Background:  The increasing incidence of HIV-1 protease inhibitor (PI) cross-resistance necessitates the better understanding of the molecular basis of resistance development to currently marketed, as well as investigational PI. The aim of this study is to investigate the differential impact of multi-drug-resistance-associated mutations involved in resistance to such drugs.

Methods:  A novel bio-informatics resistance determination (BIRD) approach was established. An analysis of our in-house database of matching phenotypic and genotypic profiles of HIV-1 clinical isolates was performed, yielding a defined set of multi-PI-resistant viruses (sources) and a corresponding set of genetically closely related PI susceptible viruses (targets). For each source we constructed a set of synthetic protease genes carrying all combinations of resistance mutations that differ between source and target, representing multiple genetic pathways. Those synthetic gene products were used for recombinant protein or virus production to build a multidimensional dataset, that helps to delineate the genetic pathways involved in HIV-1 resistance to PI.

Results:  In a proof-of-concept study, we selected 2 source viruses to construct a set comprising 40 synthetic protease gene variants carrying the primary PI mutations L33F, M46I, I50V, I84V, and L90M in a background of resistance-associated mutations—L10I, K20I, M36I, I47V, I54L, A71V, and V77I—in different combinations. Removal of the mutations L33F and A71V from their sources had a significant effect on viral growth resulting in an impaired viability. Subsequent phenotypic characterization (21 of 40 viruses) showed that L33F, I47V, and I54L in this PI resistance genetic background played a crucial role in reduced susceptibility to marketed PI (e.g., amprenavir and lopinavir). In addition we observed that the binding affinity (Surface Plasmon Resonance) between such PI and multi-drug-resistance proteases was often reduced, mainly as a result of increased dissociation rates.

Conclusions:  The BIRD approach elaborates on genetic profiles observed in the clinic and allows the assessment of the influence of all combinations of mutations in complex, PI-resistant genetic background. Future work includes measurement of binding kinetics between mutant proteases and PI. This novel approach will guide our further understanding of the molecular basis for resistance development to current and investigational PI (e.g., TMC114).