Towards identification of individual etiologies by resolving genomic and biological conundrums in patients with autism spectrum disorders

Mol Syndromol. 2013 Jun;4(5):213-26. doi: 10.1159/000350041. Epub 2013 Apr 4.


Recent genomic research into autism spectrum disorders (ASD) has revealed a remarkably complex genetic architecture. Large numbers of common variants, copy number variations and single nucleotide variants have been identified, yet each of them individually afforded only a small phenotypic impact. A polygenic model in which multiple genes interact either in an additive or a synergistic way appears the most plausible for the majority of ASD patients. Based on recently identified ASD candidate genes, transgenic mouse models for neuroligins/neurorexins and genes such as Cntnap2, Cntn5, Tsc1, Tsc2, Akt3, Cyfip1, Scn1a, En2, Slc6a4, and Bckdk have been generated and studied with respect to behavioral and neuroanatomical phenotypes and sensitivity to drug treatments. From these models, a few clues for potential pharmacologic intervention emerged. The Fmr1, Shank2 and Cntn5 knockout mice exhibited alterations of glutamate receptors, which may become a target for pharmacologic modulation. Some of the phenotypes of Mecp2 knockout mice can be ameliorated by administering IGF1. In the near future, comprehensive genotyping of individual patients and siblings combined with the novel insights generated from the transgenic animal studies may provide us with personalized treatment options. Eventually, autism may indeed turn out to be a phenotypically heterogeneous group of disorders ('autisms') caused by combinations of changes in multiple possible candidate genes, being different in each patient and requiring for each combination of mutations a distinct, individually tailored treatment.

Keywords: Autism; Copy number variation; Personalized medicine; Polygenic mode; Single nucleotide variation; Transgenic mouse models.