(393b) Where Are We in HIV Research? a Novel, Computer-Based Strategy for Predicting Resistance to HIV-1 Nnrtis

The introduction of Highly Active Antiretroviral Therapy (HAART) has had a dramatically positive effect on the natural history of HIV-1 disease in the developed world. However, incomplete suppression leading to drug resistance has often impaired the HAART efficacy. For example, resistance towards non-nucleoside HIV-1 reverse transcriptase inhibitors (NNRTIs) develops rapidly in the clinical setting. Accordingly, the optimal use of NNRTIs will require both the appreciation of the potential for the development of drug resistance, and the recognition that this problem can be avoided. Thus, an approach to the issue of rapid emergence of NNRTI resistance could be the attempt to provide sufficient levels of potent inhibitors in order to inhibit not only wild type HIV-1, but also preexisting resistant viral variants found at low levels as a result of de novo mutations during ongoing virus replication, or at high levels as a consequence of NNRTI treatment failure. In this work we developed a computational procedure for the evaluation of the free energy contribution of each residue in the HIV-1 RT in binding to several, different NNRTIs which are known to fail in the presence of given mutations. The purpose of these computations is: 1) to obtain indication for the design of resistance-evading drugs and, 2) to calculate the values of an empirical parameter, GV, which combines free energy calculations and sequence analysis to suggest possible drug resistance mutations on the RT. In practical terms, this parameter is defined as the product of a given residue contribution to the total binding free energy and the variability of that residue. This quantity could, in principle, be used in assisted resistance-evading drug design for HIV-1 RT, as well as for any other proteic viral target. The developed computational method allowed to correctly predict all the resistance mutations found in vitro and/or in vivo for a substantial number of NNRTIs. Moreover, it was also able to highlight aminoacid residues classified as sensitive to resistance. The method proposed is highly predictive in all cases considered. More importantly, for those compounds for which the X-ray structure of the relevant complex with HIV-1 RT is not available, an ab initio procedure is devised, starting from the docking of the drug within the enzyme to the prediction of the relevant site susceptible to mutation. This method can be employed for the a priori prediction of the insurgence of resistance of newly designed drugs, or to aptly modify existing drugs towards a more efficacious binding even in the presence of resistant HIV-1 RT mutants.