O presente ensaio teve como objetivo principal contribuir com a caracterizacao da arquitetura genetica do TEA a partir de uma analise de uma casuistica brasileira, que e ainda pouco estudada. Para atingirmos este objetivo, verificou-se qual a proporcao de casos de TEA causados por variantes de novo em genes de neurodesenvolvimento (genes do banco SFARI e associados ao neurodesenvolvimento descritos no banco DECIPHER). Foram avaliados 63 trios, compostos pelos genitores e probandos com diagnostico de TEA atendidos no Centro de Estudos do Genoma Humano e Celulas-Tronco (CEGH-CEL, USP). A genealogia, os dados clinicos, sexo, idade na consulta, idade parental foram coletados. O sequenciamento de exoma completo foi realizado por meio de uma colaboracao com Mount Sinai, NY, US (colaboracao com o Autism Sequencing Consortium-ACS). A identificacao de variantes De novo em genes candidatos para TEA foi realizada pelo programa LOVD (LOVD v.3.0 - Leiden Open Variation Database). Observou-se que a maior parte dos probandos eram meninos (N= 55, 86%) e a minoria apresentava antecedentes familiares de TEA (N=4, 6%). Tambem verificou-se que 40% (N=25) dos individuos apresentavam atraso no desenvolvimento da linguagem e uma pequena porcentagem apresentava comorbidades como TDAH e epilepsia (N=6, 10% e N= 2, 3%; respectivamente). A idade parental media no momento da gestacao foi proxima de 30 anos para ambos os genitores (29,7 e 32,5, da mae e do pai respectivamente). Foram identificadas nove variantes de novo patogenicas ou potencialmente patogenicas em genes candidatos: para o TEA: seis em genes SFARI (quatro variantes patogenicas nos genes NF1, TLK2, DNAH17, BRSK2 e duas variantes provavelmente patogenicas em ARHGAP5 e HUWE1) e tres em genes do neurodesenvolvimento do DECIPHER (Deciphering Developmental Disorders Study, 2015) (duas variantes patogenicas nos genes ERLIN2, ST3GAL3 e uma provavelmente patogenica no COL11A1). Ao realizar a analise de enriquecimento genico dos genes com variantes patogenicas ou potencialmente patogenicas, observamos o enriquecimento de genes de transporte intracelular de proteinas. O quadro clinico dos individuos com variantes patogenicas ou potencialmente patogenicas foi esperado, como descricoes previas da literatura. Este estudo sugere que variantes de novo sao tambem um mecanismo importante para a etiologia de TEA no Brasil, explicando a arquitetura genetica de 9,5 % dos casos. (Projeto CEPID/FAPESP 2013/08028-1).

The main objective of this essay was to contribute to the characterization of the genetic architecture of ASD based on an analysis of a Brazilian series, which is still little studied. To achieve this goal, we verified the proportion of cases of ASD caused by de novo variants in neurodevelopment genes (genes from the SFARI bank and those associated with neurodevelopment described in the DECIPHER bank). Sixty-three trios were evaluated, composed of parents and probands diagnosed with ASD treated at the Human Genome and Stem Cell Studies Center (CEGH-CEL, USP). Genealogy, clinical data, gender, age at the consultation, and parental age were collected. Whole-exome sequencing was performed through a collaboration with Mount Sinai, NY, US (collaboration with the Autism Sequencing Consortium-ACS). Identification of De novo variants in candidate genes for ASD was performed using the LOVD program (LOVD v.3.0 - Leiden Open Variation Database). It was observed that most of the probands were boys (N= 55, 86%), and the minority had a family history of ASD (N=4, 6%). It was also found that 40% (N=25) of individuals had a delay in language development, and a small percentage had comorbidities such as ADHD and epilepsy (N=6, 10% and N= 2, 3%, respectively). The mean parental age at the time of pregnancy was close to 30 years for both parents (29.7 and 32.5 for the mother and father, respectively). Nine de novo pathogenic or potentially pathogenic variants were identified in candidate genes: for TEA: six in SFARI genes (four pathogenic variants in NF1, TLK2, DNAH17, BRSK2 genes, and two probably pathogenic variants in ARHGAP5 and HUWE1) and three in genes of neurodevelopment of DECIPHER (Deciphering Developmental Disorders Study, 2015) (two pathogenic variants in the ERLIN2, ST3GAL3 genes and one probably pathogenic in COL11A1). When performing the gene enrichment analysis of genes with pathogenic or potentially pathogenic variants, we observed the enrichment of genes for intracellular protein transport. The clinical picture of individuals with pathogenic or potentially pathogenic variants was expected, as previously described in the literature. This study suggests that de novo variants are also an essential mechanism for the etiology of ASD in Brazil, explaining the genetic architecture of 9.5% of cases. (Projeto CEPID/FAPESP 2013/08028-).