The G118R plus R263K Combination of Integrase Mutations Associated with Dolutegravir-Based Treatment Failure Reduces HIV-1 Replicative Capacity and Integration

Abstract

Human immunodeficiency virus (HIV) treatment with antiretroviral regimens containing integrase strand transfer inhibitors such as dolutegravir (DTG) and bictegravir (BIC) offers high levels of protection against the development of drug resistance mutations. Despite this, resistance to DTG and BIC can occur through the development of the R263K integrase substitution. Failure with DTG has also been associated with the emergence of the G118R substitution. G118R and R263K are usually found separately but have been reported together in highly treatment-experienced persons who experienced treatment failure with DTG. We used cell-free strand transfer and DNA binding assays and cell-based infectivity, replicative capacity, and resistance assays to characterize the G118R plus R263K combination of integrase mutations. R263K reduced DTG and BIC susceptibility ~2-fold, in agreement with our previous work. Single-cycle infectivity assays showed that G118R and G118R plus R263K conferred ~10-fold resistance to DTG. G118R alone conferred low levels of resistance to BIC (3.9-fold). However, the G118R plus R263K combination conferred high levels of resistance to BIC (33.7-fold), likely precluding the use of BIC after DTG failure with the G118R plus R263K combination. DNA binding, viral infectivity, and replicative capacity of the double mutant were further impaired, compared to single mutants. We propose that impaired fitness helps to explain the scarcity of the G118R plus R263K combination of integrase substitutions in clinical settings and that immunodeficiency likely contributes to its development.

Keywords: G118R; HIV drug resistance; R263K; bictegravir; dolutegravir; integrase; integrase inhibitors.

FIG 1

Schematic representation of the strand transfer and DNA binding assays. (A) The strand transfer assay was performed with a preprocessed LTR DNA (red) that lacked a GT dinucleotide and carried a reactive hydroxyl group (-OH) at the 3′ terminus. The complete method is described in Materials and Methods. Briefly, immobilized preprocessed LTR DNA (red) was incubated with integrase (multicolor) and different concentrations of biotin-labeled target DNA (black) for 1 h at 37°C. After washes, the amount of integrated target DNA was measured via biotin (green) quantification with europium-labeled streptavidin (dark blue). Time-resolved fluorescence was plotted against the target DNA concentrations to produce pseudo-V_max (maximum integration) and pseudo-K_m, which accounted for target DNA binding affinity. (B) DNA binding assays were performed with a blunt LTR DNA (red) labeled with rhodamine (green). The complete method is described in Materials and Methods. Immobilized integrase proteins (multicolor) were incubated with various concentrations of LTR DNA for 1 h at 37°C. After washes, the amount of bound DNA was measured by quantifying fluorescence. Results were plotted against LTR DNA concentrations to produce K_d values.

FIG 2

Strand transfer activities of WT, R263K, G118R, and G118R plus R263K recombinant HIV-1 integrase proteins. (A) Relative strand transfer activity in the presence of various protein concentrations. Results from two separate experiments performed with two different protein batches (four experiments) were compiled after normalization against the mean fluorescence value (from triplicates) obtained with the WT protein at 400 nM, arbitrarily set at 100%. Means ± standard deviations are presented. (B) Relative strand transfer activity in the presence of increasing target DNA concentrations (0, 1.8, 3.75, 7, 15, 30, 60, and 120 nM). Results from three separate experiments performed with two different protein batches (six experiments) were compiled after normalization against the mean fluorescence value (from triplicates) obtained with the WT protein and 120 nM target DNA, arbitrarily set at 100%. Means ± standard deviations are presented.

FIG 3

Relative infectivity of WT, R263K, G118R, and G118R plus R263K HIV-1. TZM-bl reporter cells were infected for 24 h with various HIV-1 forms at different concentrations. Results from two separate experiments with two different viral stocks were compiled after normalization against the mean maximal and minimal RLU values (from triplicates), arbitrarily set at 100% and 0%, respectively. Infectivity curves were plotted against the log viral concentrations in RT units using GraphPad Prism v9.4.1. Means ± standard deviations are presented.

FIG 4

Multicycle replication assay results. PM1 cells were infected with WT, R263K, G118R, and G118R plus R263K HIV-1, and replication was monitored by quantifying the RT activity in the cell culture fluids 2, 4, 6, 9, 11, 13, 16, 18, and 20 days postinfection. Results from two separate experiments with two different viral stocks were compiled after normalization against the maximal RT values obtained with the WT virus after 11 days of infection, arbitrarily set at 100%. A logarithmic scale was used to allow visualization of the G118R plus R263K virus replication curve. Areas under the curves were calculated using GraphPad Prism v9.4.1, and results are presented in Table 1. Means ± standard deviations are presented.

FIG 5

Effect of G118R on BIC binding to the HIV-1 integrase catalytic site. The WT protein is shown in red and the in silico-modeled G118R mutant in green. BIC is colored by standard elements (nitrogen in blue, carbon in gray, oxygen in red, and fluorine in green). The overlap between G118R and BIC suggests a conformational clash between the substituted amino acid and the small molecule.