Efforts to find genetic causes of autism have identified hundreds of rare mutations in individuals with the disorder, and this list is anticipated to grow steadily in the next few years. A pressing question is which of the mutations are responsible for conferring a disease risk. A small number of the mutations appear likely to disrupt the function of the affected genes, and individuals with autism have a higher burden of these mutations, suggesting a causative link to the disorder. However, the majority of mutations change only a single amino acid of the protein product or are ‘silent’ according to the genetic code. These mutations occur at similar frequencies in individuals with autism and unaffected siblings, implying that most of them are probably benign.

Chaolin Zhang and his colleagues at Columbia University and Cold Spring Harbor Laboratory in New York believe that some of these mild or silent mutations may be more deleterious than one would expect. The team aims to develop a new strategy to distinguish deleterious mutations from a much larger pool of benign mutations. Instead of evaluating the impact of each mutation based on the genetic code, Zhang and his colleagues plan to identify mutations that can potentially alter interaction of RNA with a group of RNA-binding proteins. Such alterations would affect regulation of the process generating mature RNA transcripts ready for translation, and thus result in more dramatic changes in the protein products.

Taking this ‘regulation-centric’ approach may help depict a more complete picture of the complex genetic landscape underlying autism. Besides mutations in protein-coding regions of the genome, which are a major focus of many current studies, the team anticipates that this approach can be extended to interpret additional mutations located in the noncoding regions representing 99 percent of the human genome, and to identify their links to autism.