West LA Times

A new study led by UCLA researchers has unveiled the most detailed view of the complex biological mechanisms underlying autism, showing the first link between genetic risk for the disorder and observed cellular and genetic activity across different layers of the brain.

Their study is part of a second package of studies from the National Institutes of Health consortium PsychENCODE. Launched in 2015, the initiative, chaired by UCLA neurogeneticist Dr. Daniel Geschwind, is working to create maps of gene regulation across different regions of the brain and different stages of brain development. The consortium aims to bridge the gap between studies on the genetic risk for various psychiatric disorders and potential causal mechanisms at the molecular level.

“This collection of manuscripts from PsychENCODE, both individually and as a package, provides an unprecedented resource for understanding the relationship of disease risk to genetic mechanisms in the brain,” Geschwind said.

Geschwind’s study on autism, one of nine published in the May 24 issue of the journal Science, builds on decades of his group’s research profiling genes that increase susceptibility to autism spectrum disorder and defining convergent molecular changes observed in brains with autism. However, what drives these molecular changes and how they relate to genetic susceptibility at cellular and circuit levels are not well understood.

Gene profiling for autism spectrum disorder has long been limited to using bulk tissue from brains after death. These tissue studies are unable to provide detailed information such as differences in brain layer, circuit level, cell type–specific pathways associated with autism or mechanisms for gene regulation.

To address this limitation, Geschwind used advances in single-cell assays — a technique that makes it possible to extract and identify genetic information in individual cell nuclei. This technique allows researchers to navigate the brain’s complex network of different cell types.

More than 800,000 nuclei were isolated from post-mortem brain tissue of 66 individuals ranging in age from 2 to 60 years old. This included 33 individuals with autism spectrum disorder (ASD) and 30 neurotypical individuals who acted as controls. The individuals with ASD included five with a defined genetic form called 15q duplication syndrome. Each sample was matched by age, sex, and cause of death.

Through this method, Geschwind's team identified major cortical cell types affected in ASD including neurons and their support cells known as glial cells. The study found profound changes particularly in neurons connecting two hemispheres providing long-range connectivity between different brain regions and interneurons called somatostatin interneurons important for maturation and refinement of brain circuits.

A critical aspect was identifying specific transcription factor networks — interactions where proteins control when a gene is expressed or inhibited — driving observed changes. Remarkably these drivers were enriched in known high-confidence ASD risk genes influencing large differential expression changes across specific cell subtypes. This marks the first time a potential mechanism connects changes occurring in ASD brains directly to underlying genetic causes.

Identifying these complex molecular mechanisms could help develop new therapeutics for treating autism and other psychiatric disorders studied under PsychENCODE initiative.

“These findings provide a robust framework for understanding molecular changes occurring in brains with ASD — which cell types they occur in and how they relate to brain circuits,” Geschwind said. “They suggest that observed changes are downstream effects of known genetic causes providing insight into potential causal mechanisms.”

The PsychENCODE papers are presented as a collection on Science website.

The studies received grant funding from National Institutes of Health.

___