Sensitivity to Broadly Neutralizing Antibodies of Recently Transmitted HIV-1 Clade CRF02_AG Viruses with a Focus on Evolution over Time

Abstract

Broadly neutralizing antibodies (bnAbs) are promising agents for prevention and/or treatment of HIV-1 infection. However, the diversity among HIV-1 envelope (Env) glycoproteins impacts bnAb potency and breadth. Neutralization data on the CRF02_AG clade are scarce although it is highly prevalent in West Africa and Europe. We assessed the sensitivity to bnAbs of a panel of 33 early transmitted CRF02_AG viruses over a 15-year period of the French epidemic (1997 to 2012). Env pseudotyped CRF02_AG viruses were best neutralized by the CD4 binding site (CD4bs)-directed bnAbs (VRC01, 3BNC117, NIH45-46^G54W, and N6) and the gp41 membrane-proximal external region (MPER)-directed bnAb 10E8 in terms of both potency and breadth. We observed a higher resistance to bnAbs targeting the V1V2-glycan region (PG9 and PGT145) and the V3-glycan region (PGT121 and 10-1074). Combinations were required to achieve full coverage across this subtype. We observed increased resistance to bnAbs targeting the CD4bs linked to the diversification of CRF02_AG Env over the course of the epidemic, a phenomenon which was previously described for subtypes B and C. These data on the sensitivity to bnAbs of CRF02_AG viruses, including only recently transmitted viruses, will inform future passive immunization studies. Considering the drift of the HIV-1 species toward higher resistance to neutralizing antibodies, it appears necessary to keep updating existing panels for evaluation of future vaccine and passive immunization studies.IMPORTANCE Major progress occurred during the last decade leading to the isolation of human monoclonal antibodies, termed broadly neutralizing antibodies (bnAbs) due to their capacity to neutralize various strains of HIV-1. Several clinical trials are under way in order to evaluate their efficacy in preventive or therapeutic strategies. However, no single bnAb is active against 100% of strains. It is important to gather data on the sensitivity to neutralizing antibodies of all genotypes, especially those more widespread in regions where the prevalence of HIV-1 infection is high. Here, we assembled a large panel of clade CRF02_AG viruses, the most frequent genotype circulating in West Africa and the second most frequent found in several European countries. We evaluated their sensitivities to bnAbs, including those most advanced in clinical trials, and looked for the best combinations. In addition, we observed a trend toward increased resistance to bnAbs over the course of the epidemic.

Keywords: evolution; human immunodeficiency virus; neutralizing antibodies.

FIG 1

Phylogenetic analysis of the 33 CRF02_AG pseudotyped viruses included in the panel. env sequences were aligned with 141 CRF02_AG env full-length sequences from the Los Alamos database. A maximum-likelihood phylogenetic tree was constructed after exclusion of hypervariable regions. Sequences are labeled based on their geographic origins (Africa in yellow, Europe in blue, North America in red, and Asia in green). The identification numbers of each variant included in our panel and of those included in previous CRF02_AG panels are indicated.

FIG 2

Potency of bnAbs against the panel of early/acute CRF02_AG pseudotyped viruses. The heat map represents experimental IC₅₀ values. Virus isolates are represented on rows, and antibodies are represented on columns. IC₅₀ values are indicated according to the color legend on the figure, with white cells indicating concentrations above the experimental threshold of 10 μg/ml. The geometric mean IC₅₀ of each bnAb against the whole panel is indicated at the top.

FIG 3

Neutralization breadth and potency of bnAbs against the panel of early/acute CRF02_AG pseudotyped viruses. (A) Potency-breadth curves against the panel of 33 CRF02_AG pseudotyped viruses. The y axis shows the cumulative frequency of IC₅₀ values up to the concentration shown on the x axis. (B) Percentages of viruses neutralized for each bnAb at less than 1 μg/ml.

FIG 4

Major determinants for sensitivity to bnAbs targeting V1V2 glycan (V1V2g) (A), V3 glycan (V3g) (B), and CD4bs (C) among CRF02_AG Env sequences. An alignment of the Env protein sequences of the variants used in the study is depicted, focusing on key determinants associated with sensitivity to bnAbs targeting V1V2g (N156, N160, and the lysine-rich region KKQK), V3g (N332), and CD4bs (loop D, CD4 binding loop, and V5). For each targeted region, the sequences in the alignment were ordered according to IC₅₀ values of PG9, 10-1074, and VRC01, respectively. For the indicated regions, relevant N-glycosylation sites, characterized by the sequon NX(S/T), are highlighted in blue; loss of the N-glycosylation site and relevant mutations are highlighted in red. The red box marks the highly variable region in V5. The IC₅₀ values are shown according to the legend on the figure.

FIG 5

Role of the glycan at position 332 in neutralization susceptibility of CRF02_AG viruses. A comparison of susceptibilities according to the presence/absence of N332 glycan (N332+/N332−). Box plots show the distribution of IC₅₀s of each bnAb; the horizontal lines represent the 10th, median, and 90^th percentiles. Significance was tested by nonparametric two-sided Mann-Whitney tests.

FIG 6

Phylogenetic clustering does not match sensitivity to neutralization of CRF02_AG Env pseudoviruses. A hierarchically clustered dendrogram based on IC₅₀ values for each bnAb is represented vertically. The dendrogram was computed from squared Euclidean distance values using Ward’s clustering method. A maximum likelihood phylogenetic tree constructed after exclusion of hypervariable regions is represented horizontally. Each virus of the panel is placed on the grid formed by the two classifications and labeled according to the gender, ethnicity, and risk group of the individual from whom the virus was isolated. Square, MSM; circle, heterosexual male; triangle, heterosexual female; hexagon, male with unknown sex orientation; open symbols, Caucasian; black symbols: Black; gray symbols, unknown origin. The blue bracket indicates the 13 viruses belonging to the same monophyletic group.

FIG 7

Diversification of the CRF02_AG variants, over a period spanning 1997 to 2012, is associated with increased resistance to CD4bs-directed bnAbs. (A) bnAb susceptibility of CRF02_AG viruses from two distinct periods, 1997 to 2002 (n = 10) and 2008 to 2012 (n = 23), tested against 10 bnAbs. Box plots show the distribution of IC₅₀s of each bnAb; the horizontal lines represent the 10th, median, and 90^th percentiles. Significance was tested by nonparametric one-sided Mann-Whitney tests. (B) Genetic distance from a consensus CRF02_AG sequence correlates with the calendar year and increases over time. Sequences were codon aligned with the CRF02_AG consensus sequence (Los Alamos database). After exclusion of the hypervariable region, evolutionary divergence was estimated by distance matrix using MEGA software (Pearson correlation). (C, D, and E) IC₅₀ values of 3BNC117, NIH45-46^G54W, and VRC01 were positively correlated with distance, indicating an increase in resistance (Spearman correlation).