Design of SARS-CoV-2 RBD immunogens to focus immune responses toward conserved coronavirus epitopes

Abstract

SARS-CoV-2 continues to evolve, with new variants emerging that evade pre-existing immunity and limit the efficacy of existing vaccines. One approach toward developing superior, variant-proof vaccines is to engineer immunogens that preferentially elicit antibodies with broad cross-reactivity against SARS-CoV-2 and its variants by targeting conserved epitopes on spike. The inner and outer faces of the receptor binding domain (RBD) are two such conserved regions targeted by antibodies that recognize diverse human and animal coronaviruses. To promote the elicitation of such antibodies by vaccination, we engineered "resurfaced" RBD immunogens that contained mutations at exposed RBD residues outside the target epitopes. In the context of pre-existing immunity, these vaccine candidates aim to disfavor the elicitation of strain-specific antibodies against the immunodominant receptor binding motif (RBM) while boosting the induction of inner and outer face antibodies. The engineered resurfaced RBD immunogens were stable, lacked binding to monoclonal antibodies with limited breadth, and maintained strong interactions with target broadly neutralizing antibodies. When used as vaccines, they limited humoral responses against the RBM as intended. Multimerization on nanoparticles further increased the immunogenicity of the resurfaced RBD immunogens, thus supporting resurfacing as a promising immunogen design approach to rationally shift natural immune responses to develop more protective vaccines.IMPORTANCESARS-CoV-2 is the third coronavirus to cause significant human disease over the last two decades. Despite their success in preventing serious disease, current SARS-CoV-2 vaccines must be updated regularly to match the circulating strains for continued protection. Therefore, it would be advantageous to develop vaccines that protect more broadly against SARS-CoV-2, its variants, and other pre-emergent coronaviruses. This may be achieved by preferentially eliciting antibodies against conserved regions of the spike protein that decorates the virus. Toward this goal, we engineered vaccine candidates to target the conserved inner and outer domains of the Receptor Binding Domain of SARS-CoV-2, by altering the surface of the wild-type protein such that strain-specific antibodies that bind outside these regions are no longer recognized. When used in animals with pre-existing SARS-CoV-2 immunity, these molecules reduce the elicitation of variant-specific antibodies, thus providing a blueprint to alter the natural immunodominance hierarchies of SARS-CoV-2 proteins.

Keywords: RBD; SARS-CoV-2 vaccines; immunogen design; vaccine design.

Fig 1

Conservation and antigenicity of the RBD across SARS-CoV-2 VoCs and other betacoronaviruses. (a) SARS-CoV-2 spike protein domain organization. NTD: N-terminal domain; RBD: receptor-binding domain; RBM: receptor-binding motif; FCS: furin cleavage site; S2’: S2’ protease cleavage site; FP: fusion peptide; HR1 and HR2: heptad repeats 1 and 2; TM/CT, transmembrane/cytoplasmic tail. (b) Mapping of sarbecovirus amino acid sequence conservation on the RBD structure in complex with ACE2 (gray cartoon), with the binding footprints of the inner face antibody S2X259 (pink) and the outer face antibody S309 (orange) outlined. Conservation values represent trident conservation scores. (c) Binding site conservation scores across the individual RBD amino acids involved in interactions with S309 mAb (orange), S2X259 mAb (pink), and ACE2 (gray), respectively. Group mean values are shown in black. ^nsP = not significant, ^***P ≤ 0.001, ^**P ≤ 0.01, as calculated with Wilcoxon rank sum test. (d) Structural mapping of a representative set of RBM antibodies: H4 and B38 (cyan/green), inner face mAb S2X259 (pink), and outer face mAb S309 (orange) onto the SARS-CoV-2 RBD. (e) ELISA binding of diverse CoVs spikes to antibodies that target the RBM, outer, and inner face regions of the RBD. Values are calculated as the logarithm of the area under the ELISA binding curve.

Fig 2

Design and in vitro characterization of resurfaced RBD immunogens. (a) Top view of wild-type RBD and the three engineered resurfaced RBD immunogens, with mutated sites in orange and green. ACE2 binding residues are shown in gray. (b) Side view of RBD in complex with S309 (left) and S2X259 (right) antibodies. Residues modified in Resurf6 are shown in orange. ACE2 binding residues are shown in gray. (c) Sequence alignment between WA-1 RBD and the three resurfaced RBD immunogens, with modified residues delineated in boxes. Residues binding to S309 mAb, S2X259, and ACE2 are colored orange, pink, and gray, respectively. (d) Binding of WT and resurfaced RBDs to mAbs targeting diverse epitopes on RBD. Values represent the logarithm of the area under the curve. (e) Binding affinities of resurfaced immunogens to RBM, outer and inner face mAbs. (f) Melting temperature curves of WA-1 RBD (teal) and the resurfaced immunogens Resurf6 (pink), Resurf61 (orange), and Resurf64 (blue). (g) Binding of human sera from subjects naïve (blue) or with existing SARS-CoV-2 immunity acquired by vaccination (red), infection (green), or vaccination followed by infection (purple) to WT and engineered RBDs. Data are calculated as the logarithm of the area under the ELISA binding curve and are presented as the mean ± standard deviation. (h) The same data as in (g) are plotted for each subject. The cohort average is shown in black. For graphs (g and h), n = 12 independent subjects. Significance was tested using the Wilcoxon signed-rank test. ^***P ≤ 0.001 and ^*P ≤ 0.05.

Fig 3

In vivo characterization of resurfaced RBD immunogens. (a) Study design to determine the immunogenicity of monomeric resurfaced immunogens in BALB/c mice. (b) Binding of terminal sera from animals immunized with different RBD immunogens against WA-1 spike, WA-1 RBD, and the resurfaced immunogens. Data are calculated as the logarithm of the area under the ELISA binding curve and are presented as the mean ± standard deviation. (c) Binding of terminal sera from animals primed with mRNA and boosted with different RBD immunogens against diverse CoV spikes. D614G, Gamma, Delta, BA.1.1, BA.4/BA.5, BQ.1.1, XBB, XBB.1.5, XBB.1.16, and EG.5.1 refer to the respective SARS-CoV-2 variants. CoV-1 = SARS-CoV-1; SHC014 = BatCoV RsSHC014; RaTG13 = BatCoV RaTG13. Data are calculated as the logarithm of the area under the ELISA binding curve and are presented as the mean ± standard deviation. (d) Live virus neutralization of the sera from (c). (e) Blocking of terminal sera from animal groups in (c) to RBM (DH1041, DH1042, and DH1284), outer (LY-CoV1404 and S309), and inner (DH1047) face targeting mAbs. Data are presented as the mean ± standard deviation. For graphs (b–e), all groups n = 8 represent independent mice. Significance was tested using the Wilcoxon signed-rank test. ^***P ≤ 0.001, ^**P ≤ 0.01, ^*P ≤ 0.05.

Fig 4

Immunogenicity of resurfaced RBD immunogens multimerized on nanoparticles. (a) Development of mi03 nanoparticles (NP) displaying multimerized resurfaced RBD immunogens. (b) Study design to determine the immunogenicity of multimerized NPs displaying resurfaced immunogens in BALB/c mice. (c) Binding of terminal sera from animals primed with D614G mRNA spike and boosted with the RBD immunogens to diverse CoV spikes. D614G, Gamma, Delta, BA.1.1, BA.4/BA.5, BQ.1.1, XBB, XBB.1.5, XBB.1.16, and EG.5.1 refer to the SARS-CoV-2 variants. CoV-1 = SARS-CoV-1; SHC014 = BatCoV RsSHC014; RaTG13 = BatCoV RaTG13. Data are calculated as the logarithm of the area under the ELISA binding curve and are presented as the mean ± standard deviation. (d) Neutralization of terminal sera from groups in (c) to different CoV live viruses. Average group neutralization titers against each live virus. Level of detection (LOD) in red. Error bars show ± standard deviation. (e) Blocking of the terminal sera from immunization groups in (c) to RBM (DH1041, DH1042, DH1284, and B38), outer (sp1-77), and inner (DH1047) face mAbs. Data are presented as the mean ± standard deviation. For graphs (c–e), all groups n = 8 represent independent mice, except for mRNA + Resurf6-mi03 + Resurf61-mi03 (n = 7) due to a mouse death prior to terminal bleeds. Significance was tested using the Wilcoxon signed-rank test. ^***P ≤ 0.001, ^**P ≤ 0.01, ^*P ≤ 0.05.