Broadly neutralizing HIV antibodies define a glycan-dependent epitope on the prefusion conformation of gp41 on cleaved envelope trimers

Abstract

Broadly neutralizing HIV antibodies are much sought after (a) to guide vaccine design, both as templates and as indicators of the authenticity of vaccine candidates, (b) to assist in structural studies, and (c) to serve as potential therapeutics. However, the number of targets on the viral envelope spike for such antibodies has been limited. Here, we describe a set of human monoclonal antibodies that define what is, to the best of our knowledge, a previously undefined target on HIV Env. The antibodies recognize a glycan-dependent epitope on the prefusion conformation of gp41 and unambiguously distinguish cleaved from uncleaved Env trimers, an important property given increasing evidence that cleavage is required for vaccine candidates that seek to mimic the functional HIV envelope spike. The availability of this set of antibodies expands the number of vaccine targets on HIV and provides reagents to characterize the native envelope spike.

Figure 1. PGT151–158 neutralization of an indicator panel of 5 pseudoviruses

Serial dilutions of antibody were preincubated with pseudovirus for 1 hour and then added to TZM-bl cells. Three days post-infection luciferase values were measured and % neutralization was calculated. Data is one representative experiment of at least two replicate experiments and is presented as mean of two replicate wells +/− SEM.

Figure 2. PGT151 and PGT152 binding to different forms of Env

(A) PGT151 and PGT152 do not bind to JR-FL gp120 monomer as measured by ELISA. (B) PGT151 binds to native, cleaved (cl) but not to uncleaved (uncl) JR-FL E168K trimers expressed on the surface of 293T cells as measured by flow cytometry, and (C) PGT151 and PGT152 bind to cleaved soluble BG505 SOSIP.664 gp140 trimers but not to the corresponding uncleaved gp140 trimers as measured by ELISA. See also Figure S2. Data for (A) and (C) is presented as mean of two replicates +/− SEM. Data for (B) is a representative experiment of two replicate experiments.

Figure 3. PGT151-PGT158 bind complex carbohydrates

(A) PGT151 and PGT152 neutralization of JR-CSF pseudovirus generated (i) in the presence of the glycosidase inhibitor swainsonine, which prevents the formation of complex glycans by inhibiting the trimming of mannose residues from the Man-α6 arm of the GlcNAcMan₅GlcNAc₂ structure or (ii) in 293S cells (GNT1^−/− cells), a cell line that is deficient in N-acetylglucosaminyltransferase I and is unable to add a GlcNAc residue to the Man₅GlcNAc₂ structure to permit processing to complex glycans. Neutralization of JR-SCF pseudovirus by 2G12 is not affected by either (i) or (ii) since it binds exclusively to high-mannose glycans on the glycan shield of Env. Data is presented as mean +/− SEM. (B) Glycan microarray analysis reveals PGT151-PGT158 preferentially bind tetraantennary complex carbohydrates and PGT151 additionally binds a triantennary glycan as detected on the Wong glycan array. Data is presented as mean +/− SD. (C) Glycan microarray analysis using the CFG microarray additionally reveals that all PGT151 family MAbs except PGT153 preferentially bind triantennary complex carbohydrates; and (D) PGT151 and PGT152 bind a tetrantennary complex carbohydrate on the neoglycolipid microarray. The symbols for common monosaccharides are as follows: purple diamonds represent sialic acid, yellow circles represent galactose, red triangles represent fucose, blue squares represent N-acetyl glucosamine and green circles represent mannose. See also Tables S4–S5. Data for the CFG microarray is presented as mean +/− SEM.

Figure 4. PGT151 neutralization of viruses containing N611A, N637A, E647A single and double alanine substitutions

Serial dilutions of antibody (A) PGT151 and (B) PGV04, used as a control, were preincubated with pseudovirus for 1 hour and then added to TZM-bl cells. Three days post-infection luciferase values were measured and % neutralization was calculated. See also Figure S4. Data are from a representative experiment of at least two replicate experiments and are presented as mean values of two replicate wells +/− SEM.

Figure 5. ADCC by PGT151 and PGT152

(A) PGT151 and (B) PGT152 were titered for ADCC activity against target cells infected with HIV-1 NL4–3, YU2, JR-CSF and SIVmac239, and using an NK cell line (KHYG-1) expressing CD16 as the effector cell. The killing of virus-infected cells by ADCC is indicated by a loss of relative light units (RLU). SIVmac239 is used as a negative control and HIVIG is used as a positive control. Data are presented as mean values of three replicates +/− SEM.