Relative replication fitness of a high-level 3'-azido-3'-deoxythymidine-resistant variant of human immunodeficiency virus type 1 possessing an amino acid deletion at codon 67 and a novel substitution (Thr-->Gly) at codon 69

Abstract

The combination of an amino acid deletion at codon 67 (delta 67) and Thr-to-Gly change at codon 69 (T69G) in the reverse transcriptase (RT) of human immunodeficiency virus type 1 (HIV-1) is associated with high-level resistance to multiple RT inhibitors. To determine the relative contributions of the delta 67 and T69G mutations on viral fitness, we performed a series of studies of HIV replication using recombinant variants. A high-level 3'-azido-3'-deoxythymidine (AZT)-resistant variant containing delta 67 plus T69G/K70R/L74I/K103N/T215F/K219Q in RT replicated as efficiently as wild-type virus (Wt). In contrast, the construct without delta 67 exhibited impaired replication (23% of growth of Wt). A competitive fitness study failed to reveal any differences in replication rates between the delta 67+T69G/K70R/L74I/K103N/T215F/+ ++K219Q mutant and Wt. Evaluation of proviral DNA sequences over a 3-year period in a patient harboring the multiresistant HIV revealed that the T69G mutation emerged in the context of a D67N/K70R/T215F/K219Q mutant backbone prior to appearance of the delta 67 deletion. To assess the impact of this stepwise accumulation of mutations on viral replication, a series of recombinant variants was constructed and analyzed for replication competence. The T69G mutation was found to confer 2',3'-dideoxyinosine resistance at the expense of fitness. Subsequently, the development of the delta 67 deletion led to a virus with improved replication and high-level AZT resistance.

FIG. 1

Growth kinetics of recombinant HIV variants. MT-2 cells were infected with chimeric (A) or mutated (B) NL4.3. Virus growth was monitored by daily quantitation of p24 antigen in culture supernatants. (A) Open circles, Wt; closed circles, HIV_RT31980; triangles, HIV_RT31980 with an aspartic acid at RT codon 67. (B) Open circles, Wt; open triangles, recombinant HIV with Δ67; closed triangles, HIV mutant T69G/K70R/L74I/K103N/T215F/K219Q; closed circles, HIV Δ67+T69G/K70R/L74I/K103N/T215F/K219Q mutant. Results are representative of three independent experiments.

FIG. 2

Serial total CD4⁺ cells count, HIV RNA copy number, and HIV-1 RT genotypes in a patient treated with AZT monotherapy followed by AZT-ddI combination therapy (27). Particle-associated HIV-1 RNA levels in plasma were determined by the branched-DNA signal amplification assay (version 1; Chiron Diagnostics Corporation, East Walpole, Mass.) (12), the detection limit of which was 10,000 copies per ml. Sequence data were obtained from PCR-amplified proviral DNA. Boxed numbers 67, 70, 74, 103, 215, and 219 refer to amino acid changes at codons 67 (Asp to Asn), 69 (Thr to Gly), 70 (Lys to Arg), 74 (Leu to Ile), 103 (Lys to Asn), 215 (Thr to Phe), and 219 (Lys to Gln), respectively. Short arrows indicate time points when interleukin-2 was administered.

FIG. 3

Competitive replication assays of Wt and mutant Δ67+T69G/K70R/L74I/K103N/T215F/K219Q, mixed at ratios of 1:1 (A) and 4:1 (B). Infected cells were cultured in the presence (closed circles) or absence (open circles) of 10 μM AZT. Data were generated based on relative peak heights of electropherograms produced from direct DNA sequencing of virion RNA from tissue culture supernatants at the end of each passage.

FIG. 4

Replication of HIV variants containing mutations associated with in vivo growth. Recombinant mutants were constructed by site-directed mutagenesis. The mutations in the variants correspond to mutations existing in proviral DNA derived from a patient (Fig. 2). To generate viruses, MT-2 cells were infected with 1,250 TCID₅₀/3 × 10⁶ cells and cultured for 7 days. Levels of p24 antigen were measured on day 7 to determine the growth properties of each virus. Viruses were cultured in the absence (closed bars) or presence of 1 (hatched bars) or 10 (opened bars) μM AZT. Results are expressed as the percentage of growth ±/standard error of the mean compared to Wt growth in the absence of AZT. In this experiment, the p24 concentration of Wt was 644 ± 75 ng/ml. Three independent experiments were performed. Numbers 70, 215, 219, 67, 74, and 103 on the x axis indicate K70R, T215F, K219Q, D67N, L74I, and K103N, respectively.

FIG. 5

Replication properties of HIV variants. Recombinant HIV-1 mutants were constructed by site-directed mutagenesis. MT-2 cells were infected with 1,250 TCID₅₀/3 × 10⁶ cells and cultured for 7 days. Levels of p24 antigen in day 7 culture supernatant were measured to determine the growth properties of each virus. Viruses were cultured in the absence (closed bars) or presence (hatched bars) of 1 μM AZT. Results are expressed as in Fig. 4. In this experiment, the p24 concentration of Wt was 468 ± 52 ng/ml. Three independent experiments were performed. G, I, N, GI, GN, IN, and GIN denote HIV variants containing T69G, L74I, K103N, T69G/L74I, T69G/K103N, L74I/K103N, and T69G/L74I/K103N mutations, respectively. (A) Recombinant viruses with Wt amino acid (aspartic acid) at codon 67; (B) recombinant viruses with Δ67; (C) viruses containing K70R, T215F, and K219Q in addition to the indicated mutations; (D) same as panel C plus Δ67.

FIG. 6

Proposed model for the accumulation of mutations in the RT of HIV-1 during AZT-ddI combination therapy. Data points were derived from Fig. 2 and Table 1. (A) Fold increase in IC₅₀ of AZT in the absence (solid line) and presence (dashed line) of 1 μM ddI; (B) fold increase in IC₅₀ of ddI in the absence (solid line) and presence (dashed line) of 1 μM AZT; (C) percentage of Wt growth in the absence (solid line) and presence (dashed line) of 10 μM AZT. 70.215.219, 67.70.215.219, 67.69.70.215.219, Δ67+69.70.215.219, and Δ67+69.70.74.103.215.219 stand for K70R/T215F/K219Q, D67N/K70R/T215F/K219Q, D67N/T69G/K70R/T215F/K219Q, Δ67+T69G/K70R/T215F/K219Q, and Δ67+T69G/K70R/L74I/K103N/T215F/K219Q, respectively.