Model Membranes Study the Lipid-Reactivity of HIV-1 Antibodies and Vaccine Antigen

Abstract

One promising HIV-1 vaccine target is the membrane-proximal external region (MPER) of viral gp41. MPER is poorly immunogenic, however, the two rare neutralizing antibodies (NAbs), 2F5 and 4E10, bind to MPER with great neutralizing ability. Although their neutralizing mechanism represents a promising framework for the design of new HIV-1 liposomal vaccine candidates, this mechanism remains poorly understood. It is known that 2F5 and 4E10 are required to first associate with HIV-1 lipids before binding to the target MPER antigen, however, little is known about how lipid membranes contribute to NAb-antigen binding. To this end we have developed model membrane systems to study NAb and antigen lipid interactions. We first created a surface plasmon resonance (SPR) spectroscopy based assay that monitors antibody binding to thiol monolayers, which mimic the surface chemical properties of lipid membranes. Next, we focused on mimicking the lipid phase organization (i.e., domain formation) of native membranes by using supported lipid bilayers (SLBs). We used simple SLB compositions to model the liquid-disordered (Ld) and gel phases. To model the HIV-1 envelope, we used a complex SLB composition that contains an Ld and liquid-ordered (Lo) phase. To reliably create model HIV-1 SLBs, we developed an SLB formation technique that uses amphipathic, α-helical peptides as a catalyst to generate complex SLBs that have a high cholesterol content and contain multiple lipid types. For all SLB surfaces we used atomic force microscopy (AFM) to visualize membrane domains, antigen presentation, and antibody-membrane interactions.Results from experiments using thiol surfaces showed that NAb binding to hydrophobic thiol surfaces was significantly greater than that of control monoclonal antibodies. This supports the hypothesis that these NAbs embed into the hydrophobic membrane core. Our results demonstrate that 2F5/4E10 do not interact with the highly ordered gel and Lo domains in the SLB but exclusively bind to the Ld phase. This suggests that 2F5/4E10 require low membrane order and weak lateral lipid-lipid interactions to insert into the hydrophobic membrane interior. Thus, vaccine liposomes that primarily contain an Ld phase are more likely to elicit the production of lipid reactive, 2F5- and 4E10-like antibodies, compared to liposomes that contain an Lo or gel phase. In the context of liposomal antigen presentation, our results show that the presence of the MPER656 antigen can severely limit the Ld area available for antibody interactions. Subsequently, this reduces the amount of MPER656 that is accessible for 2F5/4E10 binding, since MPER656 preferentially localizes to the Ld area. If Ld forming lipid components are used in vaccine liposomes, it is important to ensure that the presence of antigen does not inhibit large-scale Ld formation.