(C) PLOS One [1]. This unaltered content originally appeared in journals.plosone.org.
Licensed under Creative Commons Attribution (CC BY) license.
url:https://journals.plos.org/plosone/s/licenses-and-copyright

------------



A vesicular stomatitis virus-based prime-boost vaccination strategy induces potent and protective neutralizing antibodies against SARS-CoV-2

['Gyoung Nyoun Kim', 'Department Of Microbiology', 'Immunology', 'Schulich School Of Medicine', 'Dentistry', 'The University Of Western Ontario', 'London', 'Ontario', 'Jung-Ah Choi', 'International Vaccine Institute']

Date: 2022-02 

The development of safe and effective vaccines to prevent SARS-CoV-2 infections remains an urgent priority worldwide. We have used a recombinant vesicular stomatitis virus (rVSV)-based prime-boost immunization strategy to develop an effective COVID-19 vaccine candidate. We have constructed VSV genomes carrying exogenous genes resulting in the production of avirulent rVSV carrying the full-length spike protein (S F ), the S1 subunit, or the receptor-binding domain (RBD) plus envelope (E) protein of SARS-CoV-2. Adding the honeybee melittin signal peptide (msp) to the N-terminus enhanced the protein expression, and adding the VSV G protein transmembrane domain and the cytoplasmic tail (Gtc) enhanced protein incorporation into pseudotype VSV. All rVSVs expressed three different forms of SARS-CoV-2 spike proteins, but chimeras with VSV-Gtc demonstrated the highest rVSV-associated expression. In immunized mice, rVSV with chimeric S protein-Gtc derivatives induced the highest level of potent neutralizing antibodies and T cell responses, and rVSV harboring the full-length msp-S F -Gtc proved to be the superior immunogen. More importantly, rVSV-msp-S F -Gtc vaccinated animals were completely protected from a subsequent SARS-CoV-2 challenge. Overall, we have developed an efficient strategy to induce a protective response in SARS-CoV-2 challenged immunized mice. Vaccination with our rVSV-based vector may be an effective solution in the global fight against COVID-19.

The COVID-19 pandemic has had unprecedented global health, economic and societal impact globally. Vaccinating the majority of the world’s population is the best way to help prevent new infections. Many vaccines have been developed to prevent various viral diseases that are currently in use around the world. This has generated a high demand for these vaccines, putting a strain on production capacity and delivery. With new variants of concern starting to dominate the human pandemic, new derivatives of the current vaccines may be necessary for continued protection from SARS-CoV-2 infection. We have developed a vaccine that uses a safe vesicular stomatitis virus-based delivery vehicle to present a key SARS-CoV-2 protein to our immune system in order to train it to recognize and prevent SARS-CoV-2 infection. Our vaccine completely protected vaccinated animals from SARS-CoV-2 infection and significantly reduced lung damage, a major hallmark of COVID-19.

Funding: Funding for this study was provided in part by CIHR of Canada, cihr-irsc.gc.ca/e/193.html, (COV-440388) to SDB, RMT, EJA, JDD, GAD, SMMH, BF, CYK and by a research contract from Sumagen, sumagen.com , to CYK. Additional funding was provided for RMT by the CIHR and Public Health Agency of Canada, Canada.ca/en/public-health.html, through the COVID-19 Immunity Taskforce (2020-VR2-173205). In addition, this research was also supported by a grant from the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), khidi,or.kr/eps, funded by the ministry of Health & Welfare, Republic of Korea to Sumagen (grant number: HQ21C0129). Na Hyung Kim, Eunsil Choi, Seungho Choo, and Sangkyun Lee are Sumagen scientists who receive salaries from Sumagen. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Herein, we developed and tested multiple rVSV vaccine candidates using different SARS-CoV-2 antigens. In an attempt to induce potent and long-lasting protection against SARS-CoV-2 infection, we developed rVSVs expressing the full-length spike (S F ) protein, S1 subunit, and receptor-binding domain of SARS-CoV-2 spike protein. These proteins were expressed in cis with or without the honeybee melittin signal peptide (msp), at the N-terminus to increase the efficiency of spike protein synthesis, glycosylation and intracellular trafficking, as well as with or without the transmembrane domain and cytoplasmic tail (Gtc) of VSV G protein at the C-terminus to enhance the incorporation of the spike protein into pseudotype VSV [ 24 ]. We found that prime-boost vaccination of mice with dual rVSV serotypes carrying modified SARS-CoV-2 S F antigens induced robust neutralizing antibodies and cell-mediated immune responses against wild-type SARS-CoV-2. Furthermore, vaccination resulted in complete survival of hACE2 transgenic mice challenged with SARS-CoV-2, limited weight loss, lack of infectious virus recovery from the lungs, and significantly reduced pathology in lung tissue.

Vesicular stomatitis virus (VSV) was developed as a live viral vaccine vector for Ebola virus infections and is in pre-clinical trials for other viruses including SARS-CoV-2, HIV-1, hantaviruses, arenaviruses, and influenza viruses [ 13 – 19 ]. For increased vaccine safety, we previously introduced three mutations into the matrix (M) gene of these vectors that resulted in reduced vector cytopathogenicity [ 20 ]. The recombinant VSV-ΔG expressing SARS-CoV-2 spike protein vaccine developed by Yahalom-Ronen [ 19 ], similar to Merck’s Ebola virus vaccine, showed protection against the SARS-CoV-2 challenge in Syrian hamsters. This approach was abandoned due to insufficient VSV vector production mediated by entry through the SARS-CoV-2 spike protein, an observation we have also described herein. Our VSV approach involves non-pathogenic vector production through VSV G-mediated entry and through the introduction of three mutations into the matrix (M) gene that drastically reduced vector cytopathogenicity [ 20 ]. We previously developed a recombinant vesicular stomatitis virus vaccine platform to prime with the Indiana serotype (rVSV Ind ) and to boost with the New Jersey serotype (rVSV NJ ) (or vice versa). This induced robust adaptive immune responses against the inserted gene products [ 20 , 21 ]. Antibodies raised against one serotype of rVSV did not neutralize the other serotype [ 22 ]. Therefore, the impact of anti-vector antibodies arising after the priming immunization was avoided or minimized, resulting in a stronger immune response to the inserted immunogen. The attenuated and replication-competent nature of the rVSV vaccine vector allows for lower levels of vaccine needed per vaccination compared to replication-incompetent vaccine vectors and thus, decreases manufacturing costs and capacity [ 20 ]. Similar viral vectors have demonstrated an excellent safety profile in human trials (reviewed in [ 23 ]).

The ability of SARS-CoV-2 to infect humans is largely due to its envelope spike protein using the human angiotensin-converting enzyme 2 (hACE2) for cell entry [ 4 , 5 ]. The spike protein is cleaved into S1 and S2 subunits and entry is initiated by binding of the receptor-binding domain (RBD) within the S1 subunit to hACE2 on the cell surface. The S2 subunit mediates viral envelope fusion with the host cell membrane. The full-length spike protein (S F ) induces neutralizing antibodies and cell-mediated immune responses, making it a favorable target antigen for vaccine development [ 6 – 9 ]. Several different vaccine platforms have used the S F to induce immunity in humans. Phase 3 clinical trials have shown varying degrees of protection from COVID-19 disease in humans by presenting the S F to induce immunity using different vaccine platforms. Two experimental mRNA vaccines, one from Pfizer/BioNTech and another from Moderna, have shown the highest efficacy to date (95% and 94% respectively) [ 10 , 11 ]. Oxford–AstraZeneca’s chimpanzee adenovirus vector-based vaccine trial has shown a 62.1%-90% efficacy depending on the dose of the priming vaccination [ 12 ]. With high global demand for COVID-19 vaccines, it has been challenging to maintain the current production capacity. This challenge, together with others such as delivery, vaccine hesitancy, and unknown longevity of protection highlights the need for additional vaccine strategies. An estimated two-thirds to three-quarters of global population will likely receive viral vector vaccines for SARS-CoV-2 prevention due to reduced cost of production and delivery compared to mRNA vaccines. If new variants of concern dominate the human pandemic in the future, new derivatives of the current vaccines may be necessary for continued SARS-CoV-2 protection and a concomitant block in viral spread. However, anti-vector immunity induced with the prime-boost delivery of the adenovirus-based vaccines such as those produced by Oxford–AstraZeneca and Johnson & Johnson may prevent reuse of these viral vectors for vaccination for new SARS-CoV-2 variants of concern in the future. Thus, SARS-CoV-2 vaccine development and testing must continue, and inexpensive viral vector-based vaccines, such as the rVSV-SARS-CoV-2 described herein, may be critical for future pandemic control.

Using a multiplex Luminex kit and Magpix, we measured the release of 24 different cytokine/chemokine in human PBMCs exposed to rVSV Ind -SARS-CoV-2-msp-S F -Gtc ( S5A Table ). At 37°C and the more permissive temperature of 31°C, there was minimal release of both Th1 and Th2-type cytokines/chemokines from PBMCs exposed to a low MOI of 0.01 despite replication of the rVSV Ind -SARS-CoV-2-msp-S F -Gtc vector and in comparison to the mitogen PMA/Ionomycin treatment ( S5A Table ). Only at the high MOI of 0.1 and the more permissive 31°C did we observe a low level release of IFN-γ, IL-12, IL-5, IL-18, and IL-28A, or mostly low levels of Th1 cytokines as expected. In the sera of macaques pre- and post-prime-boost immunization, we observed no elevations, significant differences or even trends for differences in the panel of 26 cytokines/chemokines ( S5B Table ). Based on the high level antibody responses and neutralization activity described in Table 1 and S5 Fig , we suspect an immune activation in the primary and secondary lymphoid tissues during prime-boost vaccinations but this did not lead to systemic inflammation.

There was slight reduction of inhibition of the Beta/South Africa/B.1.351 and of the Gamma/Brazil/P.1 SARS-CoV-2 by sera B and C when compared to inhibition of wild-type. In S10 Fig , we show comparable levels of SARS-CoV-2 neutralization by the monoclonal Sino-R001 anti-S NAb (panels A) with neutralization by sera of the immunized macaque B (panel E). However, whereas the monoclonal antibody had minimal cross neutralization of the variants alpha, beta and gamma ( S10F , S10G , and S10H Fig ), it is likely that multiple and/or broadly neutralizing antibodies were present in the sera of immunized macaques (panels F, G, and H). The grey box symbols in panels E to H show the minimal neutralization of the wild-type and VOCs by pre-immunization sera of macaque B.

Sera from the three macaques were all effective at neutralizing wild-type SARS-CoV-2 expressing Nano-Luciferase (nLuc activity used for quantitation) and wild-type SARS-CoV-2 in which production was measured by qRT-PCR. The neutralizing titer for 50% inhibition of the wild-type nLuc was slightly higher with sera A (1/320) and less potent than that of sera B and C at ~1/640 based on duplicate analyses ( S4 Table ). Sera B and C were then tested for neutralization of wild-type Wuhan SARS-CoV-2 as well as Alpha, Beta, and Gamma VOCs. Based on the qRT-PCR analyses of quadruplicate diluted sera assays, the wild-type SARS-CoV-2 was inhibited with slightly less sera B (1/640 to 1/1280) than observed with the wild-type nLuc. The Alpha/UK/B.1.1.7 SARS-CoV-2 showed similar sensitivity as the wild-type SARS-CoV-2 to inhibition by sera B and C of macaques prime-boost immunized with rVSV Ind -SARS-CoV-2-msp-S F -Gtc.

With preliminary findings (Figs 8 and S5 ) of high neutralizing antibody titers in mice immunized in particular with rVSV Ind -SARS-CoV-2-msp-S F -Gtc, we wanted to confirm these observations in other animal models including rabbits and macaques. Rabbit studies are ongoing. Macaques were immediately available from a terminated HIV vaccine study and three animals were IM immunized with 10 9 PFU rVSV Ind -SARS-CoV-2-msp-S F -Gtc as a prime vaccine and boost vaccinated with the same immunogen and dose at 20 days. Details of the protocols are provided in the Supplementary data. Sera utilized in these neutralization assays were collected prior to prime immunization and at 40 days following boost immunization. These three immunized macaques provided sufficient sera, unlike the limited sera availability from mice, for an extensive series of neutralizing titer analyses on sera with multiple VOCs and for subsequent isolation and analyses of neutralizing antibodies. Macaques were not challenged with SARS-CoV-2 due to limited animals and no control group.

We then performed histopathological examination of lungs from control and vaccinated mice groups at 3, 7, and 15 days post-infection (dpi). We observed no abnormal findings in the uninfected control vehicle mice at all time points ( Fig 11 and S7 and S8 Tables). We screened for inflammatory foci characterized by infiltration of inflammatory cells around blood vessels and interstitial alveolar inflammation in the lungs, which are indicators of active SARS-CoV-2 infection, in the challenged hACE2 mice [ 35 ]. In the case of inflammatory foci, we graded inflammation based on the criteria described in the materials and methods. The pathologist was blinded to mice in specific vaccination groups. Most blood vessels in the lung sections of empty vector infected mice (group 1) had prominent and heavy perivascular infiltration of inflammatory cells, which was considered to be related to the virus infection. In comparison, vaccinated mice (groups 2–4) showed very weak cell infiltration, close to normal, but there was recognizable inflammatory cell infiltration around some vessels three days after challenge ( Fig 11 and S7 and S8 Tables). No specific abnormal findings were observed in the lungs of the control vehicle (uninfected) mice (group 5).

Since S F with msp and Gtc modifications yielded the highest level of neutralizing antibodies in C57BL/6 mice, we selected it for use in a SARS-CoV-2 challenge study. We prime vaccinated hACE2 transgenic mice with Indiana serotype of rVSV Ind -msp-S F -Gtc (5x10 8 PFU). Two weeks later, we boost vaccinated with either 5x10 8 PFU rVSV of the Indiana serotype (group 2) or New Jersey serotype (group 3) harboring the same S protein insert (rVSV Ind -msp-S F -Gtc or rVSV NJ -msp-S F -Gtc) ( S6 Table ). Group 4 was the rVSV Ind prime and rVSV NJ boost with the same S insert, using 5x10 7 PFU. All vaccine constructs induced Spike(ΔTM)-specific IgG antibodies (as measured by ELISA) with 100-fold higher levels following boost vaccination ( Fig 7 , day 27). The 5x10 8 PFU dose yielded slightly elevated antibody levels as compared to the 5x10 7 PFU dose ( Fig 7 ). The binding and neutralizing antibody findings were directly correlative. The neutralizing antibody levels also increased by over 100-fold after prime and boost vaccinations and were slightly increased at the higher vaccine dose ( Fig 6 ). Four weeks after boost-vaccination, we challenged the mice intranasally with 1x10 5 PFU of SARS-CoV-2. We monitored the survival and body weight of each mouse daily ( Fig 9 ). All vaccinated hACE2 transgenic mice survived ( Fig 9C ) and no significant weight loss was observed ( Fig 9A ). Non-vaccinated control mice lost weight and three out of five animals died within nine days after the challenge ( Fig 9B and 9C ). Non-vaccinated control mice showed detectable SARS-CoV-2 in the lungs via plaque assay seven days after challenge. In contrast, no SARS-CoV-2 was detectable in the lungs of vaccinated mice ( Fig 10 ). These results showed that the rVSV-SARS-CoV-2-msp-S F -Gtc vaccine can prevent detectable SARS-CoV-2 infection in the lungs of hACE2 transgenic mice. Moreover, a prime-boost vaccination regimen with the same or different rVSV serotype generates comparable levels of neutralizing antibodies. Next, we repeated the challenge studies with a 5-fold higher amount of SARS-CoV-2 to determine if vaccinated hACE2 transgenic mice can be protected from a lethal dose of the challenge virus. We found that the rVSV-SARS-CoV-2-msp-S F -Gtc vaccine protects animals from a lethal dose of SARS-CoV-2 with a minimum amount of weight loss and lung damage ( S8 and S9 Figs). In contrast, the rVSV Ind vector alone immunized mice died with severe lung damage. These results demonstrate that the rVSV-SARS-CoV-2-msp-S F -Gtc vaccine protects transgenic mice from lethal infection.

To investigate the cellular immune responses elicited by our rVSV-SARS-CoV-2 vaccines, we tested T cell immune responses against SARS-CoV-2 Spike protein-derived peptides in an IFN-γ ELISpot assay using a peptide pool (with an 11 amino acid overlap) covering the immunodominant sequence domains of the Spike glycoprotein of SARS-CoV-2 (GenBank MN908947.3, Protein QHD43416.1) ( S7 Fig ). We analyzed splenocytes from mice immunized with the six different SARS-CoV-2 vaccine constructs. Results indicated the presence of responsive IFN-γ-secreting cells from mice in several immunized groups, with rVSV-msp-S F -Gtc immunized mice having the highest number of IFN-γ-secreting cells ( Fig 8A ). As a negative control, HIV-1 Gag peptides showed minimal IFN-γ ELISpot responses for all groups ( Fig 8B ). Overall, the pattern of T cell immune responses elicited by the six different rVSV-SARS-CoV-2 vaccines are consistent with humoral immune responses shown in Fig 5 and Table 1 .

Anti-vector mediated immunity is expected considering both the VSV-G and SARS-CoV-2-S are co-expressed on the VSV vector surface. As compared to other viral vectors for vaccinations, non-pathogenic, replicating VSV vector can achieve stronger immune responses with considerably less vaccine vector load upon immunization. Replication of these vaccines are also self-limiting because the anti-VSV G antibody response will eventually control vector replication. In S6B Fig , we show that the sera of SARS-CoV-2 infected patients and an anti-S SARS-CoV-2 neutralizing antibody did not block the replication of the Indiana serotype of rVSV-SARS-CoV-2-msp-S F -Gtc. This indicates that, despite high levels of the S F on the vector surface, only VSV-G and not SARS-CoV-2 S F mediated vector replication. In S3B Table , we show that sera obtained post rVSV Ind -SARS-CoV-2-msp-S F -Gtc vaccination of mice could neutralize the replication of VSV Ind vector but only at a low sera dilution (1:50) suggesting some level of control of vector replication following vaccination. The sera from mice primed with rVSV Ind -SARS-CoV-2-msp-S F -Gtc had higher neutralizing antibody titers to SARS-CoV-2 (a 50% neutralization titer of ~1/200; Fig 6 ) but that the anti-S NAb in this serum was likely not involved in VSV Ind inhibition (as suggested by S6B Fig ). Sera from mice primed with rVSV Ind -SARS-CoV-2-msp-S F -Gtc could not neutralize the New Jersey serotype vaccine, rVSV NJ -SARS-CoV-2-msp-S F -Gtc ( S3A Table ). Given that the neutralizing SARS-CoV-2 titers of sera increased over 100-fold upon the boost with either the New Jersey or Indiana serotype of rVSV-SARS-CoV-2-msp-S F -Gtc ( Fig 6 ), the boosting vector is clearly inducing enhanced humoral immunity (Figs 6 and 7 ) and protection from SARS-CoV-2 challenges (see below). If there was immediate antibody-mediated vaccine vector reduction based on anti-G Ind antibodies induced by the prime, we would have anticipated much higher anti-S IgG levels and sera neutralization with the New Jersey serotype vector boost than with the Indiana serotype vector boost, because the latter would be controlled by the vector mediated immunity from the prime. Instead the anti-S IgG levels ( Fig 7 ) and sera neutralization ( Fig 6 ) was the same for the homologous VSV Ind -VSV Ind and heterologous VSV Ind -VSV NJ prime-boost.

To assess the possibility that this VSV-based vaccine prevents SARS-CoV-2 infection, we measured the level of SARS-CoV-2 neutralizing antibodies in immunized mice sera on day 13 and/or day 27 sera using a FRNT 50 assay. In all immunized groups, neutralizing antibody titers were significantly increased after boost ( Table 1 and S5 Fig ). Mice immunized with rVSV-msp-S F -Gtc or rVSV-S F had the highest average neutralizing antibody titers (1/13,824 and 1/4,864 respectively) after boost compared to all other groups of immunized mice, with neutralizing titers up to a high of 1/40,960. These results indicate that the S F protein induced stronger neutralizing antibody responses than S1 or RBD+E immunogens, and modification of S F with msp and Gtc further enhanced neutralizing antibody titers. It is not clear why the full-length spike protein induced the highest level of neutralizing antibodies when compared with RBD or S1. The induction of a strong neutralizing antibody may depend on the efficiency of full-length spike protein incorporation into the pseudotype VSV. Moreover, all mice immunized with 5x10 8 PFU induced stronger neutralizing antibody responses compared to mice immunized with 5x10 7 PFU ( Table 1 ). Taken together, a vaccine regimen in mice involving prime immunization with 5x10 8 PFU of rVSV Ind -msp-S F -Gtc, followed two weeks later by boost immunization with 5x10 8 PFU rVSV NJ -msp-S F -Gtc induced highly potent neutralizing antibody responses.

Mice were prime immunized with rVSV Ind -SARS-CoV-2 and boost immunized with rVSV NJ -SARS-CoV-2 two weeks after prime-immunization. Serum was collected to determine SARS-CoV-2 S1 protein-specific antibody levels by ELISA on day 13, one day before boost-immunization, and on day 27, two weeks after boost-immunization. (A) Prime-boost vaccination schedule. (B) Spike(ΔTM)-specific IgG titer after the prime-boost vaccination with doses of 5X10 7 PFU/mouse ( C) Spike(ΔTM)-specific IgG titer after the prime-boost vaccination with doses of 5X10 8 PFU/mouse. Statistical significance was determined by two-way ANOVA with Tukey’s correction (*, p < 0.05; **, p < 0.005; ***, p< 0.001; ns, not significant). The data were presented as means with error bars of standard deviation (n = 5 mice per group). Purple box: honeybee msp, red box: VSV Gtc. VSV-Mock denotes VSV vector alone without any gene insert.

To analyze humoral immune responses towards SARS-CoV-2 S F , S1 and RBD, we vaccinated C57BL/6 mice (N = 5/vaccination group) intramuscularly with rVSV vaccine vectors at 5x10 7 PFU/mouse or 5x10 8 PFU/mouse ( S2 Table ). We prime vaccinated each mouse with rVSV Ind constructs ( Fig 5A ). Two weeks after prime immunization, we boost-immunized the mice with rVSV NJ constructs. We then collected sera on days 13 (one day before boost immunization) and day 27 (two weeks after boost-immunization) ( Fig 5A ). We detected higher levels of Spike(ΔTM)-specific IgG antibodies by ELISA in sera of mice immunized with the 5x10 8 PFU dose than with the 5x10 7 PFU dose, regardless of rVSV construct ( Fig 5B and 5C ). For both doses, Spike(ΔTM)-specific antibody levels were further increased after boost-immunization ( Fig 5B and 5C ). Overall, mice immunized with rVSV-msp-S F -Gtc or rVSV-S F produced the highest level of Spike(ΔTM)-specific antibodies.

To examine immune responses in mice, it was first necessary to purify rVSV-SARS-CoV-2 viral particles by anion-exchange chromatography. One μg of the purified rVSV-SARS-CoV-2 was analyzed by SDS-PAGE and the presence of RBD, S1, S2, and S F was determined by Western blot analysis. (A) Detection of RBD, S1, and S F on VSV particles. (B) Detection of S2 and S F on VSV particles. (C) Detection of VSV Ind and VSV NJ proteins. (D) Depicted model of pseudotype recombinant VSV virions with three different forms of SARS-CoV-2 Spike proteins. rVSV pseudotypes are formed when rVSV-SARS-CoV-2 Spike proteins are expressed with the msp at the NH 2 -terminus and VSV Gtc at the COOH-terminus. Purple box: honeybee msp, red box: VSV Gtc.

To prepare rVSV vaccine vectors for immunization, we purified rVSV containing the SARS-CoV-2 genes, as well as control viruses without SARS-CoV-2 genes, by anion-exchange chromatography. Western blot analysis revealed appreciable quantities of S F , RBD, S1, and S2 incorporated into rVSV pseudotype virions when these proteins were expressed with msp and Gtc ( Fig 4 ). In addition, our results showed that rVSV forms pseudotype virions containing both VSV G and SARS-CoV-2 S glycoproteins (Figs 4D and S4 ). Thus, purified rVSV-SARS-CoV-2-msp-S F -Gtc vaccine virus carries spike proteins of SARS-CoV-2. As expected, msp and Gtc modified S F or S1 protein had higher virion incorporation into the rVSV pseudotypes when compared to the non-modified forms. This difference was particularly apparent for rVSV NJ vaccines where the non-modified S F had no detectable incorporation of S F or S1, and barely detectable S2. Similar quantities of VSV G and M proteins were observed between the different rVSV SARS-CoV-2 vaccine constructs. The results demonstrated that purified rVSV-SARS-CoV-2 carry sufficient RBD, S1, S2, and S F proteins to induce immunity. The relative amount of infectious rVSV-SARS-CoV-2-msp-S F -Gtc vectors harboring gRNA was between 1 and 2% ( S1 Table ) as an estimate of total virus particle count. All VSV particles contain VSV gRNA based on process of rhabdovirus assembly. The ratio of infectious to total virus particles for wild type VSV is estimated to be 1 to 5%, which is similar to that observed with our VSV vaccine vectors [ 30 – 32 ]. Recent studies suggest that aggregation of infectious VSV particles can lead to co-infection of a cell resulting in a single plaque, often reducing the ratio of infectious to non-infectious particles [ 33 , 34 ]. It is important to stress that all infectious, defective-interfering, or non-infectious VSV vectors found as single particle or as an aggregate, will express the SARS-CoV-2 spike protein ( S4 Fig ) and are immunogenic (see below).

Incorporation of SARS-CoV-2 S F , S1, S2, and RBD with or without VSV Gtc into rVSV Ind particles was examined by infecting BHK-21 cells with rVSV Ind -SARS-CoV-2 at an MOI of 3. The rVSV Ind -SARS-CoV-2 infected cells were incubated at 31°C for 6 hrs. Infected cell lysates were prepared in lysis buffer (lanes 1, 2, and 5). Culture media from the infected cells was centrifuged at 500 x g for 10 minutes and supernatant was filtered through a 0.45 μm filter to remove cell debris. The filtered culture media was loaded onto 1 ml of 25% sucrose cushion and ultra-centrifuged at 150,900 x g for 3 hrs. Supernatant on top of the 25% sucrose cushion was collected to check the soluble proteins in the media (lanes 3 and 6). Pelleted samples were checked for proteins incorporated into VSV particles (lanes 4 and 7). We detected RBD, S1, and S F proteins by Western blot using an antibody against the SARS-CoV-2 RBD protein. S2 and S F proteins were detected by rabbit antibody against SARS-CoV-2 S2. (A) Detection of S F and S1 proteins in cell lysate, concentrated culture media, and virus pellet from cells infected with rVSV Ind -msp-S F -Gtc or rVSV Ind -S F . (B) Detection of S F and S2 proteins in cell lysate, concentrated culture media, and virus pellet from cells infected with rVSV Ind -msp-S F -Gtc or rVSV Ind -S F . (C) Detection of VSV Ind proteins in cell lysate, concentrated culture media, and virus pellet from cells infected with rVSV Ind -msp-S F -Gtc or rVSV Ind -S F . (D) Detection of S1 protein in cell lysate, concentrated culture media, and virus pellet from cells infected with rVSV Ind -msp-S1-Gtc or rVSV Ind -S1. (E) Detection of RBD proteins in cell lysate, concentrated culture media, and virus pellet from cells infected with rVSV Ind -msp-RBD-Gtc+E-Gtc or rVSV Ind -msp-RBD+E. (F) Detection of VSV Ind proteins in cell lysate, concentrated culture media, and virus pellet from the cells infected with rVSV Ind -msp-RBD-Gtc+E-Gtc or rVSV Ind -msp-RBD+E. Purple box: honeybee msp, red box: VSV Gtc.

Analysis of the rVSV-RBD constructs showed that while RBD protein, with and without Gtc modification was readily detected in the cell lysates, only RBD with the Gtc addition was detected in virus pellets ( Fig 3E , lane 4). This indicates that VSV Gtc is required for RBD to be incorporated into rVSV particles or released into the supernatant ( Fig 3E , lanes 3 and 6). Taken together, these data show that msp and Gtc modified SARS-CoV-2 S F , S1, and RBD proteins are efficiently incorporated into rVSV particles and are secreted from infected cells. In contrast, msp and Gtc modified S1 protein was not incorporated into rVSV efficiently ( Fig 3D ), suggesting that S2 protein is required for efficient incorporation of S1.

We then used Western blotting to analyze the incorporation of SARS-CoV-2 proteins into viral particles as pseudotype VSV, and their secretion from infected cells. We analyzed centrifuged virus pellets and the remaining supernatant fractions using antibodies to RBD, S2, or VSV and compared them to cell lysates ( Fig 3 ). In contrast to the similar levels observed in the cell lysates ( Fig 2 ), levels of S protein constituents expressed on viral particles, or within the supernatant (free of virus particles), were much higher for the S F and S1 vaccine constructs with msp and Gtc modifications (compare lane 3 & 4 with 6 & 7 in Fig 3A, 3B, 3D and 3E ). We observed similar levels of VSV G and M proteins for each of the rVSV constructs ( Fig 3C and 3F ). This indicates that differences in S protein secretion and virion incorporation were not due to VSV particle production. Results for rVSV Ind ( Fig 3 ) and rVSV NJ ( S3 Fig ) S F vaccine constructs were similar.

To check the expression of SARS-CoV-2 RBD, S1, S F , and E proteins from the rVSV Ind -SARS-CoV-2 infected cells, BHK-21 cells were infected with the virus at an MOI of 6. After six hours incubation at 37°C, cell lysates were prepared and protein expression was determined by Western blot analysis. Cell lysates were loaded in 5 μg quantity for SDS-PAGE. RBD, S1, and S F proteins were detected by rabbit antibody against SARS-CoV-2 RBD. S2 protein was detected by rabbit antibody against SARS-CoV-2 S2. E protein was detected by rabbit antibody against SARS-CoV-2 E peptides. (A) Expression of RBD, S1, and S F with and without msp and Gtc. (B) Expression of S2 with and without Gtc. (C) Expression of E protein. (D) Expression of VSV Ind N, P, M, and G proteins. Purple box: honeybee msp, red box: VSV Gtc.

Codon-optimized full-length Spike protein gene (S F ), S1 subunit gene and the receptor-binding domain (RBD) plus envelope protein genes of SARS-CoV-2 with and without 21 amino acids honeybee melittin signal peptide [(msp) NH 2 -MKFLVNVALVFMVVYISYIYA-COOH] [ 24 ] gene in the purple box, and 49 amino acids VSV G protein transmembrane domain and cytoplasmic tail [(Gtc) NH 2 -SSIASFFFIIGLIIGLFL VLRVGIYLCIKLKHTKKRQIYTDIEMNRLGK-COOH] gene in the red box were inserted into the G and L gene junction of rVSV Ind and rVSV NJ . In addition, 25- nucleotides-long VSV intergenic junctions (5´-CATATGAAAAAAACTAACAGATATC-3´), in the green box, were inserted between genes to provide transcription termination, polyadenylation and the transcription reinitiation sequences. Recombinant viruses were rescued by VSV reverse genetics [ 20 ]. pT7: Bacteriophage T7 promoter for DNA-dependent RNA polymerase. N: VSV Nucleocapsid Protein gene. P: VSV Phosphoprotein gene. M: VSV Matrix protein gene. G: VSV Glycoprotein gene. L: VSV Large protein, RNA-dependent RNA polymerase gene. l: Leader region in the 3´-end of the VSV genome. t: Trailer region in the 5´-end of the VSV genome. HDV: Hepatitis delta virus ribozyme encoding sequences. T7δ: Bacteriophage T7 transcriptional terminator sequences. nt: nucleotides. aa: amino acids.

To construct SARS-CoV-2 vaccines, we individually inserted codon-optimized genes of the receptor-binding domain (RBD), the N-terminal half of the Spike protein (S1), and full-length S (S F ) coding sequences of SARS-CoV-2 between the VSV glycoprotein (G) and polymerase (L) genes in both the New Jersey serotype and Indiana serotype of rVSV vectors ( Fig 1 ). We designed the rVSV-RBD constructs to also encode the SARS-CoV-2 envelope (E) protein because we previously found that RBD plus E induced better neutralizing antibodies against MERS-CoV than RBD alone. We used Western blotting to compare expression levels of S F , S1, S2, RBD and E proteins from the various constructs described in Fig 1 in rVSV Ind -infected BHK-21 cells. S1 and S2 proteins were generated after cleavage of S F by the cellular protease furin [ 25 , 26 ]. We detected robust expression of RBD ( Fig 2A ), S F ( Fig 2A ) and S1 cleavage product ( Fig 2A ) using an anti-RBD antibody, and of S F and S2 proteins ( Fig 2B ) from the VSV-S F and VSV-msp-S F -Gtc constructs using the anti-S2 antibody. The anti-S2 antibody detected its binding domain less efficiently in the S F compared to the S2 cleavage product. We noted a moderate level of E protein expression in cells infected with rVSV carrying RBD and E protein genes using the anti-E antibody (VSV-msp-RBD+E, Figs 2C and S1C ). For unknown reason, we did not detect expression of E protein with the VSV-msp-RBD-Gtc+E-Gtc construct. As expected, the addition of VSV Gtc to RBD and S1 increased their protein molecular mass by 5.5 kDa ( Fig 2A lanes 2–5). In contrast, we did not observe a difference in levels of S protein constituents in the cell lysate with the addition of msp or Gtc. All cell lysates contained a similar amount of VSV proteins ( Fig 2D ). Expression levels of S F , S1, S2, and RBD proteins in rVSV NJ infected cells ( S1 Fig ) were very similar to levels in rVSV Ind infected cells ( Fig 2A, 2B and 2C ).

Discussion

This study was designed to: a) develop an effective SARS-CoV-2 vaccine for sustainable and prolonged delivery of antigenic forms of the S protein, and b) to induce effective adaptive immunity for protection against a SARS-CoV-2 challenge. The use of live attenuated rVSV as a vaccine vector has several advantages. Replication-competent forms of rVSV have demonstrated safety and immunogenicity in multiple clinical trials [36,37]. Vaccination with live rVSV stimulates both humoral and cellular immunity to generate long-lasting immunity [38–42]. Human VSV infections are extremely rare, therefore pre-existing immunity to the rVSV vector will be low. In addition, rVSV can be grown to high titers in cell culture, allowing for rapid and efficient biomanufacturing of large vaccine stocks. The VSV vector is nonpathogenic in humans and has been previously employed in a vaccine that delivered the Ebola Glycoprotein (GP) as an immunogen to induce protective responses (reviewed in [43]). VSV lacking its own G glycoprotein is able to replicate in vaccinated humans due to the expression and pseudotyping of the virus particles with Ebola GP. However, unlike the Ebola GP, which utilizes a variety of host cell attachment factors (i.e., C-type lectins, T-cell immunoglobulin and mucin domain 1, and Tyrosine kinase receptor Axl) and macropinocytosis for entry (reviewed in [44]), SARS-CoV-2 primarily uses the cell surface ACE2 protein that is expressed differentially in many different cells of the body [45–49]. The recombinant VSV-ΔG-spike vaccine developed by Yahalom-Ronen [19] showed protection against SARS-CoV-2 challenge. This vaccine uses a G protein gene deleted single serotype VSV vector that resembles Merck’s Ebola virus vaccine. The ΔG VSV vector-based vaccine produces a low yield of virus which poses a major challenge for large quantity vaccine production. Our VSV-based SARS-CoV-2-msp-S F -Gtc vaccine utilizes the VSV G protein for vaccine production on an industrial scale. VSV G also allows for vector replication following vaccination, which is necessary for effective and prolonged antigen presentation. Our rVSV vectors (described below) contain mutations that we and others have shown to attenuate cytopathogenicity and neurovirulence [20,38,42,50–54]. In our studies herein measuring 24 cytokine/chemokines, human PBMCs exposed to or macaques vaccinated with replicating VSV-based SARS-CoV-2-msp-S F -Gtc did not induce considerable PBMC activation or systemic immune activation in non-human primates. Thus, in addition to the vector attenuation and eventual immune clearance, the VSV-based SARS-CoV-2-msp-S F -Gtc will induce a robust S-specific immune response but is unlikely to result in systemic immune activation. In addition, our results showed homologous and heterologous boost produce equally robust immune responses. This suggests that anti-vector immunity induced by prime vaccination did not limit the boost induced by homologous rVSV. It is also possible that anti-spike immunity in vaccinated or COVID-19 convalescent individuals could neutralize rVSV infectivity resulting in reduced immunogenicity. However, our results (S6 Fig) indicate that sera containing high levels of anti-spike neutralizing antibodies were unable to neutralize our rVSV-SARS-CoV-2-msp-S F -Gtc vaccine. This suggests that anti-spike immunity is unlikely to present a barrier to efficacy of the rVSV vaccine, likely due to sufficient VSV G protein-mediated entry.

We have used avirulent M gene mutants of two different serotypes of VSV for our COVID-19 vaccine development. We have recently demonstrated the efficacy of prime-boost vaccination against the Zika virus [21]. We demonstrated that rVSV Ind -ZIKV-E prime vaccination followed by rVSV NJ -ZIKV-E boost vaccination induced highly potent neutralizing antibodies and T cell responses, and protected type 1 interferon receptor knockout (Ifnar-1-) mice from a lethal dose of the ZIKV challenge. We used the same VSV vectors for this present study. We show that the addition of honeybee melittin signal peptide (msp) and the addition of the transmembrane domain of the VSV G protein (Gtc) to the spike protein of SARS-CoV-2, made pseudotype VSV that carry the spike protein more efficiently, and triggered strong immune responses. In this study we also analyzed the safety of our M gene mutants of VSV Ind (GML) and VSV NJ (GMM). We found these two M gene mutants are non-cytolytic and non-lethal when we injected highly sensitive type 1 interferon receptor knockout transgenic (Ifnar-1-) mice. All VSV Ind (GML) and VSV NJ (GMM) injected transgenic mice have survived and had normal weight gain [21].

Regardless of the levels of the S protein derivatives expressed on the rVSV particles, our vaccine is dependent on the quality and quantity of the elicited immune response towards the S protein. In C57BL/6 as well as hACE2-transgenic mice, rVSV expressing the msp-S F -Gtc induced the most significant level of anti-S protein antibodies after priming and boosting immunization. Anti-S antibody titers of over 105 and nearly 106 were observed with low (5x107 PFU) and high (5x108 PFU) dose priming and boosting immunizations, which appears to be among the highest reported in mice as compared to other published SARS-CoV-2 vaccines. Likewise, we observed strong IFN-γ lymphocyte responses to pooled S peptides, the highest with the 5x108 PFU priming and boosting immunization dose. The high levels of anti-S antibody response in immunized mice recapitulated in the hACE2 transgenic mice vaccinated with a prime-boost of the rVSV msp-S F -Gtc construct. More importantly, this vaccine also provided protection from SARS-CoV-2 challenge as discussed below. We suspect that the high level expression and processing of S F driven by an infectious replication-competent virion in an infected antigen-presenting cell (APC) will lead to good MHC I presentation. Furthermore, because these virus vectors are displaying the S F protein on uptake by APC as well as taking up secreted S F protein, this should lead to cross-presentation of T cell epitopes by APC on MHC I. Thus, this should further enhance a CTL response. The display of S F on the virion, as well as the uptake of secreted S F will also lead to a stronger more efficient MHC II presentation by APC. The CD4 helper responses augment CTL, but more importantly provide robust support for B cell responses that include augmented production of neutralizing antibody, especially after the boost. Of note, the S F protein with the msp and Gtc induced higher neutralizing antibody titers and a higher IFN-γ response when compared with the same S F protein without exogenous signal peptides and transmembrane domain.

A boost with the same vaccine vector can potentially lead to activation of a VSV vector-based immune response that eliminates or reduces effectiveness of the boost. One possibility for the higher efficacy and protection from SARS-CoV-2 with the “unintentional arm” of the Oxford/AstraZeneca vaccine trial may relate to the lower dose SARS-CoV-2 S vaccine used in the priming immunization. The lower dose of priming may have resulted in a reduced anti-vector immune response that in turn did not reduce the boost vaccination with the same adenovirus vector. The Gameleya Institute vaccine (Sputnik V) approach involved the use of adenovirus 5-SARS-CoV-2 S priming immunization followed by a boost with adenovirus 26-SARS-CoV-2 S in attempts to reduce vector-mediated immunity. Published reports suggest that the dual Adenovirus vector approach of Gameleya was more effective at preventing infection than the Oxford/AstraZeneca standard dose priming/standard dose boosting approach (91.6% versus 62.1%) using the higher dose prime/boost [11,55]. Prior to the results of these trials, we optimized S protein antigen expression from a dual VSV serotype vaccine vector system involving rVSV Ind and rVSV NJ , different serotypes that do not result in cross neutralization by antibodies. In the vaccination studies of hACE2 transgenic mice, a high titer prime with the rVSV Ind followed by a high titer boost with the rVSV NJ induced the same production of anti-S F binding antibodies and same levels of sera neutralization of SARS-CoV-2 as did prime-boost with just the rVSV Ind expressing the same msp-S F -Gtc. Likewise heterologous prime-boost with rVSV Ind /rVSV NJ also resulted in the complete survival of the SARS-CoV-2 challenged hACE2 transgenic mice, with limited weight loss, and lacking infectious virus recovery from the lungs, and significantly reduced abnormal pathology in lung tissue. A 10-fold decrease in neutralizing antibodies by reducing the heterologous prime-boost dose (5x107 PFU versus 5x108 PFU) still resulted in complete protection against SARS-CoV-2 challenge. These findings suggest that heterologous versus homologous rVSV prime-boost may not be superior for SARS-CoV-2 protection in hACE2-transgenic mice. However, based on the Gameleya Institute and Oxford/AstraZeneca vaccine Phase III trials, there may be concern of vector-mediated immunity reducing protective efficacy that was not borne out in the small animal studies.

Many countries are experiencing a fourth wave of SARS-CoV-2 infection, mainly due to the delta variant of SARS-CoV-2 and a few vaccine makers have proposed to provide a second boost (third vaccination) in order to protect vaccinees from the variant coronavirus infection. Because our rVSV-SARS-CoV-2-msp-S F -Gtc vaccine induces strong neutralizing antibodies, our vaccine can be used as an effective boost vaccine as well after immunization with other vaccines. Sera from prime/boost immunized macaques with VSV-based SARS-CoV-2-msp-S F -Gtc showed strong neutralization of the “wild-type” Wuhan SARS-CoV-2, as observed with the sera obtained from vaccinated mice. Since more sera were available from immunized macaques, we also showed strong cross neutralization of the alpha, beta and gamma variants. Studies are now underway with the delta variant. Although there are differences between immunization protocols and immunogens, our prime/boost with rVSV-SARS-CoV-2-msp-S F -Gtc had higher levels of binding antibodies to S and higher neutralization titers from sera in mice and macaques than did the prime/boost vaccination studies performed with mice and macaques using the adenovirus vector vaccine, ChAdOx1-S [56]. However, these vaccines were not compared head-on in the same study. Finally, we believe that there would be no problem using our rVSV-SARS-CoV-2-msp-S F -Gtc as a boost vaccine following previous vaccination. Our vaccine may be very useful to increase the level of neutralizing antibodies in people who have been vaccinated with other vaccines.

[END]

[1] Url: https://journals.plos.org/plospathogens/article?id=10.1371/journal.ppat.1010092

(C) Plos One. "Accelerating the publication of peer-reviewed science."
Licensed under Creative Commons Attribution (CC BY 4.0)
URL: https://creativecommons.org/licenses/by/4.0/


via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/