Evolution of DC-SIGN use revealed by fitness studies of R5 HIV-I variants emerging during AIDS progression

Background: At early stages of infection CCR5 is the predominant HIV-1 coreceptor, but in approximately 50% of those infected CXCR4-using viruses emerge with disease progression. This coreceptor switch is correlated with an accelerated progression. However, those that maintain virus exclusively restricted to CCR5 (R5) also develop AIDS. We have previously reported that R5 variants in these "non-switch virus" patients evolve during disease progression towards a more replicative phenotype exhibiting altered CCR5 coreceptor interactions. DC-SIGN is a C-type lectin expressed by dendritic cells that HIV-1 may bind and utilize for enhanced infection of T cells in trans. To further explore the evolution of the R5 phenotype we analyzed sequential R5 isolates obtained before and after AIDS onset, i.e. at the chronic stage and during end-stage disease, with regard to efficiency of DC-SIGN use in trans-infections. Results: Results from binding and trans-infection assays showed that R5 viruses emerging during end-stage AIDS disease displayed reduced ability to use DC-SIGN. To better understand viral determinants underlying altered DC-SIGN usage by R5 viruses, we cloned and sequenced the HIV-1 env gene. We found that end-stage R5 viruses lacked potential N-linked glycosylation sites (PNGS) in the gp120 V2 and V4 regions, which were present in the majority of the chronic stage R5 variants. One of these sites, amino acid position 160 (aa160) in the V2 region, also correlated with efficient use of DC-SIGN for binding and trans-infections. In fitness assays, where head-to-head competitions between chronic stage and AIDS R5 viruses were setup in parallel direct and DCSIGN-mediated infections, results were further supported. Competitions revealed that R5 viruses obtained before AIDS onset, containing the V2 PNGS at aa160, were selected for in the transinfection. Whereas, in agreement with our previous studies, the opposite was seen in direct target cell infections where end-stage viruses out-competed the chronic stage viruses. Conclusion: Results of our study suggest R5 virus variants with diverse fitness for direct and DCSIGN-mediated trans-infections evolve within infected individuals at end-stage disease. In addition, our results point to the importance of a glycosylation site within the gp120 V2 region for efficient DC-SIGN use of HIV-1 R5 viruses.