key: cord-0741036-5029z4lq
authors: Tan, Tiong Kit; Rijal, Pramila; Rahikainen, Rolle; Keeble, Anthony H.; Schimanski, Lisa; Hussain, Saira; Harvey, Ruth; Hayes, Jack W.P.; Edwards, Jane. C.; McLean, Rebecca K.; Martini, Veronica; Pedrera, Miriam; Thakur, Nazia; Conceicao, Carina; Dietrich, Isabelle; Shelton, Holly; Ludi, Anna; Wilsden, Ginette; Browning, Clare; Zagrajek, Adrian K.; Bialy, Dagmara; Bhat, Sushant; Stevenson-Leggett, Phoebe; Hollinghurst, Philippa; Tully, Matthew; Moffat, Katy; Chiu, Chris; Waters, Ryan; Gray, Ashley; Azhar, Mehreen; Mioulet, Valerie; Newman, Joseph; Asfor, Amin S.; Burman, Alison; Crossley, Sylvia; Hammond, John A.; Tchilian, Elma; Charleston, Bryan; Bailey, Dalan; Tuthill, Tobias J.; Graham, Simon P.; Malinauskas, Tomas; Huo, Jiandong; Tree, Julia A.; Buttigieg, Karen R.; Owens, Raymond J.; Caroll, Miles W.; Daniels, Rodney S.; McCauley, John W.; Huang, Kuan-Ying A.; Howarth, Mark; Townsend, Alain R.
title: A COVID-19 vaccine candidate using SpyCatcher multimerization of the SARS-CoV-2 spike protein receptor-binding domain induces potent neutralising antibody responses
date: 2020-08-31
journal: bioRxiv
DOI: 10.1101/2020.08.31.275701
sha: caf84222e36b53b61351c95a5a7c3723f36123d1
doc_id: 741036
cord_uid: 5029z4lq

There is dire need for an effective and affordable vaccine against SARS-CoV-2 to tackle the ongoing pandemic. In this study, we describe a modular virus-like particle vaccine candidate displaying the SARS-CoV-2 spike glycoprotein receptor-binding domain (RBD) using SpyTag/SpyCatcher technology (RBD-SpyVLP). Low doses of RBD-SpyVLP in a prime-boost regimen induced a strong neutralising antibody response in mice and pigs that was superior to convalescent human sera. We evaluated antibody quality using ACE2 blocking and neutralisation of cell infection by pseudovirus or wild-type SARS-CoV-2. Using competition assays with a monoclonal antibody panel, we showed that RBD-SpyVLP induced a polyclonal antibody response that recognised all key epitopes on the RBD, reducing the likelihood of selecting neutralisation-escape mutants. The induction of potent and polyclonal antibody responses by RBD-SpyVLP provides strong potential to address clinical and logistic challenges of the COVID-19 pandemic. Moreover, RBD-SpyVLP is highly resilient, thermostable and can be lyophilised without losing immunogenicity, to facilitate global distribution and reduce cold-chain dependence.

computationally engineered dodecahedron based on an aldolase from a thermophilic 132 bacterium 20, 25 . The SpyTag-RBD can be efficiently conjugated to the SpyCatcher-mi3 133 VLP, with 93% display efficiency reached after 16 h ( Figure 1B ). This corresponds to 134 an average of 56 RBDs per VLP. We saw no sign of VLP aggregation following 135 coupling and the RBD-SpyVLP is homogeneous, as shown by a uniform peak of the 136 hydrodynamic radius (RH) at 20.7 ± 4.2 nm in dynamic light scattering (DLS) (Figure  137 1C). For immunisation, we chose a conjugation ratio that leaves minimal free RBD 138

(1:1 molar ratio), which corresponds to ~64% display efficiency or around 38 RBD per 139 VLP ( Figure 1B) . 140 141

To confirm the antigenicity of RBD-SpyVLP, we performed a series of binding assays. 144

Binding to RBD-SpyVLP was tested using a panel of novel monoclonal antibodies 145 (mAbs) some of which are strongly neutralising, from COVID-19-infected donors 26 , 146 that bind to at least three independent epitopes on the RBD. We included the 147 published conformation-specific mAb (CR3022) 27 , a nanobody-Fc fusion VHH72-Fc 148 28 , and a dimeric human ACE2-Fc 29 , for which there are published structures. All 149 tested mAbs and ACE2-Fc bound strongly to the RBD-SpyVLP (Figure 2A) , showing 150 that a broad range of epitopes on the RBD-SpyVLP are exposed and correctly folded. 151

An anti-influenza neuraminidase mAb (Flu mAb), used as a negative control, showed 152 no binding to RBD-SpyVLP, confirming the specificity of the assay (Figure 2A) . 153 154 We then tested the stability of RBD-SpyVLP, to determine its resilience and likely 155 sensitivity to failures in the cold-chain 30 . The unconjugated SpyCatcher003-mi3 VLP 156 had previously been shown to be highly thermostable as a platform for antigen display 157 18 . For conjugated RBD-SpyVLP, we tested its solubility following storage for two 158 weeks at -80, -20, 4 or 25 °C in Tris Buffered Saline (TBS). We then centrifuged out 159 any aggregates and analysed soluble protein by SDS-PAGE with Coomassie staining. 160

We found no significant change in the soluble fraction following storage at 4 °C (n=3, 161

Kruskal-Wallis and Dunn's post-hoc test, p>0.05), with only a 12% decrease after 162 storage for two weeks at 25 °C ( Figure 2B ) and no degradation was observed at 25 °C 163 ( Figure S2A ). We further analysed the integrity of the sample with ELISA against the 164 conformation-dependent CR3022 mAb and observed no loss of antigenicity under these storage conditions ( Figure 2C ). We next assessed the resilience of SpyVLP to freezing, challenging RBD-SpyVLP with multiple rounds of freeze-thaw. 167

Even after five rounds of freeze-thaw, there was no significant loss of soluble RBD-168 SpyVLP ( Figure 2D ) or CR3022 recognition ( Figure 2E ) (n=3, Kruskal-Wallis and 169

Dunn's post-hoc test, p>0.05) and no degradation was observed ( Figure S2B ). After 170 reconstitution following lyophilisation, we saw a minimal change in soluble protein for 171 RBD-SpyVLP (91.5±3.8% of the initial value (mean±SD) ( Figure 2F ), which was not 172 statistically significant (n=3, Mann-Whitney U test, p>0.05). There was also no 173 difference in terms of binding of RBD-SpyVLP to a panel of mAbs or ACE2-Fc 174 recognising non-overlapping footprints on the RBD ( Figure 2G ). Overall, RBD-SpyVLP 175 showed a high level of resilience. to the VLP group ( Figure 3B) . 193 194 We then tested RBD-SpyVLP in a second mouse strain (BALB/c) with the same 195 dosage regimen to confirm the immunogenicity ( Figure 3A and B). In BALB/c, the 0.1 196 µg and 0.5 µg RBD-SpyVLP groups showed higher levels of RBD-specific antibody 197 (EPT: 0.1 µg: 1:8,406, p<0.01 and 0.5 µg: 1:16,636, p<0.05) ( Figure 3A ) and spike-198 was detected post-boost at day 35 (IC50 1:10 to 1:50) and reduced at day 56 (IC50 1:5 266 to 1:25) in all three groups ( Figure 4B ). ACE2 blocking activities in all three groups 267 were significantly higher than for sera from convalescent humans (5 µg RBD Kruskal-Wallis test on day 56) ( Figure 4C ). Similarly, no difference in neutralising 276 activity was detected in all three groups against live virus at day 21 ( Figure 4D ). 277

Neutralisation was detected on both day 35 and 56 in all groups, with ND50 levels 278 between 1:6,000 to 1:11,000 on day 56 and no significant difference between the three 279 groups (p>0.05, Kruskal-Wallis test on day 56 data) ( Figure 4D ). Nasal and oral 280 secretions were also tested for the presence of neutralising antibodies: here RBD-281 specific IgG, but notably not IgA, was detected in all three groups with no significant 282 difference between groups post-boost for both nasal and oral secretions (p>0.05, 283

Kruskal-Wallis followed by Dunn's Post-hoc) ( Figure 4E & F) . 284

We assessed the level of RBD-specific B cells in peripheral blood. The predominance 286 of an IgG response following the booster immunisation was confirmed in all three 287 groups by assessment of RBD tetramer labelling of IgM, IgG and IgA B cells in 288 peripheral blood and IgG ELISpot assay ( Figure S3A & B). In each case, RBD-specific 289

IgG B cells peaked in peripheral blood shortly after the boost ( Figure S3A ). We also 290 performed longitudinal analysis of CD4 + and CD8 + T cell responses following 291 immunisation. Intracellular cytokine staining of S-peptide-stimulated peripheral blood 292 mononuclear cells (PBMCs) demonstrated a T cell IFN-γ response, slightly larger in 293 the CD4 + than CD8 + T cell pool, which peaked 7 days after prime and boost 294 immunisations ( Figure S3C ). A concern regarding RBD-based vaccines is whether the immune response will be 300 focussed to a single site on the antigen, because of the relatively small size of the RBD, 301 potentially leading to a narrow response that would be sensitive to immune escape 35, 36 . 302

A recent report showed that passive immunisation with a single mAb led to escape 303 mutants of SARS-CoV-2, whereas a cocktail of two neutralising antibodies to independent 304 epitopes prevented emergence of neutralisation-escape mutations, demonstrating the 305 importance of a polyclonal antibody response 37 . We assessed sera from mice and pigs 306 immunised with RBD-SpyVLP for antibody responses that target multiple RBD epitopes 307 using a competition ELISA against four different mAbs: FI-3A, FD-11A, EY6A and S309, 308 that target three non-overlapping epitopes on the RBD with FD-11A and S309 showing 309 overlap as defined by competition ELISA (see diagram in Figure 2A pigs showed a trend of partial competition against all four mAb, but this was not 315 statistically significant because of inter-animal variation ( Figure 5B ). When responses of 316 individual pigs were compared to their pre-immune sera 2 out of 3 animals in each dosing 317 group showed significant competition for the antibodies FI-3A and S309 that defined 318 independent neutralising epitopes compared to their preimmune sera ( Figure S5 ). These 319 results show that RBD-SpyVLP does not have an immunodominant epitope and does not 320 induce a highly focussed antibody response, making it a vaccine candidate that is likely 321 to resist the generation of neutralisation-escape mutants. 322 323 324

In the current study we investigated an RBD-based VLP vaccine candidate for COVID-326 19 based on SpyTag/SpyCatcher technology which was used to assemble RBDs into 327 the mi3 VLP via the formation of an irreversible isopeptide bond 19 . We showed RBD-328

SpyVLPs to be strongly immunogenic in mice and pigs, inducing high titre neutralising 329 antibody responses against wild type SARS-CoV-2 virus. This study confirms that the 330 RBD is the key immunogenic domain for eliciting neutralising monoclonal antibodies 331 against SARS-CoV-2, in line with studies showing that highly neutralising antibodies 332 isolated from convalescent patients bind to the RBD 16, 26, 38, 40-46 . We showed that RBD-SpyVLPs are recognised by a panel of mAbs isolated from convalescent patients 334 26 binding to various epitopes on the RBD (Figure 2A ). This distributed reactivity shows 335 that all of the epitopes that could potentially induce protective antibodies to RBD are 336 present in RBD-SpyVLPs. 337

We detected negligible antibody responses in mice vaccinated with equivalent doses 339 (0.1 µg or 0.5 µg) of purified RBD alone, but strong responses to the RBD when 340 displayed on the VLP (Figure 3) . Previous studies showed that RBD from SARS-CoV 341 and SARS-COV-2 can induce neutralising antibodies in animal models but typically 342 after administration of much higher doses (e.g. ~50 to 100 µg) and with frequent 343 dosing 13, 14, 47 . On the other hand, we showed that high titre neutralising antibody 344 responses can be detected in two strains of mice immunised with relatively low doses 345 of RBD-SpyVLP (to ND50 ~500 to 2,000). These results confirm the enhanced 346 immunogenicity of RBD when displayed on SpyVLPs. Sera from mice immunised with 347 both 0.1 µg or 0.5 µg of RBD-SpyVLP exhibited high levels of antibody against SARS-348

CoV-2 RBD and full-length spike glycoprotein and ACE2 blocking activity ( Figure 3B , 349 C and D). All of these responses were higher than the levels found in plasma from Surprisingly, no increase in antibody titre was observed in pigs that received a higher 361 dose of antigen (50 µg RBD-SpyVLP). There was a trend that the 50 µg RBD-SpyVLP 362 group generated a more rapid and higher response post-prime but the antibody 363 response between the 5 µg and 50 µg RBD-SpyVLP groups were identical post-boost. 364

This suggests a threshold effect. 365

Since RBD-SpyVLPs induce antibody responses that target multiple epitopes on the 367 RBD the chance of selecting neutralisation-escape mutants should be greatly reduced. 368

Circulating SARS-CoV-2 stains are constantly mutating and the likelihood of 369 persistence of the virus in the human population is high 48, 49 . 370

We observed differences in the levels of ACE2 blocking in sera from immunised mice 372 and pigs ( Figure 3C and 4B). Sera taken from mice immunised with RBD-SpyVLP had 373 at least one order of magnitude higher ACE2 blocking activity compared to serum from 374 pigs, despite neutralisation titres being comparable ( Figure 3D and 4D ). This suggests 375 that mice and pigs may produce distinct antibody responses against the vaccine 376 candidate, although they were equally potent in neutralising live viruses. Surprisingly, 377 ACE2 blocking in both mild and critical/severe convalescent humans who had natural 378 infection were also low compared to the sera from immunised mice ( Figure 3D ). RBD-SpyVLP can also be co-displayed with antigens from other pathogens such as 422 the HA and NA from influenza virus. We have recently shown HA and NA to be highly 423 immunogenic in mice after formulation as a SpyVLP 18 . This assembly could potentially 424 provide protection against both SARS-Cov-2 and influenza viruses. Testing resilience 425 of the vaccine candidate, we found that RBD-SpyVLP is stable at ambient temperature, 426 resistant to freeze-thaw, and can be lyophilised and reconstituted with minimal loss in 427 activity ( Figure 2B -G) or immunogenicity ( Figure 3F ). This resilience may not only 428 simplify vaccine distribution worldwide, especially to countries where cold-chain 429 resources are incomplete, but also reduce the overall vaccine cost by removing cold-430 chain dependence. We are currently investigating cheaper and more scalable 431 alternatives to produce RBD-SpyVLP to cope with the global demand for a SARS-432

CoV-2 vaccine. Collectively, our results show that the RBD-SpyVLP is a potent and

The SpyTag-RBD expression construct ( Figure S1 and Addgene deposition in progress) (see Figure S1 ). pET28a-SpyCatcher003-mi3 443 Samples were centrifuged for 30 min at 16,900 g at 4 °C to pellet possible aggregates. 562

Before each measurement, the quartz cuvette was incubated in the instrument for 5 563 min to stabilise the sample temperature. Samples were measured at 125-250 µg/mL 564 total protein concentration. 30 µL of sample was measured at 20 °C using an 565

Omnisizer (Victotek) with 20 scans of 10 s each. The settings were 50% laser intensity, 566 15% maximum baseline drift, and 20% spike tolerance. The intensity of the size 567 distribution was normalised to the peak value and plotted in GraphPad Prism 8 568 (GraphPad Software).

To prepare the RBD-SpyVLP for vaccination at 125 µg/mL (based on SpyTag-RBD 572 concentration), 5 µM SpyTag-RBD was conjugated with 3.33 µM or 5 µM of 573 SpyCatcher003-mi3 in TBS pH 8.0 at 4 °C for 16 h. The reaction was centrifuged for 574 30 min at 17,000 g at 4 °C to remove potential aggregates. RBD-SpyVLP was 575 aliquoted and stored at -80 °C. Matching non-conjugated SpyTag-RBD or 576

SpyCatcher003-mi3 VLP were diluted in the same buffers and incubated and 577 centrifuged in the same way. RBD-SpyVLP (125 µg/mL) was diluted to 4 µg/mL (0.1 578 µg dose) or 20 µg/mL (0.5 µg dose) in the same buffer freshly before immunisation. 

A cell-based ELISA as described previously 26 was used to determine the anti-spike 634 glycoprotein antibody response in the mouse sera and convalescent plasma. Briefly, 635

full-length SARS-CoV-2 spike glycoprotein cDNA using a lentiviral vector. MDCK-637

at 37 °C. Mouse sera was diluted as above and 50 μL was transferred to the washed 639 plates seeded with MDCK-Spike cells for 1 h at RT. Human plasma was pre-incubated 640 with MDCK-SIAT1 cells as described above, before dilution and adding to MDCK-641

Spike for 1 h at RT. Parallel plates seeded with MDCK-SIAT1 for background 642 subtraction was done as described above for human plasma. For mouse sera, 50 μL 643 of a secondary Alexa Fluor 647 goat-anti mouse antibody (1:500) (Life Technologies 644 A21235) was then added for 1 h at RT. For human sera, 50 μL of a secondary Alexa 645

Fluor 647 goat-anti human antibody (1:500) (Life Technologies A21455) was used. 646

Plates were then washed with PBS and 100 μL of PBS/1% formalin was added to each 647 well. Fluorescence signal was read on a Clariostar plate reader and the EPT titre was 648 calculated as described above. Triton-X-100 in PBS stained to visualise virus plaques, as described previously for the 684 neutralisation of influenza viruses 61 , but using a rabbit polyclonal anti-NSP8 antibody 685 (Antibodies Online; ABIN233792) and anti-rabbit-HRP conjugate (Bio-Rad) and 686 detected using HRP on a TMB based substrate. Virus plaques were quantified and 687 ND50 for sera was calculated using LabView software as described previously 61 . 

Lentiviral-based SARS-CoV-2 pseudoviruses were generated as described previously 804 34 . Briefly, HEK293T cells were seeded at a density of 7.5 × 10 5 in 6-well plates before 805 were then washed with PBS and blocked using cRPMI for at least 1 h at 37 °C, 5% 896 CO2, 95% humidity. Blocking solution was then removed, and PBMC were added at a 897 density of 5 × 10 5 /well for antigen-specific response or at 5 × 10 4 /well for wells 898 assigned to total IgG (positive control). The plates were then incubated for 18 h at 899 37 °C, 5% CO2, 95% humidity. Media was removed and cells were lysed with cold 900 distilled water, followed by three PBS washes as before. To measure total IgG, 50 901 μL/well of biotinylated anti-IgG mAb (clone MT424-biotin, Mabtech) was added at 0.5 902 µg/mL. To assess antigen-specific responses 50 μL/well of biotinylated SARS-CoV-2 903 RBD was added at 2.5μg/mL. As a negative control, 50 μL/well of biotinylated Nipah 904 G protein was added to the relevant wells at 2.5 μg/mL. All antigens were diluted in 905 PBS with 0.5% (v/v) FCS. An additional set of negative control wells were also 906 prepared by adding 50 μL/well PBS with 0.5% (v/v) FCS. Each condition was tested 907 in triplicate. Plates were incubated for 2 h at RT, before washing five times with PBS. 908

Following this, 50 μL/well of streptavidin-alkaline phosphatase (streptavidin-ALP) 909 enzyme conjugate (Mabtech) (diluted 1:1,000 in PBS with 0.5% (v/v) FCS) was added 910 to each well and plates were incubated for 1 h at RT (protected from light). 911

Streptavidin-ALP was removed, and plates were washed another 5 times with PBS, 912 followed by addition of 50 μL/well BCIP/NBTplus substrate (Mabtech), neat. Plates 913 were left for 30 min, until distinct spots developed. Finally, development was stopped 914 by addition of 150 μL/well of 4 °C distilled water followed by rinsing both the front and 915 back of the plates with copious tap water. Plates were air-dried, before spots were 916 counted using a CTL ImmunoSpot Analyzer (Cellular Technologies). 917 918

All statistical analyses were performed using GraphPad Prism 8 (GraphPad Software). 920

Statistical differences were analysed using either Mann-Whitney U test or Kruskal- 

A Novel Coronavirus from Patients with Pneumonia in China

979 2. World Health Organization

Coronaviridae Study Group of the International Committee on Taxonomy of, V. The 982 species Severe acute respiratory syndrome-related coronavirus: classifying 2019-nCoV 983 and naming it SARS-CoV-2

Author Correction: A new coronavirus associated with human respiratory 985 disease in China

Structure, Function, and Antigenicity of the SARS-CoV-2 Spike 987 Glycoprotein

Structural basis of receptor recognition by SARS-CoV-2

Structural basis for the recognition of SARS-CoV-2 by full-length human 991 ACE2

A pneumonia outbreak associated with a new coronavirus of probable 993 bat origin

Harnessing Nanoparticles for 995 Immunomodulation and Vaccines. Vaccines (Basel)

Receptor-binding domains of spike proteins 997 of emerging or re-emerging viruses as targets for development of antiviral vaccines

Prospects for a MERS-CoV spike vaccine. Expert Rev 1000 Vaccines

Farzan M The SARS-CoV-2 receptor-binding domain elicits a potent 1003 neutralizing response without antibody-dependent enhancement. bioRxiv (2020)

A vaccine targeting the RBD of the S protein of SARS-CoV-2 induces 1005 protective immunity

Human neutralizing antibodies elicited by SARS-CoV-2 infection

Isolation of potent SARS-CoV-2 neutralizing antibodies and 1009 protection from disease in a small animal model

Enhancing Humoral Responses Against HIV Envelope Trimers via 1011

Overcoming Symmetry Mismatch in Vaccine Nanoassembly with 1013 SpyCatcher003 Amidation

Plug-and-Display: decoration of Virus-Like Particles via isopeptide 1015 bonds for modular immunization

Nanoscaffold To Enhance Plug-and-Display Vaccination

Immunization expands B cells specific to HIV-1 V3 glycan in mice 1020 and macaques

Bacterial superglue enables easy development of efficient virus-like 1022 particle based vaccines

Spy&Go purification of SpyTag-proteins using pseudo-1024

SpyCatcher to access an oligomerization toolbox

Approaching infinite affinity through engineering of peptide-protein 1026 interaction

Corrigendum: Design of a hyperstable 60-subunit protein icosahedron

Plasmablast-derived antibody response to acute SARS-CoV-2 1030 infection in humans

Human monoclonal antibody combination against SARS 1032 coronavirus: synergy and coverage of escape mutants

Structural Basis for Potent Neutralization of Betacoronaviruses by 1034 Single-Domain Camelid Antibodies

Structural and Functional Basis of SARS-CoV-2 Entry by Using Human 1036 ACE2

Vaccine 1038 provision: Delivering sustained & widespread use

MF59-adjuvanted vaccines: increased immunogenicity 1040 with an optimal safety profile

The pig: a model 1042 for human infectious diseases

Establishing the pig as a large animal model for vaccine 1044 development against human cancer

Evaluation of the immunogenicity of prime-boost vaccination with 1046 the replication-deficient viral vectored COVID-19 vaccine candidate ChAdOx1 nCoV-1047 19

Antibody Immunodominance: The Key 1049 to Understanding Influenza Virus Antigenic Drift

Focused antibody response to influenza linked to antigenic drift

Antibody cocktail to SARS-CoV-2 spike protein prevents rapid 1053 mutational escape seen with individual antibodies

Cross-neutralization of SARS-CoV-2 by a human monoclonal SARS-1055 CoV antibody

Structural basis for the neutralization of SARS-CoV-2 by an antibody 1057 from a convalescent patient

Potent human neutralizing antibodies elicited by SARS-CoV-2 infection

Potent neutralizing antibodies from COVID-19 patients define 1061 multiple targets of vulnerability

Potent Neutralizing Antibodies against SARS-CoV-2 Identified by High-1063

Throughput Single-Cell Sequencing of Convalescent Patients' B Cells

Studies in humanized mice and convalescent humans yield a SARS-1066

CoV-2 antibody cocktail

Longitudinal Isolation of Potent Near-Germline SARS-CoV-2-1068

Neutralizing Antibodies from COVID-19 Patients

Potent neutralizing antibodies against multiple epitopes on SARS-CoV-2 1070 spike

Convergent antibody responses to SARS-CoV-2 in convalescent 1072 individuals

Identification of an ideal adjuvant for receptor-binding domain-based 1074 subunit vaccines against Middle East respiratory syndrome coronavirus

Genome-wide analysis of SARS-CoV-2 virus strains circulating 1077 worldwide implicates heterogeneity

Elicitation of potent neutralizing antibody responses by designed 1080 protein nanoparticle vaccines for SARS-CoV-2. bioRxiv

1083 52. van Doremalen, N. et al. ChAdOx1 nCoV-19 vaccination prevents SARS-CoV-2 1084 pneumonia in rhesus macaques

First-in-Human Trial of a SARS CoV 2 Recombinant Spike Protein 1086 Nanoparticle Vaccine. medRxiv

Cross-sectional evaluation of humoral responses against SARS-CoV-1088

Longitudinal evaluation and decline of antibody responses in SARS-1090 CoV-2 infection

Mosaic nanoparticle display of diverse influenza virus 1092 hemagglutinins elicits broad B cell responses

Peptide tag forming a rapid covalent bond to a protein, through 1094 engineering a bacterial adhesin

Neutralization of SARS-CoV-2 by Destruction of the Prefusion Spike

Overexpression of the alpha-2,6-sialyltransferase in MDCK cells increases influenza 1099 virus sensitivity to neuraminidase inhibitors

Broad and strong memory CD4 (+) and CD8 (+) T cells induced by 1101 SARS-CoV-2 in UK convalescent COVID-19 patients

Optimisation of a micro-neutralisation assay and its application in 1103 antigenic characterisation of influenza viruses

Isolation and rapid sharing of the 2019 novel coronavirus (SARS-CoV-1106 2) from the first patient diagnosed with COVID-19 in Australia

A serological assay to detect SARS-CoV-2 seroconversion in humans. 1109 medRxiv

Site-specific biotinylation of purified proteins using BirA

Bovine Herpesvirus-4-Vectored Delivery of Nipah Virus 1113

Figure 1. SpyTag-RBD can be efficiently conjugated to SpyCatcher003-mi3 VLP

A) Schematic diagram of the RBD-SpyVLP vaccine candidate

SpyCatcher003-VLP conjugated with SpyTag-RBD. The isopeptide bonds formed 1126 spontaneously between SpyTag and SpyCatcher are indicated with red dots

Reactions 1128 were performed at 4 °C overnight and analysed using SDS-PAGE with Coomassie 1129 staining and densitometry, with the percentage of unreacted VLP shown. (C) Dynamic 1130 light scattering (DLS) characterisation of SpyTag-RBD, SpyVLP, and conjugated 1131 RBD-SpyVLP (n=3, values shown as mean±SD)

RBD-SpyVLPs are reactive to SARS-CoV-2 binders and thermostable 1135 and resilient. (A) Binding of RBD-SpyVLP to a panel of monoclonal antibodies 1136 isolated from COVID-19 recovered patients that target independent epitopes on the 1137

The boxed antibodies 1138 form groups that compete with each other and with antibodies or nanobodies with 1139 structurally defined footprints: CR3022 (PDB 6W41) (footprint 1)

S309 (footprint 3) (PDB 6WPT) and the ACE2 binding site (PDB 6M0J)

Each point represents the mean of duplicate readings and error bars represent ± 1 SD

The diagram of the RBD, created in PyMOL, shows the ACE2 binding site and the 1143 three binding footprints highlighted. (B) Solubility and (C) immunoreactivity

SpyVLP after storage for two weeks at various temperatures, determined using SDS-1145 PAGE and densitometry or ELISA (10 µg/mL CR3022 mAb). (D) Solubility and (E) 1146 immunoreactivity of RBD-VLP after freeze−thaw determined using SDS-PAGE and 1147

ELISA (10 µg/mL CR3022 mAb) after one to five cycles of freeze−thawing

SpyVLP soluble fraction, before and after lyophilisation reconstituted in the same 1149 buffer volume and (G) immunoreactivity determined using ELISA with ACE2-Fc and 1150 mAbs that target non-overlapping epitopes on the RBD. Error bars in B, E & F 1151 represent group mean (n=3)

Statistical difference in B to E was determined using Kruskal-Wallis test followed by 1153

Dunn's multiple comparison test. Statistical difference in F & G was determined using 1154

U test. n.s. = not significant

RBD-SpyVLPs induce strong antibody responses in mice that are 1160 comparable to the responses in recovered patients. C57BL/6 (red) or BALB/c 1161 (blue) mice (n=6 in each group) were dosed twice IM, two weeks apart with 0.1 µg or 1162 0.5 µg purified RBD, RBD-SpyVLP or VLP alone with AddaVax adjuvant added to all

Sera were harvested at two weeks after the first dose (open circles) and at three weeks 1164 after the second dose (closed circles). Sera were analysed in (A) ELISA against RBD 1165

ELISA against full-length spike glycoprotein, (C) in an ACE2 competition assay

and (D) in virus neutralisation assays (VNT) against wild-type SARS-CoV-2 virus

Antibody response analysed by RBD ELISA for mice dosed twice with 0

Data are 1169 presented as the group geometric means ± 95% confidence intervals. COVID-19 1170 convalescent plasma from humans with mild (open mauve circles) or critical/severe 1171 (closed mauve circles) disease were included for comparison

001 determined by Kruskal-Wallis test followed by Dunn's multiple comparison 1173 test. # p<0.05, ### p<0.001, #### p<0.0001 determined by Mann-Whitney U test to 1174 compare against convalescent human plasma. Dotted lines represent the lowest 1175 mouse sera dilutions tested

Violet 421 mAb (clone Mab11, BioLegend). Cells were analysed using a BD 842LSRFortessa flow cytometer (BD Biosciences) and data analysed using FlowJo 843 software (BD Biosciences). Total SARS-CoV-2 S-specific IFN-γ-positive responses for 844 live CD3 + CD4 + and CD3 + CD4 -CD8 + T cells are presented after subtraction of the 845 background response detected in the media-stimulated control PBMC samples of 846 each pig, prior to summing together the frequency of S-peptide pools 1-3 specific cells. 847

A biotinylated form of RBD was generated for B-cell tetramer staining assays. An RBD 850protein with a C-terminal biotin acceptor peptide (RBD-BAP) was expressed from 851 plasmid pOPINTTGNeo in Expi293 cells according to the manufacturer's instructions. 852Culture supernatants were clarified by centrifugation and purified through a 5 mL 853HisTrap FF column (GE Healthcare), using the ÄKTA Pure chromatography system 854