Identification and characterization of a broadly cross-reactive HIV-1 human monoclonal antibody that binds to both gp120 and gp41

Abstract

Identification of broadly cross-reactive HIV-1-neutralizing antibodies (bnAbs) may assist vaccine immunogen design. Here we report a novel human monoclonal antibody (mAb), designated m43, which co-targets the gp120 and gp41 subunits of the HIV-1 envelope glycoprotein (Env). M43 bound to recombinant gp140 s from various primary isolates, to membrane-associated Envs on transfected cells and HIV-1 infected cells, as well as to recombinant gp120 s and gp41 fusion intermediate structures containing N-trimer structure, but did not bind to denatured recombinant gp140 s and the CD4 binding site (CD4bs) mutant, gp120 D368R, suggesting that the m43 epitope is conformational and overlaps the CD4bs on gp120 and the N-trimer structure on gp41. M43 neutralized 34% of the HIV-1 primary isolates from different clades and all the SHIVs tested in assays based on infection of peripheral blood mononuclear cells (PBMCs) by replication-competent virus, but was less potent in cell line-based pseudovirus assays. In contrast to CD4, m43 did not induce Env conformational changes upon binding leading to exposure of the coreceptor binding site, enhanced binding of mAbs 2F5 and 4E10 specific for the membrane proximal external region (MPER) of gp41 Envs, or increased gp120 shedding. The overall modest neutralization activity of m43 is likely due to the limited binding of m43 to functional Envs which could be increased by antibody engineering if needed. M43 may represent a new class of bnAbs targeting conformational epitopes overlapping structures on both gp120 and gp41. Its novel epitope and possibly new mechanism(s) of neutralization could helpdesign improved vaccine immunogens and candidate therapeutics.

Figure 1. m43 sequence and extent of affinity maturation.

The sequence of m43 is shown along with nearest VH- and VΚ-genomic precursors for heavy and light chain V-segments, respectively. Affinity maturation changes in V-segments are indicated in green.

Figure 2. Binding of m43 and control mAbs to recombinant Envs by ELISA (A–E) and to gp41 intermediate structures by Western blot (F).

A: Binding of Fab m43 to recombinant gp140s, gp120s and gp41Fc fusion protein. B: Binding of Fabs m43 and Z13, and IgG₁ 2F5 to non-denatured (solid symbols) and denatured gp140_89.6 (empty symbols). C: Binding of Fabs m43, m44 and Z13, and mouse mAb NC-1 to recombinant gp140_89.6 in the absence (solid symbols) or presence (empty symbols) of sCD4 (2 µg/ml). D: Competition of CD4bs mAbs b12, m18 and VRC01, CD4i mAbs m16 and m9Fc fusion protein, and gp41-specific mAbs m44, m46, 2F5 and 4E10 with biotinylated IgG₁ m43 for binding to coated gp140_89.6. E: EC₅₀s of IgG₁ m43, CD4bs mAbs b12 and VRC01, CD4i mAb 17b, and V3 loop-specific mAb 39F with gp120_JRFL wild-type (WT) and V1, V2, V3 loop deletion mutants, and gp120_yu2 WT and its site mutants in CD4bs (D368R), CD4i (I420R), loop D (N279E) and V5 loop (G459E). nd: not done. F: Binding of Fabs m43 and m44 to non-reduced N35_CCG-gp41 (lane 1), N35_CCG-N13 (lane 2), N34_CCG (lane 3), gp41 core (lane 4), and 5HB (25.4 kDa) (lane 5) by Western blot.

Figure 3. Binding of IgG₁ m43 and other mAbs to functional Env trimers by flow cytometry.

A. Binding of IgG₁s m43, b12, VRC01, 2F5 and 4E10 to cleavage-competent gp160_JRFL on 293T cells. B. Binding of Abs to C8166.R5 cells five days after acute infection with the R5-tropic HIV isolates BaL or 208/K8. C. Binding of mAbs (3 µg/ml) to persistently infected H9/NL4-3 cells in the presence or absence of sCD4 (500 ng/ml). D. The ability of seven different unconjugated mAbs to alter the binding of eleven Alexa 488-labeled Abs to H9/NL4-3 cells. The results are displayed as the mean and SEM fluorescent intensity, with each graph representing one labeled Ab. The vertical line on each graph indicates the binding of the labeled Ab in the absence of unlabeled Ab. Fluorescent intensity values for unlabeled cells were 260±3, and for cells incubated with an Alexa 488-labeled isotype control 298±2.

Figure 4. Percentage neutralization of IgG₁ m43 with SHIV viruses.

Three clade C (SHIV-1157ipd3N4, SHIV-1157ipEL-p, and SHIV-2873Nip) and one clade B (SHIV-SF162) SHIV viruses were tested in a PBMC-p27 assay. IgG₁ VRC01 was used as positive control, while Fm-6 served as negative isotype control. The data shown are representative results obtained from two independent experiments.

Figure 5. Effect of sCD4 (top panel) and m43 (low panel) on binding of 17b, 2F5 and 4E10 to JRFL gp160 on 293T cells by flow cytometry.

Curves in red: 2nd Ab (PE-avidin) only; in green: 10 µg/ml biotinylated 17b, or 2F5, or 4E10; in blue: 10 µg/ml sCD4 or m43 in combination with 10 µg/ml biotinylated 17b, or 2F5, or 4E10; in brown: 50 µg/ml sCD4 or m43 in combination with 10 µg/ml biotinylated 17b, or 2F5, or 4E10.

Figure 6. Effect of m43 binding on JRFL functional Env trimer.

Western blot analysis of Env from the supernatant of 293T cells transfected with cleavage-competent gp160 JRFLdCT Env plasmid DNA. HIV-1 gp120 shedding upon binding of ligands, including sCD4, VRC01, b12, and m43, from the Env trimer on the surface of 293T cells was evaluated by probing the blot with anti-gp120 human mAb 39F (V3-specific). Each ligand concentration was 50 µg/ml.

Figure 7. Binding of m43, b12 and VRC01 to cleavage-competent (+) and cleavage-incompetent (−) JRFL and Yu2 gp160s on 293T cells by flow cytometry.

Different concentrations of mAbs used in the assay were indicated. PE conjugated to anti-human IgG, F(ab)2 were used as secondary antibody for detection.