In vitro inhibition of androstenedione 5alpha-reduction by finasteride in epithelium and stroma of human benign prostatic hyperplasia.

Finasteride is a well known steroid 5alpha-reductase inhibitor. In this context, recently we have shown that in human benign prostatic hyperplasia (BPH) finasteride inhibits the 5alpha-reduction of testosterone to dihydrostestosterone (DHT) more effectively in the epithelium as compared to the stroma. The aim of the present study was to describe in epithelium and stroma of human BPH the effect of finasteride on the 5alpha-reduction of androstenedione, that is the second main circulating androgen in men, to androstanedione. Using a finasteride concentration of 75 nM and an androstenedione concentration of 220 nM, the mean inhibition [% +/- SEM] of 5alpha-reductase activity was significantly higher in epithelium (69 +/- 2) than in stroma (52 +/- 4). Both in epithelium and stroma, this inhibition of 5alpha-reductase activity was dose-dependent and competitive. Dixon plots as well as slope replots of Lineweaver-Burk plots showed that the mean inhibition constant Ki (nM +/- SEM) was significantly lower in epithelium (10 +/- 1 and 11 +/- 2, respectively) than in stroma (33 +/- 7 and 28 +/- 4, respectively) indicating a significantly stronger inhibitory effect of finasteride in epithelium. From those mean Ki values, it follows that in human BPH finasteride inhibits equally well both the 5alpha-reduction of androstenedione to androstanedione and testosterone to DHT. Based on these inhibition studies, there is no evidence for the coexistence of substrate-specific 5alpha-reductases converting either testosterone or androstenedione. However, the striking difference in finasteride sensitivity of the 5alpha-reduction between epithelium and stroma could be due to a cell-type specific expression of structurally different 5alpha-reductases as well as to a different access of finasteride to 5alpha-reductase in epithelium and stroma where, compared to each other, the lipid environment is significantly different.