key: cord-297826-2nruf2g7
authors: Tian, Jing-Hui; Patel, Nita; Haupt, Robert; Zhou, Haixia; Weston, Stuart; Hammond, Holly; Lague, James; Portnoff, Alyse D.; Norton, James; Guebre-Xabier, Mimi; Zhou, Bin; Jacobson, Kelsey; Maciejewski, Sonia; Khatoon, Rafia; Wisniewska, Malgorzata; Moffitt, Will; Kluepfel-Stahl, Stefanie; Ekechukwu, Betty; Papin, James; Boddapati, Sarathi; Wong, C. Jason; Piedra, Pedro A.; Frieman, Matthew B.; Massare, Michael J.; Fries, Louis; Lövgren Bengtsson, Karin; Stertman, Linda; Ellingsworth, Larry; Glenn, Gregory; Smith, Gale
title: SARS-CoV-2 spike glycoprotein vaccine candidate NVX-CoV2373 elicits immunogenicity in baboons and protection in mice
date: 2020-06-30
journal: bioRxiv
DOI: 10.1101/2020.06.29.178509
sha: 
doc_id: 297826
cord_uid: 2nruf2g7

The COVID-19 pandemic continues to spread throughout the world with an urgent need for a safe and protective vaccine to effectuate herd immunity to control the spread of SARS-CoV-2. Here, we report the development of a SARS-CoV-2 subunit vaccine (NVX-CoV2373) produced from the full-length spike (S) protein, stabilized in the prefusion conformation. Purified NVX-CoV2373 S form 27.2nm nanoparticles that are thermostable and bind with high affinity to the human angiotensin-converting enzyme 2 (hACE2) receptor. In mice and baboons, low-dose NVX-CoV2373 with saponin-based Matrix-M adjuvant elicits high titer anti-S IgG that is associated with blockade of hACE2 receptor binding, virus neutralization, and protection against SARS-CoV-2 challenge in mice with no evidence of vaccine-associated enhanced respiratory disease (VAERD). NVX-CoV2373 vaccine also elicits multifunctional CD4+ and CD8+ T cells, CD4+ T follicular helper T cells (Tfh), and the generation of antigen-specific germinal center (GC) B cells in the spleen. These results support the ongoing phase 1/2 clinical evaluation of the safety and immunogenicity of NVX-CoV2327 with Matrix-M (NCT04368988).

titers that were detected 21-28 days after a single immunization (Fig.4A, right) . Mice 171 immunized with 10 μg dose of NVX-CoV2373/Matrix-M induced antibodies that blocked 172 hACE2 receptor binding to S-protein and virus neutralizing antibodies 21-28-days after 173 a single priming dose ( Fig. 4B and 4C ). Animals immunized with the prime/boost 174 regimen had significantly elevated anti-S IgG titers that were detected 7-16 days 175

following the booster immunization across all dose levels. Animals immunized with 1 176 μg and 10 μg NVX-CoV2373/Matrix-M had similar high anti-S IgG titers following 177 immunization (GMT = 139,000 and 84,000, respectively). Importantly, mice immunized 178 with 0.1 μg, 1 μg, or 10 μg NVX-CoV/Matrix-M had significantly (p ≤ 0.00001) higher 179 anti-S IgG titers compared to mice immunized with 10 μg NVX-CoV2373 without 180 adjuvant (Fig. 4A, left) . These results indicate the potential for a 10-fold or greater across all dose levels ( Fig. 4B and 4C) . showing marked protection from weight loss compared to the unvaccinated placebo 209 animals (Fig. 4E) . The mice receiving a prime and boost vaccination with adjuvanted 210 vaccine also demonstrated significant protection against weight loss at all dose levels 211 (Fig. 4F) . In addition, we compared the prime/boost regimens using 10 μg of either 212 adjuvanted or unadjuvanted NVX-CoV2373. The mice receiving the prime/boost with 213 adjuvant were significantly protected from weight loss relative to placebo mice, while the 214 group immunized with 10 μg NVX-CoV2373 alone were not protected against weight 215 loss (Fig. 4G) . These results confirm that NVX-CoV2373 confers protection against 216 SARS-CoV-2 and that low doses of the vaccine associated with lower serologic 217 responses do not exacerbate weight loss or demonstrate exaggerated illness. with increased periarteriolar cuffing. The thickened alveolar septa remain with increased 225 diffuse interstitial inflammation throughout the alveolar septa (Fig. 5) . 226

The NVX-CoV2373 immunized mice showed significant reduction in lung pathology 227 at both day 4 and day 7 post infection in a dose-dependent manner. The prime only 228 group displays reduced inflammation at the 10 μg and 1 μg dose with a reduction in 229 inflammation surrounding the bronchi and arterioles compared to placebo mice. In the 230 lower doses of the prime-only groups, lung inflammation resembles that of the placebo 231 groups, correlating with weight loss and lung virus titer. The prime/boost immunized 232 groups displayed a significant reduction in lung inflammation for all doses tested, which 233 again correlated with lung viral titer and weight loss data. The epithelial cells in the large 234 and small bronchi at day 4 and 7 were substantially preserved with minimal bronchiolar 235 sloughing or signs of viral infection. The arterioles of animals immunized with 10 μg, 1 236 μg and 0.1 μg doses have minimal inflammation with only moderate cuffing seen in the 237 0.01 μg dose, similar to placebo. Alveolar inflammation was reduced in animals that 238 received the higher doses with only the lower 0.01 μg dose with inflammation (Fig. 5) . 239

These data demonstrate that NVX-CoV2373 reduces lung inflammation after challenge 240 and that even doses and regimens of NVX-CoV2373 that elicit minimal or no detectable 241 neutralizing activity are not associated with any obvious exacerbation of the 242 inflammatory response to the virus. 243

To determine 244 the role of Matrix-M in generating T cell responses, we immunized groups of mice (N = 245 6/group) with 10 μg NVX-CoV2373 alone or with 5 μg Matrix-M in a 2-dose regimen 246 spaced 21-days apart. Antigen-specific T cell responses were measured by ELISPOT 247 and intracellular cytokine staining (ICCS) from spleens collected 7-days after the 248 second immunization (study day 28). The number of IFN-γ secreting cells after ex vivo 249 stimulation increased 7-fold in spleens of mice immunized with NVX-CoV2373/Matrix-250 M compared to NVX-CoV2373 alone as measured by the ELISPOT assay (Fig. 6A) . In 251 order to examine CD4 + and CD8 + T cell responses separately, ICCS assays were 252 performed in combination with surface marker staining. Data shown were gated on 253 CD44 hi CD62Leffector memory T cell population. Importantly, we found the frequency 254 of IFN-γ + , TNF-α + , and IL-2 + cytokine-secreting CD4 + and CD8 + T cells was 255 significantly higher (p <0.0001) in spleens from the NVX-CoV2373/Matrix-M immunized 256 mice compared to mice immunized without adjuvant ( Fig. 6B and 6C) . Further, we 257 noted the frequency of multifunctional CD4 + and CD8 + T cells, which simultaneously 258 produce at least two or three cytokines was also significantly increased (p <0.0001) in 259 spleens from the NVX-CoV2373/Matrix-M immunized mice ( Fig. 6B and 6C) . 260

Immunization with NVX-CoV2373/Matrix-M resulted in higher proportions of 261 multifunctional phenotype within both CD4 + and CD8 + T cell populations. The 262

proportions of multifunctional phenotypes detected in memory CD4 + T cells were higher 263 than those in CD8 + T cells (Fig. 6D) . 264

Type 2 cytokine IL-4 and IL-5 secretion from CD4 + T cells was also determined by 265 ICCS and ELISPOT respectively. We found that immunization with NVX-266

CoV2373/Matrix-M also increased type 2 cytokine IL-4 and IL-5 secretion (2-fold) 267 compared to immunization with NVX-CoV2373 alone, but to a lesser degree than 268 enhancement of type 1 cytokine production (e.g. IFN-γ increased 20-fold). These Tfh cells (CD4 + CXCR5 + PD-1 + ) (p = 0.01), as well as the frequency of GC B cells 278 (CD19 + GL7 + CD95 + ) (p = 0.0002) in spleens ( Fig. 7A and 7B) . Anti-S protein IgG titers were detected within 21-days of a single priming immunization 288 in animals immunized with NVX-CoV2373/Matrix-M across all the dose levels (GMT = 289 1,249-19,000). Anti-S protein IgG titers increased over a log (GMT = 33,000-174,000) 290 within 1 to 2 weeks following a booster immunization (days 28 and 35) across all of the 291 dose levels. Importantly, animals immunized with NVX-CoV2373 without adjuvant had 292 minimum or no detected anti-S IgG titer (GMT <125) after one immunization, which was 293 not boosted by a second immunization (Fig. 8A) . 294

We also determined the functionality of the antibodies. Low levels of hACE2 receptor 295 blocking antibodies were detected in animals following a single immunization with 5 or μg NVX-CoV2373 alone had no detectable antibodies that block S-protein binding to 302 hACE2 (Fig. 8B) . Neutralizing titers increased 25-to 38-fold following the second 303 immunization (GMT = 6,400-17,000) (Fig. 8C) . Animals receiving the NVX-CoV2373 304 alone had little or no detectable neutralizing antibodies (GMT <100). There was a 305 significant correlation (p <0.0001) between anti-S IgG levels and neutralizing antibody 306 titers (Fig. 8D) . The immunogenicity of the adjuvanted vaccine in nonhuman primates is 307 consistent with the mouse immunogenicity results and further supports the role of 308

Matrix-M adjuvant in promoting the generation of neutralizing antibodies and dose 309 sparing. 310

PBMCs were collected 7 days after the second immunization (day 28) and T cell CoV2373 alone or 1 µg NVXCoV2373/Matrix-M (Fig. 8E) . By ICCS analysis, 315 immunization with 5 µg NVXCoV2373/Matrix-M also showed the highest frequency of 316 IFN-γ + , IL-2 + , and TNF-α + CD4 + T cells (Fig. 8F) . This trend was also true for 317 multifunctional CD4 + T cells, in which at least two or three type 1 cytokines were 318 produced simultaneously (Fig. 8F) . Type 2 cytokine IL-4 level were too low to be 319 detected in baboons by ELISPOT analysis. We also compared the level of functional hACE2 receptor inhibiting (50% RI) titers. 325

Baboons receiving the vaccine had 8-fold higher binding and receptor inhibiting 326 antibodies (50% RI = 478, 95%CI, 161.1-794.4) compared to COVID-19 convalescent 327 serum (50% RI = 61, 95%CI, 35.7-85.5) (Fig. 9) . Therefore, NVX-CoV2373 vaccine 328 induced binding and functional antibodies in a nonhuman primate at levels comparable 329 or higher than individuals recovered from COVID-19. Collectively these results support 330 the development of NVX-CoV2373 for prevention of COVID-19 331

Here, we showed that a full-length, stabilized prefusion SARS-CoV-2 spike glycoprotein with Tris buffer containing NP-9 detergent, clarified by centrifugation at 10,000 x g for 30 398 min. S-proteins were purified by TMAE anion exchange and lentil lectin affinity 399 chromatography. Hollow fiber tangential flow filtration was used to formulate the purified 400 spike protein at 100-150 μg mL -1 in 25 mM sodium phosphate (pH 7.2), 300 mM NaCl, 401 Cumulants analysis of the scattered intensity autocorrelation function was performed 408

with instrument software to provide the z-average particle diameter and polydispersity 409 index (PDI). 410

heated from 4°C to 120°C at 1°C per minute and the differential heat capacity change 412 was measured in a NanoDSC (TA Instruments, New Castle, DE). A separate buffer 413 scan was performed to obtain a baseline, which was subtracted from the sample scan 414 to produce a baseline-corrected profile. The temperature where the peak apex is 415 located is the transition temperature (Tmax) and the area under the peak provides the 416 enthalpy of transition (ΔHcal). 417

Transmission electron microscopy (TEM) and 2D class averaging. Electron 418 microscopy was perform by NanoImaging Services (San Diego, CA) with a FEI Tecani 419 T12 electron microscope, operated at 120keV equipped with a FEI Eagle 4k x 4k CCD 420 camera. SARS-CoV-2 S proteins were diluted to 2.5 µg mL -1 in formulation buffer. The 421 samples (3 µL) were applied to nitrocellulose-supported 400-mesh copper grids and 422 stained with uranyl format. Images of each grid were acquired at multiple scales to 423 assess the overall distribution of the sample. High-magnification images were acquired 424 at nominal magnifications of 110,000x (0.10 nm/pixel) and 67,000x (0.16 nm/pixel). The 425

images were acquired at a nominal under focus of -1.4µm to -0.8µm (110,000x) and 426 electron doses of ~25 e/Å2. 427

For class averaging, particles were identified at high magnification prior to 428 alignment and classification. The individual particles were selected, boxed out, and 429

individual sub-images were combined into a stack to be processed using reference-free 430 classification. Individual particles in the 67,000x high magnification images were 431 selected using an automated picking protocol 17 . An initial round of alignments was 432 performed for each sample, and from the alignment class averages that appeared to 433 contain recognizable particles were selected for additional rounds of alignment. These 434 averages were used to estimate the percentage of particles that resembled single 435 trimers and oligomers. A reference-free alignment strategy based on XMIPP processing 436 package was used for particle alignment and classification 18 . 437

binding kinetics was determined by bio-layer interferometry (BLI) using an Octet QK384 439 system (Pall Forté Bio, Fremont, CA). Hist-tagged human ACE2 (2 μg mL -1 ) was 440 immobilized on nickel-charged Ni-NTA biosensor tips. After baseline, SARS-CoV-2 S 441 protein containing samples were 2-fold serially diluted over a range 4.7nM to 300 nM 442 range were allowed to associate for 600 sec followed by dissociation for an additional 443 900 sec. Data was analyzed with Octet software HT 101:1 global curve fit. 444

Specificity of SARS-CoV-2 S binding to hACE2 receptor by ELISA. Ninety-six well 445 plates were coated with 100 μL SARS-CoV-2 spike protein (2 μg mL -1 ) overnight at 4°C. Similarly, Baboon IFN-γ and IL-4 assays were carried out using NHP IFN-γ and Human 554 IL-4 assay kit from Mabtech using PBMC collected at day 7 following the second 555 immunization (day 28). 

Systematic Comparison of Two Animal-to-Human Transmitted Human 605 Coronaviruses: SARS-CoV-2 and SARS-CoV

Severe acute respiratory syndrome 608 coronavirus 2 (SARS-CoV-2) and coronavirus disease-2019 (COVID-19): The 609 epidemic and the challenges

Spatial-temporal distribution of COVID-19 in China and its 612 prediction: A data-driven modeling analysis

A strategic approach to COVID-19 615 vaccine R&D

Tectonic conformational changes of a coronavirus spike 617 glycoprotein promote membrane fusion

The clinical pathology of severe acute respiratory syndrome (SARS): 620 a report from China

Identification of human neutralizing antibodies against MERS-CoV 623 and their role in virus adaptive evolution

Structure, Function, and Evolution of Coronavirus Spike Proteins

The coronavirus spike protein is 628 a class I virus fusion protein: structural and functional characterization of the fusion 629 core complex

The spike 632 glycoprotein of the new coronavirus 2019-nCoV contains a furin-like cleavage site 633 absent in CoV of the same clade

Immunogenicity and structures of a rationally designed prefusion 636 MERS-CoV spike antigen

Antigenicity of the SARS-CoV-2 Spike Glycoprotein

Dipeptidyl peptidase 4 is a functional receptor for the emerging 642 human coronavirus-EMC

Host cell entry of Middle East respiratory syndrome 644 coronavirus after two-step, furin-mediated activation of the spike protein

Activation of the SARS coronavirus spike 647 protein via sequential proteolytic cleavage at two distinct sites

Angiotensin-converting enzyme 2 is a functional receptor for the SARS 650 coronavirus

Appion: an integrated, database-driven pipeline to facilitate EM 652 image processing

XMIPP: a new generation of an open-source image processing 654 package for electron microscopy

Cryo-EM structure of the 2019-nCoV spike in the prefusion 656 conformation

Characterization of spike glycoprotein of SARS-CoV-2 on virus entry 658 and its immune cross-reactivity with SARS-CoV

Improved Titers against Influenza Drift Variants with a Nanoparticle 661

Fries L et al. A Randomized, Blinded, Dose-Ranging Trial of an Ebola Virus 663

EBOV GP) Nanoparticle Vaccine with Matrix-M™ Adjuvant in 664

Healthy Adults

Anti-spike IgG causes severe acute lung injury by skewing macrophage 666 responses during acute SARS-CoV infection

Complement Activation Contributes to Severe Acute Respiratory 669 Syndrome Coronavirus Pathogenesis. mBio