Deep intronic NIPBL variants underlie unresolved Cornelia de Lange syndrome revealed by integrated epigenomic and transcriptomic analysis.

Cornelia de Lange syndrome (CdLS) is a rare congenital disorder characterised by craniofacial features, growth retardation, limb anomalies, and intellectual disability. The condition arises from pathogenic variants in genes encoding cohesin complex components or regulators, most frequently NIPBL. Despite advances in genomic testing, approximately 30% of clinically diagnosed cases remain unresolved, suggesting the presence of non-coding variants. We investigated 10 classical CdLS patients negative for pathogenic variants in NIPBL, SMC1A, SMC3, RAD21, and HDAC8. Genome-wide methylation profiling was performed to identify CdLS-specific epigenetic signatures. Episignature-positive cases underwent trio-based deep sequencing of the entire NIPBL locus while targeted NIPBL RNA sequencing was used to assess splicing. Whole-genome sequencing (WGS) was performed in episignature-negative individuals. Five of ten patients exhibited a positive CdLS episignature in blood, saliva, or skin. Deep sequencing identified de novo deep intronic NIPBL variants in all cases, and two additional variants in episignature-negative patients. Seven distinct intronic variants were identified across introns 4, 10, 21, and 32. In silico tools predicted the creation or activation of cryptic splice sites in six variants. RNA sequencing confirmed pseudoexon inclusion in three patients, leading to premature termination codons and predicted nonsense-mediated decay. Tissue-specific variation in aberrant transcript expression was observed between blood and fibroblasts. We conclude that deep intronic NIPBL variants represent an under-recognised cause of CdLS, frequently escaping detection by conventional exome analysis. Integration of methylation episignatures, deep intronic sequencing, and RNA analysis enhances diagnostic yield and provides insight into the regulatory architecture of cohesinopathies.