Postdoctoral Fellowship Final Report, SFRH/BPD/74739/2010

Autism Spectrum Disorder (ASD) is a common complex behavioural disorder with significant clinical heterogeneity and unclear etiology. Assessment protocols and diagnostic instruments are complex to implement, and not widely used by most clinical practitioners or pediatricians worldwide. It is well known that educational interventions, language and behavioural therapies may significantly improve the patient’s quality of life, and are particularly beneficial when initiated at a young age. While the contribution of genetic factors for ASD etiology has been clearly established by family and twin studies, common risk genes accounting for a high proportion of autism heritability have not yet been identified. At the same time, the increased burden of rare variants in ASD is increasingly recognized from genomic screening in large population samples. The latest genome wide association studies (GWAS) have failed to find common risk variants with a significant impact, however all have shown an excess of rare and heterogeneous structural variants designated Copy Number Variants (CNV). The significance of most of these rare CNVs for etiological diagnosis of ASD still needs to be properly addressed. The importance of rare variation in autism has also been supported by the many reports of rare mutations in synaptic candidate genes (such as SHANK, MET, NLGN, NRNX amongst others). It seems likely that both common low risk genes and rare, high penetrance, variants will converge in a restricted number of biological pathways. Our group is part of the international consortium Autism Genome Project (AGP), which gathers more than 100 scientists in 13 investigation centers. Recently it has been conducting a Genome Wide Association Study (GWAS), with more than 3000 families of autistic patients, including 342 Portuguese families. In this GW analysis, almost 40.000 CNVs were identified in patients and parents. The functional consequences and relevance of these CNVs have been analyzed, namely for ANXA1, PARK2 and DSC3 genes, and novel mutations in candidate genes identified by a PPI-based approach are being validated and targeted exome sequencing is being performed. Also, the correlation between genetic and clinical data, using bioinformatic-based data mining approaches, in a large scale sample is very important for a better characterization of the pathology. Our sample has the major advantage of a detailed clinical evaluation following the Autism Speaks project funded The Autism Trio Collection (TASC) project protocol. Finally, the fact that common risk genes for ASD have not yet been identified, indicates that different hypothesis should be addressed, mainly the possibility that epigenetic factors, such deregulated microRNAs (miRNAs) expression. We have been characterizing the expression profile of circulating miRNAs in plasma samples, as putative novel biomarkers for ASD.Fellowship SFRH/BPD/74739/2010 - Fundação para a Ciência e Tecnologia