ESPE Abstracts

Background: Inhibin consists of two homologous subunits, the α (INHA) and the βA or βB subunits (INHBA/INHBB). Both inhibins play an important role in the hypothalamic-pituitary-gonadal axis by regulating the follicle stimulating hormone levels. INHA knockout mice develop mixed or incompletely differentiated gonadal stromal tumours. In females, some INHA variants have been associated with primary ovarian failure (POF), while in males, homozygous INHA variants were recently described causing decreased prenatal and postnatal testosterone production and infertility. Similarly, variants in SF-1/NR5A1 may cause a very broad phenotype of variation of sexual development. This study aimed at understanding the combined effect of heterozygous variants in INHA [c.675T>G, Ser225Arg] and SF-1/NR5A1 [c.58G>C, Val20Leu] found in a 46,XY individual with severe undervirilization.

Methods: Genetic family analysis was performed. Different in silico tools predicted the possible impact of SF-1/NR5A1 and INHA variants on protein structure and function to classify them according to the ACMG guidelines for pathogenicity. We searched for human INHA variants in literature and in databases (ClinVar, HGMD, UniProt and Ensembl), and for interaction between SF-1/NR5A1 and INHA in different species. Using bioinformatic tools, binding sites for SF-1/NR5A1 in the 5′ flanking region of INHA were searched and used as templates to generate several promoter-reporter constructs in pGL3. Expression vectors containing the wild-type and missense SF-1/NR5A1 and INHA variants were also generated. Functional studies assessed the transcriptional activation of the INHA promoter by WT or mutant SF-1/NR5A1, while (in-)direct protein interactions between WT or variant INHA and SF1 were assessed by transient transfection experiments in different cell models.

Results: The healthy father of the index patient carried only the SF-1/NR5A1 variant. But in silico analyses suggested that both the SF-1/NR5A1 and INHA variants were pathogenic. Multiple sequence alignment showed conservative amino acids affected. We found no reported interactions between the two proteins. However, 26 INHA variants with suggested pathogenicity were reported in different databases, majority missense variants (18/26, 69%), 12 (66%) with unknown significance (VUS), although associated with adrenocortical or ovarian tumours, POF and male infertility. We found several potential SF-1/NR5A1 and CREB binding sites in the INHA promoter that are currently tested for their functional activity using corresponding promoter deletion constructs. Experiments checking INHA-SF-1 protein interaction are also ongoing.

Conclusion: The exact role of INHA in human sex development remains largely unknown. Our study suggests an interplay between INHA and SF-1/NR5A1, and possible digenic pathogenicity with genetic variants.