Structure-based design, synthesis, and biological evaluation of conformationally restricted novel 2-alkylthio-6-[1-(2,6-difluorophenyl)alkyl]-3,4-dihydro-5-alkylpyrimidin-4(3H)-ones as non-nucleoside inhibitors of HIV-1 reverse transcriptase.

Abstract

5-Alkyl-2-(alkylthio)-6-(2,6-difluorobenzyl)-3,4-dihydropyrimidin-4(3H)-ones (S-DABOs, 2) have been recently described as a new class of human immunodeficiency virus type 1 (HIV-1) non-nucleoside reverse transcriptase (RT) inhibitors (NNRTIs) active at nanomolar concentrations (Mai, A. et al. J. Med. Chem. 1999, 42, 619-627). In pursuing our lead optimization efforts, we designed novel conformationally restricted S-DABOs, 3, featuring a methyl at the benzylic carbon (Y = Me) and at the pyrimidine 5-position (R = Me). Conformational analyses and docking simulations suggested that the presence of both methyls would significantly reduce conformational flexibility without compromising, in the R enantiomers, the capability of fitting into the RT non-nucleoside binding pocket. To develop structure-activity relationships, we prepared several congeners of type 3 belonging to the thymine (R = Me) and uracil (R = H) series, featuring various 2-alkylthio side chains (X = Me, i-Pr, n-Bu, i-Bu, s-Bu, c-pentyl, and c-hexyl) and aryl moieties different from the 2,6-difluorophenyl (Ar = phenyl, 2,6-dichlorophenyl, 1-naphthyl). Moreover, alpha-ethyl derivatives (Y = Et) were included in the synthetic project in addition to alpha-methyl derivatives (Y = Me). All of the new compounds were evaluated for their cytotoxicity and anti-HIV-1 activity in MT-4 cells, and some of them were assayed against highly purified recombinant wild-type HIV-1 RT using homopolymeric template primers. The results were expressed as CC(50) (cytotoxicity), EC(50) (anti-HIV-1 activity), SI (selectivity, given by the CC(50)/EC(50) ratio), and IC(50) (RT inhibitory activity) values. In the 2,6-difluorobenzylthymine (R = Me) series, methylation of the benzylic carbon improved anti-HIV-1 and RT inhibitory activities together with selectivity. Compound 3w (Ar = 2,6-F(2)-Ph, R = Y = Me, X = c-pentyl) turned out the most potent and selective among the S-DABOs reported to date (CC(50) > 200 microM, EC(50) = 6 nM, IC(50) = 5 nM, and SI > 33 333). Assays performed on the pure enantiomer (+)-3w, much more active than (-)-3w, yielded the following results: CC(50) > 200 microM, EC(50) = 2 nM, IC(50) = 8 nM, and SI > 100 000, under conditions wherein MKC-442 was less active and selective (CC(50) > 200 microM, EC(50) = 30 nM, IC(50) = 40 nM, SI > 6666). The 2,6-difluorophenylethylthymines (R = Me) were generally endowed with higher potency compared with the uracil counterparts (R = H). In the 2,6-difluorophenyl series the best and the least performant 2-alkylthio side chains were the 2-c-pentylthio and the 2-methylthio, respectively. When the methyl at the benzylic carbon was replaced by an ethyl, activity was retained or decreased slightly, thus suggesting that the dimensions of the cavity within the RT hosting this substituent would not be compatible with groups larger than ethyl. Aryl moieties different from the 2,6-difluorophenyl (phenyl, 1-naphthyl, 2,6-dichlorophenyl) were generally detrimental to activity, consistent with a favorable electronic effect exerted by the 2,6-fluorines on a putative charge-transfer interaction between the aromatic moieties of the inhibitor and Tyr188.