Variable Expressivity of Neurodevelopmental Disturbances Due to Loss-of-Function of AP1S2

Background: Autism spectrum disorder (ASD) is a common neurodevelopmental disorder phenotypically and genetically heterogeneous. One of the most concurrent conditions in ASD individuals is intellectual disability (ID), which affects approximately 50% of the cases. Among the ASD familial cases, which represent less than 20% of the families, it has been recognized all Mendelian patterns of inheritance. The genetic factors associated with ASD familial cases are still unclear in most of them. Thus, it is essential to analyze family history and pedigree to figure out the inheritance model that best fit to each family, and, consecutively, conduct genetic analysis in order to identify the major pathogenic event leading to the phenotype.

Objectives: To investigate genetic variants that would explain the ASD and ID segregating in 4 generations in a family with an X-linked inheritance pattern.

Methods: We selected two distant related cousins (individuals IV-2 and IV-4 in Figure 1), which belong to a family (Family F8293) with a total of 3 autistic individuals (individuals III-7, IV-2 and IV-4 in Figure 1) and 4 ID affected individuals (individuals II-4, II-5, III-6 and III-8 in Figure 1), to perform whole-exome sequencing (WES) using the Nextera Rapid Capture Exome kit (Illumina) for library preparation and sequencing on a HighSeq 2500 (Illumina). The alignment, processing, variant calling and annotation were carried out on BWA, GATK and ANNOVAR. To filter possible pathogenic variants we used a MAF<0.01, adopting as references 1000g, 6500 exomes and Brazilian control 60+ databases. Sanger sequencing was used to investigate if the selected candidate variants were shared among the other members of the family.

Results: We first search for shared variants between both cousins, but we did not detect any X-linked or autosomal variant that could explain the phenotype. Next, we searched for loss of function variants (LoFs) in brain expressed genes located in the chromosome X that were exclusive of each affected individual. In individual IV-4 we identified a stop codon variant in AP1S2. We verified that the same variant was shared among all other affected individuals, except for individual IV-2. AP1S2 is already associated with syndromic mental retardation, but it had not been linked to ASD yet. No relevant LoF variant in the X-chromosome or in autosomes of IV-2 was identified.

Conclusions: Our findings show the complexity of the analysis of each ASD family and highlight the relevance of another ID-associated gene to ASD.