key: cord-0803901-1v0f2dtx
authors: Corbett, Kizzmekia S.; Edwards, Darin; Leist, Sarah R.; Abiona, Olubukola M.; Boyoglu-Barnum, Seyhan; Gillespie, Rebecca A.; Himansu, Sunny; Schäfer, Alexandra; Ziwawo, Cynthia T.; DiPiazza, Anthony T.; Dinnon, Kenneth H.; Elbashir, Sayda M.; Shaw, Christine A.; Woods, Angela; Fritch, Ethan J.; Martinez, David R.; Bock, Kevin W.; Minai, Mahnaz; Nagata, Bianca M.; Hutchinson, Geoffrey B.; Bahl, Kapil; Garcia-Dominguez, Dario; Ma, LingZhi; Renzi, Isabella; Kong, Wing-Pui; Schmidt, Stephen D.; Wang, Lingshu; Zhang, Yi; Stevens, Laura J.; Phung, Emily; Chang, Lauren A.; Loomis, Rebecca J.; Altaras, Nedim Emil; Narayanan, Elisabeth; Metkar, Mihir; Presnyak, Vlad; Liu, Catherine; Louder, Mark K.; Shi, Wei; Leung, Kwanyee; Yang, Eun Sung; West, Ande; Gully, Kendra L.; Wang, Nianshuang; Wrapp, Daniel; Doria-Rose, Nicole A.; Stewart-Jones, Guillaume; Bennett, Hamilton; Nason, Martha C.; Ruckwardt, Tracy J.; McLellan, Jason S.; Denison, Mark R.; Chappell, James D.; Moore, Ian N.; Morabito, Kaitlyn M.; Mascola, John R.; Baric, Ralph S.; Carfi, Andrea; Graham, Barney S.
title: SARS-CoV-2 mRNA Vaccine Development Enabled by Prototype Pathogen Preparedness
date: 2020-06-11
journal: bioRxiv
DOI: 10.1101/2020.06.11.145920
sha: 687ad44e3b56c0991c9d723e62b871ceff27ca68
doc_id: 803901
cord_uid: 1v0f2dtx

A SARS-CoV-2 vaccine is needed to control the global COVID-19 public health crisis. Atomic-level structures directed the application of prefusion-stabilizing mutations that improved expression and immunogenicity of betacoronavirus spike proteins. Using this established immunogen design, the release of SARS-CoV-2 sequences triggered immediate rapid manufacturing of an mRNA vaccine expressing the prefusion-stabilized SARS-CoV-2 spike trimer (mRNA-1273). Here, we show that mRNA-1273 induces both potent neutralizing antibody and CD8 T cell responses and protects against SARS-CoV-2 infection in lungs and noses of mice without evidence of immunopathology. mRNA-1273 is currently in a Phase 2 clinical trial with a trajectory towards Phase 3 efficacy evaluation.

with mRNA-1273 elicits a balanced Th1/Th2 response in contrast to the Th2-biased response 148 seen with S protein adjuvanted with alum, suggesting that mRNA vaccination avoids Th2-biased 149 immune responses that have been linked to VAERD. 150 Protective immunity was assessed in young adult BALB/cJ mice challenged with mouse-151 adapted (MA) SARS-CoV-2 that exhibits viral replication localized to lungs and nasal 152 turbinates 28 . BALB/cJ mice that received two 1 μg doses of mRNA-1273 were completely 153 protected from viral replication in lungs after challenge at a 5- (Fig. 4a) or 13-week intervals 154 following boost (Extended Data Fig. 8a ). mRNA-1273-induced immunity also rendered viral 155 replication in nasal turbinates undetectable in 6 out of 7 mice (Fig. 4b, Extended Data Fig. 8b) . 156

Efficacy of mRNA-1273 was dose-dependent, with two 0.1 μg mRNA-1273 doses reducing lung 157 viral load by ~100-fold and two 0.01 μg mRNA-1273 doses reducing lung viral load by ~3-fold 158 (Fig. 4a) . Of note, mice challenged 7 weeks after a single dose of 1 μg or 10 μg of mRNA-1273 159

were also completely protected against lung viral replication (Fig. 4c) . Challenging animals 160 immunized with sub-protective doses provides an orthogonal assessment of safety signals, 161 such as increased clinical illness or pathology. Similar to what was observed with MERS-CoV S-162 2P mRNA, mice immunized with sub-protective 0.1 and 0.01 µg mRNA-1273 doses showed no 163 evidence of enhanced lung pathology or excessive mucus production (Fig. 4d) . In summary, 164 mRNA-1273 is immunogenic, efficacious, and does not show evidence of promoting VAERD 165 when given at sub-protective doses in mice. 166

Here, we showed that 1 μg of mRNA-1273 was sufficient to induce robust neutralizing activity 167 and CD8 T cell responses, balanced Th1/Th2 antibody isotype responses, and protection from 168 viral replication for more than 3 months following a prime/boost regimen similar to that being 169 tested in humans. Inclusion of lower sub-protective doses demonstrated the dose-dependence 170 of antibody, Th1 CD4 T cell responses, and protection, suggesting immune correlates of 171 protection can be further elucidated. A major goal of animal studies to support SARS-CoV-2 vaccine candidates through clinical trials is to not only prove elicitation of potent protective 173 immune responses, but to show that sub-protective responses do not cause VAERD 3 . Sub-174 protective doses did not prime mice for enhanced immunopathology following challenge. 175

Moreover, the induction of protective immunity following a single dose suggests that 176 consideration could be given to administering one dose of this vaccine in the outbreak setting. 177

These data, combined with immunogenicity data from nonhuman primates and subjects in early 178

Phase 1 clinical trials, will be used to inform the dose and regimen of mRNA-1273 in advanced 179 clinical efficacy trials. 180

The COVID-19 pandemic of 2020 is the Pathogen X event that has long been predicted 12, 13 . 181

Here, we provide a paradigm for rapid vaccine development. Structure-guided stabilization of 182 the MERS-CoV S protein combined with a fast, scalable, and safe mRNA/LNP vaccine platform 183 led to a generalizable beta-CoV vaccine solution that translated into a commercial mRNA 184 vaccine delivery platform, paving the way for the rapid response to the COVID-19 outbreak. This 185 is a demonstration of how the power of new technology-driven concepts like synthetic 186 vaccinology facilitate a vaccine development program that can be initiated with pathogen 187 sequences alone 11 . It is also a proof-of-concept for the prototype pathogen approach for 188 pandemic preparedness and response that is predicated on identifying generalizable solutions 189 for medical countermeasures within virus families or genera 12 . Even though the response to the 190 COVID-19 pandemic is unprecedented in its speed and breadth, we envision a response that 191 could be quicker. There are 24 other virus families known to infect humans, and with sustained 192 investigation of those potential threats, we could be better prepared for future looming 193 Formulations were dialyzed against phosphate-buffered saline (pH 7.4) for at least 18 hr, 244 concentrated using Amicon ultracentrifugal filters (EMD Millipore), passed through a 0.22-μm 245 filter and stored at -20°C until use. All formulations underwent quality control for particle size, encapsulation on mRNA and < 10 EU/mL endotoxin. 248

Vectors encoding MERS-CoV S-2P 11 and SARS-CoV S-2P 22 were generated as previously week-old 288/330 +/+ mice 21 were immunized. Four weeks post-boost, pre-challenge sera were 299 collected from a subset of mice, and remaining mice were challenged with 5x10 5 PFU of a 300 mouse-adapted MERS-CoV EMC derivative, m35c4 31 . On day 3 post-challenge, lungs were 301 harvested, and hemorrhage and viral titer were assessed, per previously published methods 32 . 302

For challenge studies to evaluate SARS-CoV-2 vaccines, BALB/cJ mice were challenged with 303 10 5 PFU of mouse-adapted SARS-CoV-2 (SARS-CoV-2 MA). On day 2 post-challenge, lungs 304 and nasal turbinates were harvested for viral titer assessment, per previously published 305

Histology 307

Lungs from mice were collected at the indicated study endpoints and placed in 10% neutral 308 buffered formalin (NBF) until adequately fixed. Thereafter, tissues were trimmed to a thickness 309 of 3-5 mm, processed and paraffin embedded. The respective paraffin tissue blocks 310 were sectioned at 5 µm and stained with hematoxylin and eosin (H&E). All sections were 311 examined by a board-certified veterinary pathologist using an Olympus BX51 light microscope 312 and photomicrographs were taken using an Olympus DP73 camera. 313

Nunc Maxisorp ELISA plates (ThermoFisher) were coated with 100 ng/well of protein in 1X PBS 315 at 4°C for 16 hr. Where applicable, to eliminate fold-on-specific binding from MERS S-2P-or 316 SARS-CoV-2 S-2P protein-immune mouse serum, 50 µg/mL of fold-on protein was added for 1 317 hr at room temperature (RT). After standard washes and blocks, plates were incubated with 318 serial dilutions of heat-inactivated (HI) sera for 1 hr at RT. Following washes, anti-mouse IgG, 319 IgG1, or IgG2a or IgG2c-horseradish peroxidase conjugates (ThermoFisher) were used as 320 secondary Abs, and 3,5,3′5′-tetramethylbenzidine (TMB) (KPL) was used as the substrate to 321 detect Ab responses. Endpoint titers were calculated as the dilution that emitted an optical 322 density exceeding 4X background (secondary Ab alone). 323

We introduced divergent amino acids, as predicted from translated sequences, into the CMV/R-325 MERS-CoV EMC S (GenBank#: AFS88936) gene 33 to generate a MERS-CoV m35c4 S gene 31 . 326

To produce SARS-CoV-2 pseudoviruses, a codon-optimized CMV/R-SARS-CoV-2 S (Wuhan-1, 327

Genbank #: MN908947.3) plasmid was constructed. Pseudoviruses were produced by co-328 transfection of plasmids encoding a luciferase reporter, lentivirus backbone, and S genes into 329

HEK293T/17 cells (ATCC #CRL-11268), as previously described 33 . For SARS-CoV-2 330 pseudovirus, human transmembrane protease serine 2 (TMPRSS2) plasmid was also co-331 transfected 34 . Pseudoneutralization assay methods have been previously described 11 . Briefly, HI 332 serum was mixed with pseudoviruses, incubated, and then added to Huh7.5 cells or ACE-2-333 expressing 293T cells, for MERS-CoV and SARS-CoV-2 respectively. Seventy-two hr later, cells 334

were lysed, and luciferase activity (relative light units, RLU) was measured. Percent 335 neutralization was normalized considering uninfected cells as 100% neutralization and cells 336 infected with only pseudovirus as 0% neutralization. IC50 titers were determined using a log 337 Concatenated files shown were generated using the same number of randomly selected events from each animal across the different stimulation conditions using FlowJo software, v1

An interactive web-based dashboard to track COVID-19 403 in real time

COVID-19: 406 Emergence, Spread, Possible Treatments, and Global Burden

Rapid COVID-19 vaccine development

Structure-Based Vaccine Antigen 411 Design

Structure of RSV fusion glycoprotein trimer bound to a prefusion-413 specific neutralizing antibody

Structure-based design of a fusion glycoprotein vaccine for 416 respiratory syncytial virus

A proof of concept for structure-based vaccine design targeting RSV 418 in humans

Rapid profiling of RSV antibody repertoires from the memory B 420 cells of naturally infected adult donors

Cryo-electron microscopy structure of a coronavirus spike glycoprotein 423 trimer

Pre-fusion structure of a human coronavirus spike protein. 425

Immunogenicity and structures of a rationally designed prefusion 427 MERS-CoV spike antigen

Emerging viral diseases from a vaccinology perspective: 430 preparing for the next pandemic

Prototype pathogen approach for pandemic 433 preparedness: world on fire

A SARS-like cluster of circulating bat coronaviruses shows 435 potential for human emergence

SARS-like WIV1-CoV poised for human emergence

Novel Vaccine Technologies: Essential 439 Components of an Adequate Response to Emerging Viral Diseases

Rapid development of a DNA vaccine for Zika virus

mRNA vaccines -a new era in 444 vaccinology

Optimization of Lipid Nanoparticles for Intramuscular Administration 446 of mRNA Vaccines

mRNA structure regulates protein expression through changes in 449 functional half-life

Extended Data Figure 7. mRNA-1273 elicits Th1-skewed responses compared to S protein adjuvanted with alum. BALB/c mice were immunized at weeks 0 and 2 weeks with 1 (red

(a-b) Sera were collected 2 weeks post-boost and assessed by ELISA for SARS-CoV-2 S-specific IgG1 and IgG2a. Endpoint titers (a) and endpoint titer ratios of IgG2a to IgG1 (b) were calculated. (c-d) Splenocytes were also collected 4 weeks post-boost to evaluate IFN-γ IL-4, IL-5, and IL-13 cytokine levels secreted by T cells re-stimulated with S1 (c) and S2 (d) peptide pools, measured by Luminex. Dotted line = assay limit of detection. IgG1 and IgG2a/c (a) were compared at each dose level

Pool # 1