Antibody vs. HIV in a clash of evolutionary titans

Abstract

HIV has evolved many strategies to avoid neutralizing antibody responses, particularly to conserved regions on the external glycoprotein spikes of the virus. Nevertheless, a small number of antibodies have been evolved by the human immune system to recognize conserved parts of the glycoproteins, and therefore, have broadly neutralizing cross-strain activities. These antibodies constitute important tools in the quest to design immunogens that can elicit broadly neutralizing antibodies in humans and hence contribute to an effective HIV vaccine. Crystallographic analyses of the antibodies, in many cases in an antigen-complexed form, have revealed novel and, in some instances, remarkable structural adaptations to attain virus recognition. Antibodies, like HIV, can evolve relatively rapidly through mutation and selection. It seems that the structures of these broadly neutralizing antibodies bear witness to a heroic struggle between two titans of rapid evolution.

Fig. 1.

The trimeric Env spike of HIV-1/SIV. (a) Electron micrographs of SIV particles showing trimeric Env spikes on the surface (63). This micrograph (courtesy of Ken Roux, Florida State University, Tallahassee) shows trimers on the surface of an SIV particle expressing high levels of Env. HIV-1 Env appears to be less stable than Env of SIV, and there is likely heterogeneity in the number of Env spikes per virion. (b) Model of the Env spike based on the structure of core gp120 (11, 64), with three gp120 monomers shown in gray, pale green, and pale blue. gp41 is shown schematically as three pink tubes. Carbohydrate chains are shown in yellow, and the oligomannose cluster proposed to interact with mAb 2G12 is shown in cyan. The approximate locations of the epitopes for broadly neutralizing mAbs are indicated.

Fig. 2.

The structure of mAb b12. The structure of Fab b12 as described in ref. is shown with the CDRs presented in colors and labeled as belonging to the heavy (H) or light (L) chain. The long HCDR3 (H3) is clearly visible as an extended finger-like structure. Critical aromatic residues on the CDRs for binding to gp120 (–21) are prominent.

Fig. 3.

A model that illustrates accessibility of the coreceptor site region to CD4i antibody fragments after the engagement of the HIV-1 Env spike by CD4 (33). CD4 (yellow) on the target cell membrane engages the CD4bs on gp120 molecules to assemble and expose the coreceptor binding site. In the context of virus–target cell interaction, the site now appears accessible to single-chain Fv (scFv) and Fab fragments of CD4i antibodies but not to intact antibody molecules. In contrast for free monomeric gp120, soluble CD4 binding triggers the binding of intact CD4i antibodies.

Fig. 4.

The structure of a V3-loop peptide in the binding site of the antibody 447 (38). The CDRH3 loop (pink, numbered) forms a mixed β-sheet with the V3 loop (blue). GPGR forms the turn in the peptide structure and interacts with the base of the CDRH3 loop. Main-chain interactions dominate the interaction of the peptide with the CDRH3 loop.

Fig. 5.

A model of mAb 2G12 Fab₂ bound to the HIV-1 Env spike. The heavy chains of 2G12 are shown in dark blue and light red, and the light chains are shown in azure. The domain-swapped structure of 2G12 uncomplexed and complexed with Manα1–2Man and with Man₉GlcNAc₂ is described in ref. . The gp120 oligomannose residues important in 2G12 binding were assigned based on data from a number of different approaches (44, 45). Docking of the structure of 2G12, complexed in the conventional V_H–V_L combining sites with Man₉GlcNAc₂, onto gp120 places the GlcNac₂ groups very close to N332 and N392 (outer dark red moieties). The Man₉GlcNAc₂ group attached to N339 (middle dark red) can be readily modeled to interact with the nonconventional V_H–V_H interface region.

Fig. 6.

A model of mAbs 2F5 (56) and 4E10 (59) Fabs bound to their epitopes close to the virus membrane. The Fabs are shown as a solvent-accessible surface (gray) with their Cα trace from the heavy (blue) and light (cyan) embedded in the translucent structure. The CDRH3s of the antibodies are shown in purple. The MPER model structure (yellow) is based on connecting the actual structures of peptides observed in the corresponding crystal structures of the Fab–peptide complexes. The potential for simultaneous interaction of the antibody molecules with the viral membrane is illustrated.