Background
Neurodevelopmental disorders (NDDs) are highly diverse conditions often associated with genetic abnormalities. However, a large number of NDD cases remain undiagnosed despite thorough genetic testing using short-read sequencing and chromosomal microarray analysis. Emerging evidence indicates that structural variants (SVs) in evolutionarily new and human-specific genomic regions may be responsible for these unresolved cases.
Methods
We used Oxford Nanopore long-read DNA sequencing in six patients with unexplained NDDs who had previously tested negative for genetic mutations. Structural variants were identified and filtered based on their genomic location, regulatory potential, and expression in the central nervous system. Confirmatory fluorescence in situ hybridization (FISH) was performed. Additionally, immunohistochemical analysis of the candidate gene ANKRD20A1 was carried out to assess its spatiotemporal expression in the developing human brain. Interactome analysis was also performed to evaluate the functional connections of genes impacted by SVs.
Results
A total of twenty-six candidate SVs were found, including deletions, duplications, insertions, and inversions. Many of these SVs affect rapidly evolving gene families such as NBPF, TBC1D3, and RGPD, as well as regulatory elements in brain-expressed genes, suggesting their potential to disrupt brain development and lead to NDD. FISH analysis confirmed several large SVs. Patient-specific interactome maps showed that most RGPD-disrupted genes create extensive interactions. Immunohistochemical analysis revealed that ANKRD20A1, a candidate gene, is dynamically expressed during midfetal human cortical development, suggesting its involvement in neurodevelopmental events.
Conclusion
Our findings highlight the crucial role of long-read DNA sequencing in uncovering concealed structural variants within recently evolved genomic regions. These SVs may contribute to the development of NDDs by disrupting coding sequences, regulatory elements, or complex gene networks. This study supports integrating long-read sequencing into research workflows.
