Design and proof of concept for targeted phage-based COVID-19 vaccination strategies with a streamlined cold-free supply chain

Abstract

Development of effective vaccines against coronavirus disease 2019 (COVID-19) is a global imperative. Rapid immunization of the entire human population against a widespread, continually evolving, and highly pathogenic virus is an unprecedented challenge, and different vaccine approaches are being pursued. Engineered filamentous bacteriophage (phage) particles have unique potential in vaccine development due to their inherent immunogenicity, genetic plasticity, stability, cost-effectiveness for large-scale production, and proven safety profile in humans. Herein we report the development and initial evaluation of two targeted phage-based vaccination approaches against SARS-CoV-2: dual ligand peptide-targeted phage and adeno-associated virus/phage (AAVP) particles. For peptide-targeted phage, we performed structure-guided antigen design to select six solvent-exposed epitopes of the SARS-CoV-2 spike (S) protein. One of these epitopes displayed on the major capsid protein pVIII of phage induced a specific and sustained humoral response when injected in mice. These phage were further engineered to simultaneously display the peptide CAKSMGDIVC on the minor capsid protein pIII to enable their transport from the lung epithelium into the systemic circulation. Aerosolization of these "dual-display" phage into the lungs of mice generated a systemic and specific antibody response. In the second approach, targeted AAVP particles were engineered to deliver the entire S protein gene under the control of a constitutive CMV promoter. This induced tissue-specific transgene expression, stimulating a systemic S protein-specific antibody response in mice. With these proof-of-concept preclinical experiments, we show that both targeted phage- and AAVP-based particles serve as robust yet versatile platforms that can promptly yield COVID-19 vaccine prototypes for translational development.

Keywords: AAVP; COVID-19; SARS-CoV-2; gene delivery; phage display.

Fig. 1.

Representation of the phage- and AAVP-based vaccine candidates. Scheme of the approach used for the conception, design, and application of two strategies for immunization against the SARS-CoV-2 S protein using phage particles. Step 1: Structural analysis, selection of structurally defined epitopes and cloning steps for the generation of dual-display phage particles and AAVP encoding the full-length S protein. Step 2: Molecular engineering of single- and dual-display phage particles, and AAVP S constructs. Step 3: Functional validation and immunization studies in vivo in mice.

Fig. 2.

Identification of structural epitopes on the S protein trimer. (A) Six epitopes (orange) spanning the SARS-CoV-2 S protein were selected for display on the rpVIII protein. Four epitopes are located within the S1 subunit, epitope 1 (aa 336 to 361), epitope 2 (aa 379 to 391), epitope 3 (aa 480 to 488), epitope 4 (aa 662 to 671), and two within the S2 subunit, epitope 5 (aa 738 to 760), and epitope 6 (aa 1,032 to 1,043). These epitopes are solvent exposed in the surface representation of the predominantly closed-state conformation of the S protein trimer (gray, cornflower blue, and rosy brown) (PDB ID: 6ZP0) (63). Glycans (shown in purple) are present throughout the structure; of the six epitopes, only epitope 1 (aa 336 to 361) contains a site for glycosylation (at N343). Two cleavage sites located in the S2 subunit, S1/S2 (aa 685/686) and S2′ (aa 815/816), are represented by red arrows. (B) All of the epitopes (orange) maintain a cyclic conformation in the ribbon representation of a S protein protomer (gray); disulfide bridges (yellow) are present between the flanking cysteine residues of all epitopes except on epitope 2 (aa 379 to 391). The open-state conformation of the S protein trimer with one RBD erect displays a change in orientation of epitopes 1, 2, and 3, though all remain solvent exposed (PDB ID: 6ZGG) (64). NTD, N-terminal domain; RBD, receptor-binding domain; HR1 and HR2, heptapeptide repeat sequence-1 and -2; CT, C terminus.

Fig. 3.

Explicit-solvent simulations reveal the near-native conformation of epitope 4 relative to the full-length S protein. (A) rmsd from the S protein conformation for each epitope within the S1 subunit: epitope 1 (aa 336 to 361), epitope 2 (aa 379 to 391), epitope 3 (aa 480 to 488), and epitope 4 (aa 662 to 671). Epitope 4 frequently samples low values (∼2 Å), while the other epitopes undergo substantial conformational rearrangements (rmsd >4 Å). (B) Probability as a function of rmsd shows that epitope 4 is most likely to sample conformations that are similar (rmsd ∼2.5 Å) to the S protein conformation. (C) rmsd as a function of time for the epitopes within the S2 subunit: epitope 5 (aa 738 to 760) and epitope 6 (aa 1,032 to 1,043). Both epitopes rapidly adopt and maintain large rmsd values (D).

Fig. 4.

Immunogenicity of S protein epitopes on single-display and dual-phage particles. Five-week-old female Swiss Webster mice were immunized via SC administration of single-display phage constructs containing one of the six different epitopes expressed on rpVIII protein or the negative control insertless phage. Animals received a boost injection 3 wk after the first administration. (A) S protein-specific IgG antibodies and (B) phage-specific IgG antibodies were evaluated in sera of mice after 2 and 5 wk by ELISA (n = 3 mice per group). (C) Five-week-old female BALB/c mice were immunized via IT administration with epitope 4/CAKSMGDIVC dual-display phage particles, epitope 4 single-display phage particles, or the control insertless phage. Animals received a boost 3 wk after the first administration. S protein-specific IgG antibodies were evaluated weekly by ELISA (n = 10 mice per group). Data represent ± SEM (***P < 0.001).

Fig. 5.

Immunogenicity of the RGD-4C-AAVP S in mice. Schematic representation of the AAVP-based vaccine candidate. (A) The modified, synthetic SARS-CoV-2 S protein gene was excised from the pUC57 and cloned into the RGD-4C-AAVP-TNFΔEcoRI829. Expression of the S protein transgene cassette is driven by the constitutive CMV promoter and flanked by AAV ITRs. (B) Schematic representation of the RGD-4C-AAVP S and the control RGD-4C-AAVP empty vector (RGD-4C-AAVP S-null). (C) S protein-specific IgG antibody response in the sera of mice immunized with RGD-4C-AAVP S via different routes of administration (n = 5 mice per group) by ELISA. (D) S protein-specific IgG antibodies in sera of mice weekly immunized with RGD-4C-AAVP S or the control RGD-4C-AAVP S-null (n = 12 mice per group) via SC administration. Data represent ± SEM (**P < 0.01). (E) Tissue-specific expression of the S protein transgene in mice immunized with AAVP S 5 wk after the first immunization. Data represent ± SEM (***P < 0.001). (F) Phage-specific IgG antibody response in the sera of mice immunized with RGD-4C-AAVP S via different routes of administration (n = 5 mice per group). Phage-specific IgG antibody response in mice immunized with RGD-4C-AAVP S (G) or RGD-4C-AAVP S-null (H) at 2 and 5 wk after the first immunization. Phage-specific IgG antibody response was evaluated by ELISA in 96-well plates coated with 10¹⁰ AAVP particles per well. Tet^R, tetracycline resistance gene. Amp^R, ampicillin resistance gene. Ori, origin of replication.