Structure and Recognition of a Novel HIV-1 gp120-gp41 Interface Antibody that Caused MPER Exposure through Viral Escape

Abstract

A comprehensive understanding of the regions on HIV-1 envelope trimers targeted by broadly neutralizing antibodies may contribute to rational design of an HIV-1 vaccine. We previously identified a participant in the CAPRISA cohort, CAP248, who developed trimer-specific antibodies capable of neutralizing 60% of heterologous viruses at three years post-infection. Here, we report the isolation by B cell culture of monoclonal antibody CAP248-2B, which targets a novel membrane proximal epitope including elements of gp120 and gp41. Despite low maximum inhibition plateaus, often below 50% inhibitory concentrations, the breadth of CAP248-2B significantly correlated with donor plasma. Site-directed mutagenesis, X-ray crystallography, and negative-stain electron microscopy 3D reconstructions revealed how CAP248-2B recognizes a cleavage-dependent epitope that includes the gp120 C terminus. While this epitope is distinct, it overlapped in parts of gp41 with the epitopes of broadly neutralizing antibodies PGT151, VRC34, 35O22, 3BC315, and 10E8. CAP248-2B has a conformationally variable paratope with an unusually long 19 amino acid light chain third complementarity determining region. Two phenylalanines at the loop apex were predicted by docking and mutagenesis data to interact with the viral membrane. Neutralization by CAP248-2B is not dependent on any single glycan proximal to its epitope, and low neutralization plateaus could not be completely explained by N- or O-linked glycosylation pathway inhibitors, furin co-transfection, or pre-incubation with soluble CD4. Viral escape from CAP248-2B involved a cluster of rare mutations in the gp120-gp41 cleavage sites. Simultaneous introduction of these mutations into heterologous viruses abrogated neutralization by CAP248-2B, but enhanced neutralization sensitivity to 35O22, 4E10, and 10E8 by 10-100-fold. Altogether, this study expands the region of the HIV-1 gp120-gp41 quaternary interface that is a target for broadly neutralizing antibodies and identifies a set of mutations in the gp120 C terminus that exposes the membrane-proximal external region of gp41, with potential utility in HIV vaccine design.

Comprehensive Mapping of HIV-1 Escape from a Broadly Neutralizing Antibody

“…This bnAb has been extensively studied, using both structural (Blattner et al, 2014; Lee et al, 2016) and functional approaches (Falkowska et al, 2014; Van Gils et al, 2016; Kong et al, 2016; Wibmer et al, 2017). …”

“…These sites include residues originally mapped by mutagenesis and pseudovirus-neutralization assays: the dominantly targeted 611 glycan, the 637 glycan, and residue 647 (Figure 2B, 2C) (Falkowska et al, 2014). There was also strong differential selection at fusion-peptide sites 512 and 514, which have been mapped as part of the PGT151 epitope by structural (Lee et al, 2016) and functional (Van Gils et al, 2016; Kong et al, 2016; Wibmer et al, 2017) methods (Figure 2C). Therefore, our mutational antigenic profiling identified strong selection at epitope sites that have been identified by other approaches.…”

“…For instance, such studies mostly restrict themselves to alanine mutations at surface sites, making them biased towards certain types of escape mutations. Therefore, despite the fact that PGT151 has been a subject of multiple studies using diverse techniques (Blattner et al, 2014; Falkowska et al, 2014; Van Gils et al, 2016; Kong et al, 2016; Lee et al, 2016; Wibmer et al, 2017), our mutational antigenic profiling has yielded a far more comprehensive and unbiased map of viral escape than all this previous work.…”

Surface-Matrix Screening Identifies Semi-specific Interactions that Improve Potency of a Near Pan-reactive HIV-1-Neutralizing Antibody

“…In addition to 4E10, a number of other HIV-1 broadly neutralizing antibodies have been proposed to interact with viral membrane, including antibodies 2F5 (Ofek et al, 2004), CAP248-2B (Wibmer et al, 2017), and DH511 (Williams et al, 2017) (Figure 5C). As polyreactivity, such as indicated by binding to HEp2 cells or to cardiolipin, can negatively impact the therapeutic efficacy of these antibodies, we assessed the polyreactivity of six membrane-interacting HIV-1 broadly neutralizing antibodies and of Trp-enhanced functional variants for four of the antibodies (Figure 5D; Figure S3).…”

Serum glycan-binding IgG antibodies in HIV-1 infection and during the development of broadly neutralizing responses

“…Glycan arrays have been developed to study the glycan-binding profiles of antibodies and have proven to be particularly useful in determining the types of glycan structures that are bound by HIV-1 bNAbs [10, 11, 18–25] . In addition, the anti-glycan antibody responses in sera of non-human primates after Ad5hr-SIV vaccination and SIV infection [26] and HIV vaccine responses in rabbits have been assayed on arrays [20, 27, 28] .…”