“It's the right disorder and the right technology kind of coming together,” says State. Not to mention the fact that researchers have increasing access to samples from children with autism and their family members, which is also helping to speed research (See sidebar: The importance of biological samples).
These exome sequencing studies are a natural extension of the earlier work on copy number variations according to State, with researchers now looking across the coding genome with single-base resolution for loss-of function mutations. “We would not have had the capacity to do that previously.”A path forward with a few hints
The critical finding emerging from these autism exome studies has less to do with specific genes and more to do with validating this particular approach as a tool for understanding the genetic underpinnings of these disorders. “What our data and all of the papers showed was that if you move to a higher level of resolution, essentially you're seeing about the same picture, which is that there's an overrepresentation of these highly deleterious mutations,” acknowledges State.
Across the four studies, researchers observe de novo mutations in the genomes of children with autism and, where tested, their unaffected siblings. However, children with autism were more likely to have mutations that disrupted protein function—mutations that halted gene transcription early or splice site mutations.
Three of the four studies looked at samples from the Simons Simplex Collection, a collection of genetic samples from 2700 families with a single child with an autism spectrum disorder and unaffected parents and an unaffected sibling. Even when grouping all of the Simons Simplex Collection (SSC) families together whose genes were sequenced in these three studies, they total less than 800 out of 2700 families (Neale and his colleagues performed exome sequencing on samples from 175 individuals with autism and their parents outside the SSC). With only a handful of genes, did researchers find disruptive de novo mutations in the same gene in different individuals with autism? “We almost never saw lightning strike twice,” says Evan Eichler of the University of Washington, the principal investigator on the third Nature paper. If a proportional number of recurrent genes pop up in the rest of the SSC, exome sequencing of this collection might point to a total of 25 to 30 new genetic mutations that confer autism risk, says Geschwind.
It might be too early to make predictions about specific genes and what they mean for the biology autism. “I think most people who are cautious in this field were thrilled by those papers, but they really want to see what the whole [SSC] is going to look like,” says John Spiro, deputy scientific director for the Simons Foundation Autism Research Initiative in New York, NY.
Still, some of the findings offer hints that may develop into wider patterns. In their Nature study, State and his colleagues pulled out one gene where disruptive de novo mutations occurred in two autism probands (Neale and his colleagues also pulled out this gene while their Nature paper was under review). The gene, SCN2A, encodes a sodium channel, that is highly expressed in the brain. In addition, SCN2A mutations have been connected with epilepsy, and a variety of genetic data has pointed to overlaps between genetic risk for epilepsy and the risk for autism spectrum disorders.
Two groups, one led by Eichler at the University of Washington and the other by Neale, found recurrent mutations in CHD8, a gene encoding a chromodomain helicase which could modify transcription of a whole host of other genes. Eichler's group also found a recurrent disruptive mutation in netrin G1 (NTNG1), a gene important in axonal guidance, while Neale and his colleagues found recurrent mutations in KATNAL2, a gene whose function is not well characterized.
Researchers also see potential connections between the different observed de novo mutations. Forty percent of the mutations Eichler's team observed were in genes making up a highly interconnected pathway surrounding CHD8, and Ronemus and his colleagues used mouse studies to identify a number of mutated genes connected to Fragile X mental retardation protein (FMRP), an important regulatory of Fragile X syndrome suggesting the pathway involved in Fragile X syndrome could be central in a wider number of autism cases.