Rapid Development of SARS-CoV-2 Spike Protein Receptor-Binding Domain Self-Assembled Nanoparticle Vaccine Candidates.

The coronavirus disease pandemic of 2019 (COVID-19) caused by the novel SARS-CoV-2 coronavirus resulted in economic losses and threatened human health worldwide. The pandemic highlights an urgent need for a stable, easily produced, and effective vaccine. SARS-CoV-2 uses the spike protein receptor-binding domain (RBD) to bind its cognate receptor, angiotensin-converting enzyme 2 (ACE2), and initiate membrane fusion. Thus, the RBD is an ideal target for vaccine development. In this study, we designed three different RBD-conjugated nanoparticle vaccine candidates, namely, RBD-Ferritin (24-mer), RBD-mi3 (60-mer), and RBD-I53-50 (120-mer), via covalent conjugation using the SpyTag-SpyCatcher system. When mice were immunized with the RBD-conjugated nanoparticles (NPs) in conjunction with the AddaVax or Sigma Adjuvant System, the resulting antisera exhibited 8- to 120-fold greater neutralizing activity against both a pseudovirus and the authentic virus than those of mice immunized with monomeric RBD. Most importantly, sera from mice immunized with RBD-conjugated NPs more efficiently blocked the binding of RBD to ACE2 in vitro, further corroborating the promising immunization effect. Additionally, the vaccine has distinct advantages in terms of a relatively simple scale-up and flexible assembly. These results illustrate that the SARS-CoV-2 RBD-conjugated nanoparticles developed in this study are a competitive vaccine candidate and that the carrier nanoparticles could be adopted as a universal platform for a future vaccine development.

The ongoing COVID-19 pandemic has become a global public health crisis affecting 216 countries or regions, with over 80 million confirmed cases, including over 1.7 million deaths as of December 29, 2020 (https://covid19.who.int). SARS-CoV-2, the causative virus of COVID-19, is a β-coronavirus in the coronavirus family. Other highly pathogenetic virus strains in this coronavirus family include the SARS-CoV, MERS-CoV.[1−3] The infection from SARS-CoV-2 is characterized by person-to-person transmission[4] and a diverse clinical presentation, ranging from mild cases to a respiratory distress syndrome that can even lead to death.[5,6]

The SARS-CoV-2 adopts a similar cell-entry mechanism to that of other coronavirus family members. The SARS-CoV-2 virus relies on a spike protein on the viral membrane for host cell recognition, attachment, and membrane fusion. The coronavirus spike protein also shares a similarity in its structural appearance as a trimeric fusion protein.[7−9] As is the case in the closely related SARS-CoV, the spike protein of SARS-CoV-2 recognizes angiotensin converting enzyme 2 (ACE2) as the cell entry receptor. The key binding interface lies on the spike protein receptor binding domain (RBD), which has been confirmed by both high-resolution cryogenic electron microscopy (cryo-EM) structures and interface mutation scanning.[10,11]

On the basis of clear structural information and knowledge of the biological function of SARS-CoV-2 spike protein with its key RBD, this protein region is an ideal target for vaccine development.[8,12,13] Despite a comprehensive effort to develop RBD-based vaccines, application of the RBD subunit as a vaccine candidate is still hindered by its low immunogenicity.[14] To increase the immunogenicity, researchers propose to modify the RBD to achieve a larger antigen-carrier complex size or multimerization. This modification would complicate the overall RBD structure and prolong the production and validation process of the recombinant modified antigen.[15,16] To shorten the time needed for vaccine development in the pandemic emergency, a more compact workflow based on antigen-displaying nanoparticles is considered a highly attractive option.

A protein covalent bond linking strategy has advanced significantly in recent years, where the development has led to easier protein modification and multimerization.[17−19] Since the development of the bacteria-derived spontaneously ligating SpyTag-SpyCatcher pair in 2012, this linking system has been optimized to improve the linking efficiency and stability.[20−22] This approach can lead to quicker protein purification[23] and macromolecular assembly[24] and is versatile for protein design projects, including vaccine development.[25] Furthermore, previous work on an HBV[26] and HIV[27] vaccine using SpyTag-SpyCatcher to link the antigen to a nanoparticle scaffold demonstrated improved antigenicity compared to the unlinked monomer.

In this study, we designed SARS-CoV-2 vaccine candidates by linking the spike protein RBD to a nanoparticle via a covalent ligation strategy. By fusing SpyTag to the C-terminus of RBD, the antigen could be covalently linked to the SpyCatcher on the nanoparticle scaffold. The RBD nanoparticles elicited significantly higher titers of neutralizing antibodies in mice than monomeric RBD, which was confirmed by stronger RBD competition with both ACE2 and neutralizing antibodies. Finally, our study identified three different nanoparticle platforms with different SpyCatcher constructs at the N-terminus for the coupling of SpyTag-fused proteins, which may be applicable as a general nanoparticle capture platform for other antigens in the future.

Results and Discussion

Design and Production of RBD-Conjugated Nanoparticles

Previous studies have demonstrated that immunization with the receptor binding domain of SARS-CoV-2 spike protein formulated with aluminum hydroxide adjuvant elicited higher titers of neutralizing antibodies in mice than the extracellular domain protein (ECD), S1-subunit protein (S1), or S2-subunit protein (S2).[28] The RBD amino acid sequences from 24 representative SARS-CoV-2 strains isolated in different countries were aligned and found to be highly conserved (Figure S1). In the present study, we focused on the RBD of SARS-CoV-2 S glycoprotein (Figure S2) to design an RBD-conjugated nanoparticle vaccine based on the SpyTag-SpyCatcher system. We used the previously developed shortened form of SpyCatcher, ΔN1-SpyCatcher.[20,25] Three ΔN1-SpyCatcher-nanoparticles (ΔN1-SpyCatcher-NPs) conjugation platforms were developed, including ΔN1-SpyCatcher-Ferritin, ΔN1-SpyCatcher-mi3, and ΔN1-SpyCatcher-I53–50, to display more antigen on the surface of nanoparticles (NPs) based on the formation of the isopeptide bond between the SpyTag peptide and ΔN1-SpyCatcher in vivo (Figures A,B and S3). Ferritin can self-assemble into a spherical 24-meric particle and form an octahedral nanocage.[29] The computationally designed and optimized mi3 NP protein based on KDPG aldolase, containing the C76A and C100A mutations to avoid potential disulfide bond-mediated heterogeneity, can self-assemble into a dodecameric cage scaffold with 60 total subunits in multiple display positions on the NP surface.[25,30] The I53–50 NPs were an icosahedral nanoparticle assembled from two components, 20 copies of trimeric I53–50A1.1PT1 and 12 copies of pentameric I53–50B.4PT1.[31]

Figure 1

Construction and structural characteristics of RBD-conjugated NPs. (A) Schematic representation of RBD-conjugated NP design. The image shows ideal NPs with a full valency of RBD. Each NP is shown in accordance with the displayed palette of the following charts. (B) Construction of target protein expression plasmids for different expression systems, namely, E. coli and HEK293F cells. (C) Reduced SDS-PAGE of the RBD monomer, RBD-conjugated NPs and nonbonded NPs. A high covalent bond linking efficiency is achieved as the blot of RBD monomer and unlinked NP scaffold disappear in the lane of RBD-conjugated NPs. (D) SEC of RBD monomer, RBD-conjugated NPs, and nonbonded NPs on a Superose 6 increase 10/300GL column. Peak forward shifts of retention are observed after ligation of RBD-SpyTag with ΔN1-SpyCatcher-NPs. (E) DLS of the RBD monomer, RBD-conjugated NPs, and nonbonded NPs. Increased hydrodynamic diameters of NPs after ligation are shown.

Previous studies have demonstrated that immunization with the RBD of SARS-CoV S protein expressed in mammalian cells could elicit more potent neutralizing antibody responses in mice, which provided complete protection following infection with SARS-CoV compared with those expressed in insect cells or Escherichia coli.[32] The RBD-SpyTag (residues 319–541) protein was expressed by the transient transfection into HEK293F cells and purified by nickel-nitrilotriacetic acid (Ni-NTA) affinity chromatography, followed by size-exclusion chromatography (SEC). As shown in Figure C,D, the purified RBD-SpyTag fusion protein was uniform and pure, demonstrated by a clear single blot in sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and a single major peak in the SEC chromatogram. As shown in Figure C,D, ΔN1-SpyCatcher-NPs, ΔN1-SpyCatcher-Ferritin NP, ΔN1-SpyCatcher-mi3 NP, and ΔN1-SpyCatcher-I53–50 NP were expressed in E. coli and purified by Ni-NTA affinity chromatography followed by SEC. The high yield and production quality of the protein was evident in the SDS-PAGE gel and SEC chromatogram. Furthermore, with the preparation of precursor proteins assembly of the RBD-conjugated nanoparticles, we purified ΔN1-SpyCatcher-Ferritin, ΔN1-SpyCatcher-mi3, and ΔN1-SpyCatcher-I53–50A1.1PT1 proteins. RBD-SpyTag was incubated at a final concentration of 50 μM with an eightfold molar excess of ΔN1-SpyCatcher-NPs overnight for the in vitro conjugation reaction and then purified by SEC to remove unreacted RBD monomers and unlinked nanoparticles. RBD-conjugated Ferritin, mi3, and I53–50 NPs were efficiently incorporated into ΔN1-SpyCatcher-NPs, as shown by the presence of a single band for each and uniform increases of their molecular weights (from ∼35 to 72 kDa) on the reducing SDS-PAGE gel (Figure C). We further confirmed this conclusion by the peak’s forward shifts in the chromatogram (Figure D). The results suggest that the RBD-SpyTag was conjugated with ΔN1-SpyCatcher-NPs at a high efficacy. To further examine the conjugation efficiency, we performed a ratio-dependent assembly and SDS-PAGE to detect the conjugation output. The results showed that the amount of remaining ΔN1-SpyCatcher-nanoparticle scaffold after conjugation was significantly reduced with the increase of added RBD-SpyTag, especially at a more than fourfold excess over the ΔN1-SpyCatcher-scaffold (Figure S4A). Overall, we used the SpyTag-SpyCatcher system to guarantee a flexible and highly efficient production of SARS-CoV-2 RBD-conjugated nanoparticles. Moreover, due to the independent expression of the RBD as the antigen and the nanoparticle scaffold, the construction and production of proteins could be achieved using different optimized expression systems.

Structural Characterization of the SARS-CoV-2 RBD-Conjugated Nanoparticles

To determine the stability of the RBD monomer and RBD-conjugated nanoparticles, we performed SDS-PAGE analysis. We stored the RBD monomer, RBD-conjugated nanoparticle scaffold (I53–50A1.PT1), and RBD-conjugated nanoparticles at 37, 25, 4, and −80 °C for two weeks to assess if there is any degradation or depolymerization. The SDS-PAGE profiles indicated that all of the tested protein preparations remained highly stable after two weeks of storage at all tested temperatures, similar to the RBD monomer and RBD nanoparticles (Figure S4B).

We next observed the structural characteristics of the RBD-conjugated nanoparticles using negative staining electron microscopy (EM). As shown in Figure A, the RBD conjugated with Ferritin, mi3, and I53–50 nanoparticles was visible on the surface of the monodispersed particles. As shown by the EM graphs, a buried exterior surface of nanoparticles could be observed among the RBD-conjugated NPs, especially for RBD-Ferritin NP. The hydrodynamic diameters of the RBD monomer, ΔN1-SpyCatcher-NPs, and RBD-conjugated NPs were further measured using dynamic light scattering (DLS). As shown in Figure B, the particle characteristics of the unconjugated NPs and RBD-conjugated NPs were examined, whereby both presented a uniform distribution of particle sizes. Moreover, consistent with the results of a negative-staining EM, the DLS analysis indicated that the hydrodynamic diameter of the RBD-conjugated nanoparticles was larger than that of the unconjugated NPs.

Figure 2

Assembly validation and physical evaluation of the nanoparticles. (A) Negative-staining EMs of unlinked nanoparticles and RBD-conjugated NPs. (B) Detailed information on DLS and nano DSF results. Rd: Hydrodynamic diameter. PDI: polydispersity index; a PDI lower than 0.2 indicates a uniform particle size. Tm1: the first melting temperature. Tm2: the second melting temperature. Tagg: the aggregation temperature. The melting and aggregation temperatures were calculated using the analysis software of the nanoDSF system.

To further explore the physical stability of the nanoparticles, nano differential scanning fluorimetry (nanoDSF) was performed on the RBD monomer, SpyCatcher-NPs, and RBD-conjugated NPs. Detailed thermostability parameters were obtained (Figure B). The similar Tm1 of RBD and the RBD-conjugated NPs indicated that the overall RBD structure was not affected by the conjugation to the nanoparticle. The higher Tm1 of mi3-NP after conjugation with RBD may be ascribed to the RBD-buried exterior surface, which may have strengthened the structural stability of the conjugation of the RBD with the nanoparticles. No aggregation was observed for the RBD monomer, RBD-Ferritin NP, and RBD-mi3 NP during the thermal denaturation process (Figure B). However, RBD-I53–50 NP underwent aggregation at ∼70 °C, which is close to the aggregation temperature of RBD-free unloaded I53–50 NP and significantly higher than the Tm1 of the RBD monomer. These results demonstrated that, under general conditions or in a conventional vaccine storage environment at 4 °C, the designed conjugated vaccine maintained a similar stability to that of the RBD monomer. The production process may be beneficial for commercial production and distribution.

Validation of the In Vitro Antigenicity of SARS-CoV-2 RBD-Conjugated Nanoparticles

We next expressed and purified recombinant human ACE2 (hACE2) ectodomain and an RBD-specific neutralizing antibody (CB6). We characterized the antigenicity of RBD-conjugated NPs by detecting their binding affinity and kinetics with the receptor and antibody. The CB6 neutralizing antibody was isolated from a convalescent COVID-19 patient. It recognized an epitope that overlaps with the hACE2-binding site of the RBD, a critical characteristic enabling it to neutralize authentic SARS-CoV-2.[33] Enzyme-linked immunosorbent assay (ELISA) profiles showed that the RBD-SpyTag monomer and the three RBD-conjugated NPs bound the hACE2 and the CB6 antibody in a dose-dependent manner. Analogously to the soluble RBD-SpyTag monomer, the three RBD-conjugated NPs bound to purified hACE2. These results suggest that the conformation of the RBD monomer was retained on the conjugated nanoparticles (Figure A). Notably, the binding affinity of the CB6 antibody for the three RBD-conjugated NPs was significantly higher than for the soluble RBD-SpyTag monomer (Figure B). Biolayer interferometry (BLI) was then applied to further examine the binding kinetics of the RBD-conjugated NPs. As illustrated in Figure B,D, the measured binding affinity constants (KD) of the RBD monomer and the three RBD-conjugated NPs, RBD-Ferritin NP, RBD-mi3 NP, and RBD-I53–50 NP, with the hACE2 receptor were 4.34 × 10–9, 1.74 × 10–8, <1.0 × 10–12, and 1.00 × 10–9 M. Because of continued binding during the dissociation phase, the dissociation rate of RBD-mi3 NP was significantly lower than that of the RBD monomer and the other two nanoparticles. The binding kinetics of RBD-conjugated NPs to the CB6 antibody were also measured. The binding of the CB6 antibody to the three RBD-conjugated NPs was significantly stronger than what was observed with the RBD monomer (Figure C,D). These results indicated that the three RBD-conjugated NPs may elicit a higher affinity to specific BCR targeting than the RBD of SARS-CoV-2. All the in vitro experiments indicated that the RBD-conjugated nanoparticles had advantages in terms of both stability and antigenicity compared with the soluble RBD monomer. Consequently, we conducted animal experiments to further verify their efficacy and immune response in BALB/c mice in vivo.

Figure 3

Antigenicity characterization of RBD monomer and RBD-conjugated NPs. (A) ELISA assay of ACE2 and CB6 antibody binding capability. Statistical analysis of binding titers between RBD monomer and the three RBD-NPs was performed using two-way ANOVA corrected using Dunnett’s test. (B, C) BLI kinetic assays of RBD monomer and RBD-NPs. (D) Detailed information on the BLI assay. KD: binding affinity constant calculated as Kon/Kdis; smaller values generally indicate a stronger binding ability. Kon: association rate. Kdis: dissociation rate.

Immunogenicity of RBD-Conjugated Nanoparticles in BALB/c Mice

We compared the immunogenicity of the three RBD-conjugated NPs and soluble monomeric RBD. Mice were immunized with 5 μg of monomeric RBD or a corresponding amount of RBD-mi3 NP, RBD-Ferritin NP, or RBD-I53–50 NP to yield an equimolar dose of RBD formulated with 50% (v/v) AddaVax or Sigma Adjuvant System (SAS) adjuvant at weeks 0, 2, and 4 (Figure A). Phosphate-buffered saline (PBS) was used for the negative control group. As expected, after the primary immunization, no binding antibody response was detected in the groups immunized with monomeric RBD with either AddaVax or SAS as the adjuvant. However, mice immunized with RBD-Ferritin, RBD-mi3, or RBD-I53–50, formulated with AddaVax adjuvant, elicited 71.8- to 168.4-fold higher binding antibody responses (reciprocal half-maximal effective concentration (EC50): 103.8±0.4, 103.9±0.2, and 104.2±0.2, respectively) than that of RBD (reciprocal EC50: 102.0). Similar to the AddaVax adjuvant, RBD-Ferritin NP, RBD-mi3 NP, and RBD-I53–50 NP formulated with SAS adjuvant induced virtually the same antigen-specific binding antibody response (reciprocal EC50: 104.1±0.3, 104.0±0.2, and 104.3±0.2, respectively) after the primary immunization. The RBD-specific antibody titers were substantially increased among the groups immunized with monomeric RBD and the three RBD-conjugated NPs with a first and second boost immunizations. The ELISA profiles showed that the three RBD-conjugated NPs induced significantly higher RBD-specific binding antibody titers compared to the monomeric RBD during all immunizations, and the results were generally similar to both adjuvants (Figure B,C). In order to explore the detailed immune response during immunization, we further evaluated the IgG subtypes of the elicited antibodies, and the results showed that a similar trend of antibody titer among different immunization groups could be observed regardless of the subtypes of IgG. Moreover, the IgG1/IgG2a ratio was greater than 1 throughout the whole immunization process, indicating a Th2-favored antibody response (Figure B,C).[34]

Figure 4

Immunogenicity of RBD monomer and RBD-conjugated nanoparticles. (A) Schematic flow diagram of the animal immunization procedure. (B) (C) ELISA of serum antibody titers of mice immunized with immunogen combined with either AddaVax (B) or SAS (C) as the adjuvant. The statistical significance of the difference between RBD monomer and RBD-NPs was calculated using two-way ANOVA corrected using Dunnett’s test, with the monomer as control group. * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001. (D) BLI serum competition assay against ACE2 or CB6 antibody of sera from mice immunized with RBD monomer or RBD-NPs in conjunction with AddaVax. Rc represents the binding signal of ACE2 or CB6 at each dilution level. Ro represents the serum-free binding signal of ACE2 or CB6. (E) Heatmap overview of the competition assay. The competition level is represented by the (Ro – Rc)/Ro ratio. A brighter color indicates stronger competition against the receptor ACE2 or the neutralizing antibody CB6 at each dilution level.

In addition to the antibody production intensity, the neutralization ability of the generated antibodies is another critical factor impacting the immunization quality. Thus, for further evaluation, the immunogenicity of RBD-conjugated NPs was studied using a serum competition assay based on BLI. The sera of mice immunized using the antigen with the adjuvant after the second boost were retrieved and mixed within each group for a general evaluation. Serially diluted sera were contacted with the RBD monomer on the biosensor. It was observed that the sera of animals immunized with the RBD-conjugated NPs more effectively blocked the binding of ACE2 and the CB6 antibody to the RBD at each dilution than those elicited by immunization with the RBD monomer (Figure D). The noncompeting binding curve height Ro and competing binding curve of each dilution level Rc were used for quantitative analysis. The competition levels of the antibodies from the animals immunized with the RBD-conjugated NPs were 4- to 16-fold higher than those elicited using monomeric RBD (Figure E). Furthermore, as the number of copies of RBD presented on the surface increased, a stronger competition could be observed when comparing the RBD-Ferritin NP with RBD-I53–50 NP and RBD-mi3 NP. A stronger relative competition level may indicate a longer occupation of the spike protein RBD on the virus.

The titers of neutralizing antibodies were determined in vitro using SARS-CoV-2 pseudovirus and authentic SARS-CoV-2 virus neutralization assays. In the SARS-CoV-2 pseudovirus-based assay, the neutralization titers of sera from animals immunized with the RBD-conjugated nanoparticles formulated with AddaVax adjuvant after the second boost were 10- to 120-fold greater than those from animals immunized with the monomeric RBD. We also observe these results when immunized with antigen formulated with the SAS adjuvant (Figure A). We studied the SARS-CoV-2 virus to detect the sera’s neutralization activity using the focus reduction test and directly observing the cytopathic effect (CPE). As shown in Figure B, 90% focus reduction neutralization antibody titers (FRNT90) of post-second boost sera from mice immunized with RBD-Ferritin NP, RBD-mi3 NP, and RBD-I53–50 NP formulated with AddaVax adjuvant (FRNT90: 104.1±0.5, 104.0±0.6, and 104.1±0.3, respectively) were 10- to 40-fold higher than in those immunized with monomeric RBD (FRNT90: 102.1±0.8). Similar to AddaVax, the three RBD-conjugated nanoparticles formulated with the SAS adjuvant also elicited a significantly higher FRNT90 than the RBD monomer (Figure B). Additionally, we also performed the CPE-based microneutralization assay to better understand the difference in neutralization ability between the groups (Figure C). After the first prime, we observed a small amount of neutralizing effect for all groups. As the immunization procedures progressed, the neutralizing effect of the sera from animals immunized with the RBD-conjugated NPs was much higher than in the monomeric RBD group with either adjuvant. A comparative serum-neutralizing activity could be observed after the first boost for the RBD-conjugated groups, while an equal strength of neutralization could not be observed until the second boost in the monomeric RBD group. Moreover, the neutralization activities of sera from the nanoparticle groups were nearly 10-fold higher than those of the monomeric RBD group. Note that RBD-Ferritin NP showed a relatively inferior effect compared to RBD-mi3 NP and RBD-I53–50 NP when we made a parallel comparison between the groups. Overall, the RBD-conjugated NPs displayed a significantly stronger immunization efficacy than the monomeric RBD regardless of adjuvant formulation. This result was in accordance with the in vitro experiments, which indicated a stronger antigenicity of the nanoparticle design. The overwhelming neutralization effect of the sera elicited by RBD-conjugated NPs, especially for the higher valency groups (RBD-mi3 NP and RBD-I53–50 NP), indicate that, with the increase of valency of RBD on the nanoparticle surface, a more favorable immunization effect could be induced.

Figure 5

Neutralization activity of sera from mice immunized with RBD monomer or RBD-conjugated nanoparticles. (A) SARS-CoV-2 pseudovirus neutralization assay showing the NT90. (B) SARS-CoV-2 live virus neutralization assay showing the FRNT90. The statistical significance of the difference of neutralizing titers of mice immunized with immunogen combined with either AddaVax or SAS as the adjuvant was calculated using the unpaired two-tailed nonparametric Mann–Whitney U test. * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001 (C) Table of SARS-CoV-2 live virus neutralizing titers determined by the induced CPE. A deeper red color represents a higher dilution ratio.

Cellular Responses to the Immunization

We explored the difference between the immune cell level post immune response to antigen and determined the T cell immune response elicited by RBD-conjugated NPs. We analyzed the germinal centers (GCs) B cells, T follicular helper (Tfh) cells, and immune cells containing intracellular cytokines. Analysis of the responding cells in draining lymph nodes following by a second boost immunization indicated that there is no significant difference in Tfh and GC cell responses among animals immunized with the three RBD-conjugated NPs compared to monomeric RBD (Figure S5A,B). Consistent with the RBD-dimer results from a previous study,[16] there were no substantial increases in the T cell responses of collected draining lymph nodes and spleen when we compared with the monomeric RBD immunized mice. This conclusion was demonstrated by a flow cytometry analysis of RBD-conjugated NPs immunized mice (Figure S5C,D and Figure S6).

Various SARS-CoV-2 vaccine candidates have been demonstrated and validated in preclinical or clinical trials.[35−38] Vaccine strategies including the live-virus platform,[39,40] viral vector vaccine using vesicular stomatitis virus[41] or adenovirus,[42] lipid nanoparticle-encapsulated mRNA,[43] or whole inactivated virus[44,45] are examined. A strategy that involves a multivalent presentation of antigen on protein nanoparticles was regarded as one of the most rapidly developing methods for vaccine design.[46,47] However, the increased structural redundancy during the de novo design of antigen presentation on the multimeric component challenges our ability to conceive a structural design versus production and utility.[48] We report the design of RBD-based nanoparticles using a covalent bond-linking strategy for designing antigen-nanoparticle vaccines.

A spike protein is the focused antigen candidate for a vaccine design, leading to the selection of different protein subsection used (full ectodomain S protein, S1 segment, and RBD) based on the costructure and complementary ACE2 receptor.[8,13,49] Despite a full display of available antigenic sites, full-length S protein bears the uncertainty of a prospective immune response due to increasing evidence of versatile mutations[50] and an unpredictable presenting efficacy of neutralizing epitope.[51] As the crystal structure of RBD in complex with ACE2 has been solved, growing attention was placed on the RBD as the primary antigen for vaccine design, and variable strategies were adopted to enhance its immunogenicity. These include dimerization, nanoparticle formulations, or the simple combined use of a conventional adjuvant.[12,13] Among all the strategies, the multivalency of antigen and the enlargement of antigen size received the most effort, because the increased antigen size and antigen saturation of BCR prolongs antigen retention and presentation, promoting the recognition of antigenic epitopes on the immunogen.[52] Multimeric RBD display can exist for the monomer, but it was necessary to introduce components to the original RBD sequence to form a commutative bond between RBDs or to add an additional scaffold to initiate multimerization. We require a skilled structure-guided modification on a corresponding antigen and iterations of ideal-to-real test production.[37,53] We explored a covalent bond-linking strategy, a comparatively more compact method during rapid development of SARS-CoV-2 vaccine, and confirmed the viability by a comprehensive evaluation.

Conclusions

In this study, we designed and tested three RBD-conjugated nanoparticles using a potentially universal strategy for a rapid and general vaccine design. The covalent bond linking would be performed easily by an incubation of RBD-SpyTag and ΔN1-SpyCatcher-NPs, yielding fully multivalent RBD-conjugated NPs with high structural uniformity and stability and little sacrifice in assembling efficiency. Both monomer RBD and RBD-conjugated NPs underwent further antigenicity inspection, and results showed that multivalent RBD-conjugated NPs exhibited stronger affinity to receptor ACE2 and neutralizing antibody CB6. These results might indicate a favorable BCR affinity to elicit a higher titer of antibody. We immunized BALB/c mice with monomer RBD and RBD-conjugated NPs adjuvanted with AddaVax or SAS. Serum anti-RBD antibody titers of RBD-conjugated NPs were significantly higher than those of monomer RBD regardless of adjuvant used. A neutralizing assay of pseudovirus or authentic SARS-CoV-2 virus proved that elicited neutralizing antibody titers of RBD-conjugated NPs were higher than those of the monomer, demonstrated by the competition assay of immunized mice sera to ACE2 or CB6 antibody. Moreover, the potential universality of the nanoparticle-conjugation platform and competitive immunization efficacy will, hopefully, enable the rapid manufacturing and broad application of next-generation vaccines.

Methods

Cells and Viruses

Vero-E6 (clone E6) and Vero kidney epithelial cells from African green monkeys, as well as the HEK293T human embryonic kidney epithelial cells, were purchased from ATCC. The HEK293F cells were purchased from Life Technologies and maintained in Union 293 medium (Union-Biotech) at 37 °C and 120 rpm, in a humidified atmosphere comprising 5% CO2. The HEK293T cells stably expressing human angiotensin-converting enzyme 2 (HEK293T-hACE2) were deposited in our lab. All adherent cells were grown in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin at 37 °C, in a humidified atmosphere comprising 5% CO2. All cell lines were confirmed to be free of mycoplasma contamination. Both SARS-CoV-2 strains used in this study were isolated from COVID-19 patients in Guangzhou (GenBank: MT123290 and GISAID: EPI_ISL_413859).

Mice

Specific pathogen-free (SPF) six- to eight-week-old female BALB/c mice were obtained from Beijing Vital River Laboratory Animal Technology Co., Ltd. All animal experiments were approved by the ethics committee of Sun Yat-sen University Cancer Center (approval no. L102042020000A).

Gene Synthesis and Plasmid Construction

The SARS-CoV-2 RBD construct (residues 319–541, GenBank: MN908947) for preparation of the RBD-based nanoparticle vaccine were codon-optimized for human cells and synthesized by GenScript and further cloned into the mammalian expression vector VRC8400 with an N-terminal Kozak consensus sequence, signal peptide for protein secretion, and a C-terminal octa-histidine tag followed by a 13-residues SpyTag[20] using the BamHI restriction sites. The SARS-CoV-2 RBD construct for the ELISA was basically the same as above but without the 13-residue SpyTag. The sequence encoding hACE2 (residues 19–615, GenBank: NM_021804.2) was synthesized by GenScript and cloned into VRC8400 with an N-terminal Kozak consensus sequence and signal peptide, as well as a C-terminal octa-histidine tag.

The IgG heavy and light chain variable genes of CB6 mAb (GenBank: MT470196 and MT470197) were codon-optimized for human cells, synthesized by GenScript, and cloned into the antibody expression vector.

The ΔN1-SpyCatcher-mi3 construct (GenBank: MH425515) with the C76A and C100A mutations based on pentameric I3–01(60-mer)[25,30] was codon-optimized for E. coli, synthesized by GenScript, and cloned into a modified pET28a+ vector with an N-terminal hexahistidine tag. The ΔN1-SpyCatcher sequence was N-terminally fused to trimeric Ferritin (24-mer)[29] or trimeric I53–50A1.1PT1 (60-mer)[31] using a (GGS)4 spacer, codon-optimized for E. coli, synthesized by GenScript, and then cloned into a modified pET28a+ vector with an N-terminal hexahistidine tag or a C-terminal hexahistidine tag, respectively.

The I53–50B.4PT1 sequence was codon-optimized, synthesized by GenScript, and then cloned into a modified pET28a+ vector with a C-terminal hexahistidine tag.

Protein Expression and Purification in HEK293F Cells

Plasmid DNA was extracted from E. coli DH5a using the NucleoBond Xtra Maxi kit, according to the manufacturer’s protocol. The SARS-CoV-2 RBD monomer and hACE2 proteins were produced in HEK293F cells. The HEK293F cells were cultured in Union 293 medium at 37 °C, 80–90% humidity, and 5% CO2, with rotation at 120 rpm for expansion. Then, the cells were transiently transfected with 1 mg of expression plasmid per liter using poly(ethylenimine) (PEI) MAX (Polysciences) reagent at a density of 1.0 × 106 cells/ml in fresh Union 293 medium. After 5 d of culture, the cells were removed by centrifugation at 8000g and 4 °C for 1 h, and the supernatant was filtered using a Steritop cartridge (0.22 μm pore size; Guangzhou Jet Bio-Filtration Co., Ltd.) and concentrated to one-tenth volume using a tangential flow filtration system (10 kDa molecular weight cutoff, Millipore). Then, the concentrated supernatant was purified by immobilized metal-affinity chromatography with Ni-NTA resin (Roche) in a WET FRED gravity flow column (IBA). The target protein was eluted with buffer comprising 50 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), pH 7.3, 300 mM imidazole, and 300 mM NaCl. The eluate was concentrated and further purified via size-exclusion chromatography using a Superose 6 Increase 10/300 GL gel filtration column (GE Healthcare) in a buffer comprising 50 mM HEPES, pH 7.3, and 300 mM NaCl. Fractions of the target peak were pooled and concentrated using a centrifugal cartridge (10 kDa cutoff, Millipore) and stored at 4 °C for further use.

The CB6 antibody was expressed and purified as previously reported.[33] Briefly, HEK293F cells were transiently transfected with the plasmids encoding the IgG heavy and light chain genes at a ratio of 5:6. The culture supernatant was collected 5 d after transfection, centrifuged, purified by affinity chromatography with protein A resin (GenScript), and buffer-exchanged to 50 mM NaPO4, 150 mM NaCl, pH 7.3 using a HiTrap Desalting column (GE Healthcare) on an ÄKTA pure chromatography system (GE Healthcare).

Protein Expression and Purification in E. Coli

The modified pET28a+ expression plasmids encoding the ΔN1-SpyCatcher-mi3, ΔN1-SpyCatcher-Ferritin, ΔN1-SpyCatcher-I53–50A1.1PT1, and I53–50B.4PT1 fusion proteins were individually introduced into Rosetta (DE3) competent cells (TIANGEN). After overnight incubation at 37 °C on terrific broth (TB) agar plates supplemented with 50 μg/mL kanamycin and 33 μg/mL chloramphenicol, a single positive colony was transferred into 10 mL of TB medium with 50 μg/mL kanamycin and 33 μg/mL chloramphenicol and grown overnight at 37 °C with shaking at 220 rpm. The resulting seed culture was added to a 3 L baffled conical shake flask containing 1 L of TB medium with 50 μg/mL kanamycin and grown at 37 °C with shaking at 150 rpm. When the OD600 value of the culture reached 0.6–0.8, isopropyl thiogalactoside (IPTG) was added to a final concentration of 1 mM for induction at 20 °C for 16–20 h with shaking at 150 rpm. The bacterial cultures were harvested by centrifugation at 2450g and 20 °C for 15 min.

For the purification of ΔN1-SpyCatcher-mi3, cell pellets were resuspended in a lysis buffer (250 mM Tris, pH 8.5, 300 mM NaCl, 30 mM imidazole, 50 μg/mL deoxyribonuclease (DNase), 0.75% 3-((3-cholamidopropyl) dimethylammonio)-1-propanesulfonate (CHAPS), 5 mM MgCl2, and ethylenediaminetetraacetic acid (EDTA)-free protease inhibitor cocktail [Roche]) and lysed using a high-pressure cell homogenizer (Union-Biotech) at a pressure of 800 MPa. The cell debris was removed by centrifugation at 40 000g for 1 h; the supernatant was filtered using a Steritop cartridge (0.22 μm pore size) and loaded onto a gravity flow column containing Ni-NTA resin. The target protein was eluted with 50 mM HEPES, pH 8.0, 300 mM NaCl, 300 mM imidazole, and 0.75% CHAPS. The eluate was concentrated to 1 mL for SEC on a Superose 6 Increase 10/300 GL gel filtration column (GE Healthcare) pre-equilibrated with 50 mM HEPES, pH 8.0, and 300 mM NaCl. Peak fractions were analyzed using SDS-PAGE to determine whether the target protein was collected. After a concrete confirmation of purity and yield of protein, the fractions were pooled, concentrated, and stored at 4 °C.

For the purification of ΔN1-SpyCatcher-Ferritin and ΔN1-SpyCatcher-I53–50A1.1PT1 fusion proteins, cell pellets were resuspended in a lysis buffer (50 mM HEPES, pH 7.3, 300 mM NaCl, 30 mM imidazole, 1 mM dithiothreitol (DTT), 0.75% CHAPS, 50 μg/mL DNase, 5 mM MgCl2, and EDTA-free protease inhibitor cocktail [Roche]). Purification was analogously performed, with a modified elution buffer used (50 mM HEPES, pH 7.3, 300 mM NaCl, 300 mM imidazole, 1 mM DTT, and 0.75% CHAPS) and an equilibration buffer for SEC (50 mM HEPES, pH 7.3, 300 mM NaCl).

Endotoxin was removed from all purified protein preparations using the ToxinEraser Endotoxin Removal Kit (GenScript) according to the manufacturer’s instructions. Residual endotoxin was quantified using the ToxinSensor Chromogenic LAL Endotoxin Assay Kit (GenScript), and no more than 0.1 EU/mL of endotoxin was detected.

Preparation of RBD-Conjugated Nanoparticles

To assemble the RBD-conjugated nanoparticles, RBD-SpyTag was incubated with ΔN1-SpyCatcher-NPs to form a covalent peptide bond. Purified ΔN1-SpyCatcher-Ferritin or ΔN1-SpyCatcher-I53–50A1.1PT1 protein at a subunit concentration of 50 μM was incubated with a 1:8 molar excess of RBD-SpyTag overnight at room temperature in 50 mM HEPES, pH 7.3, 300 mM NaCl. For ΔN1-SpyCatcher-mi3, 50 μM subunit was mixed with a 1:8 molar excess of RBD-SpyTag overnight at room temperature in 50 mM HEPES, pH 7.5, 300 mM NaCl, with 5% glycerol. In order to separate the conjugated nanoparticles from the empty NPs and monomeric RBD, all preparations were purified using SEC on a Superose 6 Increase 10/300 GL gel filtration column pre-equilibrated with 50 mM HEPES, pH 7.3, and 300 mM NaCl. Fractions were collected for SDS-PAGE analysis. The protein of interest was selected, concentrated, and stored at 4 °C.

SDS-PAGE Analysis

SDS-PAGE was performed as described previously.[54] Briefly, purified protein was combined with an appropriate amount of 5× loading buffer, heated at 95 °C for 5 min, and separated on a 12% polyacrylamide Tris-glycine gel for 30 min at 300 V. The gels were stained with Coomassie brilliant blue (Beyotime), destained with 30% methanol, 10% glacial acetic acid in double-distilled water, until a clear background appeared, and recorded using the ChemiDoc system (BioRad).

Dynamic Light Scattering

DLS analysis was performed to characterize the hydrodynamic diameter and polydispersity index of RBD-conjugated nanoparticles and individual nanoparticles using a Zetasizer Ultra instrument (Malvern PANalytical). Briefly, purified proteins were centrifuged at 16 250g and 4 °C for 10 min to remove any aggregates, diluted to a concentration of 0.5 mg/mL in PBS, and loaded into the disposable solvent-resistant microcuvette. The particle distribution of the purified protein was determined by measuring the intensity of the light scattered by the sample using the avalanche photodiode detector placed at a measurement angle of 173° at 25 °C. Each sample was measured in triplicate, and the average values of the hydrodynamic diameter and polydispersity index of the sample were recorded and analyzed using the manufacturer’s software (Malvern PANalytical).

Negative Staining Electron Microscopy

Approximately 5 μL aliquots of purified nanoparticles at a concentration of 0.05–2 mg/mL were applied to freshly glow-discharged 300-mesh copper grids and incubated for 1 min. Excess liquid was blotted with filter paper. The grids were washed twice with double-distilled water, blotted, negatively stained with freshly prepared 2% (w/v) uranyl acetate for 45 s, and air-dried. The grids were imaged using an FEI Tecnai T12 transmission electron microscope (FEI) operating at 120 kV. The digital micrographs were obtained at 150 000× magnification.

Nano Differential Scanning Fluorimetry (NanoDSF)

The NanoDSF System was used to measure the thermostability and aggregation of purified protein using a Prometheus NT.48 instrument (NanoTemper Technologies). A 10 μL aliquot of sample diluted to a concentration of 0.5 mg/mL was applied to the quartz capillary cassette and placed into the card slot. The scan temperature was increased linearly from 20.0 to 95 °C at a scan rate of 1 °C/min. The thermal transition midpoint (Tm) and aggregation results from start to finish (Tagg) were reported and analyzed in PR.ThermControl software (NanoTemper Technologies). Three replicates were measured for each sample.

Enzyme-Linked Immunosorbent Assay (ELISA)

ELISA was performed to examine the binding ability of the purified SARS-CoV-2 RBD protein to the receptor ACE2 and the RBD-specific CB6 antibody. Purified RBD monomer and RBD-conjugated NPs were diluted to a concentration of 1 μg/mL to coat 96-well microplates (Corning) (100 μL/well) in triplicate overnight at 4 °C. The plates were washed three times with 0.05% Tween 20 in PBS (PBS-T) and blocked with ELISA blocking buffer (2% gelatin, 5% casein, and 0.1% ProClin 300 in PBS) for 1 h at 37 °C. The plates were then incubated with sequentially 1:5 diluted ACE2 starting from 50 μg/mL in ELISA blocking buffer. After 1 h of incubation, the plates were washed with PBS-T five times, and 100 μL of hACE2-specific rabbit antibody (Sino Biological Inc.) diluted at a ratio of 1:5000 was added to the wells and incubated at 37 °C for 1 h. The plates were washed with PBS-T five times, and a 1:5000 dilution of Horseradish Peroxidas (HRP)-conjugated goat antirabbit IgG antibody (Promega) was added and incubated for 45 min at 37 °C. After the final round of washing to remove all unspecifically bound antibodies, the substrate 3,3′,5,5′-tetramethylbenzidine (TMB, Sangon Biotech) was added to initiate the chromogenic reaction. After 15 min at room temperature, the reaction was stopped by adding 50 μL of 2 M H2SO4 to each well, after which the absorbance at 450 and 630 nm was immediately recorded using a SpectraMax Plus plate reader (Molecular Devices).

For the CB6 antibody binding assay, the coating, incubation, and chromogenic reaction were the same, but the primary antibody binding to the antigen was not required, and HRP-conjugated goat antihuman IgG antibody (Promega) was used as the secondary antibody to detect the binding of the CB6 antibody to the antigens.

Biolayer Interferometry (BLI)

A BLI analysis was performed on an Octet Red 96 instrument (ForteBio) at 30 °C with shaking at 1000 rpm. Signals were collected at the default frequency of 5.0 Hz.

Kinetic Assay

First, the Streptavidin (SA) biosensors (ForteBio) were preincubated in PBS (ThermoFisher) containing 0.05% Tween 20 (Sigma-Aldrich), the assay buffer used throughout the whole procedure, for 15 min. To couple the RBD protein to the biosensors, the EZ-link-Sulfo-NHS-biotin biotinylation kit (ThermoFisher) was used to biotinylate the RBD/ACE2/CB6 antibody as follows:

Step1: Calculation of the amount of biotinylation agent used.

Calculate the millimoles of biotin reagent to add to the reaction for a 20-fold molar excess.

Calculate the microliters of 10 mM biotin solution to add to the reaction.

Step2: Biotinylation.

Add the calculated amount of 10 mM biotin solution to the RBD/ACE2/CB6 protein in PBS (5 mg/mL, 200 μL) and incubate the reaction system at room temperature for 30 min.

Step3: Desalting.

Equilibrate the PD-10 desalting column (GE Pharmacia) with 10 mL of PBS. After the equilibration, the reaction solution is added to the column and washed with 400 μL of PBS.

To perform the kinetic assay, after 60 s of baseline recording, ACE/CB6-biotin protein diluted with the buffer was captured on the SA sensor at 2 μg/mL for 120 s. Then, a twofold lower molar concentration of the RBD monomer or different RBD-conjugated NPs was contacted with the biosensor for 180 s, followed by a 300 s disassociation and three rounds of regeneration with 10 mM glycine-HCl, pH 1.5. The curves were analyzed using ForteBio data analysis software. Raw curves aligned at association were adjusted with baseline signals before a 1:1 binding model fitting was performed. Then, a global fit to all binding curves was conducted to calculate the overall kinetic parameters (KD, Kon, Kdis, etc.).

Serum Competition Assay

Sera of the mice immunized using the AddaVax adjuvant were harvested after the second boost, and an equal volume (5 μL) of serum from each mouse within the same immunization group was mixed together to obtain a mix representative of the overall characteristics of the group. To perform the competition assay, biotinylated RBD protein was captured on the biosensor as described above at 5 μg/mL. Then, a twofold dilution of the mice sera mixtures from each group and control PBST were loaded to the biosensor to saturate the RBD for 300 s. After the end of serum loading, 400 nM ACE2 or CB6 antibody was associated with the biosensors for 300 s to detect the competitive binding signal under the saturation of each dilution level of mouse serum. The sensors were then regenerated with 10 mM Glycine-HCl, pH 1.5. Real-time signal data were collected, and the competition behavior was recorded as the ACE2/CB6 binding signal of different curves. Binding signal data were retrieved from the curve to calculate the Ro, representig the saturated noncompeting binding curve height, and Rc, representing the saturated competing binding curves of each dilution level. Relative competition levels of each serum dilution were calculated as (Ro – Rc)/Rc.

SARS-CoV-2 Pseudovirus Production

The SARS-CoV-2 pseudovirus was produced as described previously,[55] with minor modifications. Briefly, the gene encoding SARS-CoV-2 S protein (GenBank: QHU36824.1) with a 19-amino-acid deletion at the C-terminus was codon-optimized for human cells and cloned into the expression vector pCMV14–3 × Flag, a generous gift from Zhaohui Qian, Chinese Academy of Medical Sciences. HEK293T cells were cotransfected with the plasmids PsPAX2 and pCMV14-SARS-CoV-2 S ΔCT-3 × Flag, as well as the luciferase reporter plasmid pLenti-GFP at a ratio of 1:2:1 using PEI-MAX (mentioned above). After 5 h, the supernatant was replaced with fresh DMEM supplemented with 10% FBS. At 60 h post-transfection, the supernatant containing SARS-CoV-2 pseudovirus was harvested, centrifuged at 3000g and 4 °C for 10 min to remove cell debris, and filtered using a Steritop cartridge (0.45 μm pore size, Guangzhou Jet Bio-Filtration Co., Ltd.). Sterile PEG-8000 solution was added to the clarified supernatants and incubated for 1 h at 4 °C. The mixture was then centrifuged, concentrated, and resuspended in DMEM for a final collection of the viral particles, which were stored at −80 °C.

Immunization of BALB/c Mice

Forty-five 6–8 weeks-old female BALB/c mice were purchased from the Beijing Vital River Laboratory Animal Technology Co., Ltd., and randomly divided into nine groups. Before immunization, purified immunogen was diluted with PBS and gently mixed with an equal volume of AddaVax adjuvant (InvivoGen) or Sigma Adjuvant System (SAS, Sigma) and incubated overnight to achieve the full adsorption of antigen onto the surface of adjuvant particles at 4 °C with shaking at 40 rpm. Each group of mice received three immunizations two weeks apart through the subcutaneous route. The immunization dose was 5 μg of RBD monomer, or corresponding amounts of RBD-conjugated nanoparticle immunogens containing an equimolar ratio of RBD, as follows: RBD-mi3 (9.51 μg), RBD-Ferritin (9.34 μg), and RBD-I53–50 (11.91 μg). PBS was used as a negative control. Blood samples were harvested 10 d following each immunization and placed at 37 °C for 30 min for complete coagulation. Then, the blood samples were centrifuged at 16 250g and 4 °C for 10 min, and the upper serum layer was gently removed, incubated at 56 °C for 30 min to deactivate the complement factors and pathogens, and then stored at −20 °C until further use.

Serum ELISA

A solution comprising 1 μg/mL RBD monomer without SpyTag in PBS was precoated onto 96-well MaxiSorp ELISA microplates (Corning) overnight at 4 °C. Then, the plates were blocked with ELISA blocking buffer for 1 h at 37 °C. Mouse serum samples were fivefold sequentially diluted with blocking buffer starting at 1:100, added to the coated plates, and incubated at 37 °C for 1 h. After incubation, the plates were washed with PBS-T five times, after which 100 μL of a 1:5000 dilution of HRP-conjugated goat antimouse IgG antibody (Promega) in blocking buffer was added and incubated at 37 °C for 45 min. To analyze RBD-specific IgG isotype titers, a 1:5000 dilution of HRP-conjugated goat antimouse IgG1 and IgG2a antibody was incubated at this step. The plates were washed, developed with TMB, and quenched with H2SO4. The absorbance at 450 and 630 nm of each well was immediately recorded using a SpectraMax Plus plate reader (Molecular Devices). The results were plotted and fitted using GraphPad Prism 8, and the reciprocal EC50 values were calculated using four-parameter nonlinear regression.

Pseudovirus Neutralization Assay

The pseudovirus neutralization assay was performed to evaluate the neutralization ability of the sera of the immunized mice. Briefly, ∼1.75 × 104 HEK293T-hACE2 cells were seeded into 96-well culture plates and incubated overnight at 37 °C in a humidified atmosphere comprising 5% CO2. The sera were fourfold serially diluted with complete DMEM medium starting at 1:20 and preincubated with an equal volume of SARS-CoV-2 pseudovirus at 37 °C for 2 h. The mixtures were then added to the HEK293T-hACE2 cells for infection. After 2 h of incubation, the mixtures were replaced with fresh DMEM medium containing 2% FBS and 1% penicillin and streptomycin and incubated for 48 h at 37 °C, 5% CO2. The wells treated with only medium or virus without serum were included as negative and positive controls in each plate, respectively. Afterward, the cells were lysed with lysis buffer, and luciferase activity was immediately measured using the Dual-Glo luciferase assay system (Promega). The inhibition rate (%) was calculated as {1 – (serum signal – blank control signal)/(virus signal – blank control signal)} × 100. The 90% neutralization antibody titers (NT90) were determined using four-parameter nonlinear regression in GraphPad Prism 8.

Authentic SARS-CoV-2 Virus Neutralization Assay

All microneutralization assays with authentic SARS-CoV-2 virus were performed in a BSL-3 facility. Two methods, the authentic SARS-CoV-2 virus-induced cytopathic effect and focus reduction neutralization test, were used to evaluate the neutralizing antibody titers of the sera of immunized mice. For the CPE method, sera were fourfold serially diluted starting at 1:4 with DMEM supplemented with 2% FBS and 1% penicillin and streptomycin, mixed with equal volumes of 100 tissue culture half-infective doses (TCID50) of the SARS-CoV-2 strain 2020XN4276, and incubated at 37 °C for 2 h. Afterward, the serum-virus mixture was added to preplated Vero-E6 cells in 96-well culture plates and incubated for an additional 96 h at 37 °C in a humidified atmosphere comprising 5% CO2 to observe the CPE at 40× magnification. Wells with pure virus, pure diluted serum, or cells only were included as controls in each plate. Virus back-titration was performed in each plate. All diluted serum samples were tested in duplicate. The neutralization antibody titers of all sera were defined as the reciprocal of serum dilution that could neutralize 50% of virus infection at 4 d postinfection.

For the FRNT method, serum samples were fivefold serially diluted starting at 1:10, mixed with equal volumes comprising 100 focus forming units (FFU) of the SARS-CoV-2 strain CHN/IQTC01/2020 in 96-well culture plates, and incubated for 1 h at 37 °C. The mixtures were then added to 96-well plates preseeded with Vero-E6 cells. After incubation for 1 h at 37 °C, 5% CO2, the mixtures were removed and replaced with 100 μL of Minimum Essential Medium (MEM) containing 1.2% carboxymethylcellulose prewarmed to 37 °C and cultured for an additional 24 h. Then, the cells were fixed with 4% paraformaldehyde, permeabilized with 0.2% Triton X-100 in PBS, and incubated with rabbit anti-SARS-CoV-2 nucleocapsid protein antibody (Sino Biological, Inc.) for 1 h at room temperature, followed by a 1:4000 dilution of HRP-conjugated goat antirabbit IgG secondary antibody (Jackson ImmunoResearch Laboratories, Inc.). The signals were developed using TrueBlue Peroxidase Substrate (KPL). Foci were counted using an ELISPOT reader (Cellular Technology Ltd.). The 90% neutralization antibody titers (NT90) were defined as the reciprocal of serum dilution that could inhibit 90% FFU of virus-infection, and it was calculated using four-parameter nonlinear regression in GraphPad Prism 8.

Germinal Center and T Follicular Helper Analysis of Mouse Draining Lymph Nodes

Immunized mice were sacrificed by CO2 inhalation 12 d after the second boost immunization (40 d). To simultaneously identify RBD-specific germinal center B cells and Tfh cells in the draining lymph nodes of the mice, cell suspensions from the draining lymph nodes of the mice were stained with a fixable viability stain 780 and treated with anti-CD16/32 antibody to block the Fc receptor, followed by labeling with anti-B220-BV421, anti-IgD-PE, anti-GL7-Alexa Fluor 647, anti-CD95-FITC, anti-CD4-BV510, anti-CD44-BV786, anti-ICOS-PE-Cyanine7, anti-CXCR5-PE-CF594, and anti-PD-1-APC-R700 antibodies in PBS in the presence of 2% BSA (stain and antibodies from BD Biosciences). The fluorescence signals of the labeled samples were recorded on a CytoFLEX S flow cytometry system (Beckman Coulter).

Intracellular Cytokine Staining

The draining lymph nodes and spleens of the immunized mice were harvested and washed with RPMI 1640 medium. Then, the tissues were abraded into cell suspensions in culture medium (RPMI 1640 containing 10% FBS and 1% antibiotics) using the piston handle of a 2 mL syringe. The suspension was filtered through a 40 μm nylon mesh cell strainer (Sangon Biotech). The cells were washed with culture medium, and sterile erythrocyte lysis buffer was added (150 mM NH4Cl, 10 mM NaHCO3, 1 mM EDTA in deionized water, pH 7.4) to remove red blood cells, followed by staining with fixable viability stain 780 for 30 min at room temperature. Approximately 1.0 × 106 cells were added to the 6-wells plates and then stimulated with 15 μg/mL purified RBD monomer for 3 h at 37 °C. After incubation, GolgiStop and GolgiPlug (BD Biosciences) reagents were added to each well and incubated for an additional 15 h at 37 °C. Next, the cells were harvested and washed twice with culture medium, treated with anti-CD16/32 antibody (BD Biosciences) to block the Fc receptor, and then labeled with anti-CD3e-PerCP-Cy5.5 (BD Biosciences), anti-CD4-BV510 (BD Biosciences), and anti-CD8a FITC (BD Biosciences) antibodies in PBS in the presence of 2% BSA, after which the cells were further fixed with 4% paraformaldehyde and permeabilized with permeabilization buffer (2% BSA, 0.1% saponin, 0.05% Na3N in PBS). Finally, the cells were washed with PBS containing 2% BSA and incubated with anti-IFN-γ-PE-CY7 (BD Biosciences), anti-IL-2-APC (BD Biosciences), anti-TNF-α-PE (BD Biosciences), and control anti-IgG1 antibodies for 30 min at 4 °C. The fluorescence signal of the labeled samples was acquired using a CytoFLEX S flow cytometry system (Beckman Coulter).

Sequence Alignment and Analysis

All sequences except for the sequence of Wuhan-Hu-1, which was first identified from a COVID-19 patient in Wuhan city[5] and which was obtained from the National Center for Biotechnology Information (NCBI) database, were obtained from the Global Initiative on Sharing All Influenza Data (GISAID). The accession numbers of the RBD sequences of representative SARS CoV-2 strains isolated in different countries are as follows: Wuhan-Hu-1 (GenBank: MN908947), South_Korea/KCDC2489/2020 (EPI_ISL_514892), Thailand/NIH-2492/2020 (EPI_ISL_430841), Japan/Hu_DP_Kng_19–027/2020 (EPI_ISL_412969), India/OR-RMRC25/2020 (EPI_ISL_455308), USA/WA-UW-1762/2020 (EPI_ISL_424245), Mexico/CMX-IMSS_01/2020 (EPI_ISL_424731), Canada/ON_PHL2294/2020 (EPI_ISL_418384), Australia/VIC546/2020 (EPI_ISL_426809), Greece/218_35009/2020 (EPI_ISL_437886), Greece/218_35009/2020 (EPI_ISL_437886), Greece/218_35009/2020 (EPI_ISL_437886), Greece/218_35009/2020 (EPI_ISL_437886), Russia/SCPM-O-08/2020 (EPI_ISL_451970), Spain/Madrid_LP24_5999/2020 (EPI_ISL_428680), Sweden/20–50261/2020 (EPI_ISL_469078), Switzerland/ZH-1000477102/2020 (EPI_ISL_413019), Portugal/PT0533/2020 (EPI_ISL_454257), Scotland/CVR138/2020 (EPI_ISL_425681), Denmark/SSI-101/2020 (EPI_ISL_415646), England/20134020004/2020 (EPI_ISL_423108) and Iceland/348/2020 (EPI_ISL_424372), Nigeria/OS085-CV14/2020 (EPI_ISL_455424), and Venezuela/VEN-95072/2020 (EPI_ISL_476704).

Quantification and Statistical Analysis

The kinetic parameters of Biolayer interferometry were rendered using Octet Data Analysis software (ForteBio), and detailed curve fitting methods can be found in the materials and methods. Statistical analyses of all experimental results were performed using GraphPad Prism 8.01 software. Except for the results of flow cytometry, which were expressed as a percentage of positive cells, all other results were presented as means ± standard error of measure (SEM). The methods used to assess the statistical significance of differences between groups are indicated in the figure legends or corresponding method descriptions.