MOLECULAR MODELING STUDIES AND SAR OF ANTI-HIV AGENTS

点击翻译按钮获取中文摘要

To date, no systematic studies of the structure-activity relationships (SAR) of anti-HIV nucleosides using modern theoretical methods have been published. However, it has been shown that in the solid state active analogs exhibit moderate to extreme S-type furanose conformations (pseudorotational angle P up to 215degrees). Following this lead, using principles of conformational analysis, one can rationalize the activity or inactivity of other nucleoside analogs (including many not yet studied by X-ray methods) in terms of the effects of substituents on the furanose ring conformation. Such an analysis shows that S-type conformations are stabilized (relative to N-type) in all the active ribose analogs, whereas this effect is absent in most inactive compounds. A working hypothesis based on correlations from NMR studies is that in extreme S-type conformations (P>180degrees) the +sc orientation of 05' is destabilized, increasing access to the ap orientation, which may be more favorable for activity. The "gauche effect" is a major determinant of the N/S conformational equilibrium, and a significant correlation can be made between activity of 3' substituted THY analogs and the gauche/trans conformational preference of the substituents (as determined by MNDO calculations on substituted ethanols). Preliminary results also indicate that low energy conformers determined by molecular mechanics (MM) correlate well with observed crystal structures for anti-HIV nucleosides, but careful parameterization is required for accurate determination of relative energies. Ab initio and semiempirical quantum mechanics calculations will be used to develop MM parameters to accurately model the gauche and anomeric effects in nucleosides (generally lacking in most current force fields) and to determine conformational energetics and electronic structures of nucleosides. The preliminary results described here will be extended by comprehensive studies of active and inactive compounds (e.g. BCH-189, the activity of which support our conformational hypothesis), examining such variables as conformation, steric bulk,, hydrophobicity, molecular electrostatic potentials and hydrogen bonding. The stated working hypothesis will be tested, amplified and/or modified, and emphasis will be placed on the development of models with testable predictive value. Results will be made generally available to assist researchers working in the area of AIDS antiviral drug development.