Background:  During the course of HIV-1 infection the phenomenon of co-receptor switch from viruses using CCR5 as co-receptor (R5) to viruses using CXCR4 as co-receptor (X4) is associated with disease progression. To determine a role for positive selection in X4 emergence and identify amino acid sites in the gp120 associated to co-receptor use efficiency, we investigated in vivo evolution of X4 HIV-1 phenotype within children who developed AIDS during the first year of their life and did not receive HAART.

Methods:  We used high-resolution phylodynamic analysis to track the intra-patient emergence of X4 virus populations in longitudinal peripheral blood mononuclear cells (PBMC), and lymphoid and non-lymphoid tissues (brain, lung, lymph nodes, spleen, and thymus) collected post mortem. The env gp120 hypervariable domains V1, V2, and V3 were analyzed.

Results:  Reconstruction of ancestral HIV-1 envelope sequences showed that HIV-1 quasispecies undergo multiple population bottlenecks. Bottlenecks that occurred during the early years of infection developed independent of co-receptor switch, but led to the emergence of viral populations with positively selected amino acid in V1, V2, and C2, as well as in V3, domains. In contrast, bottlenecks that appeared late in the course of infection were associated with a switch in co-receptor use from R5 to X4, reflecting additional amino acid substitutions in V3. The emergence/evolution of X4 HIV-1 variants is a multi-step temporally ordered process requiring the initial fixation of positively selected mutations in V1, V2, and C2, without which high-charge V3 mutants cannot survive. X4 viruses were amplified in the thymus and exported from the thymus to other lymphoid organs, such as spleen and lymph nodes.

Conclusions:  Phylodynamic analysis is a cutting-edge tool to explain HIV-1 intra-patient evolutionary dynamics. Our study is the first detailed mapping of the evolutionary patterns of HIV-1 in different anatomic tissues. The observed episodic positive selection occurred before the emergence of X4 strains, in absence of HAART, and in immunodepressed patients. Our data demonstrate that viral entry efficiency into target cells can be another major factor driving the selection of fitter HIV-1 variants over time and the emergence of X4 viral strains. R5 and X4 co-receptor-blocking agents may be extremely efficient in avoiding the emergence of the X4 quasispecies during the early stage of the disease, and controlling HIV-1 infection in advanced disease stages.