(4R,5S)/(4S,5R)-4,5-Bis(4-hydroxyphenyl)-2-imidazolines 1-7 were synthesized by the reaction of the methoxy-substituted (1R,2S)/(1S,2R)-1,2-diarylethylenediamines 1b-7b with triethyl orthoformate and subsequent ether cleavage with BBr(3). All compounds were tested for estrogen receptor (ER) binding in a competition experiment with [(3)H]-estradiol and for gene activation in a luciferase assay using ER positive MCF-7-2a breast cancer cells stably transfected with the plasmid ERE(wtc)luc. The relative binding affinities of the 2-imidazolines were very low (RBA < 0.1%). Nevertheless, 4-7 possessed full agonistic activity in the luciferase assay. The relative transcription potency increased in the order 5 (2,2'-I) < 6 (2,6-Cl(2), 2'-F) < 4 (2,2'-Cl) < 7 (2,6-Cl(2), 2'-Cl). These data together with spectroscopic and molecular modeling studies were used to investigate the preferred binding mode adopted by the imidazoline ligands. The 1,2-diarylethane pharmacophor takes a Z-stilbene-like structure with pseudoaxially oriented phenyl rings at the planar heterocyclic ring. Because of this unusual spatial structure, the (4R,5S)/(4S,5R)-4,5-bis(4-hydroxyphenyl)-2-imidazolines have to be assigned to a second class of estrogenically active compounds (type II estrogens). In contrast to type I estrogens, e.g., estradiol (E2), diethylstilbestrol (DES), and meso-hexestrol (HES), which are connected to His 524 in the binding site, type II estrogens are very likely H-bonded to Asp 351 in a hydrophobic side pocket.

(4R,5S)/(4S,5R)-4,5-Bis(4-hydroxyphenyl)-2-imidazolines bearing 2,2'-H (3a), 2,2'-Cl (3b), 2,2',6-Cl (3c), and 2,2'-F (3d) substituents in the aromatic rings were C2-alkylated (5a-i), N-alkylated (7, 7a-c), and N,N'-dialkylated (9a-c). The synthesis started from the diastereomerically pure (1R,2S)/(1S,2R)-1,2-diamino-1,2-bis(4-methoxyphenyl)ethanes 1a-d, which were cyclized to the imidazolines 2a-d and 4a-i with triethylorthoesters or iminoethers. Ether cleavage with BBr(3) yielded the (4R,5S)/(4S,5R)-4,5-bis(4-hydroxyphenyl)-2-imidazolines 3a-d and 5a-i. The N-alkylation and N,N'-dialkylation of 2b, employed for obtaining 7a-c and 9a-c, were performed prior to the ether cleavage with alkyl iodine in dry THF. By use of HPLC, the influence of the substitution patterns in the aromatic rings and alkyl chains at the C2- or N-atoms on the hydrolysis rate of the imidazolines was studied under in vitro conditions. It appeared that only imidazolines with C2- or N-alkyl substituents show sufficient stability to interact as heterocycles with the estrogen receptor (ER). The resulting gene activation was monitored in a luciferase assay using ERalpha-positive MCF-7-2a breast cancer cells stably transfected with the plasmid ERE(wtc)luc. It is interesting to note that C2-alkylation led to a strong reduction or even a complete loss of activity whereas N-alkylation improved the estrogenic profile. The (4R,5S)/(4S,5R)-N-ethyl-4,5-bis(2-chloro-4-hydroxyphenyl)-2-imidazoline 7b has proven to be the most active compound in this structure-activity relationship study (EC(50) = 0.015 microM).

No abstract

Previously, we reported on the synthesis and estrogen receptor (ER) interaction of imidazoles, which had to be 1-alkyl-4, 5-bis(2-halo-4-hydroxyphenyl) substituted for a high relative binding affinity (RBA >1 %). This led to the assumption that a shielding of the polar heterocyclic system is a prerequisite for ER binding. In continuation of this study we synthesized 2, 4, 5-tris(4-hydroxyphenyl)imidazoles with Cl-or F-atoms in the ortho-positions of the aromatic rings and evaluated whether they mediate sufficient hydrophobicity for ER interaction. 2-(2, 6-Dichloro-3/4-hydroxyphenyl)-4, 5-bis(2-halo-4-hydroxyphenyl)imidazoles were synthesized by reaction of the respective methoxy-substituted benzil with either the 2, 6-dichloro-4-methoxy-or the 2, 6-dichloro-3-methoxybenzaldehyde in ammonium acetate solution. The required ether cleavage was performed subsequently with BBr(3). In the competition experiment with [(3)H]estradiol the imidazoles with the a C2-standing (2, 6-dichloro-4-hydroxyphenyl) ring showed an RBA >0.02 %, but did not activate the luciferase gene in estrogen receptor positive MCF-7-2a breast cancer cells stably transfected with the plasmid ERE(wtc)luc. In the test for antagonistic potency only the 2-(2, 6-dichloro-4-hydroxyphenyl)-4, 5-bis(4-hydroxyphenyl)imidazole 3 antagonized the effects of 1 nM estradiol slightly. From these data, it can be concluded that a C2-standing 2, 6-dichloro-4-hydroxyphenyl ring is not appropriate to optimize the ER interaction of 4, 5-(4-hydroxyphenyl)imidazoles.

4,5-Bis(4-hydroxyphenyl)imidazoles with 2,2'-H (1),2,2'-F (2),2,2'-Cl (3), and 2,2'6-Cl (4) substituents in the aromatic rings were synthesized by oxidation of the respective methoxy-substituted (R,S)/(S,R)-4,5-diaryl-2-imidazolines with MnO2 and subsequent ether cleavage with BBr3. N-alkylation of 1 and 3 with ethyl iodide yielded the compounds 5 and 6. The imidazoles were characterized by NMR spectroscopy and tested for estrogen receptor binding in a competition experiment with [3H]estradiol using calf uterine cytosol. Gene activation was verified in a luciferase assay using estrogen receptor positive MCF-7-2a cells stably transfected with the plasmid ERE(wtc)luc. All halide substituted imidazoles competed with estradiol for the binding site at the estrogen receptor. The N-ethyl derivative 6 showed the highest relative binding affinity of 1.26 %. Treatment of MCF-7-2a cells, however, did not lead to gene activation. The relative activation of 6 amounted only to 10% at 1 microM compared to E2 (100%).