INFLUENCE OF ESTROGEN STRUCTURE ON NUCLEAR-BINDING AND PROGESTERONE-RECEPTOR INDUCTION BY THE RECEPTOR COMPLEX

The relationship between steroid structure, estrogen receptor (ER) binding affinity, nuclear binding of the ER complex, and induction of progesterone receptor (PgR) have been examined. The level of ER in membrane-free homogenates of MCF-7 cells was found to be 10.0 +/- 0.5 fmol/mug of DNA by utilizing an enzyme immunoassay (EIA). However, only 2.5 +/- 0.2 fmol of ER complex/mug of DNA was bound by nuclei during maximal stimulation of PgR synthesis (2.9 +/- 0.2 fmol of PgR/mug of DNA; measured by EIA) following a pulse with 10(-10) M E2. Except at micromolar concentrations, estratriene was an ineffective estrogen. The addition of a hydroxyl group to either position 3 or position 17beta of estratriene yielded ligands which were capable of causing nuclear binding and processing of ER as well as PgR induction. D-ring regioisomers of estradiol (E2) had lower affinity for receptor than E2. However, receptor complexed with these estrogens was fully capable of binding to nuclear material, undergoing processing, and inducing PgR. On the other hand, A-ring regioisomers of E2 displayed significant differences in their ability to mediate nuclear binding of receptor complex and induction of PgR. Although 1-hydroxyestratrien-17beta-ol was weakly bound by ER, this dihydroxyestrogen was capable of bringing about nuclear binding and processing of ER and the stimulation of PgR synthesis. In contrast, 2- and 4-hydroxyestratrien-17beta-ol, which caused extensive nuclear binding of ER (5-7 fmol/mug of DNA), were incapable of significant PgR induction. Provided that the A-ring hydroxyl group was positioned correctly (3beta) on 5alpha-androstanediols or 5-androstenediol, an aromatic ring was not required for nuclear binding of the ER complex and stimulation of PgR synthesis. With the exception of 2- and 4-hydroxyestratrien-17beta-ol, induction of PgR by structurally altered estrogens correlated with the affinity of ligand for ER. Electrostatic models generated from this data were found to be useful in the characterization of electronegative isopotential regions of the estrogen (or androstanediol) molecules which were important in modulating the gene regulatory properties of ER.