By: David Shifrin, PhD
Science Writer, Filament Life Science Communications

A couple of months ago, this blog published a post on two recent papers dealing with new advances in research on the genetics of autism spectrum disorders (ASD). In one, the authors were able to identify autism with high accuracy using gene expression levels. In the second, a group in the UK investigated the heritability of ASD.

Now, a multi-institutional group, led by researchers at the University of Washington in Seattle, has added another exciting set of data to the field of autism genetics.

The study is titled “Excess of rare, inherited truncating mutations in autism,” and was published in early May in Nature Genetics. As with any good title for a scientific manuscript, this one provides a pretty good summary of the paper. Even so, there are interesting aspects to the research that are worth spending some time on.

A growing number of mutations have been associated with ASD over the past several years. These include both de novo and inherited mutations, and cover virtually every type of genetic lesion. In general, de novo mutations seem to cause more severe forms of ASD than do inherited mutations. In this paper, first author Niklas Krumm went after both de novo and inherited changes, with the intent of “assess[ing] the relative impact” of both.

The premise of this study is rooted in the fact that while de novo mutations are significant drivers of ASD, and “have provided a rich source for understanding pathogenic genes and neurobiological mechanisms of ASD,” they account for only a minority of genetically-caused cases of ASD. Thus, state the authors, “This suggests that other genetic factors contribute to ASD, including both rare and common inherited genetic variation.”

Additional questions arise from the massive gender differences in autism cases. As many as 80% of diagnoses are in males, although new research is beginning to suggest that the ratio may be significantly lower than 4:1. (Indeed a group in Europe called Autism in Pink is working to bring cases of autism in females to light, and to support women with ASD.) The discrepancy has led some investigators to suggest that there are genetic ASD risk factors that are only penetrant in or transmitted to males. Krumm and colleagues therefore wanted to test whether copy number variants (CNV) or single-nucleotide variants (SNV), which may be differentially transmitted in males and females, actually are.

To pull all this together, the authors looked at almost 9,000 exomes that had been previously sequenced. Over 1.3 million transmitted variants fell out of this analysis. Through a series of subsequent analyses of a limited subset of SNVs (which itself was pulled from a predicted set of ~1,500 variants), a total of 21 de novo mutated genes were confirmed. Importantly, these were genes that were mutated in multiple individuals.

Again working with SNVs, the group next looked at potential transmission disequilibrium between affected individuals and their siblings. Interestingly, they found no difference “in the overall burden within proband-sibling pairs.” However, they reanalyzed the data working with the assumption that mutations with a high probability of causing problems in a gene (likely gene-disrupting, or LGD, variants) would cause more problems when they occurred in genes that were “intolerant to deleterious mutations.” Which only seems logical. This reassessment revealed a significant difference in the transmission of LGD mutations in autistic children. The more intolerant a gene was to mutation, the greater the damage, particularly when mutations accumulated. When the inherited burden was higher, the odds ratio of having a more severe form of ASD also increased.

The authors also looked at CNVs. Microarray and hybridization techniques revealed that, again not surprisingly, there was a dramatically higher number of de novo CNVs in affected individuals than their non-autistic siblings (2.4x higher). Additionally, the CNVs that occurred in children with ASD were on average larger and in genes with lower tolerance to mutation.

By using the pooled results from CNV and SNV analysis, Krumm et al discovered several new candidate autism genes. This is where the power of next generation and large-scale bioinformatics can be used to drill down and find specific issues. Indeed, the top three hits were all genes with “brain-specific expression or […] identified neural functions,” strongly supporting a putative role in the development of ASD.

Finally, the group looked at the risk for ASD conferred by different types of mutations. Four categories of mutations were identified: de novo CNVs; de novo SNVs; rare, inherited CNVs; and private likely gene disrupting CNVs. Each of the four increased risk for ASD, with de novo CNVs more than doubling the risk. Additionally, LGD SNVs inherited maternally by a male with ASD made the highest contribution to the disease. This finding was consistent with previous work showing the same thing for CNVs. As the authors put it, this result reveals “bias in transmission from mothers to their sons.”

In the end, the authors conclude, “our results provide some of the first genetic evidence that private, inherited SNVs that truncate proteins are enriched in autism probands. Remarkably, this effect is only observed for truncating SNVs that disrupt genes intolerant to functional variation […]”

The field has a long way to go to fully understand the causes of autism and the genetic and molecular pathways involved. Each newly discovered set of mutations, along with their origin and potential patterns of heritability, edges us closer to a comprehensive picture of the disease(s), uncovering potential therapeutic targets and providing just a bit more hope that the complex issue of ASD can be solved.


Wed. June 10, 2015