Synthesis of estradiol analogues based on the A-CD steroid ring system

Abstract

This thesis describes the synthesis of a series of estrogen analogs related to estradiol but lacking the B ring of the natural hormone. Those compounds are prepared by methods developed previously in our lab and variations thereof. Typically this involved coupling of the enatinomerically pure CD-ring ketone A with the lithio derivatives B derived from several substituted suitable protected 4-bromophenols. The compounds thus obtained have been evaluated as potential estrogen receptor agonists. Molecular Mechanics calculations, carried out by Prof. Wright's group at Carleton University indicate that the preferred conformation of compounds such as C is one in which the plane of the aromatic ring is essentially perpendicular to that made by the carbon skeleton of the CD ring system. However their calculations strongly suggested that in the receptor this angle is reduced to 21 +/-5°. This compares with the approximately 10° difference in these planes in the natural hormone estradiol. In order to test whether compounds in which the dihedral angle between the aromatic ring and the CD ring system is fixed at positions other than 10 or 20 ° the A-CD, the spiro derivatives D were prepared. In the spiro-5 derivative (n = 1) the planes are essentially perpendicular (90°). When n = 6 as in compound D [spiro-6] there is more flexibility and the dihedral angle can be as small as 60 °. The relative binding affinity [estradiol = 100] for the spiro-5 compound is RBAalpha = 2.5, RBAbeta = 10.7. These data are comparable to the parent A-CD analogue E [RBAalpha = 1.5, RBAbeta = 21]. The RBAs for the 6-spiro compound were somewhat smaller [RBAalpha = 0.3, RBAbeta = 3.5]. These results indicate that either the estrogen receptor can accommodate almost equally well the two compounds with different shapes or that the calculations indicating the E rotates from its preferred conformation in which the ring A is perpendicular to the plane of the CD rings to a conformation in which the two planes in the presence of the receptor is in error. The RBAs for the compounds in the families C and F [R = H, Cl and CF3] are smaller than those observed for the saturated analogs E. The eventual goal of the research carried out by the Durst/Wright groups was to prepare compounds which cannot be metabolized to form the reactive ortho-quinones which have been implicated as one of the triggers of estrogen related breast cancers. The A-CD compounds represented by the C, E and F families can in principle be metabolized to ortho-quinonones, perhaps more slowly in cases where the substituent is an electron withdrawing group. In contrast, no quinone formation is possible when the phenol is replaced by a pyrazole ring as in G. The compound G was synthesized; the relative binding assays were: RBAalpha = 0.04 and RBAbeta = 0.06. Although the binding is quite low relative to estradiol, the result is considered quite encouraging since it shows that pryazole can be a surrogate for the phenol. The low RBAs observed for G is probably due to the much shorter distance (∼9A) between the NH and OH in G, as compared to the ideal 11 A found in estradiol. It is predicted that inserting an atom, for example, CH2 or O or S between the pyrazole and the CD ring in compound G will significantly increase the binding affinity to the estrogen receptors.* *Please refer to dissertation for diagrams.