Selection pressure on HIV-1 envelope by broadly neutralizing antibodies to the conserved CD4-binding site

Abstract

The monoclonal antibody (MAb) VRC01 was isolated from a slowly progressing HIV-1-infected donor and was shown to neutralize diverse HIV-1 strains by binding to the conserved CD4 binding site (CD4bs) of gp120. To better understand the virologic factors associated with such antibody development, we characterized HIV-1 envelope (Env) variants from this donor and five other donors who developed broadly neutralizing antibodies. A total of 473 env sequences were obtained by single-genome amplification, and 100 representative env clones were expressed and tested for entry and neutralization sensitivity. While VRC01 neutralizes about 90% of the genetically diverse heterologous HIV-1 strains tested, only selective archival Env variants from the VRC01 donor were sensitive to VRC01 and all of the Env variants derived from the donor plasma were resistant, indicating strong antibody-based selection pressure. Despite their resistance to this broadly reactive MAb that partially mimics CD4, all Env variants required CD4 for entry. Three other CD4bs MAbs from the same donor were able to neutralize some VRC01 escape variants, suggesting that CD4bs antibodies continued to evolve in response to viral escape. We also observed a relatively high percentage of VRC01-resistant Env clones in the plasma of four of five additional broadly neutralizing donors, suggesting the presence of CD4bs-directed neutralizing antibodies in these donors. In total, these data indicate that the CD4bs-directed neutralizing antibodies exert ongoing selection pressure on the conserved CD4bs epitope of HIV-1 Env.

Figures

Fig 1

Phylogenetic trees of HIV-1 envelope sequences. (A) Neighbor-joining tree showing the phylogenetic clustering of Env protein sequences from six HIV-1-infected donors. A total of 473 gp160 sequences from the six subtype B-infected donors were aligned with the reference sequence HXB2. The tree was constructed based on sequence distance and unrooted and then rooted at HXB2 for visualization. The gp160 sequences were derived at a single time point for donors 18, N90, N26, and B7B5; at three time points for donor 1; and at four time points for donor 45. The donor identification is indicated at each donor's major branch node. Representative sequences indicated by a dot were cloned and expressed. A predominant sequence from donor 1 is shown in light blue from the 1995 time point (indicated by a bracket); one such sequence was cloned as indicated with an identification number. (B) Maximum-likelihood tree of envelope sequences from donor 45. A total of 186 gp160 nucleotide sequences from four temporal samples were aligned with the reference sequence HXB2. The tree was constructed as unrooted and then rooted at HXB2 for visualization. The gp160 sequences are color coded as follows: red, 2001 provirus; green, 2001 plasma; blue, 2006 plasma; purple, 2009 plasma. The major branch colors follow the color of most of the sequences on the branch. Representative sequences indicated by a dot and an identification number were cloned and expressed. The horizontal branch scale is indicated for each tree.

Fig 2

Comparisons of V1V2 lengths (left) and numbers of gp120 PNGS (right) among the subtype B sequences from the HIV-1 database (B_database), from chronic infection (B_chronic), and from the six study donors as one group or individually. The blue horizontal bars indicate the group means, and the actual mean values are indicated at the top of the columns. The number of sequences for each group or individual is indicated in parentheses below the group or donor identification. One-way ANOVA P values are indicated, with the asterisks indicating groups that are significantly different (P < 0.05) from both the B_database and B_chronic groups in subsequent Dunnett's multiple-comparison tests.

Fig 3

Analysis of neutralization sensitivity of Env variants taken from donor 45 at three time points to autologous serum IgG (left panels) and autologous CD4bs MAbs VRC01, VRC03, VRC06, and VRC06b (right panels). The horizontal bars indicate the group geometric means. The three symbols highlighted in red in the 2001 Env plots indicate the archival Env variants derived from proviral DNA.

Fig 4

Comparisons of VRC01 (A) and CD4-Ig (B) neutralization sensitivities of Env variants derived from the six study donors indicated on the x axis. The blue horizontal dashed line indicates an IC50 of 50 μg/ml, which was used as the cutoff for neutralization sensitivity. The values below the donor designations are the numbers and percentages of Env variants resistant to VRC01 or CD4-Ig neutralization. (C) Plot of log-transformed CD4-Ig neutralization IC50s for VRC01-sensitive (n = 40) and VRC01-resistant (n = 60) Env variants (left). The mean of each group is indicated by a blue horizontal bar. The log-transformed VRC01 and CD4-Ig neutralization IC50s are plotted for each individual Env variant (right, n = 100). The P values and the corresponding statistical analyses are indicated.

Fig 5

Entry of nine donor 45 Env variants derived from the 2009 plasma into CD4+ cell line Cf2Th-CD4.CCR5 (left) and CD4− cell line Cf2Th-synCCR5 (right).