Role of minority populations of human immunodeficiency virus type 1 in the evolution of viral resistance to protease inhibitors

Abstract

Human immunodeficiency virus type 1 (HIV-1) drug resistance results from the accumulation of mutations in the viral genes targeted by the drugs. These genetic changes, however, are commonly detected and monitored by techniques that only take into account the dominant population of plasma virus. Because HIV-1-infected patients harbor a complex and diverse mixture of virus populations, the mechanisms underlying the emergence and the evolution of resistance are not fully elucidated. Using techniques that allow the quantification of resistance mutations in minority virus species, we have monitored the evolution of resistance in plasma virus populations from patients failing protease inhibitor treatment. Minority populations with distinct resistance genotypes were detected in all patients throughout the evolution of resistance. The emergence of new dominant genotypes followed two possible mechanisms: (i) emergence of a new mutation in a currently dominant genotype and (ii) emergence of a new genotype derived from a minority virus species. In most cases, these population changes were associated with an increase in resistance at the expense of a reduction in replication capacity. Our findings provide a preliminary indication that minority viral species, which evolve independently of the majority virus population, can eventually become dominant populations, thereby serving as a reservoir of diversity and possibly accelerating the development of drug resistance.

FIG. 1.

Quantification of the proportion of plasma virions expressing the V82A and L90M resistance mutations. At the indicated times after virologic treatment failure with protease inhibitor (PI), the proportion of cDNA sequences expressing the V82A (A) and L90M (B) resistance mutations was quantified by using sequence-selective real-time PCR as described in Materials and Methods. Each symbol represents the results for a different patient. The V82A mutation was detected by genotyping for only four of the seven patients evaluated. The symbols corresponding to patients A to C, whose cases were explored in greater detail by the analysis of clones, are indicated in panel B.

FIG.2.

Evolution of the protease gene in patients failing retroviral treatment. At the indicated times after treatment failure, RNA was extracted from plasma, and the viral protease gene was amplified by RT-PCR and cloned. Clones with or without the L90M mutation were identified and sequenced. On the top of each panel, the viral load and the percentage of total sequences expressing the L90M mutation are shown. Below, the phylogenetic trees obtained by DNAPARS are shown, along with the corresponding protease genotypes. The results for patients A, B, and C are shown in the corresponding panels.

FIG.2.

Evolution of the protease gene in patients failing retroviral treatment. At the indicated times after treatment failure, RNA was extracted from plasma, and the viral protease gene was amplified by RT-PCR and cloned. Clones with or without the L90M mutation were identified and sequenced. On the top of each panel, the viral load and the percentage of total sequences expressing the L90M mutation are shown. Below, the phylogenetic trees obtained by DNAPARS are shown, along with the corresponding protease genotypes. The results for patients A, B, and C are shown in the corresponding panels.

FIG.2.

Evolution of the protease gene in patients failing retroviral treatment. At the indicated times after treatment failure, RNA was extracted from plasma, and the viral protease gene was amplified by RT-PCR and cloned. Clones with or without the L90M mutation were identified and sequenced. On the top of each panel, the viral load and the percentage of total sequences expressing the L90M mutation are shown. Below, the phylogenetic trees obtained by DNAPARS are shown, along with the corresponding protease genotypes. The results for patients A, B, and C are shown in the corresponding panels.

FIG. 3.

Replicative capacities of recombinant viruses containing protease sequences present at different times in the evolution of resistance. DNA fragments corresponding to the protease present in the predominant viral species at different times in the evolution of resistance for patient B (top panel) and patient C (bottom panel) were cloned into pNL-4.3XCS. Viruses produced after transfection of HeLa cells were recovered and used to infect in quadruplicate P4 cells (2 ng of p24/well) cultured in 96-well plates, and the induction of β-galactosidase was determined by a colorimetric assay 36 h later. The results for individual experiments, expressed as the percentage of wild-type virus from the same experiment, are shown by open symbols. Solid triangles represent the mean ± the standard error of the mean for all determinations. Viruses for patient B (top) had the following resistance genotypes: 1, L24I M46I I54V A71V V82T (dominant at months 27 and 33); 2, L24I M46I I54V A71V V82T I93L (month 39); and 3, M46I I54V A71V V82T L90M I93L (month 43). Virus genotypes for patient C (bottom) were as follows: 1, I54V A71V V82A I93L (month 24); 2, I54V A71V V82A I84V I93L (month 34); and 3, I54V A71V V82A L90M I93L (month 37). In two cases, polymorphisms were detected in the virions expressing the following resistance genotypes: patient B, 43 months, 55K/R, and patient C, 34 months, 37D/E. Only the most abundant population was studied here (55K and 37D, respectively). WT, wild type.

FIG. 4.

Resistance to protease inhibitors of recombinant viruses containing protease sequences present at different times in the evolution of resistance. Recombinant viruses were produced as described in the Fig. 3 legend and used to transfect HeLa cells. At 18 h after transfection, HeLa cells were treated with trypsin, and aliquots were seeded into 96-well plates containing serial threefold dilutions of saquinavir (SQV) or nelfinavir (NFV). After 24 h, aliquots of supernatant corresponding to 2 ng of p24 (based on the amount of p24 in the supernatant of HeLa cells cultured with medium alone) were used to infect in quadruplicate P4 cells cultured in 96-well plates, and the induction of β-galactosidase was determined by a colorimetric assay 36 h later. The results for each virus were fitted to a sigmoid curve with variable slope, and the IC₅₀ and IC₉₀ were determined. The results are presented as the mean ± the standard error of the mean of the fold increase in IC₅₀ (A and B) or IC₉₀ (C and D) compared to that of wild-type virus (patient B, n = 6 experiments; patient C, n = 5 experiments for saquinavir and n = 4 experiments for nelfinavir). The viruses from patient B (A and B) and patient C (C and D) are as described in the Fig. 3 legend. The mean IC₅₀ and IC₉₀ values ± the standard error of the mean for wild-type virus were, respectively, 10.9 ± 5.6 and 110.3 ± 70.1 ng/ml for nelfinavir and 1.7 ± 1.51 and 8.1 ± 3.9 ng/ml for saquinavir.

FIG. 5.

Dose-response curves for the inhibition of viral replication of recombinant viruses from patient C by saquinavir (SQV) and nelfinavir (NFV). The effect of serial dilutions of saquinavir and nelfinavir on the replication of wild-type virus (WT [gray triangles]), virus 2 (○), and virus 3 (•) was evaluated as described in the Fig. 4 legend. The data for each virus from a single experiment were fitted to sigmoid curves with variable slopes. At each of the indicated drug concentrations, the replication, expressed as a percentage of that observed for wild-type virus in the absence of drug, was determined. The values from five (saquinavir) or four (nelfinavir) independent experiments were used to generate the sigmoid curves shown in the figure.