Damaging stop gain/loss and frameshift mutations in autism subjects outline impairment in neuronal migration and adhesion pathways

  1. Swati Agarwala1,
  2. K.C. Shyamala2,
  3. Prakash Padakannaya3 ,
  4. Nallur B. Ramachandra1

Authors Affiliation(s)

  • 1Genetics and Genomics Lab, Department of Studies in Genetics and Genomics, University of Mysore, Manasagangotri, Mysuru-06, Karnataka, INDIA
  • 2Department of Speech Language Pathology, All India Institute of Speech and Language, Mysuru-06, Karnataka, INDIA
  • 3Department of Studies in Psychology, University of Mysore, Manasagangotri, Mysuru-06, Karnataka, INDIA

Can J Biotech, Volume 1, Special Issue, Page 94, DOI: https://doi.org/10.24870/cjb.2017-a81

Presenting author: swati@genetics.uni-mysore.ac.in

Abstract

Genetic heterogeneity makes it challenging to identify causal-genes responsible for autism pathogenesis. Till date, research studies report only a handful of high confidence genes for autism. There is a need to identify damaging genomic-variants, predisposing an individual towards autism manifestation. Of special interest is stop gain/loss mutations found in the exome. Such variants are prevalent, having an estimated number of 100-200 occurrences per human-genome. Stop-gains and frameshifts may lead to functional consequences. Based on stringent inclusion-exclusion criteria, the study recruited 150 autism subjects of Indian origin, of which 13 were used for WES. To understand the nature and possible consequences of these variants, we first analyzed their characteristics at the genome-level. Genome-wide analysis of more than 30000 variants provided statistical-significance to identify sequence-specific features for severity and to build a pathogenicity score. This sequence-based pathogenicity score was then applied to the analysis of variants in autism susceptibility. Several damaging stop gain/loss mutations encompassing autism genes CDH5, DDX23, CLDN5, and DPP3 were identified with protein truncations ranging from 20-70%. Loss of function mutations disrupted protein domains involved in various autism related pathways such as neuronal migration, synaptogenesis,and neuronal adhesion. Mutations were identified with previous evidences for neuronal migration and adhesion pathways in Drosophila sp, C. elegans and mice models. Homozygosity mapping analysis to identify risk-homozygous-haplotypes showed evidence of recessive polymorphisms in GIGYF1, SERPINE1, and EPHB6. Recessive alleles were identified across all the samples while polymorphisms in FOLH1, BCKDK, CDH11, and CTCF were specific. Mutations in language-specific genes, GCFC1 and MRPL19 were associated with autism phenome. A novel autism candidate gene CLDN5 that physically interacts with genes involved in various autism pathways was identified.CLDN5 belongs to the leukocyte-transendothelial-migration pathway and elevated in autism cortex, impairing the blood brain barrier leading to compensatory gene expression and protein accumulation. This on-going study identified several damaging mutations specific for autism in Indian population, adding to the growing body of mutational spectrum. Validations through Sanger sequencing and allele specific PCR is being done for the mutations identified.