The Molecular Roots of Autism: New Directions and New Methods

Several years ago, a young child and his parents came to the Duke Center for Autism and Brain Development asking for help. The son had been diagnosed with autism, and the family was requesting a comprehensive assessment of his condition.  

As part of the assessment, doctors at the Duke Autism Clinic requested a genetic evaluation for the child. Specifically, they requested whole genome sequencing, which identified a rare mutation in a single gene that could help explain their son’s specific type of autism. The family then decided to make a generous philanthropic gift to help Duke scientists take what they found in those genetic tests and move the results into the lab for further study.

Scott Soderling, PhD, the George Barth Geller Distinguished Professor for Research in Molecular Biology at Duke University, leads the team that took on this project. The Soderling Lab conducts research involving gene editing, machine learning, and proteomics (the study of how cell proteins interact and function), with the goal of uncovering the biological bases of autism.

Last year, Soderling’s team reported that they had successfully modeled the child’s genetic mutation in mice, confirming a small alteration in the protein SCN2A as one potential cause of his form of autism.

This year, the team has dived deeper into autism’s molecular roots. With continued philanthropic support from the family, researchers developed a new method—combining CRISPR genome engineering with proteomics—to discover how the SCN2A protein works in neurons. Using this new approach, they sought to determine how one amino acid change within the SCN2A protein alters how it interacts with other proteins to do its job.

To carry out these investigations, Soderling’s team first extracted neurons from their mouse model and developed a cell-based assay that allowed them to record the firing activity of SCN2A neurons in 96 wells. Doing so showed them that one particular amino acid change (each protein has several thousand amino acids) changes how that SCN2A protein interacts with others in the cell to do its job: that one alteration effectively reduced the firing rate of neurons by approximately half. “That’s a pretty dramatic change,” Soderling says.

The team then tested a hypothesis. Could dysregulation in three specific proteins be driving the reduced firing rate of the mutated neurons? Sure enough, when Soderling and his team genetically boosted the level of those three underperforming proteins, they found that they were able to fire at the usual faster rate. “It completely rescued the cellular phenotype,” Soderling says.

The implications of this preliminary study are twofold. One, it provides proof of concept that philanthropy can drive research in game-changing ways that traditional grants might be slower to accomplish. “With funding from our donor, we were able to do a high-risk project and develop a new method that has wide-ranging applications,” says Soderling.

Two, that donor’s generosity has opened a potential path forward for restoring neurons to their full firing potential. Already, the team is working to bring the results they found in the assay back into animal models. They’re also exploring how artificial intelligence (AI) can help design new proteins to target the alteration they found.

Soderling is excited about the prospects for future research. The pipeline his team developed for mapping biology to neurological condition can be applied to genetic alterations beyond the specific autism experienced by this one family, he says. The new method could eventually inform neuroscientific understandings at the broadest level.

Even more than the scientific prospects, however, he’s grateful to the family for putting a human face on the research effort.

“Our team had the family over for lunch recently,” he says. “We met the son. Most of our scientists spend all day in the lab, working with mice. To be able to meet the person who inspired this study was incredibly meaningful for all of us.”

Moving forward, the Soderling lab has hired computational scientists for new AI endeavors and launched collaborations with other experts in the field. The family’s donation also helped facilitate the career of a young scientist who will dedicate his research to understanding autism and the brain: Yudong Gao, PhD, the Duke postdoctoral fellow who helped conduct the preliminary study, has now launched his own research lab at Baylor University. It’s all a testament to the family’s dedicated engagement.

“The NIH often doesn’t fund us to develop new methods to ask a question,” Soderling says. “That's a risky investment. Your hypothesis might be wrong, and you might not get the method to work. That’s why philanthropy is so important.”  

With the family’s ongoing involvement, the future of autism research is looking brighter than ever. “We’re so grateful for their willingness to pursue this research,” Soderling says. “It’s been a transformational partnership.”