Design, syntheses, and evaluation of novel 1,1‐dihalo‐2,3‐diphenylcyclopropanes as potential cyclooxygenase‐2 (COX‐2) inhibitors with analgesic‐antiinflammatory activity

Abstract

AbstractA group of (Z)‐ and (E)‐1,1‐dihalo‐2‐(4‐substituted‐phenyl)‐3‐phenylcyclopropane [(Z)‐10, (E)‐11] stereoisomers having a variety of substituents (H, Br, Cl, F, NO2, SO2Me) at the para‐position of the C‐2 phenyl ring in conjunction with either two chloro or bromo substituents at C‐1 were synthesized for in vivo evaluation as analgesic and antiinflammatory (AI) agents, and as potential selective cyclooxygenase‐2 (COX‐2) inhibitors. This group of compounds (10‐11) exhibited significant analgesic activity since 4% NaCl‐induced abdominal constriction was reduced by 44–73% at 30 min, and 48–77% at 60 min, post‐drug administration relative to the reference drugs aspirin and celecoxib (58 and 32% inhibition at 30 min post‐drug administration) for a 50 mg/kg intraperitoneal dose. In the 1,1‐dichloro group of compounds, a Cl or MeSO2 substituent at the para‐position of the C‐2 phenyl ring generally provided superior analgesic activity. The most active analgesic compound, (E)‐1,1‐dichloro‐2‐(4‐methanesufonylphenyl)‐3‐phenylcyclopropane (11h) inhibited abdominal constriction by 72 and 77% at 30 and 60 min post‐drug administration, respectively. AI activities, determined using the carrageenan‐induced rat paw edema assay, showed that this class of (Z)‐10 and (E)‐11 compounds exhibited AI activities in the inactive‐to‐moderate activity range (1.5–45% inhibition) for a 50 mg/kg oral dose. The AI potency order, with respect to the para‐substitutent on the C‐2 phenyl ring, for the (Z)‐10 compounds was NO2 > MeSO2 ≈ H ≥ Cl, and for the (E)‐11 compounds was H ≥ MeSO2 > Cl ≈ Br. (E)‐1,1‐dibromo‐2‐(4‐methanesufonylphenyl)‐3‐phenylcyclopropane (11l), which was the most active AI compound, reduced inflammation by 45 and 37% at 3 and 5 h post‐drug administration, respectively. The (E)‐11 stereoisomer was generally a more potent AI agent than the corresponding (Z)‐10 stereoisomer. In vitro COX‐1 and COX‐2 inhibition studies showed that (E)‐1,1‐dichloro‐2‐(4‐nitrophenyl)‐3‐phenylcyclopropane (11c) inhibited COX‐1 (IC50 = 278.8 μM) and COX‐2 (IC50 = 80.5 μM) for a COX‐2 selectivity index of 3.5, whereas (E)‐1,1‐dichloro‐2‐(4‐methanesulfonylphenyl)‐3‐phenylcyclopropane (11h) was a more potent inhibitor of COX‐1 and COX‐2, but it was more selective for COX‐1 (COX‐1 IC50 = 0.59 μM, COX‐2 IC50 = 3.04 μM). A molecular modeling (docking) study for (E)‐1,1‐dichloro‐2‐(4‐methanesulfonylphenyl)‐3‐phenylcyclopropane (11h) on the active site of the human COX‐2 isozyme shows it binds in the center of the active site with the 1,1‐dichloro substituents oriented in the direction of the mouth of the channel towards Arg120, and the C‐2 MeSO2 moiety oriented towards the apex of the active site with the S‐atom of the MeSO2 substituent positioned about 6.56 Å inside the entrance to the secondary pocket (Val523) of COX‐2. In contrast, the corresponding (Z)‐10h stereoisomer assumes a different position in the COX‐2 binding site where the S‐atom of the MeSO2 moiety is near (4.02 Å) the Ser530 OH, but a much greater distance from the COX‐2 secondary pocket (Val523). The results from these docking studies are consistent with the observation that (E)‐11h is an inhibitor of both COX isozymes, whereas the (Z)‐10h stereoisomer is an inactive COX inhibitor (COX‐1 IC50 > 100 μM, COX‐2 IC50 > 200 μM). Drug Dev. Res. 55:79–90, 2002. © 2002 Wiley‐Liss, Inc.