Background:  We previously generated R5 HIV-1 variants by serial passage that are resistant to the CCR5 antagonist maraviroc (MVC, UK-427,857). To characterize the molecular mechanism of this resistance we performed site-directed mutagenesis on envelope (Env) derived from an MVC-resistant (MVCres) variant of HIV-1 strain CC1/85 and assessed the susceptibility of the mutants to MVC, both in PHA-stimulated peripheral blood lymphocytes (PBL) and an Env pseudovirus entry assay.

Methods:  Sequencing analysis of Env following serial passage of the MVCres CC1/85 strain in the absence of compound suggested that the molecular determinants of resistance were localised in the V3 loop. A 2.56-kb fragment encompassing Env was amplified from the MVCres variant and site-directed mutagenesis was performed on the region encoding the V3 loop using QuickChangeXL. Mutated Env fragments were subsequently cloned into pNL4-3. Virus was recovered in the tissue culture supernatant of HEK 293T cells, 48 hours after transfection with the Env recombinant NL4-3 site-directed mutagenesis. Virus susceptibility to MVC, aplaviroc (GW873140), enfuvirtide and efavirenz was assessed in PHA-stimulated PBL, using p24 levels in the cell culture media 7 days after acute infection. The susceptibility of Env-pseudoviruses to MVC was also measured in a single cycle replication assay using U87CD4⁺CCR5⁺ cells.

Results:  The MVCres Env recombinant NL4-3 clone exhibited high-level resistance to MVC but remained sensitive to aplaviroc, confirming that it required CCR5 for entry into PBL. Back mutation of either 316T or 323V (numbering from HxB2), to A and I, respectively, resulted in a virus that gave an incomplete dose response curve (plateau). When both mutations were replaced, dose response curves similar to the MVC-sensitive parental virus were obtained. A high concentration of MVC blocked the activity of aplaviroc on MVCres variants, consistent with the compounds binding to overlapping sites on CCR5 and suggesting that MVCres viruses have acquired the ability to use MVC-occupied receptors for entry.

Conclusions:  Reduced MVC susceptibility was characterised by dose response curves that did not reach 100% inhibition. These results support the view that resistance to CCR5 antagonists is characterised by selection of variants that recognize inhibitor:receptor complexes, and indicate that parameters other than fold changes in IC₅₀ may be important for identifying resistance in patients.