A nanoparticle-based COVID-19 vaccine candidate elicits broad neutralizing antibodies and protects against SARS-CoV-2 infection

A vaccine candidate to SARS-CoV-2 was constructed by coupling the viral receptor binding domain (RBD) to the surface of the papaya mosaic virus (PapMV) nanoparticle (nano) to generate the RBD-PapMV vaccine. Immunization of mice with the coupled RBD-PapMV vaccine enhanced the antibody titers and the T-cell mediated immune response directed to the RBD antigen as compared to immunization with the non-coupled vaccine formulation (RBD + PapMV nano). Anti-RBD antibodies, generated in vaccinated animals, neutralized SARS-CoV-2 infection in vitro against the ancestral, Delta and the Omicron variants. At last, immunization of mice susceptible to the infection by SARS-CoV-2 (K18-hACE2 transgenic mice) with the RBD-PapMV vaccine induced protection to the ancestral SARS-CoV-2 infectious challenge. The induction of the broad neutralization against SARS-CoV-2 variants induced by the RBD-PapMV vaccine demonstrate the potential of the PapMV vaccine platform in the development of efficient vaccines against viral respiratory infections.


BACKGROUND
The ongoing COVID-19 pandemic, caused by the SARS-CoV-2 virus, has highlighted once more the critical role of vaccines in limiting the spread of infectious diseases 1 . Traditionally, vaccines against viral diseases are made of live-attenuated or inactivated pathogens that necessitate growing the pathogen in confinement facilities, which is time consuming and challenging to upscale 2 . The traditional approaches led to production of COVID-19 vaccines that were only partially efficient (around 50% efficacy). On the other hand, the latest recombinant technologies were shown to be faster and more effective 2,3 . Novel vaccine platforms emerged allowing faster design, vaccine production, and shortening the time to enter clinical trials 2,3 . The fastest technologies are based on the utilization of nucleic acids (DNA or mRNA), including viral vectors, encoding the SARS-CoV-2 spike (S) protein as the vaccine antigen 4 . Following immunization, S is produced by the cells of the patients and then recognized by the immune system, leading to the desired protection through the induction of neutralizing antibodies to the S antigen 2,3 . The adenovirus vector 5 , used to transport the DNA encoding for the S antigen to the nucleus of the patient's cells was one of the first vaccine available against SARS-CoV-2 6 .
However, a slight increase in the rate of intracranial venous thrombosis (ICTV) and thrombocytopenia in adults >70 years were reported in patients receiving the ChAdOx-1-S vaccine based on an adenovirus of simian origin [7][8][9] . No correlation was detected with mRNAbased vaccines, which are considered as a safer alternative 10 . The mRNA vaccines emerged as a fast and effective approach with a protection >90% 11-16 and is the favored vaccine in Europe and North America 2,17 . The mRNA-based vaccines are made of a modified mRNA encoding for the S protein of the SARS-CoV-2 ancestral strain that is embedded into a lipidic nanosphere used to transfect cells after injection for inducing production of the S protein 11,12 in vaccinated patients.
J o u r n a l P r e -p r o o f Journal Pre-proof Despite being well positioned to respond to pandemic scenarios, the use of mRNA vaccines during the COVID-19 pandemic revealed some caveats, such as adverse reactions induced after immunization exceeding those induced by current seasonal vaccines, the need for deep freezing at -80°C 18,19 for distribution, and a rapidly declining immunity after immunization 20 .
Considering that SARS-CoV-2 infections will remain a recurring problem for years to come, pursuing the development of novel vaccine platforms is very important to continue improving the quality, stability and safety of the vaccines of the future 21 .
Vaccines based on the use of adjuvanted recombinant proteins 6,11,12,[22][23][24] have also been developed and made available at a later stage of the pandemic. They will play an important role to better control the COVID-19 pandemic considering the shortage of SARS-CoV-2 vaccines 2,17 .
They could also be used in a heterologous prime-boost immunization approach with mRNAbased vaccines that could potentially improve the protection against new emerging SARS-CoV-2 variants 25,26 .
The use of protein-based nanoparticle vaccine platforms to enhance and modulate antigen immunogenicity has emerged as a promising approach for controlling pandemic outbreaks 27,28 . successful in vaccine development [37][38][39][40][41] and may play an important role in the vaccine of tomorrow.
Recently, the use of plant-virus-based nanoparticles as a vaccine platform [42][43][44][45][46] has emerged as an interesting and effective alternative for the design of vaccines against coronavirus. Presentation of the protein antigen at the surface of the nanoparticle stabilized the antigen and improved the antigen-specific immune response [38][39][40]43,44,46 .
In this report, we propose the use of a novel rod-shaped nanoparticle made of the papaya mosaic virus (PapMV) coat protein (CP) and a synthetic ssRNA named PapMV nano. This nanoparticle has great potential for the development of vaccines because it has been shown to (i) be safe in human 47 , (ii) trigger innate immunity through the stimulation of toll-like receptors (TLR) 7/8 34 , (iii) be easily engineered with a vaccine antigen to enhance its immunogenincity 32,[37][38][39][40][48][49][50] , and (iiii) present vaccine antigens to immune cells in a repetitive and highly ordered manner that is ideal for activation of the B cells specific to the antigen 51 . Exploiting the PapMV nano technology, we designed a candidate vaccine to SARS-CoV-2 using the RBD antigen produced and purified from human cells as the vaccine antigen 52,53 .

Mice immunization and viral challenge
The immunogenicity of RBD-PapMV vaccine candidate was assessed in female BALB/c mice (Charles River, USA). Mice, 10/group, were immunized twice by intramuscular injection, 21 days apart, with RBD-PapMV (100 μg, including 5.1 μg of coupled RBD), RBD (5.1 μg) + PapMV nano (100 μg) or formulation buffer consisting of 10 mM Tris-HCl. Mice were bled before the first immunization for obtaining naïve sera. Twenty days after the first injection, 5 mice per group were euthanized, and blood was extracted by cardiac puncture for ELISA and neutralization assay. The remaining five mice/group were bled at day 39 for assessing the antibody titer, and were euthanized at day 42, when blood was extracted through cardiac puncture. The spleen was harvested for the ELISPOT assay.

Antibody titration by ELISA
The antibody titers directed towards RBD in the sera of BALB/c mice were assessed by enzyme linked immunosorbent assay (ELISA) as described previously 37 . Results are expressed as antibody endpoint titers greater than 3-fold OD 450nm of the background value consisting of a pool of pre-immune sera.

IFN-γ detection by ELISPOT
The frequency of interferon gamma (IFN-γ)-secreting cells in the spleen of BALB/c mice was measured by enzyme-linked immunospot (ELISPOT) assay. ELISPOT assay was performed using an IFN-γ murine ELISPOT Kit (Abcam, UK) as described by the manufacturer.
Splenocytes from immunized and control mice were stimulated with RBD (25 μg/ml), Concanavalin A (5 μg/ml, positive control -Invitrogen, USA), or culture medium in 96-well culture plates (Corning, USA) for 72 h at 37 C, 5% CO 2 . Then, activated cells were transferred to MultiScreen-IP filter plate (MilliporeSigma, USA) coated with IFN-γ capture antibody and incubated for 48 h at 37 C, 5% CO 2 . Finally, plates were revealed following the manufacturer's instructions.

Neutralization of SARS-CoV-2 variants
The capacity of mice sera to neutralize SARS-CoV-2 infection was assessed in vitro using a microneutralization assay (MNA) previously described 54 . The SARS-CoV-2 ancestral, Delta, and J o u r n a l P r e -p r o o f

Journal Pre-proof
Omicron variants, as well as the permissive cell line Vero.E6 were kindly provided by the INSPQ. In brief, sera from immunized mice were heat inactivated at 56 C for 1 h. Then, mice sera and a monoclonal antibody (MAb) against the Spike protein (Absolute Antibody, UK, 3 µg/ml), used as a positive control, were diluted 1/10 followed by 1/3 serial dilutions. Each dilution was incubated with 20 TCID 50 of SARS-CoV-2 for 1 h at room temperature and used to infect Vero.E6 cell monolayers. Then, the infectious mixture was removed, and cells were incubated for 72 h with media containing sera or MAb at the same dilutions 54 . Cells not exposed to the virus were the negative controls ("no virus"), while cells treated with virus alone were the positive control of infection ("virus only"). Finally, the cell monolayers were fixed with formaldehyde 4%. The amount of virus was assessed by immunostaining of the viral nucleoprotein (N) with an anti-N polyclonal antibody (Rockland) from rabbit, followed by an anti-rabbit IgG secondary antibody conjugated to peroxidase (Jackson Immunoresearch, USA).
The optical density measured at 450 nm from samples and controls was used to calculate the percentage of infection using the following formula: 100 -((OD x -Average of "no virus" wells) / (Average of "virus only wells" -Average of "no virus wells") * 100), where X corresponds to the absorbance for each sample. Non-linear regression curve fit analysis was performed to calculate the ID 50 values.

Assessment of viral titer in lungs and nasal turbinates
SARS-CoV-2 titers in the lung and nasal turbinate of infected K18-hACE2 transgenic mice were assessed by end-point titration as previously described 55 . Lungs and nasal turbinate were harvested at 2 and 5 d.p.i. and homogenized. The cytopathic effect for each sample was recorded J o u r n a l P r e -p r o o f Journal Pre-proof and the TCID 50 were computed using the Reed-Muench method 56 . Titers were expressed as log 10 of TCID50/g 55 .

Assessment of pro-inflammatory cytokines in the lung
The levels of pro-inflammatory cytokines interleukin (IL)-6, tumor necrosis factor (TNF)-alpha and the chemokine KC/GRO (CXCL 1 ) in the lungs of SARS-CoV-2 infected K18-hACE2 mice was measured using a V-Plex custom cytokine panel (Meso Scale Discovery -MSD, USA), according to the manufacturer's instructions. The concentration of cytokines in lung homogenate was calculated by interpolation of the calibrator's curves as described by the manufacturer.

Lung histopathology
For assessing the pulmonary tissue inflammatory response histologically after SARS-CoV-2 (ancestral strain) challenge, lungs of K18-hACE2 transgenic mice were extracted at 5 d.p.i. The left lung was fixed with 10% buffered formalin, embedded in paraffin, sectioned at 4 µm thickness, and stained with hematoxylin-eosin. The histopathological inflammatory scores (IS) were determined by a pathologist, Dr. C. Couture, with experience in pulmonary pathology (see supplementary table 1 for the criteria used to assess the IS). A semi-quantitative scale was used to score bronchial/endobronchial, peribronchial, perivascular, interstitial, pleural and intraalveolar inflammation. Capillary vascular congestion and pulmonary edema were also evaluated, and the inflammatory cellular infiltrate was characterized to determine if the inflammation was acute (neutrophilic) or chronic (lymphohistiocytic). The results were expressed as total pulmonary inflammatory scores.

Statistical analysis
Data was analyzed with Graph Pad PRISM 7.0b software (GraphPad Software, Inc., USA). Oneway ANOVA test followed by Tukey's test was used to compare statistical differences among groups of mice regarding the antibody titers, the IFN-γ secretory response, and the viral titers in the lungs and nasal turbinate. Non-linear regression analysis using an inhibition curve fit model was performed to calculate the ID 50 values of the MNA neutralization curves. The differences among groups regarding the percentage of viral inhibition at each dilution of the neutralization curves were assessed using two-way repeated measures ANOVA, with Tukey's as the post-test.
Finally, a two-way ANOVA and Tukey's post-test were performed to compare the levels of each cytokine in lung homogenates, as well as the total inflammation scores derived from the histology analysis. Differences were considered significant for p< 0.05.

Generation of RBD-PapMV vaccine
The coupling of the RBD antigen, from the SARS-CoV-2 ancestral strain, to the PapMV nanoparticle was performed using SrtA, a bacterial transpeptidase that induces the formation of a covalent link between two proteins through recognition of the LPETGG donor and G-repeat acceptor motifs 57 . Coupling of the RBD to the PapMV nano using SrtA results in a nanoparticle surface decorated with RBD protein (schematic in Fig. 1A). The resulting coupled protein comprised PapMV CP linked to the RBD protein, as shown by the appearance of a 62 kDa fusion protein on SDS-PAGE (Fig. 1B). Based on the intensity of the bands on the SDS-PAGE, we assessed that approximately 9% of the PapMV CP subunits were coupled with the RBD (Fig. 1B lane 2). Immunoblotting confirmed that the RBD-PapMV coupled protein contains the PapMV CP (Fig. 1C) and the RBD (Fig. 1D).
The mean length of the RBD-PapMV was measured by DLS. The average size of RBD-PapMV nanoparticles, 80nm in length, was indistinguishable from PapMV nano ( Fig. 2A). A heat denaturation curve was performed using the DLS to compare the stability of the PapMV nano with that of the RBD-PapMV vaccine candidate (Fig. 2B). Interestingly, the point of inflection leading to denaturation and aggregation of the nanoparticles was increased by 5°C with the  Fig. 2A, B and C).

Assessment of the humoral response induced by the RBD-PapMV vaccine
BALB/c mice were immunized twice intramuscularly with either RBD-PapMV, RBD+PapMV or formulation buffer. Mice sera was harvested at days 20 and 39 (schematic of the schedule in Fig. 3A). The levels of IgG1, related with T helper (Th) 2 response, and IgG2a, related with Th1 response 58,59 , were assessed by ELISA against the RBD antigen. After one immunization, the IgG2a titers to RBD were 32-folds higher with the RBD-PapMV vaccine as compared to the RBD+PapMV uncoupled formulation (Fig. 3B). Likewise, the boost with the vaccine candidate was superior to the uncoupled control by 64-folds (Fig. 3C). For both conditions, the boost immunization led to higher IgG2a titers (Fig. 3BC). The total IgG and the IgG1 titers against RBD also reflected significantly higher titers for RBD-PapMV after prime and boost when compared to uncoupled formulation (supplementary Fig. 2). The mean ratio IgG2a/IgG1 after one or two doses of the RBD-PapMV vaccines were 1.39 and 1.33, respectively, indicating a preference towards the Th1 response.

Assessment of the T cell mediated immune response induced by the RBD-PapMV vaccine.
To assess the capacity of RBD-PapMV vaccine to elicit a T cell-mediated immune response, an ELISPOT assay was performed to measure the secretion of IFN-by the CD4+ and CD8+ T cells after stimulation with the RBD antigen (Fig. 3D). The results revealed that the coupled vaccine (RBD-PapMV) is 5.8 folds more efficient than the uncoupled formulation in inducing J o u r n a l P r e -p r o o f Journal Pre-proof the proliferation of RBD-specific CD8+ and CD4+ T cells. It also confirmed that the RBD-PapMV vaccine triggers a Th1-type immune response.

Microneutralization of SARS-CoV-2 variants
A microneutralization assay (MNA) was conducted to determine the capacity of sera from vaccinated mice to inhibit the infection of Vero.E6 cells by SARS-CoV-2. The MNAs were validated using a commercial monoclonal antibody directed to the RBD protein (supplementary Figure 3).
After only one immunization, a 1:10 sera dilution from mice immunized with the RBD-PapMV vaccine was able to completely neutralize SARS-CoV-2 infection (ancestral strain) (Fig. 4A). and 85% inhibition, respectively, after which, the efficacy decreased rapidly (Fig. 4D). The sera generated by the uncoupled formulation led to efficient inhibition only with sera diluted 10-and 30-fold.
The MNA assay provides a read-out of the efficacy of virus neutralization that is influenced by two factors: (i) the amount of antibodies in the sera and (ii) their affinity for RBD. Therefore, it is possible to determine a constant of affinity for such antibodies, called the neutralization score per unit of titer (NS), by dividing the neutralization efficacy by the antibody titer. The ID 50 value of each curve, representing the inhibitory dilution at which 50% of the viral neutralization is attained, constitutes a convenient indicator of neutralization efficacy. Using this value, the formula would be: ID 50 / antibody titer = NS. High NS corresponds to serum with a higher neutralization efficacy per titer of antibodies (IgG2a titers). This is an indicator of effectiveness of the antibodies found in the sera to neutralize infection ( Table 1). The NS ratio between one and two immunizations of RBD-PapMV is 321.5/5.7 = 56.4, which reveals that for the same amount of anti-RBD antibodies, the serum of mice immunized twice is 56-folds more efficient in neutralizing SARS-CoV-2 infection than those immunized once. RBD-PapMV injected twice is also (54-folds) more efficient in neutralization of the ancestral strain than the uncoupled formulation. With the Delta variant, the RBD-PapMV is (333/3) 111-folds more efficient in the neutralization than the uncoupled formulation, and (68/13) 5.2-folds with Omicron. groups, mice were immunized twice with PapMV nano or the formulation buffer. Mice were challenged with the SARS-CoV-2 ancestral strain by intranasal instillation at day 42. On days 44 and 47, four mice per group were sacrificed to harvest the lungs and nasal turbinates (NT) and assess the viral titers (Fig. 5A). Viral titers were determined by end-point dilution of lung and NT homogenates in Vero.E6 cell monolayers (Fig. 5B). At 2 and 5 d.p.i., one dose of RBD-PapMV reduced the viral titer in the lungs as compared with the buffer control (Fig. 5B).
Notably, two doses of the RBD-PapMV vaccine reduced the viral titers below the limit of detection for 4 out of 4 animals at day 2 p.i., and 3 out of 4 animals at day 5 p.i. (Fig. 5B). In the NTs, only the animals immunized twice with RBD-PapMV showed a decrease in the virus titers at day 2 p.i.

Lung histopathology and cytokine profile in lungs of infected animals
Lungs of infected mice with SARS-CoV-2 were harvested at day 5 post-infection. Inflammation, when present, was lymphohistiocytic (chronic) in nature and predominantly interstitial and perivascular in distribution. Mice immunized with the buffer or the naked PapMV had an average total inflammation score (TIS) of 2.7 and 4.2, respectively, which was significantly higher than mice immunized with one or two doses of the RBD-PapMV vaccine that showed a TIS 0.5 and 0, respectively (Fig. 6AB). As expected, uninfected mice (negative controls) showed no sign of inflammation (TIS 0).
Additionally, inflammation was assessed by measuring the levels of pro-inflammatory cytokines (IL-6, TNF- and KC/Gro) in lung homogenates at 5 d.p.i. The groups immunized with one or two doses of the RBD-PapMV vaccine presented levels of IL-6 (<30pg/mL) comparable to the J o u r n a l P r e -p r o o f non-infected animals, while the other infected groups (Buffer and PapMV) had high levels (80 and 165pg/mL, respectively) of IL-6 ( Supplementary Fig. 4).

DISCUSSION
Pandemics constitute the most severe manifestation of a viral outbreak. With the increasing frequency of these events, there is a need to develop versatile vaccine platforms able to strongly stimulate the immune system in a safe manner. These platforms should be readily available, stable for easy transportation, and made with globally accessible technology. The development of PapMV nano technology is an attempt to respond to these needs.
In this study, we demonstrate that the coupling of the SARS-CoV-2 RBD to the surface of PapMV nano leads to a more effective protection as compared with the uncoupled formulation.
Coupling to PapMV nano significantly increased the amount and quality of antibodies directed to the RBD antigen, as revealed by micro-neutralization assays. The coupling of RBD to PapMV nano increased the efficacy of neutralization of the ancestral and the Delta variant by more than 100-folds as compared with the non-coupled formulation. The engagement of the TLR7 by the RBD-PapMV vaccine may act as a co-stimulatory signal that enhances and accelerates the antibody class switch toward the production of the IgG2a subclass and the production of higher affinity antibodies to the RBD antigen 34,61 . PapMV nano was reported to accelerate germinal center formation, and to promote affinity/avidity maturation of specific IgG and isotype switching to IgG2a subclass 62 . Recently, these events were linked with the stimulation of TLR7, which is consistent with our data 61 .
To our knowledge, the PapMV nano is one of the few vaccine technologies available that are capable to stimulate the TLR7/8. Stimulation of TLR8 is a main advantage since these receptors are abundant in monocytes, macrophages and dendritic cells that are important players in the RBD-PapMV nanoparticles were shown to be stable at 4 ± 3 °C for 1 month without loss of integrity (supplementary Fig. 3). This stability of the vaccine preparation is critical to ensure its distribution, as it is easier to distribute a vaccine that does not need to be frozen, an advantage over mRNA-based vaccines.
Overall, the PapMV nano technology is very promising for the development of effective vaccines and it is distinct among the other protein-based nanoparticles based on its capacity to stimulate the TLR 7/8 33,34 , which is responsible for the trigger of a broad neutralizing immunity that is needed for protections against novel variants. We foresee this technology to be very useful