ESPE Abstracts

Background: Loss-of-function variants in the NR5A1 gene are frequent causes of 46,XY differences of sex development (DSD). To date, the Human Gene Mutation Database contains 291 NR5A1 variants, most of which are missense (69%). Mouse models demonstrated the effect of loss of SF-1 on sex development, but the interactome of SF-1 is huge and an explanation for the broad phenotype is still missing. Controversies exist as: a) in vitro transactivation assays of NR5A1-regulated genes show contradictory results for several variants, b) in vitro modelling of heterozygous variants does not show a dominant negative effect, c) several variants are present in combination with other DSD-related gene variants of unknown significance. We aimed to describe the genotype in the largest possible international SF-1/NR5A1 cohort to allow future identification of phenotype-genotype clusters.

Methods: After recruiting cases through the I-DSD registry and through contacting researchers from previous publications on SF-1, collaborators entered genotype data in a REDCap database. Gene variants were identified by single gene analysis, next generation sequencing (NGS) and/or comparative genomic hybridization array. We performed variant classification according to guidelines from the American College of Medical Genetics and Genomics. Collaborators of SF1next study group are listed here: https://www.dropbox.com/s/wza2bud5naw2ftd/SF1next%20study%20group.pdf?dl=0.

Results: Among 107 individuals (90% 46,XY, 8% 46,XX, 1% 47,XXY, 1% 47,XYY), we found 78 different NR5A1 variants, of which 28 were novel (11/28 missense, 8/28 delins, 3/28 nonsense, 2/28 silent, 2/28 small deletions or insertions). All but 6 were heterozygous (72/78, 92%). Variants were scattered throughout the whole NR5A1 gene, mostly in the DNA-binding (33/78, 42%) and the ligand-binding domain of the protein (28/78, 36%). We classified 59/78 (76%) variants as pathogenic/likely pathogenic and 16/78 (20%) as variants of unknown significance. Additional gene variants (e.g. DHX37) were identified in 26/107 (24%) individuals using various methods (52% single-gene analysis, 33% NGS).

Conclusion: Novel variants are still identified in the NR5A1 gene. Further studies are required to explain the broad phenotype of SF-1 individuals. Current targeted approaches have limitations as they may not detect additional gene variants. In the future, individuals who have loss-of-function SF-1 variants may have to be re-evaluated with newer methods, e.g. NGS.