VelcroVax: a "Bolt-On" Vaccine Platform for Glycoprotein Display

Abstract

Having varied approaches to the design and manufacture of vaccines is critical in being able to respond to worldwide needs and newly emerging pathogens. Virus-like particles (VLPs) form the basis of two of the most successful licensed vaccines (against hepatitis B virus [HBV] and human papillomavirus). They are produced by recombinant expression of viral structural proteins, which assemble into immunogenic nanoparticles. VLPs can be modified to present unrelated antigens, and here we describe a universal "bolt-on" platform (termed VelcroVax) where the capturing VLP and the target antigen are produced separately. We utilize a modified HBV core (HBcAg) VLP with surface expression of a high-affinity binding sequence (Affimer) directed against a SUMO tag and use this to capture SUMO-tagged gp1 glycoprotein from the arenavirus Junín virus (JUNV). Using this model system, we have solved the first high-resolution structures of VelcroVax VLPs and shown that the VelcroVax-JUNV gp1 complex induces superior humoral immune responses compared to the noncomplexed viral protein. We propose that this system could be modified to present a range of antigens and therefore form the foundation of future rapid-response vaccination strategies. IMPORTANCE The hepatitis B core protein (HBc) forms noninfectious virus-like particles, which can be modified to present a capturing molecule, allowing suitably tagged antigens to be bound on their surface. This system can be adapted and provides the foundation for a universal "bolt-on" vaccine platform (termed VelcroVax) that can be easily and rapidly modified to generate nanoparticle vaccine candidates.

Keywords: HBcAg; Junín virus; VLP; platform; vaccine.

FIG 1

Generation of HBcAg and VelcroVax in Pichia pastoris. (A) X-ray crystal structure of a HBcAg dimer (PDB: 1QGT; reference 29). The locations of the C-terminal end of monomer 1 and N-terminal end of monomer 2 and the major immunodominant regions (MIR) are indicated. (B) Organization of HBcAg, tandem HBcAg, and VelcroVax constructs with amino acid positions indicated. Representation depicts MIR, arginine-rich repeat (RRR), glycine-serine linking sequence (GS) and the insertion site of a SUMO Affimer within the MIR of the first of the fused HBcAg monomers. (C) Anti-HBcAg Western blot of gradient purified HBcAg and VelcroVax particles produced in Pichia pastoris, probed with 10E11, is shown; n = 3.

FIG 2

Structural characterization of VelcroVax VLPs. (A and B) Full and sectional isosurface representations of density maps for T = 4 (A) and T = 3* (B) VelcroVax VLPs, filtered by local resolution, shown at the same contour level and colored according to the same radial coloring scheme. In each case, an expanded view of an individual asymmetric unit (T = 4–I1 symmetry; T = 3*–C5 symmetry) and corresponding atomic models are shown. For the T = 4 asymmetric unit, the VelcroVax atomic model (green) is overlaid with the cryoEM structure of wt HBcAg (gray, PDB: 7OD4; reference 30).

FIG 3

Affimer density in low-pass filtered VelcroVax VLP reconstructions. (A and B) Sections of local resolution-filtered density maps for (A) T = 4 and (B) T = 3* VelcroVax VLPs following application of a 10-Å low-pass filter. Amorphous Affimer density (orange highlight) is visible above VelcroVax four-helix bundles. (C) Atomic model for a single VelcroVax monomer (green) with a SUMO-Affimer homology model (orange) manually positioned above the four-helix bundle, indicating the expected position of the Affimer based on the density shown in panels A and B.

FIG 4

Generation of JUNV gp1 and interaction with VelcroVax. SUMO-tagged JUNV gp1 was produced in HEK293T cells and partially purified before processing through a final round of SEC. (A) Representative SEC elution profile for recombinantly derived JUNV gp1. (B) Reducing Coomassie-stained SDS-PAGE of SEC-purified JUNV gp1 with pertinent molecular mass standard sizes indicated in kDa. (C) ELISA was used to assess binding of HBcAg or VelcroVax to SUMO-tagged JUNV gp1. Particles coated on plates were subsequently incubated with JUNV gp1 and probed with anti-JUNV gp1 clone OD01-AA09, followed by incubation with anti-mouse HRP. Plates were incubated with OPD, and the OD was read at 492 nm and graphed as mean ± SEM; n = 3 in duplicate.

FIG 5

Reactive antibody titers and isotypes. (A) Antisera generated by immunization of mice with HBcAg and JUNV gp1 or VelcroVax and JUNV gp1 were assessed for endpoint titers with HBcAg, VelcroVax, and JUNV gp1. Sera were assessed at dilutions between 1:250 and 1:8,000 (HBcAg and VelcroVax) (n = 7 in duplicate) or 1:250 to 1:4,000 (JUNV gp1) (n = 7 twice in duplicate) and graphed as mean and SEM. Dotted line indicates the limit of detection (1:250). Titers were considered positive if the mean OD₄₉₂ was ≥0.2 (approximately 2× the OD of preimmune serum). (B) Sera were subsequently assessed for isotype-specific reactivity with HBcAg, VelcroVax, and JUNV gp1. Sera were assessed at 1:125 dilution and graphed as mean and SEM; n = 7 in duplicate.