Automated docking analysis provides new insights on protease inhibitor selection for antiretroviral therapy in HIV-1 infection

Antiretroviral drugs may target four key proteins on the HIV life cycle: reverse transcriptase, protease, transmembrane envelope glycoprotein and integrase. Protease inhibitors (PI) were introduced in clinical use in 1995 immediately leading to the emergence of new drug resistant variants. Due to the high mutation rate of the HIV reverse transcriptase and the strong selective pressure imposed by protease inhibitors there are numerous genetic variants of the HIV protease at the individual and population level. Selection of which protease inhibitors to include in a new therapeutic regimen may be well informed by sequencing analysis of the pol gene. However, in heavily treated patients the increasing genetic complexity of the resistant proteases limits the usefulness of genotypic resistance testing to inform therapeutic decisions. The main goal of this study was to develop an in silico method that allows the choice of PIs based on the structural differences found between drug resistant and drug susceptible proteases. Three dimensional (3D) models of the HIV-1 protease were produced for eight sequences with known PI resistance mutations selected from the Stanford HIV Resistance Database and for one control protease without resistance mutations. Ligand-protein automated docking with flexible ligand was subsequently performed for the binding site of each of the proteases with some of the PIs commercially available in Portugal. Using as an example the resistant mutation 48V to analyze 1) protease-PI binding affinity, 2) complexity of the pathway the drugs take to reach the binding site, 3) number of occurrences with free binding energies below zero and their distribution, we show that amprenavir and tipranavir are the PIs of choice to inhibit the replication of HIV strains bearing this mutation. Our preliminary results suggest that this new method may help to inform difficult therapeutic decisions in heavily treated HIV-1 patients.