Autism Spectrum Disorders: From Candidate Genes to Candidate Ontology Terms

Genetic heterogeneity for a multifactorial disease such as autism would imply that any two patients are unlikely to share the same susceptibility loci. Since different susceptibility loci may affect the same function and since functions may be considered at different levels of analysis, the following question arises: at what phenotypic levels is convergence attained by different genes in autism ? This is an important question to answer in order to shed light on subcellular, cellular or multicellular structures and functions possibly involved in the pathogenesis of autism. Among the various attempts made to answer this question it is worth mentioning the use of the best available genetic information to focus on the formation, structure and function of the glutamatergic synapse. How could we use this working scheme in a more efficient and hopefully fruitful way ? The recent advent of genomic technology offers an unprecedented opportunity to study the biological structure of autism deep into its molecular roots. For example, recent work on structural genomics has led to the identification of many Copy Number Variations in the gene-rich regions of the genome detected in a large number of patients. Interesting advances have also been made by analyzing the global gene-expression profiles in blood cells of autistic patients. The large amount of data generated by these studies is likely to contain precious information about the pathogenesis of autism. In this paper we review the literature on the genomic studies in ASD and present the results of a preliminary study showing how candidate biological processes for ASD can be identified from microarray data using the Gene Ontology database. We also discuss the importance of the transition from a strategy based on the search for candidate genes to the search for candidate ontology terms to expand our current understanding on the biological processes impaired in ASD