18807
Integrating Analytical Methods to Identify Rare Variants Associated with ASD in High-Risk Utah Pedigrees
Objectives: Integrate genome-wide Shared Genomic Segments (SGS) analysis with the pedigree extension of the Variant Annotation, Analysis, and Search Tool (pVAAST) to increase power to detect rare genetic variation contributing to ASD risk in high-risk ASD pedigrees.
Methods: High-risk pedigrees were identified through the Utah Population Database. The Familial Standardized Incidence Ratio (FSIR) was used to find pedigrees with significantly more cases than expected compared to similar background pedigrees matched for age, sex, and size. DNA was collected through blood draws on affected and unaffected family members. A total of 196 samples were genotyped (Illumina OmniExpress12v1.1) on four of the largest high-risk pedigrees, including 44 affected relatives, 4 relatives with broad autism phenotype, and 146 unaffected relatives (siblings, parents, grandparents). Genotyping data was used to identify genomic sharing using SGS analysis. A subset of 88 samples from these pedigrees underwent exome sequencing (Agilent SureSelect and Illumina GAIIx). Variants in significant SGS regions were functionally annotated and analyzed with pVAAST to identify the likelihood of the variants contributing to disease risk. Pedigree variants were compared to background exomes sequenced through the 1000 genomes project, and were selected to match ethnicities with the pedigrees.
Results: In pedigree 25002 (ASD FSIR=5.56, p<0.0001), we identified a 12Mb region on chromosome 15 that was significantly shared between family members diagnosed with autistic disorder, compared to unaffected family members (wpSGS p<0.001). Subsequent pVAAST analysis of exomic variants in this region identified 4 genes most likely to contribute to risk for autistic disorder. The variants in these genes were shared between multiple affected family members, and include rs35697691 in mitogen-activated protein kinase 6 (MAPK6, p=0.048), two variants, rs1867380 and rs142159680 in aquaporin 9 (AQP9, p=0.0087), rs55686434 in myosin 5C (MYO5C, p=0.037), and rs16976466 in protogenin (PRTG, p=0.041).
Conclusions: The complexity and heterogeneity of autism spectrum disorder require innovative strategies for identifying genetic risk factors. Here, we combined two pedigree based analyses, SGS and pVAAST. First, SGS identified a significantly shared region of interest at 15q21-15q22. Then, pVAAST interrogated rare variants from sequenced exomes to identify 4 genes with evidence for ASD risk. The 4 genes, MAPK6, AQP9, MYO5C, and PRTG are expressed in brain. Of particular interest are PRTG and MAPK6. PRTG has been previously linked to attention-deficit hyperactivity disorder and reading disabilities. Variants in MAPK6 has been previously associated with autism, as well as other neurological disorders like bipolar disorder and epilepsy. This analysis strategy has been effective at identifying regions and genes involved in genetic risk for ASD, and increase understanding of the genetic etiology of ASD.