There’s been a lot of study of copy number variants (CNV) in autism over the last decade or so. Far from beating a dead horse, a recent publication in Neuron by a large group of researchers reinvestigates CNV in the Simons Simplex Collection, reporting on newly recognized trends in mutation penetrance according to CNV size. As a disclaimer, simplex cases are typically representative of non-inherited forms of autism, and therefore mutation trends may be and likely are different in multiplex and multigenerational patterns of inheritance. However, this recent study may pinpoint some important general trends that could lend to the study of other patterns of inheritance.
Once again, as with previous studies, Sanders et al. (2015) found that increased occurrence of CNVs of all sizes were associated with autism in contrast to unaffected siblings. Although there didn’t appear to be differences in specific mutations and risk loci between males and females, it was estimated that a higher percentage of de novo (new) CNV and smaller loss-of-function (LoF) mutations occur in females (~17%) than in males (~10%). In addition, across the entire simplex collection, the investigators estimate that there are approximately 200 de novo CNV loci that confer notable autism risk.
However, probably one of the most interesting findings of the study was in showing how the level of autism risk differs according to the type of mutation. The researchers looked at large de novo CNVs that covered multiple genes, smaller de novo CNVs that comprised only the deletion, duplication, or translocation of a single gene, and other small LoF mutations such as missense and silent point mutations. They found that the highest risk for autism associated with smaller single-gene mutations and that there was considerable agreement between single-gene CNVs and missense point mutations. Meanwhile, larger CNVs that comprised more than a single gene tended instead to include several risk genes, which individually each confered less risk for autism.
I find these results particularly promising for what it suggests regarding potential mutational trends in multiplex autism. Rare mutations occur more frequently in simplex cases. In contrast, this suggests that multiplex cases may be genetically more complex, involving multiple lesser-risk loci. Given the results in this study, it’s possible that the larger CNV loci identified here may be prime areas to search for common variants that predispose towards inherited autism risk.
A good idea for a future multiplex study, no?
Correlation does not equal causation 🙂
Why presume that any of these CNVs play a causative, or even a ‘hightened risk’, role? The assaults on the genome and activation of transposable elements could simply be reflections of, and happening in parallel to other (pathogenic) goings on.
True. Although in my more recent work (which isn’t yet published) I’m finding that transposable element content and other structural features of autism-risk genes tend to be indicative of specific gene functions (e.g., developmental regulation), which would subsequently suggest that the functions of these genes are actually important in autism risk. Hoping to get that paper out in a few months. 🙂
The study identified six loci as ASD risk genes 7q11.23, 15q11q9-13 and 22q11.2. Where do these de novo mutations come from?The research community needs to focus on genetic mutations that arise before conception. Molina et al examined sperm mutations of healthy volunteers focusing on three of the identified ASD risk gene loci identified in the study. They reported sperm mutations for all three of the mutations they examined 7q11.23, 15q11q9-13 and 22q11.2, in every healthy volunteer, these are the same three (out of six) risk loci reported in the Sanders et al study.