A VLP vaccine platform comprising the core protein of hepatitis B virus with N-terminal antigen capture

Abstract

Nanoparticle presentation systems offer the potential to develop new vaccines rapidly in response to emerging diseases, a public health need that has become increasingly evident in the wake of the COVID-19 pandemic. Previously, we reported a nanoparticle scaffold system termed VelcroVax. This was constructed by insertion of a high affinity SUMO binding protein (Affimer), able to recognise a SUMO peptide tag, into the major immunodominant region of VLPs assembled from a tandem (fused dimer) form of hepatitis B virus (HBV) core protein (HBc). Here we describe an alternative form, termed N-VelcroVax, a VLP vaccine platform assembled from a monomeric HBc protein (N-anti-SUMO Affimer HBc 190) with the Affimer inserted at the N-terminus. In contrast to the tandem form of VelcroVax, N-VelcroVax VLPs were expressed well in E. coli. The VLPs effectively bound SUMO-tagged Junín virus glycoprotein, gp1 as assessed by structural and serological analyses. Cryo-EM characterisation of N-VelcroVax complexed with a SUMO-Junín gp1 showed continuous density attributable to the fused Affimer, in addition to evidence of target antigen capture. Collectively, these data suggest that N-VelcroVax has potential as a versatile next generation vaccine scaffold.

Keywords: ClearColi; HBcAg; Junín virus; Platform; Vaccine; Virus-like particle.

Fig. 1. Characterisation of N-VelcroVax.

(A) X-ray crystal structure of HBc monomer indicating the N-terminus, C-terminus and major immunodominant region (MIR) (PDB: 1QGT; ref. [48]). (B) wt HBc 190 construct (contains no anti-SUMO-Affimer) and N-VelcroVax, indicating the N-terminal insertion of the anti-SUMO-Affimer. (C) X-ray crystal structure of an Affimer selected against human SUMO protein (PDB: 5ELJ). Loop 1 and Loop 2 are the variable regions [49]. (D) Small-scale expression of HBc VLPs: Western Blot of wt HBc 190 and N-VelcroVax expressed in ClearColi BL21 (DE3) E. coli cells detected with mouse monoclonal anti-HBc (10E11). The figure is a representative example of three separate experiments.

Fig. 2. Purification and characterisation of N-VelcroVax VLPs.

(A) Coomassie blue staining (upper panel) and Western blot (lower panel) of gradient purified N-VelcroVax particles, expressed in ClearColi BL21 (DE3) E. coli cells detected with mouse monoclonal anti-HBc (10E11). (B) Negative stain TEM analysis of N-VelcroVax. The VLPs were stained using 2 % uranyl acetate. The scale bar shows 200 nm. The figure is a representative example of three separate experiments.

Fig. 3

Evaluating the binding of N-VelcroVax to SUMO-Junín gp1. N-Velcro-Vax VLPs and SUMO-Junín gp1 were mixed and incubated overnight before separation on a sucrose density gradient. Gradient fractions were analysed by Western blot using either mouse monoclonal anti-HBc (10E11, A) or anti-Junín gp1 (NR 2567, B). The figure is a representative example of three separate experiments.

Fig. 4

Evaluation of the interaction between N-VelcroVax and SUMO-Junín gp1 by ELISA. N-VelcroVax VLPs and SUMO-Junín gp1 were mixed in molecular ratios ranging from 1:1 to 1:5 and binding determined by ELISA. Anti-HBc 10E11 (1:1000) and anti-Junín gp1 NR 2567 (1:32,000) were used to detect HBc VLPs and SUMO-Junín gp1 respectively. Anti-mouse antibody conjugated with HRP was added as secondary antibody. TMB chromogenic substrate was used to detect HRP. The optical density at 450 nm (OD 450 nm) is represented in arbitrary units (n = 3).

Fig. 5

Characterisation of N-VelcroVax VLPs by cryo-EM. Density maps for T = 4 (left) and T = 3 (right) N-VelcroVax reconstructions, filtered according to local resolution. Each density map is shown as a complete isosurface representation and central cross-section at both low (approx. 2σ, Affimer densities visible) and high (approx. 4σ, Affimer densities not visible) contour level. Cross-sectional views show density, potentially nucleic acid bound non-specifically to the interior surface of the VLPs. Maps are coloured by radial distance (Å).

Fig. 6. Structural characterisation of N-VelcroVax VLPs in complex with SUMO-gp1.

(A) Representative micrographs from negative stain TEM analysis of N-Vel-croVax alone (left), SUMO-gp1 alone (centre), or N-VelcroVax:SUMO-gp1 following on-grid interaction (right). Scale bars show 200 nm. (B) Isosurface representations of cryo-EM density maps for the T = 4 configuration of unliganded N-VelcroVax VLP (left) and N-VelcroVax:SUMO-gp1 (right), shown at ∼2 σ and coloured radially. Affimers show a subtle change in relative positioning, as indicated by the white arrow. (C) Asymmetric reconstruction of particles from a single focussed class following focussed classification of N-VelcroVax:SUMO-gp1 T = 4 particles, showing additional low-resolution density corresponding to bound SUMO-gp1. All focussed classes are shown in Fig. S3. (D) Density from an asymmetric reconstruction of T = 4 N-VelcroVax:SUMO-gp1 following focussed classification (purple) is suggestive of a reorientation of the Affimer by approximately 30° compared to its position in the unliganded N-VelcroVax density map (grey), as indicated by the dashed arrow. Fitted atomic models (orange – Affimers, green – remainder of N-VelcroVax) are shown for illustrative purposes.