Recombinant subunits of SARS‐CoV‐2 spike protein as vaccine candidates to elicit neutralizing antibodies

Abstract Objectives The spike protein has been reported as one of the most critical targets for vaccine design strategies against the SARS‐CoV‐2 infection. Hence, we have designed, produced, and evaluated the potential use of three truncated recombinant proteins derived from spike protein as vaccine candidates capable of neutralizing SARS‐CoV‐2 virus. Methods In silico tools were used to design spike‐based subunit recombinant proteins (RBD (P1), fusion peptide (P2), and S1/S2 cleavage site (P3)). These proteins were checked for their ability to be identified by the anti‐SARS‐CoV‐2 antibodies by exposing them to COVID‐19 serum samples. The proteins were also injected into mice and rabbit, and the antibody titers were measured for 390 days to assess their neutralization efficiency. Results The antibodies that existed in the serum of COVID‐19 patients were identified by designed proteins. The anti‐spike antibody titer was increased in the animals injected with recombinant proteins. The VNT results revealed that the produced antibodies could neutralize the cultured live virus. Conclusion Truncated subunit vaccines could also be considered as robust tools for effective vaccination against COVID‐19. Using a combination of in silico, in vitro, and in vivo experiments, it was shown that the injection of spike‐based truncated recombinant proteins could stimulate long‐lasting and neutralizing antibody responses.

the practiced treatment protocols. 2,6 Coming to grips with the life cycle and pathogenesis, mechanism of the SARS-CoV-2 would bring about insights into the proper strategies to develop vaccines. SARS-CoV-2 is a single-stranded RNA virus. It belongs to the Sarbecovirus subgroup of the Betacoronavirus group and the subfamily of the Orthomyxoviridae. Coronaviruses are divided into four groups including α, β, δ, and γ CoV. 7,8 The size of the viral genome is between 26 and 32 kb, which is considered as one of the largest RNA viruses. 9 The diameter of these viruses is about 60-140 nm. 10 Twothirds of the viral RNA transcribes for the pp1a, pp1ab, and 16 nonstructural proteins. The rest of the genome encodes for structural proteins. 11 The main structural protein includes the spike (S), nucleocapsid (N), envelope (E), and membrane (M) proteins, which are encoded by the 3' end of the viral genome. 11,12 SARS-CoV-2 enters the host cells through the angiotensinconverting enzyme 2 (ACE2) receptor, 13,14 which resides in the human lower respiratory tract. It is also known as the cellular receptor for SARS-CoV. 15,16 The S glycoprotein on the surface of the SARS-CoV-2 can bind to the ACE2 receptor on the surface of human cells. 17 This protein is a trimmer protein that belongs to class 1 viral glycoproteins. 18 It has been reported that the S proteins are the most important structural protein of the SARS-CoV-2 to enter the target cells. The S glycoprotein plays an essential role in viral binding, fusion, and entry into the host cell. 19 This protein has two significant subunits known as S1 and S2 subunits. The S1 subunit identifies the host cell, while the S2 subunit is responsible for the fusion of the virus to the host cells. [20][21][22] The N-terminal domain (NTD) and the receptor-binding domain (RBD) are the functional domains of the S1 subunit. The RBD encompasses a receptor-binding motif (RBM) that is conserved among most strains of coronaviruses. 23 The RBD sequence from the SARS-CoV shares 72% similarity with the RBD from SARS-CoV-2. 24,25 Prior studies have revealed that the RBD can form a tight interaction with ACE2 protein, which initiates the infection process. 26 The interaction between the SARS-CoV-2 and the ACE2 receptor mainly occurs between the RBM from the S protein and the N terminus region of the ACE2. This interaction leads to endocytosis of the virus. 26 The interaction between the RBD and the ACE2 receptor stimulates structural changes in the S2 subunit.
The exerted changes play an essential role in the fusion between the viral envelope and the host cell membrane. The S2 subunit of the S protein consists of several regions, including the membraneanchoring region, the fusion peptide (FP), the heptad repeat (HR) 1 and 2. 24 Inside the endosome, the S1 subunit would be cleaved off and the FP would be exposed. The FP locates itself inside the host membrane. The S2 then bends over to bring HR1 and HR2 together.
This causes membrane fusion and release of the viral genome in the cytoplasm of the host cell. 25,27,28 According to WHO, tens of vaccines are under investigation worldwide at various stages. Various vaccine development strategies have been practiced in different settings preclinical and clinical trials, including the DNA, RNA, recombinant protein, viral vector, and the attenuated or deactivated viral vaccines. Safety considerations and the wide variety of antigen variants are potential challenges ahead of efficient vaccine development. The pivotal role of the S protein in the pathogenesis of the SARS-CoV-2 confirms that it can be the principal antigenic agent for stimulation of the host immune system and production of neutralizing antibodies. 26 Prior studies have shown that vaccines made from the S protein can stimulate the immune system and induce humoral and cellular responses. 1 The ongoing studies regarding the design of SARS-CoV-2 vaccines are already focused on the S protein using different platforms.
In light of these observations, we aimed to design vaccine candidates based on the mechanism of S protein action and bioinformatics analyses. In this regard, three antigenic regions of the S protein were selected as vaccine candidates to elicit humoral immunization against SARS-CoV-2, which may induce neutralizing antibodies. The immunogenicity of these vaccine candidates was evaluated using in silico, in vitro, and in vivo studies.

| ME THODS
All procedures were performed according to the ethical guidelines of

Faculty of Medical Sciences Tarbiat Modares University (TMU) and
National Institute of Genetic Engineering and Biotechnology (code of ethics:1399.015).

| Sequence retrieval
The sequence of the S protein was retrieved from the NCBI database at https://www.ncbi.nlm.nih.gov/. The obtained sequence was used to perform a BLAST search at https://blast.ncbi.nlm.nih. gov/Blast.cgi. The PSI-BLAST (Position-Specific Iterated BLAST) tool of the protein BLAST was employed to find the highly similar protein sequences. Multiple sequence analysis was performed on the sequences obtained from the BLAST search. The potential glycosylation sites on the S protein were predicted using NetNGlyc-1.0 software at http://www.cbs.dtu.dk/servi ces/NetNG lyc/. The glycosylation analyses would ensure the exclusion of glycosylated regions within the vaccine sequence. Since the prokaryotic expression system is unable to make accurate glycosylation on the produced antigens, the humoral responses against these regions would be rendered ineffective due to residing glycosylation on the spike protein. The S protein sequence was also searched for the existence of a signal peptide using SignalP-5.0 software at http:// www.cbs.dtu.dk/servi ces/Signa lP/.

| Sequence analyses of the selected regions
The ProtParam software at https://web.expasy.org/protp aram/ was used to predict the physicochemical properties of the selected regions. The potential glycosylation sites on the selected regions were predicted using NetNGlyc-1.0 software. The antigenicity of the selected sequences was predicted by Vaxijen-2.0 software at http:// www.ddg-pharm fac.net/vaxij en/VaxiJ en/VaxiJ en.html. The allergenicity of the regions was predicted by Algpred software at http:// crdd.osdd.net/ragha va/algpr ed/. The toxicity of the selected regions was predicted by ToxinPred software at http://crdd.osdd.net/ragha va/toxin pred/.

| Recombinant expression of the candidate vaccines
The protein sequence of the selected regions was reverse transcribed to the DNA sequences by the ExPASY translate tool at http://web.expasy.org/trans late/. The Jcat tool at http://www.jcat.
de/ was employed to optimize the DNA sequences for high levels of protein expression (the E. coli codon usage bias was used for the optimization). The EcoR1 and XhoI restriction sites were selected to insert the designed genes within the pET28a expression vector.
This design would grantee the expression of His tag sequence at the N terminus of the proteins. The final genes were ordered for chemical synthesis and subsequent subcloning by the GENERAY Biotechnology Company. The synthesized genes (within the pET28a expression vector) were transformed into E. coli BL21 (DE3) using the standard CaCl 2 method. Colony PCR using the universal T7 primers was employed to confirm the transformation. The protein expression of the transformed was performed using the same method employed in our previous study. 29 The expression of the protein was

| Protein purification and Western blotting
Protein purification and Western blot analysis were performed using the protocol adapted from our previous study. 29 Briefly, the Ni + -NTA resin-packed columns (Qiagen) and pH gradients were used to purify the expressed proteins following the procedures provided by the manufacturer. The SDS-PAGE (4% stacking gel and 15% separating gel) was used to analyze the genes' expression and the purity of the eluted fractions. The purification fraction (with the pH of 5.2), which contained a single protein band of the recombinant protein, was selected for the following experiments. The protein content of the eluted fractions was measured using standard Bradford assay. Since the proteins formed inclusion bodies, they were denatured using standard denaturing conditions by 8 M urea. The purified protein samples were finally dialyzed (for protein renaturation) to remove the urea and increase the pH from 5.2 to 7.4. A standard Western blot analysis was performed to assess the yield of the purification step. In this regard, primarily the proteins were resolved on the gel and was transferred onto the nitrocellulose membrane (Whatman Schleicher and Schuell). Ultimately, an anti-His tag antibody conjugated with horseradish peroxidase (HRP) was added to the nitrocellulose membrane in order to visualize the reactive bands. 30

| Circular dichroism (CD) spectroscopy
Since the complete removal of the urea could lead to protein precipitation and aggregation, the urea was gradually removed by dialysis.

| Animal immunization and production of polyclonal antibody
The study conduction was adhered to the principles of the declara- and bleeding was performed every month. To determine the antibody titer, the indirect ELISA method was employed as explained before.

| ELISA test on patient's serums
To assess the ability of the candidate vaccine proteins to interact with the antibody available in the serum of the patient (all study participants received a full explanation of the study and were obtained a written informed consent prior to their inclusion in the study), an ELISA test was designed and developed, and employed. Briefly In addition, a conventional rapid immuno-chromatography test was used to observe the interaction of purified antibody with the antigen coated onto the membrane in such assays.

| Antibody purification
The elicited antibodies should be purified for virus neutralization assay. Antibody purification for all serum samples was done using affinity chromatography on a Protein A Agarose column (PAO9-R5, ABT Company). The purification was performed according to the manufacturer's instructions. Briefly, saturated ammonium sulfate was gradually added to rabbit sera at the final concentration of 33%.
They were then stirred on ice for one hour. After 25 min of centrifugation at 9000 g, the supernatant was removed and led on the protein A column. The column was washed by PBS, and elution buffer (glycine-HCl, pH = 2.5) was added to elute the desired protein, neutralized using carbonate buffer to adjust the pH, concentrated via Amicon ultrafiltration device cutoff concentrators (10 kDa), measured the protein content by standard Bradford protein assay, and was determined by SDS-PAGE (4% stacking gel and 12.5% separating gel.

| Preparation of cells and virus stock
To perform the neutralization assays, the African green monkey

| Cytopathic effect (CPE) based neutralization assays
The CPE-based neutralization assays were carried out in 96-well microtiter plates in triplicate. Three rabbit sera (immunized by P 1 , P 2 , and P 3 antigens), the serum-free DMEM culture media (as a control showing the conserved and variable regions of the S protein. 32 The variability in the RBM sequence was also evidently analyzing the MSA results. There were 17 asparagine residues predicted to be Nglycosylated throughout the S sequence. The glycosylation was less condensed at the RBD region and the region connecting the S1 and S2 parts. The sequence spanning the residues 1-15 was predicted to be a signal.

| Selection of regions for vaccine candidates
Given the properties of the S protein and considering the mechanism of S protein to fuse the virus into the host cells, three areas, including the RBD (P 1 ), fusion peptide (P 2 ), and S1/S2 cleavage site (P 3 ) of the spike protein, were selected to be the candidate vaccine antigens. Their molecular weight was calculated to be 11,477.82 Da, 13,528.71 Da, and 25,000 Da for P 1 , P 2 , and P 3 antigens, respectively. The selected vaccine candidates were all predicted to be stable according to their instability index and have a high estimated half-life within mammalian cells.
Analyzing the properties of these antigens, there were no asparagine residues with possible glycosylation, all three vaccine candidates were predicted to be antigens, no allergenicity effects were expected to be inclined by these vaccine candidates, and there were no regions with significant potential of toxicity throughout their sequences.

| Protein expression
The corresponding genes for P 1 , P 2 , and P 3 vaccine candidates were optimized according to the E. coli codon usage bias, and the unwanted sequences, which could affect the optimal protein expression, were omitted. DNA sequencing and enzymatic digestion on the subcloned vector confirmed Gene cloning. The colony PCR confirmed that the vectors are transformed into the E. coli BL21 (DE3) host. The results of the protein expression have shown that P 1 , P 2 , and P 3 proteins are overexpressed and can travel to the expected molecular weight on SDS-PAGE gel (12 kDa for P 1 , 14 kDa for P 2 , and 25 kDa for P 3 ) ( Figure 1A). It has been revealed that the best condition for the expression of the proteins is adding 1mM of IPTG and shaking for 4 h at 37°C.

| Protein purification and Western blot analysis
The expressed proteins for P 1 , P 2 , and P 3 were purified by Ni + -NTA resin-packed columns following the denaturation and renaturation procedure following dialysis (pH change and urea alternation).
Running the purified protein samples on the SDS-PAGE gel indicated that the unwanted proteins moieties were removed from the samples, and the purified proteins could travel to their expected molecular weight ( Figure 1B). The Western blot analysis has also confirmed the identity of the purified antigens using the anti-His tag antibody ( Figure 1C).

| CD analyses
The results of the CD analyses for the P 1 , P 2 , and P 3 vaccine candidates were listed in Table 1, which indicated that all three vaccine candidates were folded to discernible secondary structures with different ratios.

| Antibody production
After performing the immunization regiment on both rabbits and mice groups, their serum samples were evaluated for antibody elicitation. The results of indirect ELISA tests for both animal groups indicated that antibodies were raised against P 1 , P 2 , and P 3 protein candidates (Figure 2A,B). The results of these indirect ELISA tests confirmed the immunogenicity of the P 1 , P 2 , and P 3 protein candidates within both rabbits and mice groups without any lethal consequences. The results have also confirmed long-lasting antibody response for P 1 , P 2 , and P 3 protein candidates and showed that the antibody titer for all three rabbit groups (injected with P 1 , P 2 , and P 3 proteins) remains higher than the control group up to 390 days after injection. Similarly, the antibody titer was higher than the control group for mice groups (injected with P 3 , P 1 , P 2 , and a mixture of three proteins) up to 390 days after the second injection. However, a decreasing trend for antibody titer was detected in the test results after 390 days compared to that of the previous days.

| ELISA on serum sample of COVID-19 patients
Using the P 1 , P 2 , and P 3 protein as the capture antigen in an ELISA may show the ability of these antigens to interact with antibodies produced within the serums of COVID-19 patients. Our results indicated that P 1 , P 2 , and P 3 vaccine candidates are capable of interacting with antibodies raised within the serums from COVID-19 patients (Figure 3). A positive interaction result could be construed as functional and structural similarities between the P 1 , P 2 , and P 3 vaccine candidates and their corresponding sequences within the whole virus structure.

| Antibody purification and CPE-based neutralization assays
The affinity chromatography on a Protein A column managed to purify the IgG antibodies from the rabbit serum samples ( Figure 4A).
The neutralization assay determines the effect of neutralizing antibodies based on the observation of cell morphology. This method is reported to be the first and most frequently used neutralization assay in the SARS research. 33 The results indicated that it takes 5 days for the infected Vero E6 cells to start forming visible CPE including dissociated cell patterns. It was evident that the 1:10 dilution of the rabbit antibody (equivalent to 50 µg) immunized by the P 1 , P 2 , and P 3 vaccine candidates could neutralize the 1:1000, 1:100, and 1:10 dilution of viral stock. The 1:100 concentration of the rabbit serum immunized by P 1 vaccine candidate could neutralize up to the 1:1000 concentrations of viral stock ( Figure 4B).

| DISCUSS ION
Employing an integrative approach, we aimed to produce a safe and inexpensive vaccine against the SARS-CoV-2. In this regard, the subunit vaccine platform is based on S protein. Previous studies on SARS-CoV, MERS-CoV, and recent SARS-CoV-2 have revealed that given the critical role of S protein in the mechanism behind the virus entry into the host cells, it can be deemed as the best candidate for vaccine development efforts. Usually, the whole S protein (or its subunits) or the RBD region is used for the design of a subunit vaccines against the SARS-CoV-2. The vaccine is usually injected along with a suitable adjuvant to get the best immunization results. The  It has been reported that with a gradual increase in urea concentration from 6 to 10 M, the beta sheets will be destroyed, the alpha helixes will remain stable and may even increase, and the random coils will remain intact. 50,51 The obtained CD results for the P 1 , P 2 , and P 3 vaccine candidates have confirmed these structural changes during urea removal. This could be construed as protein refolding to its native structure upon urea removal.
Yuxian et al. 52 have also used a recombinant fusion protein To have an efficient immune response against the SARS-CoV-2, the elicited antibodies should exhibit the ability to neutralize the viral particle. The Virus Neutralization Test (VNT) method is a susceptible and specific test to check for the presence of neutralizing antibodies against the target virus. This method is also practiced as the gold standard method to analyze the presence of neutralizing antibodies against SARS-CoV-2. Previous studies have demonstrated that the S protein components such as S2, S1, and especially RBD hold a profound potential for production of neutralizing antibodies. These antibodies can block the virus binding to the ACE2 receptor and its membrane fusion. 31 Our VNT results have also confirmed that the elicited antibodies against the designed proteins are highly persistent and capable of neutralizing the cultured live virus. This property could be construed as the highly promising potential of these proteins as vaccine candidates.
F I G U R E 2 (A) Serum antibody titer for rabbits immunized with P 1 , P 2 , and P 3 proteins compared to the control group in 30, 70, 130, and 390 days after injection. (B) The serum antibody titer for mice groups immunized with P 1 , P 2 , P 3 , and a mixture of three proteins compared to the control group in 70, 120, and 390 days after injection. All tests were performed duplicate F I G U R E 3 The ELISA results for anti-SARS-CoV-2 antibody detection in the serum of COVID-19 patients using the P 1 , P 2 , and P 3 proteins as capture antigens compared to the healthy subject

ACK N OWLED G M ENT
We gratefully thank the National Institute of Genetic Engineering and Biotechnology for the financial support, grant no 99/266.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data are available on request.

R E FE R E N C E S F I G U R E 4 (A) Purification of antibody.
The fractions of antibody elution from the protein A column (lanes 1-3: elusion fractions for P 1 , lane 4: molecular weight, lanes 5-7: elusion fractions for P 2 , lanes 8-10: elusion fractions for P 3 ). (B) Different concentrations of virus neutralization via the antibodies produced by P 1 , P 2 , and P 3 vaccine candidates (1: is the 1:10 serum and 1:10 virus dilution, 2: is the 1:10 serum and 1:100 virus dilution, 3: is the 1:10 serum and 1:1000 virus dilution, 4: is the 1:100 serum and 1:1000 virus dilution) compared with the control group