key: cord-0791977-e670kcd8
authors: Sullivan, Edward; Sung, Po-Yu; Wu, Weining; Berry, Neil; Kempster, Sarah; Ferguson, Deborah; Almond, Neil; Jones, Ian M.; Roy, Polly
title: SARS-CoV-2 Virus like Particles produced by a single recombinant baculovirus generate potent neutralizing antibody that protects against variant challenge
date: 2021-11-30
journal: bioRxiv
DOI: 10.1101/2021.11.29.470349
sha: b40b38c17cce3b2a0fa588cb44e1c0a45a1a74ce
doc_id: 791977
cord_uid: e670kcd8

The Covid-19 pandemic caused by SARS-CoV-2 infection has highlighted the need for the rapid generation of efficient vaccines for emerging disease. Virus-like particles, VLPs, are an established vaccine technology that produces virus-like mimics, based on expression of the structural proteins of a target virus that can stimulate strong neutralizing antibody responses. SARS-CoV-2 is a coronavirus where the basis of VLP formation has been shown to be the co-expression of the spike, membrane and envelope structural proteins. Here we describe the generation of SARS-CoV-2 VLPs by the co-expression of the salient structural proteins in insect cells using the established baculovirus expression system. VLPs were heterologous ∼100nm diameter enveloped particles with a distinct fringe that reacted strongly with SARS-CoV-2 convalescent sera. In a Syrian hamster challenge model, a non-adjuvanted VLPs induced neutralizing antibodies to the VLP-associated Wuhan S protein, reduced virus shedding following a virulent challenge with SARS-CoV-2 (B.1.1.7 variant) and protected against disease associated weight loss. Immunized animals showed reduced lung pathology and lower challenge virus replication than the non-immunized controls. Our data, using an established and scalable technology, suggest SARS-CoV-2 VLPs offer an efficient vaccine that mitigates against virus load and prevents severe disease.

Disease 2019 . Since its initial identification in Wuhan, China, the virus has spread worldwide resulting in, to date, over 250 million confirmed cases and 5 million deaths. Genetic analysis of SARS-CoV-2 has shown that it shares 79.5% sequence homology with SARS, which emerged in 2003 (1) .

SARS-CoV-2 is a member of the Betacoronavirus genus within the family Coronaviridae. The particle consists of a single copy of the 29.9kb positive sense single stranded RNA genome associated with the viral nucleocapsid protein (N) which is surrounded by an envelope derived from the host cell plasma membrane, gained at the time of virus budding, embedded with 3 further viral structural proteins, the Spike (S), Membrane (M) and Envelope (E) proteins (2) . In addition to the structural proteins, coronaviruses express a large number of non-structural proteins, essential for replication but not present in the virus particle (3) .

The most abundant of the Coronavirus (CoV) envelope-associated proteins is the M protein, a triple membrane-spanning glycoprotein that controls the conformation of the viral envelope (4) . The M protein oligomerizes to form lattices on the membranes of the ER-Golgi intermediary compartments, resulting in membrane distortion (3) . In addition, M directly interacts with the other coronavirus structural proteins enabling their recruitment into the nascent virus particle, such that the M protein is considered the primary driver of coronavirus assembly (2, 5) . The envelope associated E protein is a small, 12kDa protein, which some studies have shown to be a viroporin (5) . Very little E is incorporated into nascent virions, the majority remaining the ER and Golgi (6) but its absence impairs virus budding and leads to an attenuated phenotype (5, 7, 8) . Co-expression of E and M in isolation is sufficient to cause the formation and release of virus-like particles (VLPs) (5, 9, 10) , which can also incorporate the S protein if it is expressed in the same cells (9, 11, 12) . The trimeric S protein is a type 1 transmembrane glycoprotein responsible for receptor binding and membrane fusion (13) . S binds to angiotensin conversion enzyme II (ACE2) on the host cell surface via the outermost S1 domain (14, 15) leading to serine protease TMPRSS2 dependent cell entry driven by the S2 domain (16) . S is also the immunodominant antigen of SARS-CoV-2 and antibodies that block the interaction with S1 with ACE2 are neutralizing (17) . All currently approved vector-based or RNA vaccines for Covid-19 rely solely on S for their protective responses (18).

Previously, we described VLPs for SARS-CoV following expression of the requisite structural proteins in insect cells using a recombinant baculovirus (9) . VLPs offer multivalent presentation of the S protein to stimulate humoral immunity as well as uptake by antigen presenting cells for the generation of T-cell responses (19, 20) and VLP based vaccines for other viruses, produced by the baculovirus insect cell system, are currently in use, proving evidence of scalability and acceptability (21, 22).

Here we report the generation of SARS-CoV-2 VLPs based on the co-expression of S, M and E proteins in insect cells. Purified VLPs presented as heterogeneous vesicular structures which were strongly recognized by convalescent human sera. Small mammal immunization trials generated neutralizing antibody that protected against subsequent variant live virus challenge. Our data demonstrate that SARS-CoV-2 VLPs produced by an established and scalable technology are a feasible vaccine for Covid-19.

A codon optimized sequence encoding the S protein of the original Wuhan SARS-CoV-2 isolate (QHD43416.1) was obtained from Integrated DNA Technologies as described (23) 

Spodoptera frugiperda (Sf9) cells were used for generation of recombinant baculoviruses and for subsequent virus amplification. For production of VLPs Trichoplusia ni (Tnao38) cells (26) were used.

Virus was grown at an MOI of 0.01 and infected cells were incubated for up to 6 days before harvesting the supernatant. Virus infection for protein expression and VLP assembly employed an MOI of 5 and incubation for 4 hours, after which the cells were pelleted and re-suspended in fresh culture medium.

Infected cells were then incubated for a further 72 hours. The resultant supernatant was harvested, clarified by centrifugation twice at 4500 x g, 4°C for 20 minutes, and the VLPs then pelleted by centrifugation through a 25% sucrose cushion, at 100,000 x g, 4°C for 100 minutes. VLPs were resuspended in PBS and purified by centrifugation on a 20-60% sucrose gradient at 100,000 x g, 10°C for 18 hours. The gradients were fractionated and the presence of S protein assessed by dot blot with CR3022 (Absolute Antibody). The pixel density in each dot was assessed using Image J (27).

Samples were resolved on Bolt TM 4-12% Bis-Tris Plus gradient gels (Invitrogen). Gels were either stained with Coomassie brilliant blue or subjected to Western blot using appropriate antibodies: rabbit polyclonal anti-SARS-CoV-2 S antibody (Abcam ab272504) or rabbit polyclonal anti-SARS-CoV/CoV2 M antibody (Novus Biologicals NB100-56569).

Peak fractions from the sucrose gradient, identified by western blot with an anti-S antibody, were diluted 4-fold in PBS and concentrated to 10% of their original volume using spin filters with a cut-off of 1MDa. Carbon coated formvar grids were floated on droplets of the concentrated samples for 5 mins followed sequentially by 5 mins on a droplet of water and then on 1% uranyl acetate. The grids were blotted and allowed to dry before examination in a JEOL 2100 Plus microscope operating at 200kV.

For the tests of VLP antigenicity, 96-well Nunc Maxisorb plates (Thermo Fisher Scientific) were coated with 0.5μg of purified VLPs per well diluted in 50μl of carbonate coating buffer (15mM Na2CO3, 36mM NaHCO3, pH9.6) and incubated overnight at 4°C. The plates were washed with PBS containing 0.05% Tween-20 (PBST) and blocked with Sea Block blocking buffer (Thermo Fisher Scientific) for 1 h at room temperature. 2-fold serial diluted human sera in blocking buffer were added to the plates and incubated for 1 h at room temperature. After three washes, the bound antibody was detected with goat anti-human Ig (γ chain specific)-HRP conjugated secondary antibody (Thermo Fisher Scientific) diluted 1:10,000 in blocking buffer. The assay reaction was developed with the 1-step Ultra TMB-ELISA substrate (Thermo Fisher Scientific). The reaction was left to develop for 5 min and then stopped with 2M H2SO4. The optical density was determined at 405nm using a plate reader. A surrogate marker of the neutralizing antibody response following immunisation, the cPass™ SARS-CoV-2 Neutralization Antibody Detection Kit (Genscript, L00847), was used as directed to detect antibodies that bind competitively to the receptor binding domain (RBD) of SARS-CoV-2.

Five female Golden hamsters were administered 300µL (10g) of non-adjuvanted VLP preparation under the skin on two occasions, 4 weeks apart. A control group of 5 hamsters did not receive any treatment. Two weeks after the second immunization hamsters were challenged intranasally with 50µL (1. 

20µL of the VTM sample was added to duplicate cultures of VeroTMPRSS2 cells either neat or diluted 1:2 in VTM. After 3 days the supernatant was removed, and the cells fixed for 30 minutes with 4% neutral buffered formalin followed by staining with 0.5% methyl violet. Where cell death was observed the presence of SARS-CoV-2 in cell culture supernatant was verified by qRT-PCR as described above.

A sample was considered positive by culture if at least one of the four replicate wells showed destruction of the cell monolayer and was positive for SARS-CoV-2 by qRT-PCR.

The BioSensor SARS-CoV-2 Ag Kit (Oxford Biosystems) was used in accordance with manufacturer's instructions to process swab samples from hamsters. Sample were added to lateral flow buffer and applied to the device and the output scored according to the WHO scoring system (https://extranet.who.int/pqweb/vitro-diagnostics/performance-evaluation). The LFDs were imaged and the relative pixel density in the test to control lanes was measured using Image J (27).

Intact lungs were collected at post-mortem, fixed for 72 hours at room temperature in 10% neutral buffered formalin (Sigma) and embedded in paraffin wax (VWR) in accordance with standard histological processes. Four micron thick sections were mounted on poly-L-lysine coated slides (VWR), de-waxed with xylene (Thermo Fisher Scientific) and re-hydrated via graded ethanol:water solutions (Thermo Fisher Scientific). Hematoxylin and eosin sections were evaluated for pathological changes associated with disease and assigned a score for each variable by two veterinary pathologists, independently. The samples were scored blinded.

Immunohistochemical staining was performed using the Leica Bond Polymer Refine staining system (Leica Microsystems DS9800). Onboard de-waxing was performed in accordance with the standard Leica Bond protocol and staining undertaken using IHC Protocol F with the following adaptations:

additional non-specific block prior to primary antibody incubation (10% normal horse serum (Biorad), 

A recombinant baculovirus driving the expression of SARS-CoV-2 S and M proteins from the polyhedrin promoter and the SARS-CoV E protein from the p10 promoter was constructed as described ( Figure   1 ). Individual recombinant viruses expressing S, M or M+E were also constructed. The SARS-CoV and SARS-CoV-2 E proteins share 97% homology and are functionally exchangeable for the purposes of VLP production. Moreover, only a low level of E is incorporated into virus particles and it is not Other proteins present at lower levels in the VLP preparation were insect cell membrane proteins present at the sites of VLP budding, as previously described by others (32).

Fractions from the S monoclonal antibody strongly reactive ~35% sucrose fraction were coated to TEM grids and analyzed by negative stain electron microscopy. Among the many heterogeneous vesicles present were vesicles of ~100nm diameter that displayed a distinct fringe of ~10nm in depth that were typical of the crown-like spikes which define coronavirus morphology (12, 33) . Baculoviruses were absent from most fields of view consistent with their sedimentation to the bottom of the gradients used ( Figure 4) .

To assess relevant antigenicity, VLPs present in the peak fraction were coated to ELISA plates and probed with antibody positive convalescent sera previously screened using baculovirus expressed SARS-CoV-2 S1 protein (23). Multiple convalescent sera but not a naïve serum controls reacted strongly with immobilized VLPs ( Figure 5A ). For SARS-CoV-2 spike, extensive pepscanning of convalescent sera has shown that reactivity with linear epitopes is relatively poor (34, 35) suggesting that that VLPs used here display S in an antigenic form suitable for conformational antibody binding.

To ensure that the sera used were not biased for insect cell expressed S1 protein and bound VLP displayed S preferentially, for example as a result of the high mannose glycan content of insect cell expressed proteins (36) , VLP coated plates were also probed with a wholly different patient sera cohort pre-screened by peptide array (37) . As before, reactivity was high for sera pre-scored as positive for infection but not for the serum pre-scored as negative ( Figure 5B) . Thus, insect cell derived SARS-CoV-2 VLPs are strongly antigenic.

As the composition, appearance and immune recognition of the VLP preparation derived from insect cells infected with the VLP recombinant baculovirus was consistent with previous studies of CoV VLPs (12, 38, 39) , insect cell derived VLPs bearing the SARS-CoV-2 S protein were assessed as a candidate vaccine in a hamster challenge model (40) . Infection of Syrian hamsters with SARS-CoV-2 shows replication of the virus in the lungs with pathology similar to that reported for COVID-19 and a neutralizing antibody response that protects from subsequent infection (41, 42) . Hamsters were vaccinated with gradient purified VLPs (n = 5, dose = 10g per immunization) and boosted with the and scored for clinical signs as described (40) . Neutralizing antibody titers rose within 4 days of challenge in the VLP immunized group and later also in the control group ( Figure 6B) . A single animal culled from each group at two days post challenge showed higher anti-RBD antibody titers in the VLPimmunized animal when compared with the control ( Figure 6C ).

The oral swabs of all animals from day 1 post challenge were positive for virus by genomic RNA RT-PCR but at day 7 post challenge were lower in the VLP immunized groups when compared with the controls ( Figure 7A ) and this difference was significant in the sampled nasopharyngeal swabs (p=0.044, one-tailed t-test) ( Figure 7B ). Virus antigen positivity by LFD of the oral swabs was strongly positive for both immunized and control groups at day 2, reduced at day 4, with a trend towards lower antigen levels in the immunized animals ( Figure 7C (two animals from each group at day 2) & D (four animals from each group at day 4)) and absent by day 7 to the end of the experiment. Culturable virus was recovered from oral swabs throughout days 1 to 4 post-challenge with no discernible difference between the groups but not from any animal thereafter (not shown).

Both groups lost weight for 5 days post challenge, but this was arrested in the VLP group from day 6 whereas the control group continued to lose weight for a further 3 days. Although all animals recovered by 14 days post challenge, the difference in weight loss between the groups at 5-8 days post challenge was statistically significant (p<0.05, ANOVA post-hoc Bonferroni) with the immunized animals recovering quicker (Figure 8 ).

Pathology scoring of lung inflammation and disease revealed that the group that received the VLPs had markedly reduced bronchiolar and alveolar inflammation compared with control animals at 2 days post challenge, a feature which was retained to the end of the monitoring period at 10 days postchallenge (Figure 9 upper panels). Immunohistochemistry for the presence of the S protein in inflamed lung tissue showed clumps of S positive syncytial cells were present in both groups but that the overall level of S protein was much higher in the non-vaccinated animals when compared to those immunized with the VLPs at both of the time points tested (Figure 9 lower panels).

The emergence of SARS-CoV-2 into the human population, most likely by spill over from a zoonotic reservoir (43) , has led to the rapid development and licensing of several efficacious vaccines including those based on inactivated virus, adenovirus vectors, or direct nucleic acid (18, 44, 45) . Other experimental vaccines, including VLP based vaccines have also been reported (38, 46, 47) . Based on a previously described VLP of SARS CoV (9) we report here the development and test of a VLP for SARS-CoV-2 based on the co-expression of the S, M and E proteins using the baculovirus expression system.

VLPs were shown to consist of the requisite CoV structural proteins, which were assembled into vesicle-like structures as shown by EM analysis. Further, these VLPs reacted with SARS-CoV-2 infected patient sera from two different serum sets, confirming their antigenicity and the combination of Western blot positivity, visualization and antigenic reactivity led to their use in trial immunizations. As an immunogen, non-adjuvanted SARS-CoV-2 VLPs generated a serum response in all immunized animals that competed with a standard RBD antibody-S protein interaction, consistent with neutralizing activity, which was boosted further following live virus challenge. No adverse events were recorded following immunisation consistent with extensive previous tests of insect cells expressed material as vaccines (22, 48) . Immunisation with the SARS-CoV-2 VLPs did not prevent replication of the challenge virus but virus titers in the oral cavity, and particularly in the nasopharyngeal samples, were reduced when compared with the control group. The SARS-CoV-2 VLPs used here were assembled using the Wuhan S protein but the challenge SARS-CoV-2 virus stock was the B1.1.7 S variant that carries the N501Y residue change, which has been reported to improve ACE-2 binding and partly evade the neutralizing antibody response (49, 50) . In addition, the VLPs was administered parenterally while the challenge virus was applied directly to mucosal surfaces in the nasal passage. Fractions were diluted to reduce the sucrose and re-concentrated before adsorption to carbon coated grids. The grids were strained with 2% uranyl acetate. Peak fractions (cf. Figure 3 ) of the VLPs showed multiple vesicle like structures among which are many with fringe like projections on their surface (arrowed). Two typical fields of the same adsorbed sample are shown. 

Discovery of a novel coronavirus associated with the recent pneumonia outbreak in humans and its potential bat origin

The molecular biology of coronaviruses

Membrane binding proteins of coronaviruses

A structural analysis of M protein in coronavirus assembly and morphology

Coronavirus envelope protein: current knowledge

Coronavirus envelope (E) protein remains at the site of assembly

A severe acute respiratory syndrome coronavirus that lacks the E gene is attenuated in vitro and in vivo

Absence of E protein arrests transmissible gastroenteritis coronavirus maturation in the secretory pathway

Efficient assembly and release of SARS coronavirus-like particles by a heterologous expression system

Subcellular location and topology of severe acute respiratory syndrome coronavirus envelope protein

Immune responses against severe acute respiratory syndrome coronavirus induced by virus-like particles in mice

SARS-CoV-2 viral budding and entry can be modeled using BSL-2 level virus-like particles

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

Functional assessment of cell entry and receptor usage for SARS-CoV-2 and other lineage B betacoronaviruses

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

SARS-CoV-2 neutralizing antibody structures inform therapeutic strategies

Two linear epitopes on the SARS-CoV-2 spike protein that elicit neutralising antibodies in COVID-19 patients

Expression-systemdependent modulation of HIV-1 envelope glycoprotein antigenicity and immunogenicity

Preexisting and de novo humoral immunity to SARS-CoV-2 in humans

Construction of SARS-CoV-2 Virus-Like Particles by Mammalian Expression System

Safe and effective twoin-one replicon-and-VLP minispike vaccine for COVID-19: Protection of mice after a single immunization

SARS-CoV-2 spike D614G variant confers enhanced replication and transmissibility

Simulation of the Clinical and Pathological Manifestations of Coronavirus Disease

Animal models for COVID-19

The origins of SARS-CoV-2: A critical review

Vaccine development and technology for SARS-CoV-2: current insights

The Current Status of COVID-19 Vaccines

A COVID-19 vaccine candidate using SpyCatcher multimerization of the SARS-CoV-2 spike protein receptor-binding domain induces potent neutralising antibody responses

Two-Component Nanoparticle Vaccine Displaying Glycosylated Spike S1 Domain Induces Neutralizing Antibody Response against SARS-CoV-2 Variants

The baculovirus expression vector system: A commercial manufacturing platform for viral vaccines and gene therapy vectors

Escape from neutralizing antibodies by SARS-CoV-2 spike protein variants

SARS-CoV-2 spike variants exhibit differential infectivity and neutralization resistance to convalescent or post-vaccination sera

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

Rapid assessment of SARS-CoV-2 evolved variants using virus-like particles

An Overview of Influenza Viruses and Vaccines. Vaccines (Basel)

Virus-like particles exhibit potential as a pan-filovirus vaccine for both Ebola and Marburg viral infections

Characterization of HIV-1 virus-like particles and determination of Gag stoichiometry for different production platforms

Vaccine potential of Nipah virus-like particles