Autism spectrum disorders (ASD) are a heterogeneous group of disorders that are characterized by abnormalities in communication and social behaviors. ASD typically manifest in the first 3 years of life and have a strong heritability showing 2-6% concordance rates in siblings of affected individuals, ~10% in dizygotic twins and over 90% in monozygotic twins. Recent genome wide association studies (GWAS) utilizing high-density single nucleotide polymorphisms (SNPs) have revealed extensive genetic heterogeneity and structural rearrangements across the genomes of affected individuals. A vast majority of the copy number variants (CNVs) occur in the intergenic regions, thereby supporting the notion that most of the functional variation underlying ASD affects gene regulation rather than the protein-encoding sequences. Current findings support a model that common variants with a moderate effect together with rare and de novo mutations in genes involved in neuronal development and signaling underlie ASD susceptibility.
Objectives:
In order to identify these genes and their associated pathways we (I) performed pathway based analysis of genome-wide association signals for ASD subtypes; (II) prioritized ASD-associated loci using pathway-based computational approaches; (III) initiated resequencing of over 100 loci in ASD subjects from the Autism Genetics Research Exchange (AGRE) collection.
Methods:
In order to prioritize candidate ASD loci for resequencing we selected a list of 188 genes (46.55 MB) from the literature. Since a majority of CNVs associated with ASD have been found in the intergenic region, we included the entire flanking region (up to the next gene) for each of the genes in the list (total of 164 MB) constituting 188 genomic regions of interest (GROI). For each gene in the GROI we mapped conserved non-coding regions, small non-coding and alternatively spliced RNAs, and methylation marks in several cell types on CNV - ASD susceptibility regions. These data facilitated refinement of the list of genomic regions associated with ASD. A tiling array with short genomic regions (total 3.5MB) representing these ASD candidate loci will be used to capture DNA for resequencing using Illumina Genome analyzer.
Results:
We identified highly conserved sequences within these GROI and annotated potential enhancers and chromatin modification marks. For example, p300, a trancriptional coactivator is critical for embryonic development. We used p300 ChIP-Seq data and found an excess of forebrain (4%) and midbrain (3%) p300 binding sites compared to limb-bud enhancers (2%) in our GROI. Furthermore, we showed that 20 % of conserved sequences in non-protein coding regions, corresponded to RNA detected in a published RNA-Seq experiment.
Conclusions:
We are nearing completion of the custom array design which will be used to screen 100 affected ASD subjects from AGRE. The resulting sequence data will be comprehensively analyzed for non-synonymous SNPs in exonic regions and their effects on protein structure/function. Furthermore, we expect to identify additional causative variants in potential regulatory elements in non-protein coding regions, and fine map boundaries for previously detected and newly discovered CNVs.