Search

With the insight generated by the availability of X-ray crystal structures of various 5,6-dihydropyran-2-ones bound to HIV PR, inhibitors possessing various alkyl groups at the 6-position of 5,6-dihydropyran-2-one ring were synthesized. The inhibitors possessing a 6-alkyl group exhibited superior antiviral activities when compared to 6-phenyl analogues. Antiviral efficacies were further improved upon introduction of a polar group (hydroxyl or amino) on the 4-position of the phenethyl moiety as well as the polar group (hydroxymethyl) on the 3-(tert-butyl-5-methyl-phenylthio) moiety. The polar substitution is also advantageous for decreasing toxicity, providing inhibitors with higher therapeutic indices. The best inhibitor among this series, (S)-6-[2-(4-aminophenyl)-ethyl]-(3-(2-tert-butyl-5-methyl-phenylsulfa nyl)-4-hydroxy-6-isopropyl-5,6-dihydro-pyran-2-one (34S), exhibited an EC50 of 200 nM with a therapeutic index of > 1000. More importantly, these non-peptidic inhibitors, 16S and 34S, appear to offer little cross-resistance to the currently marketed peptidomimetic PR inhibitors. The selected inhibitors tested in vitro against mutant HIV PR showed a very small increase in binding affinities relative to wild-type HIV PR. Cmax and absolute bioavailability of 34S were higher and half-life and time above EC95 were longer compared to 16S. Thus 34S, also known as PD 178390, which displays good antiviral efficacy, promising pharmacokinetic characteristics and favorable activity against mutant enzymes and CYP3A4, has been chosen for further preclinical evaluation.

Several small, achiral nonpeptide inhibitors of HIV-1 protease with low micromolar activity were identified by mass screening of the Parke-Davis compound library. Two of the compounds, structurally similar, were both found to be competitive and reversible inhibitors [compound 1, 4-hydroxy-3-(3-phenoxypropyl)-1-benzopyran-2-one: Ki = 1.0 microM; compound 2, 4-hydroxy-6-phenyl-3-(phenylthio)-pyran-2-one: Ki = 1.1 microM]. These inhibitors were chosen as initial leads for optimization of in vitro inhibitory activity based on molecular modeling and X-ray crystallographic structural data. While improvements in inhibitory potency were small with analogues of compound 1, important X-ray crystallographic structural information of the enzyme-inhibitor complex was gained. When bound, 1 was found to displace H2O301 in the active site while hydrogen bonding to the catalytic Asps and Ile50 and Ile150. The pyranone group of compound 2 was found to bind at the active site in the same manner, with the 6-phenyl and the 3-phenylthio occupying P1 and P1', respectively. The structural information was used to develop design strategies to reach three or four of the internal pockets, P2-P2'. This work led to analogues of diverse structure with high potency (IC50 < 10 nM) that contain either one or no chiral centers and remain nonpeptide. The highly potent compounds possess less anti-HIV activity in cellular assays than expected, and current optimization now focuses on increasing cellular activity. The value of the HIV-1 protease inhibitors described is their potential as better pharmacological agents with a different pattern of viral resistance development, relative to the peptide inhibitors in human clinical trials.