 gene. That is a subset, not the whole group.
There is also a basic gap between mouse brain science and human therapy. The RT circuit and its calcium channel-driven burst firing may work in a similar way across species. Human brains are still far more complex. What calms a too-active RT in a mouse, and does so cleanly, may have different or broader effects in a human brain. Z944 would need to pass clinical trials that test it in autistic people. Those trials do not currently exist. Until they do, it could not be considered a treatment for anything other than a seizure disorder.
What should families take from this study?
For families and autistic people following research, the most grounded takeaway is about how the circuit works. Scientists now have evidence that a specific brain region, the reticular thalamic nucleus, may act as a central switch. That is for multiple autism-associated traits in this genetic model. That is a more precise target than researchers have had before in this area of study. Precision is what makes drug development possible.
The practical implication for right now is not a new treatment. Z944 is not available for autism. The DREADD method is a lab research tool, not a clinical therapy. The overlap between seizure disorders and autism has long been observed and poorly understood. It now has a plausible circuit-level reason. That reason points toward the RT as a possible target for future autism therapies. If you or someone you care for is autistic and also managing a seizure disorder, this research is worth discussing with a neurologist, a doctor who treats brain conditions. That is not because there is a new treatment to ask for. It is because those two neurological profiles can be closely linked. That understanding is beginning to get sharper. This is not a reason to start, stop, or change any medication or treatment on your own.
Disclaimer: The author is not a licensed medical professional. The information provided is for general informational and educational purposes only and is based on research from publicly available, reputable sources. It is not intended to constitute, and should not be relied upon as, medical advice, diagnosis, or treatment. Always consult a licensed physician or other qualified healthcare provider regarding any medical condition, symptoms, or medications. Do not disregard, avoid, or delay seeking professional medical advice or treatment because of information contained herein.
AI Disclaimer: This article was created with the assistance of AI tools and reviewed by a human editor.