Diagnosing Cornelia de Lange syndrome and related neurodevelopmental disorders using RNA-sequencing

Abstract

PurposeNeurodevelopmental phenotypes represent major indications for children undergoing clinical exome sequencing. However, 50% of cases remain undiagnosed even upon exome reanalysis. Here we show RNA sequencing (RNA-seq) on human B lymphoblastoid cell lines (LCL) is highly suitable for neurodevelopmental Mendelian gene testing and demonstrate the utility of this approach in suspected cases of Cornelia de Lange syndrome (CdLS). MethodsGenotype-Tissue Expression project transcriptome data for LCL, blood, and brain was assessed for neurodevelopmental Mendelian gene expression. Detection of abnormal splicing and pathogenic variants in these genes was performed with a novel RNA-seq diagnostic pipeline and using a validation CdLS-LCL cohort (n=10) and test cohort of patients who carry a clinical diagnosis of CdLS but negative genetic testing (n=5). ResultsLCLs share isoform diversity of brain tissue for a large subset of neurodevelopmental genes and express 1.8-fold more of these genes compared to blood (LCL, n=1706; whole blood, n=917). This enables testing of over 1000 genetic syndromes. The RNA-seq pipeline had 90% sensitivity for detecting pathogenic events and revealed novel diagnoses such as abnormal splice products in NIPBL and pathogenic coding variants in BRD4 and ANKRD11. ConclusionThe LCL transcriptome enables robust frontline and/or reflexive diagnostic testing for neurodevelopmental disorders.