Virus population homogenization following acute human immunodeficiency virus type 1 infection

Abstract

Understanding the properties of human immunodeficiency virus type 1 (HIV-1) variants capable of establishing infection is critical to the development of a vaccine against AIDS. Previous studies of men have shown that the HIV-1 env gene is homogeneous early in infection, leading to the suggestion that infection is established by a single transmitted variant. However, we report here that all of eight homosexual men evaluated beginning 3.7 to 9 weeks following onset of symptoms of acute infection harbored diverse virus populations in their blood, with median genetic distances averaging 1.08% in the env C2V5 region and 0.81% in the gag p17 gene. Within another 4.7 to 11 weeks, the variant lineage in env became more homogeneous, while gag sequences continued to diversify. Thus, the homogenization that has been reported to characterize acute infection is actually preceded by the replication of multiple virus variants. This early selective process focuses on viral properties within Env but not Gag p17. Hence, the viral homogeneity observed early in HIV-1 infection results from a selective process that occurs during the establishment of infection.

FIG. 1.

Initial heterogeneity and outgrowth of more-homogeneous env gene populations over the first 3 months of HIV-1 infection. (a) HMA of HIV-1 env C2-V5 coding sequences in plasma samples from eight patients. Heteroduplexes (arrows) were visualized as slow migrating bands in 5% polyacrylamide gels stained with ethidium bromide (reverse video image shown). (b and c) HTA of HIV-1 env C2-V5 coding sequences using plasma (first lane of each pair) and the corresponding PBMC (second lane of each pair) from four patients (A to D) from panel a. Second-round PCR products were reannealed with ³²P-labeled env variant probes. The time of origin (after onset of symptoms) of driver sequences are indicated at the top of the lanes. The first column of each panel (P) shows the probe reannealed to itself.

FIG. 2.

Phylogenetic reconstruction of HIV-1 env (C2V5 of the gp120 region) (a) and gag p17 (b) sequences. Sequences obtained at the earliest time point (filled boxes and circles) and at a later time point (open boxes and circles) are shown (see Fig. 1b for time points used). Arrows indicate nodes at which MRCAs were inferred.

FIG. 3.

Distribution of all pairwise genetic distances (diversity) for patients A through H (panels a to h, respectively) among p17 gag gene sequences (triangles) and C2-V5 env sequences (circles) as a function of number of days after onset of symptoms of acute HIV-1 infection. Regression lines are indicated for gag (solid lines) and env (broken lines).

FIG. 4.

Divergence from the inferred MRCA for patients A through H (panels a to h, respectively) among p17 gag gene sequences (triangles) and C2-V5 env sequences (circles) as a function of number of days after onset of symptoms of acute HIV-1 infection. Regression lines are indicated for gag (solid lines) and env (broken lines).