The mutational and phenotypic spectrum of TUBA1A-associated tubulinopathy

Abstract Background The TUBA1A-associated tubulinopathy is clinically heterogeneous with brain malformations, microcephaly, developmental delay and epilepsy being the main clinical features. It is an autosomal dominant disorder mostly caused by de novo variants in TUBA1A. Results In three individuals with developmental delay we identified heterozygous de novo missense variants in TUBA1A using exome sequencing. While the c.1307G > A, p.(Gly436Asp) variant was novel, the two variants c.518C > T, p.(Pro173Leu) and c.641G > A, p.(Arg214His) were previously described. We compared the variable phenotype observed in these individuals with a carefully conducted review of the current literature and identified 166 individuals, 146 born and 20 fetuses with a TUBA1A variant. In 107 cases with available clinical information we standardized the reported phenotypes according to the Human Phenotype Ontology. The most commonly reported features were developmental delay (98%), anomalies of the corpus callosum (96%), microcephaly (76%) and lissencephaly (agyria-pachygyria) (70%), although reporting was incomplete in the different studies. We identified a total of 121 specific variants, including 15 recurrent ones. Missense variants cluster in the C-terminal region around the most commonly affected amino acid position Arg402 (13.3%). In a three-dimensional protein model, 38.6% of all disease-causing variants including those in the C-terminal region are predicted to affect the binding of microtubule-associated proteins or motor proteins. Genotype-phenotype analysis for recurrent variants showed an overrepresentation of certain clinical features. However, individuals with these variants are often reported in the same publication. Conclusions With 166 individuals, we present the most comprehensive phenotypic and genotypic standardized synopsis for clinical interpretation of TUBA1A variants. Despite this considerable number, a detailed genotype-phenotype characterization is limited by large inter-study variability in reporting.