Decoding Autism on Chromosome 16

Why do individuals carrying the same genetic variant manifest such variable clinical outcomes – especially in the realm of neurodevelopmental disorders, such as autism? More than two dozen regions in the genome have been identified that, when deleted or duplicated, lead to neurodevelopmental disorders. The chromosome 16 deletion is one of the most frequent causes of autism, accounting for about 1 percent of all affected individuals. It has also been strongly linked with other phenotypes including obesity, epilepsy, and intellectual disability. The 16p11.2 deletion encompasses about 25 genes, of which several have important roles in nervous system development. Although single causative genes have been identified for classical deletion syndromes, such as Smith-Magenis syndrome, in the context of the 16p11.2 deletion, we were convinced that there could not be a single gene causing all of the above phenotypes; there had to be multiple genes with common functionality. So we began testing the effect of reducing the expression of individual genes – and pairwise combinations of genes – on neurodevelopmental phenotypes in flies (1).

So far, we have only mapped a general landscape of what could be going on within 16p11.2 and other copy number variants (CNVs) with variable phenotypic expression. The next step is to dissect each of these interactions in more sophisticated systems and map them back to specific sets of symptoms. In general, mapping specific genes for structural defects within CNVs has not been overly difficult because of the straightforward nature of identifying these phenotypes. For example, the TBX6 gene accounts for the scoliosis phenotype observed a small subset of individuals with the 16p11.2 deletion. In contrast, identifying specific gene combinations with neuropsychiatric effects and correlating them with severity will be challenging because it will involve mapping genes and their interactions in combination with everything else in the genomes of individuals with the deletion. In the end, it all comes down to pathways and how genes “talk” to one another within networks.

We now plan to map gene interactions within CNVs and identify common pathways and their cellular mechanisms. Identifying the functional correlates might provide us some clues as to which genes, clusters, or specific pathways to target. Although we may develop some treatments for specific symptoms by repurposing drugs used to target similar cellular pathways, for others we might have to take what we find in flies to more sophisticated systems representative of human biology (such as mouse models, induced pluripotent stem cells, and organoids). Within the next five to 10 years, I would like to see deeper, quantitative phenotyping of clinical features in thousands of affected individuals with disease-associated CNVs, and more clarity of the molecular mechanisms underlying the patient phenotypes. I also hope that significant progress is made on identifying successful treatment strategies.

To get our discoveries into the clinic, we need to have constant interactions between clinicians, scientists, and affected families. While we map specific genes, interactions, and molecular pathways to specific clinical features, clinicians can use this information to identify individuals with subtypes of these disorders, which could inform prognosis. And with more detailed phenotypic and molecular profiles, it could help with customized treatment and management strategies.