HIV-1 evolution in gag and env is highly correlated but exhibits different relationships with viral load and the immune response

Abstract

Objective: To evaluate relationships between HIV-1 evolution, including immune evasion, and markers of disease progression during chronic infection.

Design: HIV-1 evolution and disease progression markers were evaluated over approximately 5 years of infection among 37 Kenyan women from a prospective, seroincident cohort. Evolution was measured in two genes, gag and env, which are primary targets of cellular and humoral immune responses, respectively.

Methods: Proviral HIV-1 gag and env sequences were obtained from early and chronic infection when plasma viral load and CD4 cell counts were available. Human leukocyte antigen types were obtained to identify changes in gag cytotoxic T-lymphocyte epitopes. The breadth of the neutralizing antibody response was measured for each woman's plasma against a panel of six viruses. Tests of association were performed between virus evolution (diversity, divergence, and ratio of nonsynonymous to synonymous divergence), markers of disease progression (viral load and CD4 cell count), and immune parameters (gag cytotoxic T lymphocyte epitope mutation and neutralizing antibody breadth).

Results: HIV-1 gag and env diversity and divergence were highly correlated in early and late infection. Divergence in gag was strongly correlated with viral load, largely because of the accumulation of synonymous changes. Mutation in gag cytotoxic T-lymphocyte epitopes was associated with higher viral load. There was evidence for adaptive evolution in env, but the extent of env evolution was only weakly associated with neutralizing antibody breadth.

Conclusion: Our results indicate that HIV-1 evolution in gag and env is highly correlated but exhibits gene-specific differences. The different immune pressures on these genes may partly explain differences in evolution and consequences for HIV-1 disease progression.

Fig. 1. Correlations between evolution in gag and env during early and chronic infection

The median (and range) percentage diversity and divergence in each gene (gag and env) in early and chronic infection is shown for the 37 women in this study. Lines with horizontal arrows and P values indicate the results of comparisons of diversity and divergence between genes (Spearman’s correlation test). Within each gene, lines with vertical arrows and P values indicate the results of comparisons of diversity between time points and comparisons of diversity and divergence at the chronic infection time point (Spearman’s correlation test). In all tests, rho was at least 0.44.

Fig. 2. Associations between gag and env divergence and plasma viral load and CD4 cell count during chronic infection

Plasma viral load at the chronic infection time point (available for 33/37 individuals) is plotted against each individual’s median percentage divergence in gag (a) and env (b). CD4 cell count measured within 6 months of the chronic infection time point (available for 28/37 individuals) is plotted against each individual’s median percentage divergence in gag (c) and env (d). Each point represents the viral load and mean percentage divergence for sequences from one individual. Comparisons were performed using the Spearman’s correlation test, and results are shown at the top of each graph.

Fig. 3. Associations between viral load during chronic infection and the number and percentage of gag cytotoxic T lymphocyte epitopes changed

Plasma viral load measured at the chronic infection time point is plotted against the number of amino acid changes in gag CTL epitopes (a) and against the percentage of total published gag CTL epitopes that contain at least one amino acid change (b). Each point represents one individual. Comparisons were performed using the Spearman’s correlation test, and results are shown at the top of each graph. CTL, cytotoxic T lymphocyte.

Fig. 4. Example of neutralization assay with plasma from a representative subset of 24 women in this study

This table lists the IC₅₀, the reciprocal dilution of plasma that neutralizes 50% of virus, for each plasma sample (rows) screened against each of six panel viruses (columns). The highest plasma dilution tested was 1: 100, and samples that were not neutralized at this dilution were given the reciprocal IC₅₀ value of 50. IC₅₀ values that are higher than the assay median for each virus are highlighted in gray. For each plasma sample, the neutralizing antibody breadth score was calculated as the number of panel viruses (out of six) neutralized at a IC₅₀ higher than the assay median. This assay was repeated in triplicate for each plasma sample from 36 out of 37 individuals, and the average of three scores was used in later analyses. IC₅₀, 50% maximal inhibitory concentration; NAb, neutralizing antibody.