Stanford study shows autism-like behaviors can be switched off in mice

New research from Stanford University suggests that an overactive brain circuit deep in the thalamus may contribute to behaviors linked to autism. In a mouse model of the condition, suppressing activity in this region reduced symptoms such as social withdrawal, repetitive behavior, and heightened sensitivity to sensory input. The findings, published in Science Advances, point to the thalamus as a potential target for future therapies.

Autism spectrum disorder, or ASD, is a developmental condition marked by challenges in social interaction, communication differences, and restricted or repetitive behaviors. Many individuals also experience co-occurring conditions such as epilepsy, anxiety, or hyperactivity. While most research into autism has focused on areas of the cerebral cortex involved in language and social cognition, less attention has been paid to the deeper structures that relay sensory signals to the brain’s outer layers.

The thalamus is one such structure. Often described as the brain’s relay center, it sits deep in the brain and transmits sensory and motor signals to the cortex.

“We are very interested in the function of a part of the brain that is important for sensing the environment (sights, sounds, touch, etc) but also engaged in producing sleep. The thalamus is located at the center of the brain, above the spinal cord, but below the cerebral cortex. It is a place that relays (sends along) signals from the outside world where it can be processed and understood in the cortex,” said study author John Huguenard, a professor of neurology at Stanford University School of Medicine.

“The thalamus is capable of behaving erratically, and does so in some forms of epilepsy, which we have studied extensively. Some of the treatments for this type of epilepsy work by turning down activity of the thalamus. Given that many autism patients experience epilepsy, we wondered whether the thalamus may play a role.”

Free daily newsletter

Surrounding part of the thalamus is the reticular thalamic nucleus, or RT, a shell-like structure made up of inhibitory neurons. It acts as a filter, regulating what sensory information is passed on to the cortex. The RT also helps orchestrate brain rhythms that regulate sleep, attention, and arousal. Disruption in this region has been associated with several psychiatric and neurological disorders, including schizophrenia, depression, and epilepsy.

The researchers were drawn to the RT partly because of its role in epilepsy. Individuals with autism are significantly more likely to have seizures than the general population, and the RT has long been known to generate seizure-like brain activity when overactive. This connection led the researchers to ask whether the same kind of hyperactivity in the RT might also contribute to the broader set of behaviors observed in autism.

To explore this idea, the researchers used mice lacking the gene Cntnap2, which is strongly linked to autism in humans. These Cntnap2 knockout mice display a range of autism-relevant features, such as social avoidance, repetitive grooming, heightened responses to sensory stimuli, and a tendency toward seizures. Using a variety of experimental techniques, the scientists recorded the activity of RT neurons in these animals and compared it to normal mice.

They found that RT neurons in the autism-model mice were hyperactive. These cells showed increased burst firing, a pattern of rapid activity known to amplify inhibitory signals in the brain. This heightened activity was linked to elevated oscillations in thalamocortical circuits, the loops that connect the thalamus and cortex. These rhythmic bursts have been associated with both seizures and abnormal sensory processing.

Fiber photometry, a technique for tracking real-time activity in live animals, showed that RT neurons in the Cntnap2 knockout mice fired more often in response to light, sound, and social interaction. Even when no stimulus was present, RT neurons in the autism-model mice were more active than those in control animals. This pattern suggested that the RT was not only more sensitive to input, but also more likely to produce spontaneous activity that could disrupt brain function.

“Most previous studies have focused on circuits within the the cerebral cortex, where higher order functions are carried out,” Huguenard told PsyPost. “These results suggest that the brain circuitry that is affected in autism might be a larger network involving both cortex and thalamus.”

Having identified this hyperactivity, the researchers next asked whether reducing RT activity could improve the animals’ behavior. They tested two approaches.

In the first, they used a drug called Z944, originally developed to treat seizures. This drug blocks T-type calcium channels, which are responsible for the burst firing of RT neurons. When the researchers gave the drug to the Cntnap2 knockout mice, it significantly reduced RT activity. The treated mice showed fewer repetitive behaviors, moved around less excessively, and spent more time engaging with other mice — behaviors more similar to typical mice.

In the second approach, the researchers used chemogenetics, a method that allows scientists to switch specific neurons on or off using designer drugs. By targeting RT neurons for inhibition, they were able to quiet the overactive circuit and again observed improvements in social behavior and reductions in hyperactivity and repetitive grooming. In contrast, when they artificially increased RT activity in typical mice, those animals began to display behaviors similar to the autism model mice, including social withdrawal and increased repetitive actions.

“We were surprised with degree to which the animals improved with treatment,” Huguenard said. “This suggests that key aspects of autism-related behavior may depend strongly on the thalamus.”

These results indicate that the RT is not just associated with autism-related behaviors but may actively drive them. Suppressing its activity reversed core features of the condition in mice, while increasing its activity was enough to induce them.

“The thalamus is an important part of the brain circuitry that can contribute to autism related behaviors, demonstrated so far in mice,” Huguenard explained. “Secondly, because some antiepileptic drugs target the thalamus, for example some calcium channel blockers, these provide a way to test the proposed role of the thalamus in autism behaviors. Our results suggest that this targeted approach can improve in animals autism-related behaviors.”

While the findings are promising, the researchers note several limitations. Most importantly, this study was conducted in mice, and it remains to be seen whether the same mechanisms apply in humans. The Cntnap2 gene represents just one of many genetic pathways involved in autism, and the findings may not generalize to all forms of the condition. The RT is also a small but complex structure, and the study did not examine whether subgroups of RT neurons play distinct roles in different behaviors.

In addition, while the seizure medication Z944 showed positive effects in this mouse model, it has not been approved for use in people with autism. Whether it could be safe and effective in humans, or whether more targeted versions of it could be developed, is a question for future studies.

“Before we get too much farther towards treatments, there are a few central questions we want to address,” Huguenard noted. “The main one is whether our findings apply more broadly to a range of ailments in ASD, or might they be specific to what we found in one genetic form of ASD, that of loss of function of Cntnap2 gene/protein. If the therapeutic effects of targeting the thalamus (with Z944 or other compounds) show improvements in a variety of forms/models of mouse ASD, then that would begin to support the idea of preclinical trials.”

The study, “Reticular thalamic hyperexcitability drives autism spectrum disorder behaviors in the Cntnap2 model of autism,” was authored by Sung-soo Jang, Fuga Takahashi, and John R. Huguenard.