Structure/Function Relationships for a Novel Inhibitor of Human 3β-Hydroxysteroid Dehydrogenase 1
In women, 3β-HSD1 is selectively expressed in breast tumors, mammary glands, and placenta, whereas 3β-HSD2 is expressed in the adrenal glands and ovaries. As a novel approach to prevent spontaneous preterm labor, selectively inhibiting 3β-HSD1 may cease placental estradiol production when administered near term. Late in gestation, human placenta metabolizes large amounts of fetal adrenal DHEA-S to produce increasing estradiol levels via a biosynthetic pathway catalyzed by various enzymes, including 3β-HSD1, which stimulates the expression of many estrogen-dependent molecules to promote uterine contractility. Since placental 3β-HSD1 also produces progesterone and androstenedione, 17α-hydroxyprogesterone caproate, a long-acting tocolytic progestin, would need to be co-administered with a 3β-HSD1 inhibitor to promote uterine quiescence until appropriate fetal maturation is attained. By exploiting previously identified key amino acid differences between human 3β-HSD1 and 3β-HSD2, a selective inhibitor of 3β-HSD1 was designed to control placental estradiol production without significantly affecting adrenal or ovarian steroidogenesis. Our novel inhibitor 2,16-dicyano-4,5-epoxy-androstane-3,17-dione (DiCN-AND) was synthesized from androstenedione via a four-step process. Inhibition studies with DiCN-AND resulted in the competitive inhibition of 3β-HSD1 (Ki = 4.7 μM) and noncompetitive inhibition of 3β-HSD2 with a 6.5-fold higher Ki (30.7 μM), which was similar to inhibition profiles for trilostane (3β-HSD1: Ki = 0.1 μM, competitive; 3β-HSD2: Ki = 1.6 μM, noncompetitive). For the S194G-1 mutant, trilostane inhibited noncompetitively (Ki = 0.67 μM), whereas DiCN-AND inhibited competitively with a 6.3-fold higher Ki (29.5 μM) than measured for 3β-HSD1 (Ki = 4.7 μM). Since DiCN-AND noncompetitively inhibited 3β-HSD2 that has G194 and P195 instead of S194 and R195, this suggests that R195 in 3β-HSD1 and S194G-1 is required to competitively bind DiCN-AND in the substrate-binding site. Additional evidence supporting the importance of R195 to competitively bind inhibitors with greater affinity to 3β-HSD1 is the kinetic data obtained for R195P-1 with DiCN-AND, which resulted in a similar inhibition profile to 3β-HSD2 with DiCN-AND inhibiting noncompetitively (Ki = 41.3 μM). This new data presents strong evidence that R195 in 3β-HSD1 is essential to the shift from competitive inhibition to noncompetitive and development of 3β-HSD1-specific inhibitors that may lead to potential new treatments to prevent spontaneous preterm labor.
Pham, Jenny Hong