Could a Simple Eye Reflex Test Assess Autism Spectrum Disorder?

A simple eye-tracking device might help in the diagnosis of autism spectrum disorder (ASD), a developmental disability that impacts how individuals interact and communicate with the world.

Researchers from the University of San Francisco (UCSF) found that children carrying a variant of the SCN2A gene, which is associated with severe ASD, had a hypersensitive vestibulo-ocular reflex (VOR).

Monitoring this reflex using a camera mounted on a helmet, the researchers could distinguish SCN2A-ASD children from their neurotypical siblings.

The study, led by Dr. Kevin Bender, associate professor in residence in the Department of Neurology, is published in Neuron.

The SCN2A gene and autism

SCN2A encodes a voltage-gated sodium channel – NaV1.2 – that is distributed widely across different brain regions and is involved in the generation of action potentials.

“SCN2A is now recognized as the single gene most likely to be affected in a child with ASD, accounting for somewhere between 0.3% and 1% of all cases,” said Bender. “Those two statements sound at odds – how can a gene be both the most likely and yet only account for such a small fraction?”

That’s because ASD is complex and can be driven by a wide variety of genes, in addition to environmental risk factors. “So even though it’s a small fraction, dysfunction in SCN2A is still a major risk factor. In fact, it’s considered the leading ‘single gene’ risk factor (i.e., a single gene is affected, instead of a microdeletion of a strip of DNA affecting multiple genes or something like fragile X),” Bender emphasized.

In an effort to understand the association between SCN2A variants and ASD, Bender and colleagues had previously studied SCN2A’s role in brain regions such as the hippocampus, the striatus and the cortex, areas associated with functions such as communication and social skills.

“While these regions are important for some aspects of behavior, there has generally been an issue in identifying robust behavioral differences between neurotypical mice and those that are heterozygous for SCN2A (matching the human condition),” he explained.

The researchers, including Dr. Chenyu Wang, a graduate student in Bender’s lab and the first author of the study, have turned their attention to the cerebellum. Often nicknamed “the little brain”, the cerebellum is involved in processes related to movement and balance control. Nestled beneath the cerebral hemispheres, it almost looks as though it is a separate part of the brain.

“SCN2A expression is quite high in the cerebellum, and behaviors known to be controlled by the cerebellum have been well characterized,” said Bender.

The behaviors described by Bender include the VOR, which is modulated by cerebellar circuits.

“VOR is a reflex where head movements (even in the dark) drive compensatory eye movements in the opposite direction to the head,” said Bender. “This is done solely with the sensory input provided from the semicircular canals that make up your vestibular sensory apparatus. The cerebellum is important for adjusting VOR gain to levels that make sense to stabilize visual scenes. This is done with a little bit of feedback/error correction via visual input, and the associated eye reflex, called optokinetic reflex (OKR), which relies not on vestibular input, but on movement in a visual scene. Both work together to try to stabilize images on the retina.”

Dr. Guy Bouvier, a cerebellum expert and former postdoc at UCSF, and the study’s co-senior author, had the equipment required to test the VOR. First, the researchers focused on mice.

Mice carrying the SCN2A variant had an “unusually sensitive” VOR. When moved in one direction, their eyes perfectly compensated by moving in the opposite direction. However, the neural circuits within the cerebellum, which usually refine the reflex, got stuck, meaning the mice could not focus on a moving object while their head was moving.

“After seeing some initial results in mouse models, we considered whether we could extend our study to children,” said Bender. Asking questions to children with ASD-associated alterations in SCN2A function is not an easy task; quite often they have little-to-no verbal repertoire and intellectual disabilities. “Asking them to perform any behavior requires a bit of thought and patience,” said Bender.

The VOR is advantageous in this context, though. It’s a reflex that happens without training or command. “This type of reflexive behavior is therefore ideal for testing in this population, and we wanted to know if it could be useful as an indicator of alterations in the brain of kiddos that we know have SCN2A dysfunction. If that were true, maybe we could use the same test in the future in a diagnostic way.”

Children with SCN2A-ASD have a hypersensitive VOR

Families from the FamilieSCN2A Foundation, a major advocacy group for children with SCN2A variants in the US, were recruited to the study. A total of 5 children with SCN2A-ASD and 11 of their siblings that did not have ASD participated.

Bender and colleagues mounted an eye-tracking camera onto a helmet that was placed on the children’s heads. They then rotated the children left and right to the beat of a metronome. In children with SCN2A-ASD, the VOR was hypersensitive, just like the SCN2A mice. Bender and colleagues could distinguish the children from their neurotypical siblings by simply measuring how much their eyes moved in response to the head rotation.

“The amplitude of the eye movements, relative to the head movement, is called VOR gain. Normally, this gain is at or below 1.0 [e.g., abs(degrees of eye movement) / abs(degrees of head movement)]. At very slow head movement speeds, VOR gain is typically quite low, with values from 0.4 to 0.7.  What we found in SCN2A mice and humans was that, even at these low movement speeds, VOR gain was closer to 1.0,” Bender explained.

“This is, in many ways, a superpower that the SCN2A mice and SCN2A kiddos have. Going in, I didn’t even think that the vestibular system was capable of driving VOR at 1.0 at these low head movement speeds.”

Continue reading below...

Listicle

Applications of CRISPR Technology

Read more

CRISPR technology restores reflex back to normal

Using an approach called CRISPR activator (CRISPRa), developed by the Ahituv Lab at UCSF, the researchers could test what happens when the SCN2A gene is upregulated in mice.

“The CRISPRa approach has a number of therapeutic advantages, including the ability to work for a number of variants found in children, and working with the endogenous gene regulatory machinery in a way that upregulates SCN2A – and only SCN2A – in places where SCN2A is transcribed. With this, we were able to ask if we could restore VOR back to normal,” said Bender. “We found that we could, provided we intervened early enough.”

When 30-day-old SCN2A mice were treated, their VOR was less rigid but still sensitive. When 3-day-old mice were treated, however, the VOR was restored to normal. “This would suggest that VOR is a good behavioral readout for this type of therapeutic.

Translating new findings on SCN2A and the VOR to the clinic

It’s important to recognize that alterations in VOR likely have nothing to do with the core symptoms of ASD, Bender emphasized: “Rather that these changes may occur in parallel with other alterations in the brain.”

The future goal for the researchers is to translate this approach to the clinic. There are a number of hurdles to overcome before than can happen. Firstly, the study needs to be replicated in a larger cohort, as the trials comparing mice to humans in this study were small in size. Nonetheless, Bender is hopeful. “One of the big questions for us is to expand this work to other ASD-associated genes,” he said.

“We’ve seen this relationship between cerebellum-dependent reflexive eye movements and SCN2A loss – is this a relationship that exists only between SCN2A and VOR, or would similar relationships occur if we examined other ASD-associated genes? If that were true, this might be extended to be a useful metric that could help a wider range of people,” Bender concluded.

Dr. Kevin Bender was speaking to Molly Campbell, Senior Science Writer for Technology Networks.

About the interviewee

Dr. Kevin is an Associate Professor in Residence in the Department of Neurology, and holds an Endowed Chair in Honor of the Gallo Family. He received his PhD from UC San Diego, where he worked with Dan Feldman (now at UC Berkeley). His thesis focused on understanding the cellular mechanisms of cortical map plasticity, which is how circuits rewire to represent altered sensory input. Following graduate training, Kevin moved to Portland, Oregon for postdoctoral training with Larry Trussell at the Oregon Health and Science University and Vollum Institute. There, Kevin identified new cellular mechanisms involved in initiating and modulating neuronal output. At UCSF, Kevin is involved heavily in the Neuroscience Graduate Program, directing NS201A and recently finishing a seven-year stint as co-chair of admissions, among other duties. Outside the lab, he enjoys hiking and mountain biking around the Bay Area.

Reference: Wang C, Derderian KD, Hamada E, et al. Impaired cerebellar plasticity hypersensitizes sensory reflexes in SCN2A-associated ASD. Neuron. 2024. doi: 10.1016/j.neuron.2024.01.029