Endocrine Abstracts

Background: Diagnostic yield for 46,XY DSD remains low at about 50%. Whole-genome sequencing (WGS) can detect coding and non-coding variants, offering a more comprehensive approach than exome sequencing or panels to increase the likelihood of identifying potential pathogenic DNA changes missed by routine diagnosis. Here we report a family from Vietnam with compound heterozygous mutations in the CYB5A gene, which encodes a CYP17A1 co-factor essential for 17,20-lyase activity.

Methods: Clinical manifestations and hormone test were examined. WGS was performed on peripheral blood DNA extract from two affected siblings with 46,XY DSD. DNA sequencing data were processed using an in house bioinformatics pipeline. Variant calling, annotation, and prioritization were conducted to identify rare or novel variants potentially associated with DSD. Minigene RNA splicing assays in HEK293 cells assessed the functional impact of a deep intronic CYB5A variant. Variants were curated according to ACMG guideline.

Results: Two siblings presenting with 46,XY DSD, with female external genitalia, no uterus and bilateral inguinal cryptorchidism. Blood test showed increased 17OHP and low testosterone. WGS identified compound-heterozygous variants in CYB5A: a paternally inherited missense variant (p.Val34Glu) and a maternally inherited deep intronic deletion (c.129+862_129+863del), located 861 bp downstream of exon 1 which is predicted impact on splicing by SpliceAI. Both siblings carried both variants. Val34Glu is classified as likely pathogenic (ACMG). The intronic deletion is pathogenic by (ACMG) and SpliceAI predicts. Minigene assays confirmed new cryptic acceptor and donor sites created by the intronic variant produce a pseudoexon between exon 1 and exon 2, introducing a premature stop codon which results to nonsense-mediated decay. The variant is classified as pathogenic.

Conclusion: Only three validated CYB5A variants have been reported previously in DSD – all homozygous. The elevated 17-OHP and reduced testosterone in our patients are consistent with impaired CYB5A-dependent 17,20-lyase activity. A single heterozygous CYB5A mutation would not be diagnostic. We show that compound-heterozygous CYB5A variants, including the deep intronic deletion, likely underlie the DSD phenotype in this family. Under appreciated and underdiagnosed are the combined effect of compound-heterozygous variants, especially ‘hidden’ deep intronic defects. Our findings underscore the value of WGS and functional assays in validating clinically relevant non-coding variants.