Search

As part of our investigation into the structure−activity relationship of a novel class of aromatase inhibitors, C19 steroids having no oxygen function at C-3, we tested aromatase inhibition activity of polar diol compounds 4,19-dihydroxyandrost-5-en-17-ones (25 and 27) and 6,19-dihydroxyandrost-4-en-17-ones (36 and 37). 4α,19-Diol 25 was synthesized from tert-butyldimethylsilyoxyandrost-4-ene steroid (9) through its OsO4 oxidation, giving the 4α,5α-dihydroxy derivative 12, as a key reaction. Acetylation of 5β,6α-dihydroxy-19-acetate 30 and its 5α,6β-analogue 31 followed by dehydration with SOCl2 and alkaline hydroxysis gave 6α,19-diol 36 and its 6β-isomer 37, respectively. The stereochemistry of a hydroxy group at C-4 of compound 25 and that at C-6 of compounds 36 and 37 were determined on the basis of ¹H NMR spectroscopy in each case. 4β,19-Diol 27, previously synthesized, was identified as an extremely powerful competitive inhibitor of aromatase (Ki = 3.4 nM). In contrast, its 4α,19-dihydroxy isomer 25 and other series of diol compounds, 6,19-dihydroxy-4-en-17-one steroids, were moderate to poor competitive inhibitors (Ki = 110−800 nM). Through this series of analyses, it was concluded that hydrophilic interaction of a 4β,19-diol function with the active site of aromatase plays a critical role in the tight binding of 3-deoxy-5-ene steroids.

Series of 6α- and 6β-alkyl-substituted androst-4-en-17-ones (18 and 19) and their 17β-reduced derivatives (14 and 15) (alkyl:  methyl, ethyl, n-propyl, n-pentyl, n-octyl) were synthesized and evaluated as aromatase inhibitors. Androst-4-en-17-ones having an oxygen function (hydroxy, acetoxy, or methoxy group) at C-6α and C-6β (4 and 5) were also tested for their abilities to inhibit aromatase. All of the steroids studied inhibited human placental aromatase in a competitive manner. The inhibitory activities of the 6α- and 6β-methyl-17-keto steroids 18a and 19a (Ki = 3.1 and 5.3 nM, respectively) as well as the 6β-alcohol 5a (Ki = 6.0 nM) were high, and their apparent Ki values were lower than that of the parent 6-unsubstituted 3-deoxy steroid 1 (Ki = 6.8 nM). Elongation of the methyl group decreased affinity for aromatase in relation to carbon number of the alkyl chain in each series, in which the 6α-alkyl steroids 18 essentially had higher affinity for the enzyme than the corresponding 6β-isomers 19. The inhibitory activities of the 17β-hydroxy analogs 14 and 15 were less potent than those of the corresponding 17-keto steroids. The 6α-ethyl compound 18b, the 6α-oxygenated derivatives 4, and the 6β-acetoxy and 6β-methoxy analogs 5b and 5c were powerful inhibitors (Ki = 12−24 nM). The methyl steroids (18a and 19a) produced “type I” difference spectra upon interaction with aromatase. These results along with molecular modeling with the PM3 method suggest that compounds 18a and 19a may produce a thermodynamically stable enzyme−inhibitor complex in the hydrophobic binding pocket with a limited accessible volume. A carbonyl group at C-17 of the 6-alkylandrost-4-enes is essential for the tight binding. Moreover, the binding pocket also tolerates a polar hydroxy group at the 6β-position rather than at the 6α-position.

Series of 6α- and 6β-alkylandrosta-1,4-diene-3,17-diones (3 and 4) were synthesized and evaluated as time-dependent inactivators of aromatase in human placental microsomes to gain insights to the structure−activity relationship of varying the 6-n-alkyl substituents (C-1−C-7) to the time-dependent inactivation activity. All of the inhibitors synthesized were powerful to good competitive inhibitors of aromatase, with apparent Ki's ranging from 4.7 to 54 nM. The 6β-ethyl (4b) and 6β-n-pentyl (4e) compounds were the most potent among them (Ki = 4.7 and 5.0 nM for 4b and 4e, respectively). In a series of the 6α-alkyl steroids, the inhibitors 3a−d having C-1−C-4 at the 6-position as well as the 6α-n-hexyl steroid 3f caused a time-dependent inactivation of aromatase while 6α-n-pentyl (3e) and 6α-n-heptyl (3g) compounds did not. In contrast, in the 6β-alkyl steroid series, only the methyl analog 4a inactivated aromatase in a time-dependent manner, and the other alkyl steroids having more than two carbons at C-6β did not. The inactivations were prevented by the substrate androstenedione, and no significant effects of l-cysteine on the inactivation were observed in each case. These results along with molecular modeling with the PM3 method indicate that both length and stereochemistry of a straight alkyl substituent at the C-6 position of androsta-1,4-diene-3,17-dione (3h) play an important role in the cause of a time-dependent inactivation of aromatase. No significant correlation between affinity for the enzyme and the inactivation ability in the 6-alkylandrosta-1,4-diene-3,17-diones is observed.