Meta‐Analyses Support Previous and Novel Autism Candidate Genes: Outcomes of an Unexplored Brazilian Cohort

Abstract

Large genomic databases of neurodevelopmental disorders (NDD) are helpful resources of genomic variations in complex and heterogeneous conditions, as Autism Spectrum Disorder (ASD). We evaluated the role of rare copy number variations (CNVs) and exonic de novo variants, in a molecularly unexplored Brazilian cohort of 30 ASD trios (n = 90), by performing a meta‐analysis of our findings in more than 20,000 patients from NDD cohorts. We identified three pathogenic CNVs: two duplications on 1q21 and 17p13, and one deletion on 4q35. CNVs meta‐analysis (n = 8,688 cases and n = 3,591 controls) confirmed 1q21 relevance by identifying duplications in other 16 ASD patients. Exome analysis led the identification of seven de novo variants in ASD genes (SFARI list): three loss‐of‐function pathogenic variants in CUL3, CACNA1H, and SHANK3; one missense pathogenic variant in KCNB1; and three deleterious missense variants in ATP10A, ANKS1B, and DOCK1. From the remaining 12 de novo variants in non‐previous ASD genes, we prioritized PRPF8 and RBM14. Meta‐analysis (n = 13,754 probands; n = 2,299 controls) identified six and two additional patients with validated de novo variants in PRPF8 and RBM14, respectively. By comparing the de novo variants with a previously established mutational rate model, PRPF8 showed nominal significance before multiple test correction (P = 0.039, P‐value adjusted = 0.079, binomial test), suggesting its relevance to ASD. Approximately 60% of our patients presented comorbidities, and the diagnostic yield was estimated in 23% (7/30: three pathogenic CNVs and four pathogenic de novo variants). Our uncharacterized Brazilian cohort with tetra‐hybrid ethnic composition was a valuable resource to validate and identify possible novel candidate loci. Autism Res 2020, 13: 199–206. © 2019 International Society for Autism Research, Wiley Periodicals, Inc. Lay Summary We believed that to study an unexplored autistic population, such as the Brazilian, could help to find novel genes for autism. In order to test this idea, with our limited budget, we compared candidate genes obtained from genomic analyses of 30 children and their parents, with those of more than 20,000 individuals from international studies. Happily, we identified a genetic cause in 23% of our patients and suggest a possible novel candidate gene for autism (PRPF8).

Transcriptome of iPSC-derived neuronal cells reveals a module of co-expressed genes consistently associated with autism spectrum disorder

“…Duplications in this region are well described in the literature [40], and although associated with variable phenotypes and incomplete penetrance, a considerable number of individuals with this copy number variation (CNV) present ASD. This patient also carries a de novo missense variant predicted to be damaging (CADD ≥ 20) in DOCK1 (NM_001290223: c.4838C>T (p.Pro1613Leu)) [41] an ASD-candidate gene (Sfari score = 3). The other patient, F2688, is a compound heterozygote for rare potentially deleterious missense variants (NM_005045.3: c.7538C>G (p.Ser2513Cys) and c.7634C>T (p.Ala2545Val)) inherited from his parents in the Reelin gene (RELN), which is considered a high confidence ASDcandidate (Sfari = 1).…”

Complement C4 is Reduced in iPSC-Derived Astrocytes of Autism Spectrum Disorder Subjects

“…All individuals evaluated in this study (n = 7 individuals with non-syndromic ASD and n = 5 neurotypical controls) were described previously [27][28][29][30]. CGH-array and whole exome sequencing using genomic DNA from peripheral blood of the ASD subjects and their parents allowed the identi cation of known ASD pathogenic variants in two individuals: one patient harbors deleterious compound heterozygous variants in the RELN gene, and a de novo splice site variant in the CACNA1H gene; the second patient harbors a duplication of 15q11-13.…”

Cul3regulates cytoskeleton protein homeostasis and cell migration during a critical window of brain development

“…However, the molecular and cellular functions of the majority of the identified genes remain poorly understood. One of the identified high-risk ASD genes encodes the E3 ubiquitin ligase Cullin 3 (Cul3) 1-3,6-10 .…”

“…CUL3 ASD-associated genetic variants are most often de novo missense or loss of function (loF) mutations, dispersed throughout the entire gene and affecting distinct protein domains. In addition to the ASD core symptoms, patients with CUL3 de novo loF mutations can present with several comorbidities including varying levels of intellectual disability (ID), attention deficit hyperactivity disorder (ADHD), sleep disturbances, motor deficits and facial dysmorphic features 10,12,13 . The only known exception is the deletion of CUL3 exon 9 by a specific dominant splice site variant causing a severe form of pseudohypoaldosteronism type II (PHAII), featuring hypertension, hyperkalemia, and metabolic acidosis but not ASD 14,15,16 .…”