In silico and functional studies reveal novel loss-of-function variants of SRD5A2, but no variants explaining excess 5α-reductase activity.

Androgens are steroid hormones essential for human male and female development. Steroid reductases 5α (SRD5As) are key enzymes in androgen biosynthesis. Mutations in the human SRD5A2 are known to cause loss-of-function and severe 46,XY undervirilization. Gain-of-function variants have been suggested in androgen excess syndromes, but have not been found so far. Therefore we searched for gain-of-function mutations in the human SRD5A2 gene which might explain hyperandrogenic disorders such as the polycystic ovary syndrome, premature adrenarche and prostate cancer. We screened databases for candidate variants and characterised them in silico with the help of a novel SRD5A2 model. We selected 9 coding SNPs (A49T, R50A, P106L, P106A, N122A, L167S, R168C, P173S, R227Q) that have not been described in manifesting individuals, and assessed their enzyme kinetic properties in HEK293 cells. SRD5A2 activity was assessed by conversion of testosterone (T), progesterone (Prog) and androstenedione (Δ4A) to their 5α-reduced metabolites. Variants R50A and P173S showed partial activity with substrates T (34% and 28%) and Δ4A (37% and 22%). With substrate Prog variants P106L, P106A, L167S and R168C in addition showed partial activity (15% to 64%). Functional testing of all other variants showed loss-of-function. As predicted in our in silico analysis, all coding SNPs affected enzyme activity, however none of them showed gain-of-function. Thus excess 5α-reductase activity might be rather regulated at the (post)-transcriptional and/or post-translational level. However through this work seven new coding SNPs were characterised which might be of clinical relevance. It is possible that individuals carrying these SNPs show a minor phenotype that is not yet identified.