In vitro selection and characterisation of human immunodeficiency virus type-1 subtype C integrase strand transfer inhibitor resistant mutants

Abstract

The currently approved integrase strand transfer inhibitors (INSTIs), raltegravir (RAL) and elvitegravir (EVG) effectively halt HIV-1 replication but their use is limited by their low genetic resistance barrier and cross resistance. For instance, integrase amino acids N155 and Q148 represent genetic pathways selected by both drugs and are associated with considerable cross resistance to both RAL and EVG. Dolutegravir (DTG) is a second generation drug manufactured to exhibit a more robust resistance profile than RAL and EVG, and retains activity against RAL and EVG resistant isolates. Most research on drug resistance patterns have been carried out with emphasis on HIV-1 subtype B and inadequately assessed in HIV-1 subtype C. Thus, the aim of this study was to establish the drug resistance mutation profiles of HIV-1 subtype C primary virus isolates that evolve/emerge under selective pressure of the INSTIs RAL, EVG and DTG, and evaluate their impact on strand transfer. In vitro selection experiments were carried out using six primary virus isolates (three wild-type, FV, and three reverse transcriptase drug resistant, MR, viruses) grown in peripheral blood mononuclear cells in the presence of increasing concentrations of RAL, EVG and DTG, and monitored to beyond virus break-through. Viral RNA was extracted from various time points and the pol region was RT-PCR amplified and sequenced using conventional Sanger-based sequencing and next generation sequencing (Illumina MiSeq). HIV-1 subtype C FV6 wild-type and mutant recombinant integrase (generated by site-directed mutagenesis) were expressed, purified and used in strand transfer assays and surface plasmon resonance (SPR) experiments to establish the binding affinities of IN-DNA. Wild-type FV primary viruses were successfully grown in the presence of increasing concentrations of RAL, EVG and DTG, up to 266 nM, 66 nM and 32 nM, respectively. Drug resistant MR viruses were successfully grown in the presence of increasing concentrations of RAL, EVG and DTG, up to 266 nM, 16 nM and 8 nM, respectively. Sequence analysis on both platforms revealed the presence of the previously described drug resistance mutations T66IK, E92Q, F121Y, Q148R, N155H and R263K in some viruses, and additionally H114L was detected. RAL was observed to select for substitutions Q148R and N155H/H114L in isolates FV6 and MR69, respectively. EVG selected F121Y, T66I/R263K, T66K and T66I in FV3, FV6, MR69, MR81, and MR89, respectively. DTG selected substitutions E92Q and M50I in FV3 and MR81, respectively. In silico data exhibited changes in hydrophilicity, hydrophobicity and side chain changes as well as changes in polarity, and all substitutions displayed acceptable minimisation energies and distances between the atoms. Seven IN mutants were expressed and purified, and thereafter tested for efficiency in strand transfer. All mutant FV6T66I, FV6E92Q, FV6H114L, FV6F121Y, FV6Q148R, FV6N155H and FV6R263K IN enzymes demonstrated an overall loss in strand transfer capacity of 37.1%, 21.5%, 66.1%, 63.2%, 60.2%, 30.5% and 3.4%, respectively. This is the first report on loss of strand transfer activity associated with H114L. The loss in strand transfer capacity in all the mutants was not reflected by their overall binding affinities to donor DNA, as determined by surface plasmon resonance, likely attributed to the role of different residues associated with DNA and drug binding in the IN quaternary structure. In conclusion, this is the first report describing IN drug selection experiments using primary HIV-1 subtype C isolates, and a detailed genotypic and biochemical characterisation of the associated mutations.