Objective: There is evidence that gag contributes to protease inhibitor susceptibility in treatment-experienced patients. Moreover, protease inhibitor resistance-associated mutations can arise in gag in the absence of protease mutations in vitro. We wished to assess the contribution of full-length Gag to protease inhibitor susceptibility in viruses unexposed to protease inhibitors, in particular from the most common HIV-1 subtypes, namely subtype A and C.
Design: We compared the drug resistance profiles of subtype A and C cognate gag-protease (from viruses not previously exposed to protease inhibitor) to protease combined with a generic subtype B gag as in routine phenotypic testing.
Methods: We amplified gag-protease sequences from plasma-derived virus or molecular clones, and used a single cycle transfection-based drug resistance assay to compare the fold changes in the concentration of drug required to inhibit 50% of viral replication of these viruses to a generic subtype B. We made a series of chimeras to explore phenotypes further.
Results: In some cases, use of protease sequences without the cognate gag overestimated susceptibility to protease inhibitors, in particular to lopinavir. We provide evidence that gag sequences from wild-type viruses can contribute as much as 14-fold reduction in susceptibility to lopinavir, and that cognate protease can balance this by partially restoring susceptibility.
Conclusion: Our findings demonstrate the importance of considering protease inhibitor susceptibility in the context of full-length gag, particularly with respect to the range of HIV-1 subtypes circulating worldwide.