key: cord-0937972-hjrnev1g
authors: Wu, Yangtao; Huang, Xiaofen; Yuan, Lunzhi; Wang, Shaojuan; Zhang, Yali; Xiong, Hualong; Chen, Rirong; Ma, Jian; Qi, Ruoyao; Nie, Meifeng; Xu, Jingjing; Zhang, Zhigang; Chen, Liqiang; Wei, Min; Zhou, Ming; Cai, Minping; Shi, Yang; Zhang, Liang; Yu, Huan; Hong, Junping; Wang, Zikang; Hong, Yunda; Yue, Mingxi; Li, Zonglin; Chen, Dabing; Zheng, Qingbing; Li, Shaowei; Chen, Yixin; Cheng, Tong; Zhang, Jun; Zhang, Tianying; Zhu, Huachen; Zhao, Qinjian; Yuan, Quan; Guan, Yi; Xia, Ningshao
title: Sterilizing immunity against SARS-CoV-2 in hamsters conferred by a novel recombinant subunit vaccine
date: 2020-12-20
journal: bioRxiv
DOI: 10.1101/2020.12.18.423552
sha: 22820eb44ad7004b408a527a338216abfe678a15
doc_id: 937972
cord_uid: hjrnev1g

A safe and effective SARS-CoV-2 vaccine is essential to avert the on-going COVID-19 pandemic. Here, we developed a subunit vaccine, which is comprised of CHO-expressed spike ectodomain protein (StriFK) and nitrogen bisphosphonates-modified zinc-aluminum hybrid adjuvant (FH002C). This vaccine candidate rapidly elicited the robust humoral response, Th1/Th2 balanced helper CD4 T cell and CD8 T cell immune response in animal models. In mice, hamsters, and non-human primates, 2-shot and 3-shot immunization of StriFK-FH002C generated 28- to 38-fold and 47- to 269-fold higher neutralizing antibody titers than the human COVID-19 convalescent plasmas, respectively. More importantly, the StriFK-FH002C immunization conferred sterilizing immunity to prevent SARS-CoV-2 infection and transmission, which also protected animals from virus-induced weight loss, COVID-19-like symptoms, and pneumonia in hamsters. Vaccine-induced neutralizing and cell-based receptor-blocking antibody titers correlated well with protective efficacy in hamsters, suggesting vaccine-elicited protection is immune-associated. The StriFK-FH002C provided a promising SARS-CoV-2 vaccine candidate for further clinical evaluation.

The first coronavirus pandemic in human history, caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), is changing the whole landscape of global public health. To date, 218 countries and regions around the world have confirmed SARS-CoV-2 infections, with over 60 million confirmed COVID-19 cases and costing nearly 1.5 million people's lives, and the number of human infections is still rocketing up. Even though many efforts could slow down the virus transmission, it is generally believed that the ultimate tool to avert such a pandemic should be an effective and safe vaccine, with which to achieve herds immunity. Thus, the advent of one or more efficacious vaccines, having the availability, accessibility, and affordability across the globe, is essential for achieving effective control of the pandemic, which is still on-going even with new peaks of all-time high of 2 nd wave in the winter in certain countries/areas of the northern hemisphere.

The cellular entry of SARS-CoV-2 is predominantly mediated by the interaction of viral spike protein and its major receptor, the host angiotensin-converting enzyme 2 (ACE2) (1-3). Like other coronaviruses, the SARS-CoV-2 spike is a homotrimer protruding from the viral surface, and it comprises two functional subunits, namely, the S1 subunit containing the receptor-binding domain (RBD) and the S2 subunit mediating membrane fusion between the virus and the cell (4) (5) (6) . To date, several lines of evidence showed that antibodies against this spike protein could play a critical role in the immunoprophylaxis and therapeutics against COVID-19 (7) (8) (9) (10) (11) (12) . A vast array of different platforms with different molecular modality is currently being employed for vaccine development against COVID- 19 (13-19) . Subunit vaccines, particularly in the case that protein making up the subunits could be prepared using recombinant DNA technology, are of great advantage due to its proven safety and compatibility with multiple boosts if necessary (20) .

Two challenges, however, should be addressed to develop an ideal SARS-CoV-2 subunit vaccine. The first is to prepare the recombinant protein subunits in a functional form that can predominantly elicit SARS-CoV-2 neutralizing antibodies rather than nonneutralizing antibodies, as the latter may mediate antibody-dependent enhancement (ADE) for disease or infection (21) . The second is to overcome the less immunogenic drawback of the subunit protein than an intact virus or virus-like particle-based vaccines (20) . As the classical Alum adjuvant is generally apt to stimulate Th2-biased immune response, its use could be associated with vaccine-associated enhanced respiratory disease (VAERD) (22, 23) . New adjuvants that can help to generate a more balanced Th1/Th2 cellular response and to improve the immunogenicity of the non-particulate protein-based vaccine are certainly required.

In this study, through in vivo evaluation of the immunogenicity of CHO-expressed various spike derivatives, we identified the full length of the spike ectodomain (StriFK), which induced more neutralizing to non-neutralizing antibodies, as the best performing immunogen. In mice, hamsters, and non-human primates, we tested the immunostimulatory effects of different adjuvants with StriFK, specifically FH002C (an nitrogen bisphosphonate-modified zinc-aluminum hybrid adjuvant, developed in house) and Al001 (a traditional Alum adjuvant commonly used in human vaccines). Enhanced immunogenicity of StriFK was observed in combined use with FH002C compared to that with Al001 for both the functional antibodies and cellular responses. More importantly, we demonstrated the immunizations of StriFK-FH002C in hamsters provided adequate and immune-correlated protection against the infection, pathogenesis, and inter-animal transmission of SARS-CoV-2. Together, our results showed robust immune responses and protective potency elicited by a new adjuvant formulated StriFK vaccine in animals.

Using the CHO expression system, recombinant spike protein derivatives of SARS-CoV-2, including RBD, S1, S2, trimerized RBD (RBDTfd) and S-ectodomain (StriFK), were obtained ( Figure 1A ). All five purified proteins achieved acceptable purity and reacted well with antibodies in the COVID-19-convalescent human plasma in western blots ( Figure 1B ).

For adjuvant used in the immunogenicity evaluation in mice ( Figure 1C ), both the traditional Alum (Al001) and the novel FH002C adjuvants showed similar morphology and antigen adsorption capacity ( Figure S1 ).

Among all tested candidates, the StriFK-FH002C immunized mice presented the most rapid antibody response, showing detectable anti-spike, anti-RBD, and neutralization antibodies (nAb) just 1-week after administration of the 1 st dose ( Figure 1C ). Two weeks after the 1 st dose, mice receiving StriFK or S1 (with either adjuvant) all presented detectable (geometric mean titer, GMT≥100) anti-spike antibodies, whereas mice received RBDTfd-FH002C, StriFK-Al001, StriFK-FH002C, and S1-FH002C showed detectable anti-RBD antibodies. However, at this time point, only mice immunized with StriFK-FH002C presented pseudovirus neutralizing antibody (GMT=200). Compared to Al001, the new FH002C adjuvant stimulated significantly higher and more rapid antibody response in mice, regardless of the immunogen used. At Week 6, after 3-dose immunization was completed, the RBD, RBDTfd, S1, and StriFK induced similarly high levels of anti-spike and anti-RBD in combination with FH002C adjuvant.

As expected, anti-S2 antibodies were only observed at comparable levels in mice immunized with StriFK and S2 ( Figure 1D ). Sera from S2-immunized mice showed detectable nAb levels, suggesting that both S1-and S2-specific antibodies contribute to the neutralization activities of StriFK-induced antibodies. Regression analyses on the associations between the anti-spike binding titer and the neutralizing titer in mice receiving various immunogens revealed that the StriFK had a higher neutralizing-to-binding ratio (as reflected by a larger slope value) than the others ( Figure 1E ).

In a stable pool of CHO cells transfected with StriFK-expressing construct, protein expression reached over 200 mg/L ( Figure S2A ). The molecular weight of the StriFK was determined to be approximately 700-kDa by HPLC-SEC analysis ( Figure S2B) . The StriFK showed a binding affinity of 3.52 nM to human ACE2 protein ( Figure S2C ) and exhibited good reactivities to human COVID-19-convalescent plasmas in ELISA-binding assay ( Figure S2D ). The StriFK-binding activities of human COVID-19-convalescent plasmas positively correlated with their neutralizing titers against pseudotyped and authentic SARS-CoV-2 virus, also well associated with their receptor-blocking titers ( Figure S2E ).

Together, the StriFK combined with the FH002C adjuvant provided an excellent vaccine candidate for further evaluation.

To test the vaccination effect in promoting the germinal center formation, we measured the frequencies of T follicular helper (Tfh), germinal center B (GCB), and plasma cells in lymph nodes (LNs) in mice (1 week after a single immunization). In comparison to StriFK-Al001, the StriFK-FH002C significantly increased the frequencies of Tfh ( Figure 2C ), which were consistent with the more rapid antibody response and earlier antibody affinity maturation (indicated by higher IgG avidity) after StriFK-FH002C immunization ( Figure S3A ).

In mice receiving 2-dose immunizations of StriFK-FH002C and StriFK-Al001, we further measured spike-specific T cell response. In contrast to non-immunized animals, the IFNγ ELISPOT assays revealed the numbers of IFNγ-secreting cells after ex vivo antigenpeptides stimulation increased by 28.9-fold and 5.8-fold in the spleen, and 14.0-fold and 2.3-fold in lymph nodes (LNs), in mice immunized by StriFK-FH002C and StriFK-Al001, respectively ( Figure 2D and 2E). Compared to Al001, the FH002C adjuvant elicited significantly higher frequencies of spike-specific IFNγ-secreting cells in either spleen (p=0.0002) or LNs (p=0.0031).

Intracellular cytokine staining (ICS) measurements of mouse splenocytes stimulated   by spike-peptides demonstrated that the StriFK-FH002C successfully induced Th1, Th2, and CTL responses, which were evidenced by significantly higher frequencies of spikespecific T cells of IFN-γ + CD4 + (p=0.0087, Figure 2F ), IL4 + CD4 + (p=0.041, Figure 2G ) and IFN-γ + CD8 + (p=0.0022, Figure 2H ), respectively. The StriFK-Al001 also increased spikespecific IL4 + CD4 + (p=0.015, Figure 2G ) and IFN-γ + CD8 + T cells (p=0.037, Figure 2H ) in splenocytes, but it appeared less potent than StriFK-FH002C. In spike-peptide restimulated cells from LNs, although the increases of T helper cells (IFN-γ + CD4 + and IL4 + CD4 + ) were not pronounced ( Figure 2I and 2J), markedly elevated IFN-γ + CD8 + T cells were observed in both vaccinated groups ( Figure 2K ).

Consistent with T cell analysis findings, sera from StriFK-FH002C immunized animals showed significantly higher IgG2a or 2b titers ( Figure S3B ) and IgG2-to-IgG1 titer ratio ( Figure S3C ) than StriFK-Al001. Together, these results demonstrated the StriFK-FH002C could induce a more balanced Th1/Th2 immune response.

Next, we examined the immunogenicity of StriFK-FH002C in cynomolgus macaques in comparison to StriFK-Al001 and RBD-FH002C ( Figure S4 ). After the 2-dose regimen, at

Week 4 all vaccine-immunized monkeys showed detectable anti-spike ( Figure S4A StriFK-FH002C or StriFK-Al001 immunized animals slightly decreased to about 50% of the corresponding peak values. When a booster immunization was given at Week 6, the serum antibodies of animals in groups of StriFK-FH002C and StriFK-Al001 rapidly increased to comparable levels as that at Week 3. In hamsters, the StriFK-FH002C induced more than 4-fold higher neutralizing receptor-blocking antibody titers than StriFK-Al001, implying the potent immunostimulatory effect of the FH002C adjuvant. Surprisingly, the RBD-FH002C, which could generate binding and functional antibodies in mice and monkeys, showed very little even no antibody response in hamsters.

Using the intranasal SARS-CoV-2 challenging method, we investigated the protective efficacy of StriFK-FH002C in hamsters. We first tested whether a 2-shot immunization of StriFK-FH002C could protect hamsters from SARS-CoV-2 infection and its related pneumonia. In this experiment, the neutralizing antibody GMT of the six StriFK-FH002C vaccinated hamsters was 14,404 (range: 5,368-37,078). In the sham group that received FH002C-only, hamsters lost an average of 8.4%, 12.2%, and 11.7% of body weight by 3, 5, and 7 days post-infection/inoculation of virus (dpi), respectively ( Figure 4A ). In contrast, StriFK-FH002C immunized animals lost just an average of 2.0%, 2.1%, and 1.2% of body weight by the corresponding time points ( Figure 4A ), which was significantly lower than that of the sham controls (p<0.0001, two-side two-way ANOVA test) and similar to that of the uninfected (mock) group (p=0.061). All animals in the sham group exhibited obvious symptoms of weakness, piloerection, hunch back, or abdominal respiration, which was observed in neither vaccinated nor un-infected animals ( Figure 4B ). were conducted in hamster for the SARS-CoV-2 challenging study ( Figure 5A ). The neutralizing and receptor-blocking antibody titers in these StriFK-FH002C vaccinated animals were approximately 4 folds higher than that in StriFK-Al001 vaccinated animals.

The average weight loss at 5 dpi was 9.0%, -0.04% and 0.17% in the unvaccinated controls, the StriFK-FH002C and the StriFK-Al001 groups, respectively ( Figure 5B ). The weight loss was markedly lower in both vaccinated groups compared to unvaccinated controls (p<0.0001 for both comparisons) and showed no significant difference between the StriFK-FH002C and StriFK-Al001 groups (p=0.39). Infectious viruses were detected in all unvaccinated animals' lungs. Interestingly, the males showed higher viral titers than females. In the lung tissues of vaccinated hamsters, no infectious virus was detected ( Figure 5C ). Consistent with infectious virus titers, viral RNA levels were profoundly reduced (over 4 logs) in lung tissues from vaccinated hamsters compared to the controls ( Figure 5D ). In trachea and nasal turbinate tissues, StriFK-FH002C immunized hamsters exhibited more marked viral RNA reductions (over 3 logs) than those received StriFK-Al001 ( Figure 5E ). No significant gender difference was found for vaccination-mediated viral RNA reduction.

Gross lung pictures demonstrated multifocal diffuse hyperemia and consolidation in the lungs of controls, and the apparent lesions were significantly diminished in both vaccinated groups ( Figure S5A ). Histopathological examinations observed pulmonary consolidation, alveolar destruction, and diffuse inflammation in about 20-30% of lungs from female controls and 50-60% of lungs from male controls, suggesting male hamsters had more severe pneumonia than females. Comparing to controls, vaccinated animals (using StriFK-FH002C or StriFK-Al001), either males or females, only showed mild inflammation in limited areas of lungs, and had minimal pneumonia pathology in most areas of lungs ( Figure 5F ).

These results demonstrated StriFK-FH002C immunizations effectively prevented hamsters from SARS-CoV-2 virus infection and pathogenesis.

To mimic the disease transmission, we also assessed whether the StriFK-FH002C vaccine could protect inter-animal SARS-CoV-2 transmission in hamsters ( Figure 6A ). Cohoused with the infected donors, the unvaccinated control hamsters showed their maximal weight loss on Day 4 (mean: 8.5%) post-contact exposure ( Figure 6B ). In contrast, the weight loss was diminished in both vaccinated groups (StriFK-Al001 vs controls, p=0.0091;

StriFK-FH002C vs controls, p<0.0001), and it changed to a lesser degree (p=0.017) in the StriFK-FH002C group (mean: -0.22%) than that in StriFK-Al001 group (mean: 3.6%). None of the vaccinated animals presented infectious virus in their lungs at Day 5 post-exposure, whereas all control animals had detectable infectious viruses at varying titers ( Figure 6C ).

More consistent with the weight change findings, viral RNA detections revealed a lower viral load in hamsters of the StriFK-FH002C group than in StriFK-Al001 immunized animals or unvaccinated controls ( Figure 6D and 6E ). Gross lung images ( Figure S5B ) and pathological analyses ( Figure 6F ) indicated that control hamsters had extensive lung damage, with consolidated lesion and inflammatory cell infiltration across larger areas, which was quite similar to that presented in intranasal SARS-CoV-2 challenged hamsters. In contrast, both vaccines prevented tissue damage to a large degree, if any, with mild perivascular and alveolar infiltration observed in very few areas ( Figure 6F ).

Using pooled animals related to immunized hamsters showed no significant antibody increase after the SARS-CoV-2 challenge in either neutralizing antibody (p=0. 19) or receptor-blocking antibody (p=0.98), suggesting that sterilizing immunity to SARS-CoV-2 may have been conferred due to immunization in these animals. The pooled antibody data in mice, hamsters, and nonhuman primates showed that the neutralizing antibodies induced by StriFK-FH002C in animals were about 28-to 269-fold higher than that of the human COVID-19 convalescent plasmas ( Figure 7C ).

In this study, we systematically evaluated the immunogenicity and protective efficacy of a recombinant SARS-CoV-2 subunit vaccine candidate SFTK-FH002C in different animal models. This vaccine candidate was composed of CHO-expressed SARS-CoV-2 spike ectodomain protein (StriFK) and a novel adjuvant (FH002C). Comparing to other tested spike derivatives, including RBD, S1, and S2, our in vivo data demonstrated the StriFK generated a higher neutralizing-to-binding ratio ( Figure 1E ), which implied it induced more neutralizing antibodies than non-neutralizing antibodies. The FH002C is a functionalized zinc-aluminum hybrid adjuvant with a significantly improved immuneenhancing effect. Immunization with StriFK-FH002C at low doses in mice, hamsters, and monkeys rapidly generated anti-spike, neutralizing, and receptor-blocking antibodies since one-dose administration. Compared to the GMT level of human COVID-19 convalescent plasmas, 2-shot StriFK-FH002C immunization induced 28-38 folds higher neutralizing antibody and 3-shot immunization generated 47-269 folds higher neutralizing antibody in mice, hamsters, and non-human primates ( Figure 7C ).

In hamster, an animal model recapitulating severe human COVID-19 disease, StriFK- Figure   5D and Figure 6D ). Although the protective efficacy of StriFK-FH002C in non-human primates was not tested in this study, it can be expected according to the comparable levels of neutralizing and receptor-blocking antibodies induced by this vaccine in cynomolgus macaques to that in hamsters ( Figure 7C ). Apart from the highly immunogenic StriFK protein, the novel FH002C adjuvant also plays an essential role in generating high protective antibody and spike-specific T cell response levels. Due to the multiple mechanisms for effective adjuvants, novel adjuvants have focused on different complex formulation involving multiple components with different functions to form various 'adjuvant systems' (28) . In the FH002C, a drug for osteoporosis nitrogen bisphosphonate (risedronate) was loaded on the composite Alum adjuvant by harnessing the Alum's highly-phosphophilic nature. Zinc ion, one species available to the cellular environments upon adjuvant dissolution, has an immunostimulatory effect and was known to steer the immune response more towards the Th1 pathway (29, 30) . Although the precise role of nitrogen bisphosphonate was still unknown, previous studies suggested these compounds could stimulate γδ T cells expansion and act as a dendritic cell modulator due to the upstream accumulation of phosphoantigen isopentyl diphosphate as a result of inhibition of farnesyl diphosphate synthetase (31) (32) (33) . Notably, the FH002C adjuvant showed more potent effects than Al001 to promote the developments of Tfh and GCB cells and generated antibody-producing plasma cells in the early days after immunization ( Figure 2A, 2B, and 2C ). Due to the rapid differentiation of these cell populations, FH002C adjuvant-based StriFK vaccine generated a faster and stronger protective antibody response, which elicited neutralizing antibodies as early as 1-week after the initial immunization. This advantage enables the vaccine's potential application for emergency vaccination of close contacts due to high-risk exposure, aiming to quickly control outbreaks among close contacts and in the naïve population.

In addition to neutralizing antibodies, T cells are also believed to be essential to vaccine effectiveness (34) (35) (36) . The spike of SARS-CoV-2 has a subnanomolar affinity to its ACE2 receptor, and the infections may occur at the respiratory epithelium of the upper respiratory tract where the antibody concentration is low. Moreover, the functional antibody may not completely block viruses from invading cells; virus-specific T cells may play a remedial role in eradicating virus-infected cells or inhibiting viral replication and spread via non-cytotoxic effects (37, 38) . Our in vivo data showed the StriFK-FH002C immunization could induce potent spike-specific IFNγ+CD8+ and IFNγ+CD4+ T cells, which would help to establish protective immunity afforded by both arms of the immune response in vaccinated individuals.

Safety is also a crucial concern for COVID-19 vaccine development. Previous studies on RSV and measles vaccine suggested VAERD is linked to Th2-biased immune response (39) (40) (41) . A similar phenomenon was also noted in animal studies regarding SARS-CoV vaccines (39) . Therefore, an ideal COVID-19 vaccine is expected to elicit a balanced Th1/Th2 immune response. Recombinant protein vaccines adjuvanted by traditional aluminum adjuvants generally induce significantly enhanced Th2-biased immune responses but insufficient Th1 responses, which was also observed in the performance of our StriFK-Al001 vaccine in the mouse model, as indicated by the lower IgG2-to-IgG1 titer ratio ( Figure S3 ). The FH002C adjuvant is a novel tool to overcome this drawback. This vaccine candidate is under development in the pre-clinical stage and could be manufactured on a large scale, with the hope that a stronger and more sustainable immune response could be elicited using this vaccine than the natural infection.

The codon-optimized RBD cDNA of SARS-CoV-2 (GenBank: MN908947.3) was obtained by primer annealing. RBDTfd gene was produced by fusing with a trimeric foldon of the T4 phage head fibritin at C-terminal. SARS-CoV-2 S1 and S2 subunit gene referring to the full-length spike glycoprotein gene of SARS-CoV-2 (GenBank: MN908947.3) were synthesized by General Biosystems Co. Ltd (Anhui, China). Spike glycoprotein ectodomain was introduced with three amino acid mutations in furin-like cleavage site that mutating RRAR to GSAS, and fused with a trimeric foldon at C-terminal. All cDNA above were introduced with a C-terminal polyhistidine sequence and codon-optimization was performed for mammalian cell expression. These target genes were constructed into a modified PiggyBac (PB) transposon vector EIRBsMie using NEBuilder HiFi DNA Assembly Master Mix (New England Biolabs, E2621L) described previously (44) .

For initial immunogenicity evaluation in mice, the recombinant proteins of RBD, RBDTfd, S1, S2, and StriFK were transiently expressed by ExpiCHO-S cells. The plasmid expression plasmid carrying each SARS-CoV-2-associated protein was transiently transfected into ExpiCHO-S cells by using the ExpiFectamine™ CHO transfection kit according to the manufacturer's instruction (Thermo Scientific, A29129). After transfection, the cells were placed in an incubator shaker (Kühner AG, SMX1503C) at 37°C containing 8% CO2 for 24 hours, and then, transferred the cells to a shaker at 32°C containing 5% CO2 culture for 5-7 days. Subsequently, the culture supernatants of CHO cells were subjected to purification by Ni Sepharose Excel column (Cytiva). For further animal vaccination studies, polyhistidine-free StriFK proteins were produced by a stable cell pool transfected with StriFK-expressing construct. The proteins were purified from culture supernatants by ion-exchange chromatography.

Purified proteins were submitted to SDS-PAGE using SurePAGE (Genscript).

Coomassie brilliant blue staining for SDS-PAGE were performed using eStain L1 Protein Staining machine (Genscript). Gels for western blot were transferred onto the nitrocellulose membrane and reacted with COVID-19-convalescent serum (1:500 diluted), followed by washing and reaction with horseradish peroxidase (HRP)-conjugated anti-human pAb. The blots were imaged on FUSION FX7 Spectra multispectral imaging system (VILBER). The size exclusion liquid chromatography (SEC) for purified StriFK protein was performed using a high-performance liquid chromatography system (Waters Alliance HPLC) on a TSKgel G3000PWXL column.

The binding affinity of StriFK to recombinant human ACE2 protein was determined by SPR analysis. For the detection, rACE2 (mouse Fc tagged, Sino Biological) proteins were immobilized to a protein G sensorchip (Cytiva) at a level of about 500 response units (RUs) using Biacore 8000 (Cytiva). Serial dilutions of purified StriFK proteins were injected, ranging in concentration from 200 to 0.78 nM. The RU data were fit to a 1:1 binding model using Biacore TM Insight evaluation software.

Twelve plasma samples (including 11 convalescent COVID-19 plasmas and one control sample) were used to test their activities to recombinant StriFK protein. The authentic virus neutralizing titers, pseudovirus neutralizing titers, and receptor-blocking titers had been measured and described in our previous study (44) . Another 268 plasmas collected from recovered COVID-19 patients within eight weeks since illness onset were used to determine the GMT level of neutralizing antibodies (by pseudovirus virus assay) of convalescent COVID-19 patients. All of these convalescent plasmas were collected after the confirmed COVID-19 patients discharged from the First Hospital of Xiamen University.

The study was approved by the institutional review board of the School of Public Health in accordance with the Declaration of Helsinki, and written informed consent was obtained.

The appearance of FH002C and Al001 was photographed. The morphological characteristics of the two adjuvants were imaged at × 100 using the Motic AE31E inverted microscope (Motic, Xiamen, China), and were observed at a nominal ×120,000 magnification by transmission electron microscopy (JEOL, Tokyo, Japan). The particles' size and zeta potentials were measured by using LS 13320 Beckman Coulter Laser Particle Analyzer (Beckman Coulter, CA, USA) and NanoBrook Omni zeta potential analyzer (Brookhaven Instruments Corporation, NY, USA), respectively.

Al001 and FH002C are both suspensions of aluminum-based adjuvants, which were prepared in-house. Al001 is an aqueous solution of amorphous aluminum hydroxyphosphate, much like what was described for adjuvants for HBV, HPV and HEV vaccines (45) (46) (47) , with a molar ratio of P/Al:0.15. The Al content in the formulation is 0.84 mg/mL. The FH002C is a novel functionalized adjuvant, wherein the Al element was partially replaced with zinc, and the amorphous phosphophilic particles (48) were loaded with risedronate for additional immunostimulatory activity. The resulting organic-inorganic hybrid adjuvant showed similar whitish suspension morphology in aqueous solution with the widely used A1001 adjuvant (45) . Before using, the 2x adjuvants were mixed well with resuspending, then each antigen sample (in 150 mM NaCl) was mixed at 2× concentration with an equal volume of the adjuvant to achieve the final desired concentration of antigen and adjuvant in the final formulation. The vaccine formulations were mixed well with several gentle upside-and-downside turns of the vial/tube and stored at 2-8 °C until use. All the vaccine formulations containing Al001 or FH002C were prepared at least 24 hours before immunization. Based on the ELISA results using a sandwich assay with mAbs on both sides, completeness of adsorption was shown to be at least 95% for both formulations based on the antigen content in the supernatant as compared to the overall antigen in the formulation.

For antibody response evaluation, BALB/c mice were maintained in a specific pathogen-free (SPF) environment and immunized with various proteins at 1 μg/dose with Al001 or FH002C through intramuscular injection, following an immunization schedule of one priming dose at Week 0 plus two boosters at Weeks 2 and 4. Serum samples were collected at Week 0, 1, 2, 3, 4, 5, and 6 via retro-orbital bleeding to measure the antibody titers. 

Briefly, single-cell suspensions from mouse spleen (10 6 cells per well) or LNs (4×10 5 cells per well) were seed in precoated ELISPOT plate (Dakewe Biotech). Subsequently, cells were incubated with pooled peptides of SARS-CoV-2 spike (15-mer peptides with 11 amino acids overlap, cover the entire spike protein, Genscript) and cultured for 20 hours.

The detecting procedure was conducted according to the manufacturer's instructions.

Spots were counted and analyzed by using CTL-ImmunoSpot® S5 (Cellular Technology Limited). The numbers of IFN-γ secreting cells were calculated by subtracting PBSstimulated wells from spike peptide pool-stimulated wells. 

Microplates pre-coated with recombinant antigens of RBD, spike ectodomain, S1, or 

Microplates pre-coated with recombinant spike protein were provided by the Beijing Wantai company. Serial dilutions of serum samples from immunized mice were added into the wells in duplicate microplates. The plates were incubated at 37 °C for 1-hour and followed by a PBST wash cycle (5 times). Then, one of the duplicate microplates was incubated with 100 μL of 4M Urea solution and the other one was incubated with 100 μL of PBS buffer at 37 °C for 20 min. Followed by a wash cycle, HRP-conjugated anti-mouse IgG was added into the wells for incubation of 30 min at 37 °C. Finally, the plates were further washed with PBST for 5 times and followed by chromogen reaction and OD measurements. The binding activity of polyclonal antibodies in serum was defined as the dilution fold required to achieve 50% of the maximal signal (ED50). The ratio of ED50(ureatreated) to ED50(PBS-control) was used to measure the avidity of serum polyclonal antibodies to StriFK.

Serum IgG isotype-specific antibody titers were determined by a modified ELISA method. For the assay, recombinant spike coated microplates were used. The serum samples were serially diluted and reacted with plates following a similar procedure as above mentioned in IgG titer measurements. After a 1-hour incubation, the plates were washed for 5 times and subsequently incubated with HPR-conjugated goat anti-mouse IgG1, IgG2a, or IgG2b (AbD Serotec, Oxford, UK ) at 37 °C for 30-minutes. Then, the plates were washed for 5-time and followed by chromogen reaction and OD measurements. The anti-spike titers of IgG1, IgG2a, and IgG2b were determined as the dilution limit to achieve a positive result.

were produced according to our previous study (49) . The BHK21-hACE2 cells were used to perform the pseudovirus neutralization assay. Briefly, cells were pre-seeded in 96-well plates. Serially-diluted (3-fold gradient) samples were incubated with VSVpp inoculum for 1 hour. Next, the mixture was added into seeded BHK21-hACE2 cells. After a further incubation at 37 °C containing 5% CO2 for 12 hours, fluorescence images were captured by Opera Phenix or Operetta CLS High-Content Analysis System (PerkinElmer). The counts of GFP-positive cells of each well were analyzed by the Columbus System (PerkinElmer). The neutralization titer of each sample was expressed as the maximum dilution fold (ID50) required to achieve infection inhibition by 50% (50% reduction of GFPpositive cell numbers compared to controls).

A cell-based spike function-blocking test (CSBT) was used to characterize the receptor-blocking antibody titer in animal serum following the previously described procedure (44) . Briefly, the hACE2-mRuby3 293T (293T-ACE2iRb3) cells were seeded in poly-D-lysine pretreated CellCarrier-96 Black plate at 2×10 4 cells per well overnight.

Hamster or Cynomolgus monkey sera were diluted in a 2-fold serial dilution with DMEM medium containing 10% FBS. 11 μL Gamillus-fused SARS-CoV-2 spike trimer (STG) probe mixed with 44 μL of the diluted sera to a final concentration of 2.5 nM. After removing half of the medium from the pre-seeded 293T-ACE2iRb3 cells, pipette 50 μl of the mixture into the wells and incubate 1 hour at 37 °C containing 5% CO2. Subsequently, cell images were captured by Opera Phenix High-Content Analysis System in confocal mode. Quantitative image analyses were performed by the Columbus System (PerkinElmer) following the previously described algorithm (44) . The CSBT activities (receptor-blocking antibody titers) of immunized sera were displayed as ID50, calculated by 4PL curve-fitting in GraphPad Prism. 

Eight Cynomolgus monkeys were allocated randomly into four groups (one female and one male per group). Groups of cynomolgus were injected with 20 μg StriFK with Al001 or FH002C, or 20 μg RBD with FH002C adjuvant per dose via the intramuscular route for 3 doses. Another group received FH002C adjuvant-only as a sham control. All cynomolgus were injected 0.5 mL of vaccine or adjuvant-only at Weeks 0, 2, and 6. Serum samples were collected weekly for biochemical and antibody analyses, including measurement of anti-spike IgG, pseudovirus neutralizing antibody, and receptor-blocking antibody titers during the Week 1 to 10. Cynomolgus monkey immunogenicity experiment was conducted at JOINN Laboratories, Inc (Beijing).

Both intranasal challenge and direct contact challenge of SARS-CoV-2 were used in the study. For the former, hamsters were inoculated with 1×10 4 PFU of the SARS-CoV-2 virus through the intranasal route under anesthesia. For the latter, virus-carried hamsters (donors) were pre-infected with inoculation of 1×10 4 PFU of the SARS-CoV-2 virus through the intranasal route. One day later, every four donors were transferred to a new cage and were co-housed with four vaccinated or unvaccinated control animals. The hamsters were fed with a daily food amount of 7 g per 100 g of body weight. The weight changes and typical symptoms (piloerection, hunched back, and abdominal respiration) in hamsters were recorded daily since the infection or contact. Hamsters may be sacrificed for tissue pathological and virological analyses on days 3, 5, or 7 after the virus challenge. The virus challenging studies were performed in the anmial biosafety level 3 (ABSL-3) facility.

Viral RNA levels in lung, trachea, and nasal turbinate from challenged hamsters were detected by quantitative RT-PCR. Hamster tissue samples were homogenized by TissueLyser II (Qiagen, Hilden, Germany) in 1mL PBS. The RNA was extracted using the QIAamp Viral RNA mini kit (Qiagen) according to the manufacturer's instruction.

Subsequently, viral RNA quantification was conducted using SARS-CoV-2 RT-PCR Kit (Wantai, Beijing, China).

The plaque assays were performed in the biosafety level 3 (BSL-3) facility. Briefly, Vero cells were seeded in a 6-well plate at 10 5 cells per well. After overnight culture, the medium was removed, and the cells were washed twice with PBS. The supernatants of tissue lysates were serially 10-fold diluted with DMEM and were incubated with the cells.

Cell plates were then incubated for 1 hour at 37 °C containing 5% CO2. Subsequently, the tissue lysate contained medium was refreshed with a mixture of DMEM containing 1% agar.

After agar solidification, the cell plates were inverted and incubated at 37 °C containing 5% CO2 for 3 days. Subsequently, cells were fixed with 2 mL of 4% formaldehyde for 2 hours at room temperature. The cells were further stained with 0.1% crystal violet for 20 min. The viral load was calculated based on the count of plaques and the corresponding dilution factor.

The lung tissues from challenged hamsters were fixed with 10% formalin for 48 hours, then embedded in paraffin and sectioned. Next, the fixed lung sections were subjected to hematoxylin and eosin (H&E) staining. Immunohistochemical staining was performed by using the mouse monoclonal anti-SARS-CoV-2 N proteins antibody. Whole-slide images of the lung sections were captured by EVOS M7000 Images System (ThermoFisher).

The Student's t-test or Mann-Whitney U test was used for comparison of continuous variables between groups. The two-way ANOVA test was used to analyze the time-serial observations for the independent sample. Pearson test and linear regression model were used for univariate correlation analysis. Statistical significance was considered to be significant for 2-tailed p values <0.05. Statistical analyses were conducted in GraphPad Prism 8 software. Groups of BALB/c mice were immunized with different vaccine candidates (1 μg/dose) at weeks 0, 2, and 4. The proteins of RBD, RBDTfd, StriFK, S1 and S2 were used as the immunogen combined with traditional (Al001) or modified (FH002C) Alum adjuvants, respectively. Mouse serum titers of anti-spike (upper panel), anti-RBD (middle panel), and SARS-CoV-2 pseudovirus neutralizing antibody (lower panel) were displayed. (D) The serum titers of anti-S1 and anti-S2 in immunized mice at Week 6 since the initial immunization. (E) The associations between the titers of pseudovirus nAb and anti-spike binding antibody in mice received different immunogens. The linear regression curves with 95% confidence intervals were plotted. The slopes of regression curves were indicated to reflect the relative ratio of neutralizing-to-binding antibody induced by various immunogens. For panel (C) and (D), data were plotted as the geometric mean±SD.

SARS-CoV-2 Cell Entry Depends on ACE2 and TMPRSS2 and Is Blocked by a Clinically Proven Protease Inhibitor

Cryo-EM structure of the 2019-nCoV spike in the prefusion conformation

Angiotensin-converting enzyme 2 (ACE2) as a SARS-CoV-2 receptor: molecular mechanisms and potential therapeutic target

Molecular Architecture of the SARS-CoV-2 Virus

Structures and distributions of SARS-CoV-2 spike proteins on intact virions

Structure, Function, and Antigenicity of the SARS-CoV-2 Spike Glycoprotein

Potent neutralizing antibodies against multiple epitopes on SARS-CoV-2 spike

A human neutralizing antibody targets the receptor-binding site of SARS-CoV-2

Ultrapotent human antibodies protect against SARS-CoV-2 challenge via multiple mechanisms

Potent Neutralizing Antibodies against SARS-CoV-2 Identified by High-Throughput Single-Cell Sequencing of Convalescent Patients' B Cells

A human monoclonal antibody blocking SARS-CoV-2 infection

Structurally Resolved SARS-CoV-2 Antibody Shows High Efficacy in Severely Infected Hamsters and Provides a Potent Cocktail Pairing Strategy

SARS-CoV-2 mRNA vaccine design enabled by prototype pathogen preparedness

Single-shot Ad26 vaccine protects against SARS-CoV-2 in rhesus macaques

Phase 1-2 Trial of a SARS-CoV-2 Recombinant Spike Protein Nanoparticle Vaccine

Development of an inactivated vaccine candidate for SARS-CoV-2

A Thermostable mRNA Vaccine against COVID-19

A Universal Design of Betacoronavirus Vaccines against COVID-19, MERS, and SARS

SARS-CoV-2 vaccines in development

Immunological considerations for COVID-19 vaccine strategies

Antibody-dependent enhancement and SARS-CoV-2 vaccines and therapies

COVID-19 vaccines: neutralizing antibodies and the alum advantage

Rapid COVID-19 vaccine development

In Vitro and Animal Models for SARS-CoV-2 research

Pathogenesis and transmission of SARS-CoV-2 in golden hamsters

Syrian hamsters as a small animal model for SARS-CoV-2 infection and countermeasure development

Ad26 vaccine protects against SARS-CoV-2 severe clinical disease in hamsters

Key roles of adjuvants in modern vaccines

Effects of zinc deficiency on Th1 and Th2 cytokine shifts

Zinc: mechanisms of host defense

The Mevalonate Pathway Is a Druggable Target for Vaccine Adjuvant Discovery

Boning up: aminobisphophonates as immunostimulants and endosomal disruptors of dendritic cell in SARS-CoV-2 infection

Bisphosphonate-induced differential modulation of immune cell function in gingiva and bone marrow in vivo: role in osteoclast-mediated NK cell activation

Robust T Cell Immunity in Convalescent Individuals with Asymptomatic or Mild COVID-19

Broad and strong memory CD4(+) and CD8(+) T cells induced by SARS-CoV-2 in UK convalescent individuals following COVID-19

SARS-CoV-2 immunity: review and applications to phase 3 vaccine candidates

T cell responses are required for protection from clinical disease and for virus clearance in severe acute respiratory syndrome coronavirusinfected mice

Virus-specific memory CD8 T cells provide substantial protection from lethal severe acute respiratory syndrome coronavirus infection

Learning from the past: development of safe and effective COVID-19 vaccines

Immunological Lessons from Respiratory Syncytial Virus Vaccine Development

A perspective on potential antibody-dependent enhancement of SARS-CoV-2

Bisphosphonates for the prevention and treatment of osteoporosis

Structure-based design of prefusion-stabilized SARS-CoV-2 spikes

Virus-free and live-cell visualizing SARS-CoV-2 cell entry for studies of neutralizing antibodies and compound inhibitors. bioRxiv

Comparable quality attributes of hepatitis E vaccine antigen with and without adjuvant adsorption-dissolution treatment

Effect of alternative aluminum adjuvants on the absorption and immunogenicity of HPV16 L1 VLPs in mice

Optimizing the utilization of aluminum adjuvants in vaccines: you might just get what you want

Surface phosphophilicity of aluminum-containing adjuvants probed by their efficiency for catalyzing the P--O bond cleavage with chromogenic and fluorogenic substrates

Robust neutralization assay based on SARS-CoV-2 S-proteinbearing vesicular stomatitis virus (VSV) pseudovirus and ACE2-overexpressing BHK21 cells

The authors declare no competing interests. 

and StriFK-Al001 vaccine candidates in hamsters. All three vaccines were used at 10 μg/dose. Animals in Sham (FH002C-only) and RBD-FH002C groups received 2-shot immunizations at Weeks 0 and 2, whereas animals in the StriFK-Al001 group received 3shot immunizations at Weeks 0, 2, and 6. In the StriFK-FH002C group, 6 of 22 animals received 2-shot immunizations at Weeks 0 and 2, and the remaining 16 animals were immunized for 3-dose (10 μg/dose) at Weeks 0, 2, and 6. Serum samples were collected every week since initial immunization, and the anti-spike, anti-RBD, neutralizing antibody, and the receptor-blocking antibody titers were measured. The quantitative data of antibody titers were plotted as the geometric mean±SD. were measured in different pH using NanoBrook Omni zeta potential analyzer (Brookhaven Instruments Corporation, NY), and the PZC values were calculated by the instrument software. Each Zeta potential value reported is the average of three replicates. (C) Size distributions of Al001 (left panel) and FH002C (right panel). The particle size distribution of adjuvants was measured by using LS 13320 Beckman Coulter Laser Particle Analyzer (Fullerton, CA). The median particle diameter d50 (equivalent diameter when the cumulative distribution is 50%) was measured in three independent replicates, mean±SEM. Figure 5. (B) Hamsters were exposed to SARS-CoV-2 via direct contact with pre-infected donor hamsters. Related to Figure 6 . Figure S6 . Correlations between the tissue viral RNA levels and serum pseudovirus nAb (upper panels) or receptor-blocking antibody (lower panels) titers among SARS-CoV-2 challenged hamsters. The antibody titers of these hamsters at Week 7 (3 days before virus challenging study) since initial immunization were used for analysis. Tissue viral RNA levels at 5 dpi (as shown in Figure 5C and Figure 6C ) were used. All vaccinated hamsters and unvaccinated controls involved in experiments related to Figures 5 and 6 were included and pooled for analysis. Lung-1 and Lung-2 indicated lung tissues collected from tissues near and away from the pulmonary hilum, respectively.