The objective of this research is to define the structure/function relationships of human type 1 and type 2 3?-hydroxysteroid dehydrogenase/isomerase (3?-HSD 1 and 3?-HSD2). In placenta, 3?-HSD 1 catalyzes the conversion of pregnenolone to progesterone and utilizes fetal dehydroepiandrosterone (DHEA) to produce androstenedione that is further metabolized to 17?-estradiol, which participates in the cascade of events that precede labor. 3?- HSD1 is also selectively expressed in the mammary gland and breast tumors as a key enzyme in the production of estradiol from DHEA. During the current grant period, we determined that purified 3?-HSD1 utilizes substrates and binds an inhibitor (epostane) with 14-fold higher affinity than 3 ?-HSD2, the isoenzyme expressed in human adrenals and gonads.
The first aim of the grant is to characterize the amino acids responsible for catalysis, substrate and coenzyme utilization in two isoforms using mutagenesis to identify exploitable differences. Our homology model of enzyme structure has targeted potentially critical residues that perform these functions and identified a mutant that may stably bind substrate without performing catalysis for crystallography. Because the interaction, of subunits also appears to be involved in the differences between 3?-HSD 1 and 3?-HSD2, targeted residues in the predicted subunit interface of the homodimer are mutated so that monomeric forms of 3 ?-HSD 1 and 3 ?-HSD2 are expressed, purified and characterized to determine the role of subunit interactions. Using human breast tumor MCF-7 Tet-Off cells that we have transfected with vectors encoding either 3 ?-HSD 1 or 3 ?-HSD2, kinetic studies are performed to determine if membrane-bound 3 ?-HSD 1 can be selectively inhibited without affecting 3 ?-HSD2 activity. Our cytosolic form of microsomal 3?-HSD 1 with a deleted membrane-domain has produced enzyme crystals, and the second aim is to produce diffraction-quality crystals to obtain diffraction data and a tertiary/quaternary protein structure. Creation of a second cytosolic form of 3 ?-HSD1 that contains substituted hydrophobic residues in membrane domains predicted by our model is also proposed, followed by production of a cytosolic form of the 3 ?-HSD2 enzyme. After diffraction data are obtained of 3?-HSD 1 and 3 ?-HSD2, the structures of the two isoenzymes are compared when the proteins are in the 3beta-HSD (with bound substrate) and isomerase (with bound NADH) conformations. These studies may ultimately produce new treatments for the prevention of premature birth and the treatment of hormone-sensitive breast cancer while leaving adrenal steroidogenesis intact. ? ?
