An alphavirus-derived replicon RNA vaccine induces SARS-CoV-2 neutralizing antibody and T cell responses in mice and nonhuman primates

A prime-only replicating RNA vaccine delivered with lipid inorganic nanoparticles induces neutralizing antibody and T cell responses against SARS-CoV-2 in aged mice and macaques.


INTRODUCTION
Severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) first emerged in December 2019 and within 3 months, Coronavirus Disease 2019 , caused by SARS-CoV-2 infection, was declared a worldwide pandemic (1)(2)(3). Coronaviruses are enveloped, single-strand positive-sense RNA viruses with a large genome and open reading frames for four major structural proteins: Spike (S), envelope, membrane, and nucleocapsid. The S protein mediates binding of coronaviruses to angiotensin converting enzyme 2 (ACE2) on the surface of various cell types including epithelial cells of the pulmonary alveolus (4)(5)(6). Protection is thought to be mediated by neutralizing antibodies against the S protein (7,8), as most of the experimental vaccines developed against the related SARS-CoV incorporated the S protein, or its receptor binding domain (RBD), with the goal of inducing robust, neutralizing antibody responses (9)(10)(11). Indeed, previous reports have shown that human neutralizing antibodies protected mice challenged with SARS-CoV (12)(13)(14) and Middle East respiratory syndrome (MERS)-CoV (15) suggesting that protection against SARS-CoV-2 could be mediated through anti-S antibodies. Additionally, SARS vaccines that drive Type 2 T helper cell (Th2) responses have been associated with enhanced lung immunopathology following challenge with SARS-CoV, whereas those with a Type 1 T helper cell (Th1)biased immune response are associated with enhanced CORONAVIRUS An alphavirus-derived replicon RNA vaccine induces SARS-CoV-2 neutralizing antibody and T cell responses in mice and nonhuman primates protection in the absence of lung immunopathology (16,17). Therefore, an effective COVID-19 vaccine will likely need to induce, at the very least, Th1-biased immune responses comprised of SARS-CoV-2-specific neutralizing antibodies.
Nucleic acid vaccines have emerged as ideal modalities for rapid vaccine design, requiring only the target antigen's gene sequence. Other advantages include no need for pathogen culture (inactivated or live attenuated vaccines) or scaled recombinant protein production. In addition, nucleic acid vaccines avoid the problem of pre-existing immunity that can dampen immunogenicity of viral vector vaccines. Recently, clinical trials were initiated with messenger RNA (mRNA) vaccines formulated with lipid nanoparticles (LNPs) and a DNA vaccine delivered by electroporation (18). However, mRNA and DNA vaccines may not be able to induce protective efficacy in humans after a single immunization since, similar to inactivated and recombinant subunit protein vaccines, they typically require multiple administrations over an extended period of time to become effective (19). Virus-derived replicon RNA (repRNA) vaccines were first described in 1989 and have been delivered in the forms of virus-like RNA particles (VRP), in-vitro transcribed (IVT) RNA, and plasmid DNA (20)(21)(22)(23). In repRNA vaccines, the open reading frame encoding the viral RNA polymerase complex (most commonly from the Alphavirus genus) is intact but the structural protein genes are replaced with an antigen-encoding gene (20,(24)(25)(26). Whereas conventional mRNA vaccines, like those initiated in recent clinical trials, are translated directly from the incoming RNA molecules, introduction of repRNA into cells initiates ongoing biosynthesis of antigen-encoding RNA that results in dramatically increased expression and duration that enhances humoral and cellular immune responses (27). In addition, repRNA vaccines mimic an alphavirus infection in that viral-sensing stress factors are triggered and innate pathways are activated through Toll-like receptors and retinoic acid inducible gene (RIG)-I to produce interferons, pro-inflammatory factors and chemotaxis of antigen-presenting cells, as well as promoting antigen cross-priming (28). As a result, repRNA acts as its own adjuvant, eliciting more robust immune responses after a single dose, relative to conventional mRNA which typically requires multiple, 1,000-fold higher doses (29). An effective vaccine to stop a pandemic outbreak like COVID-19 would ideally induce protective immunity rapidly and after only a single dose thus reducing the load on manufacturing at scale due to a requirement for fewer and lower doses. Given that repRNA vaccines often require only a single administration to be effective (30), they offer considerable potential to meet this need.

Vaccine design and formulation
Building on experience with the attenuated Venezuelan equine encephalitis virus (VEEV) TC-83 strain (22,(30)(31)(32)(33)(34), we generated repRNAs incorporating sequences from the SARS-CoV-2 Spike (S) protein, including full length S (repRNA-CoV2S) (Fig. 1A). Using immunofluorescence and Western blot we demonstrated efficient expression of the v5-tagged S protein in BHK cells (Fig. 1B,C). Then, utilizing convalescent serum collected 29 days after onset of COVID-19 as an immunodetection reagent, we demonstrated endogenous expression of S protein in BHK cells, reactive with natural SARS-CoV-2 immune sera (Fig. 1C). Next, we evaluated the ability of repRNA-CoV2S to rapidly generate antibody and T cell responses in mice when formulated with a Lipid InOrganic Nanoparticle (LION) emulsion designed to enhance vaccine stability and intracellular delivery of the vaccine.
LION is a highly stable cationic squalene emulsion with 15 nm superparamagnetic iron oxide (Fe3O4) nanoparticles (SPIO) embedded in the hydrophobic oil phase. Squalene is a known vaccine adjuvant (35,36) and SPIO nanoparticles have a long history of clinical use in MRI contrast and intravenous iron replacement therapy; the unique nonlinear magnetic properties of SPIOs have also been leveraged for new uses in a range of imaging, targeting and therapy applications (37)(38)(39)(40)(41)(42). A key component of LION is the cationic lipid 1,2-dioleoyl-3-trimethylammonium propane (DOTAP), which enables electrostatic association with RNA molecules when combined by a simple 1:1 (v/v) mixing step (Fig. 1D). LION has an intensity-weighted average diameter of 52 nm (PDI = 0.2) measured by dynamic light scattering. The formulation is colloidally stable for at least 3 months when stored at 4 and 25°C (Fig. 1E). When mixed, electrostatic association between anionic repRNA and cationic DOTAP molecules on the surface of LION promotes immediate complex formation, as confirmed by an increase in particle size to an intensity-weighted average diameter of 90 nm detected by dynamic light scattering (Fig. 1F). Gel electrophoresis analysis of LION-formulated repRNA molecules extracted by phenol-chloroform treatment after a concentrated RNase challenge showed substantial protection from RNase-catalyzed degradation compared to unformulated repRNA (Fig. 1G). To evaluate short-term stability of the vaccine, we evaluated repRNA integrity and complex stability on 1, 4 and 7 days after mixing. LION maintained full integrity of the repRNA molecules (Fig. 1H) and complex size (Fig. 1I) at all time points. Immunogenicity of LION/repRNA-CoV2S in C57BL/6 mice A single site intramuscular immunization of C57BL/6 mice (n=5/group) with 10 or 1 μg of LION/repRNA-CoV2S induced 100% seroconversion by 14 days post-immunization and robust anti-S IgG with mean binding concentrations of 200 and 109 μg/ml, respectively, and partial seroconversion (2 out of 5) at a 0.1 μg dose, increasing to 4 out of 5 by day 28 ( Fig. 2A). All three dose groups, 10, 1, and 0.1 μg, responded to a boost immunization, reaching anti-S mean IgG  2A). Both the 10 and 1 μg prime-only doses induced neutralizing antibodies with mean 50% inhibitory concentrations (IC50) of 1:643 and 1:226, respectively, significantly higher than titers induced by the 0.1 μg dose (p=0.03 and p=0.02, respectively), as measured by the pseudovirus neutralization assay (SARS-CoV-2 Wuhan-Hu-1 pseudotype) (Fig. 2B). While all doses induced Th1-biased immune responses indicated by higher IgG2c responses when compared to IgG1 (Fig. 2C), there was a trend toward higher doses inducing greater Th1biased responses as indicated by higher IgG2c:IgG1 ratios (Fig. 2D). Given the potential role for T cells to contribute to immune protection, as seen with SARS and MERS (43)(44)(45), especially in the presence of waning antibody and memory B cell responses, we also evaluated T cell responses to LION/repRNA-CoV2S in mice. On day 28 this same cohort of mice received a second immunization and 12 days later, spleens and lungs were harvested and stimulated with an overlapping 15-mer peptide library of the S protein; IFN-γ responses were measured by enzyme-linked immune absorbent spot (ELISpot) assay. Mice receiving a 10, 1, or 0.1 μg prime/boost exhibited robust splenic T cell responses with mean IFN-γ spots/10 6 cells of 1698, 650, and 801, respectively (Fig. 2E). Robust T cell responses were also detected in the lung and were similar between groups with mean IFN-γ spots/10 6 cells of 756, 784, and 777, respectively (Fig. 2F). Interestingly, analysis of the specificity of peptide response showed a biased response toward the S1 domain of the S protein in the spleen (Fig. S1A), whereas responses in the lung were more broadly distributed between the S1 and S2 domains of the S protein (Fig. S1B).

Immunogenicity of LION/repRNA-CoV2S in young and aged BALB/c mice
The elderly are among the most vulnerable to COVID-19, but immune senescence in this population poses a barrier to effective vaccination. To evaluate the effect of immune senescence on immunogenicity, we next administered 10 or 1 μg of LION/repRNA-CoV2S in 2-, 8-, and 17-month old BALB/c mice and measured anti-S IgG concentrations at 14 days after prime and 12 days after a booster immunization. Significantly lower antibody responses were observed in the 17-month old mice at both doses following a prime immunization (Fig. 3A), when compared to 2-and 8-month old mice (p=0.001 and p=0.03, for the high dose, respectively, and p<0.0001 at both age groups for the low dose). However, following a boost immunization, the 17-month old mice developed significantly higher antibody responses relative to their prime responses (p=0.001 and p<0.0001, for the 10 and 1 μg doses, respectively), approaching the post-boost responses seen in the 8and 2-month old mice; antibody responses in the 17-month old 1 μg dose group were still significantly lower than the 2month-old counterpart (p=0.05) (Fig. 3A). No differences were observed between the 2-and 8-month old mice (Fig. 3A). At day 12 following the boost immunization, spleens were harvested and stimulated with an overlapping 15-mer peptide library of the S protein, and IFN-γ responses were measured by ELISpot assay (Fig. 3B). Mice receiving a 10 or 1 μg prime/boost exhibited robust splenic T cell responses with mean IFN-γ spots/10 6 cells of 1179, 1328, and 946 in the 10 μg groups and 458, 409, and 1231 in the 1 μg groups for the 17-, 8-, and 2-month old mice, respectively (Fig. 3B). No significant differences were observed between age groups, however, more variability in T cell responses were observed in the 17month old mice relative to the 8-and 2-month old mice. Interestingly, although BALB/c mice tend to develop a more Th2 immune-biased response following vaccination (46), LION/repRNA-CoV2S induced ratios of IgG2a:IgG1 greater than 1 (Fig. 3C, D) in all age groups of BALB/c mice, indicating a Th1-biased immune response. Given that severe, lifethreatening COVID-19 appears to be more common among elderly individuals, irrespective of type of T helper cell response, and that severe SARS is associated with skewing toward Th2 antibody profiles with an inadequate Th1 response (16,17,43), the ability of LION/repRNA-CoV2S to induce strong Th1-biased responses in 8-and 2-month old mice, even in the Th2-biased BALB/c strain, is potentially a promising finding.

Safety and immunogenicity of LION/repRNA-CoV2S in pigtail macaques
Having achieved robust immunogenicity with LION/repRNA-CoV2S in mice, we then immunized pigtail macaques (Macaca nemestrina) to determine if the vaccine was capable of inducing strong immune responses in a nonhuman primate model that more closely resembles humans in the immune response to vaccination. Three macaques received a prime-only LION/repRNA-CoV2S intramuscular 250 μg dose, administered in 5 injection sites, at week 0 and two macaques received a 50 μg prime dose at week 0 and a boost at week 4, administered into a single intramuscular injection site (Fig. 4A). Blood was collected 10, 14, 28, 42, 56, and 70 days post vaccination to monitor vaccine safety and immunogenicity. The animals were observed daily by veterinary staff for appetite, urine/fecal output and quality, attitude/activity, and no abnormalities were observed. Additionally, the vaccine injection sites were monitored daily for evidence of skin erythema, swelling, discharge, rash, and ulceration for 14 days post vaccination, and no adverse reactions were observed in any of the animals at the injection sites. Furthermore, there were no abnormalities in weight or temperature in the animals ( Fig. S2A-B), and serum chemistries revealed no abnormal findings, except for mild azotemia (mildly elevated blood urea nitrogen and creatinine) in 1 animal at 14 days post vaccination (Fig. S2C). Complete blood counts for all 5 macaques were unremarkable (Fig. S2D).  Table S1). However, larger group sizes will be needed to confirm this finding.
ELISpot analyses of peripheral blood mononuclear cells (PMBCs) collected 10, 14, 28 and 42 days post immunization, showed that all 5 macaques developed T cell responses to SARS-CoV-2 S protein by day 28 (Fig. 5A). Peak responses varied between days 28-42 and were independent of the vaccine dose. Similar to findings in our vaccinated mice, the T cell response was directed mostly toward the S1 and RBD regions of SARS-CoV-2 S protein (Fig. 5B). Further evaluation of the T cell response by intracellular cytokine staining revealed modest increases in cytokine-producing CD8 + T cells ( Fig. 5C and Fig. S5). Frequencies of IFN-γ producing CD4 + or CD8 + T cells showed an increasing trend in some animals by 42 days post vaccination, although changes were not significant (Fig.  5C). Finally, modest increased frequencies of cytokine or IFNγ producing S-specific effector memory CD8 + T cells were observed by 42 days post vaccination (Fig. 5C,D and Fig. S6).

DISCUSSION
RepRNA vaccines against a variety of infectious diseases and cancers have been shown to be safe and potent in clinical trials (47)(48)(49)(50). The cell-free and potentially highly scalable manufacturing process for repRNA vaccines when used with effective synthetic formulations, such as LION, present further benefits over mRNA vaccines. The two-vial approach with one vial containing the LION formulation and the other vial containing the repRNA vaccine provides a big manufacturing and distribution advantage over formulations that require complex processes to encapsulate the RNA into lipid nanoparticles within a single vial. Instead, the repRNA vaccine and LION formulation can be scaled-up and stockpiled separately and then combined on site prior to use. Here, we show that the LION/repRNA-CoV2S vaccine induced robust neutralization of SARS-CoV-2 in vaccinated mice and pigtail macaques, with a mean PRNT80 titer in all 5 macaques of 1:197. Recently, serum neutralizing titers, measured as the IC50 titer that neutralized SARS-CoV-2 by 50% tissue culture infectious dose (TCID50), were reported in rhesus macaques that were either re-infected (51) or challenged after vaccination with an inactivated SARS-CoV-2 vaccine (52). In the former report, IC50 titers as low as 1:8 were associated with protection from re-infection while in the latter, IC50 titers as low as 1:50 were associated with reduced viral load and protection from lung pathology. In another study, serum neutralizing titers, measured by 50% reduction in luminescence output from a recombinant SARS-CoV-2 virus expressing luciferase, were reported in rhesus macaques that were immunized with a variety of DNA vaccine candidates (53) and titers between 1:60 and 1:200 correlated with complete protection from disease. These data suggest that a 250 μg prime-only dose or a 50 μg prime/boost immunization with the LION/repRNA-CoV2S vaccine induced concentrations of neutralizing antibodies that would be sufficient to protect nonhuman primates from infection and disease. Studies are in progress to develop a pigtail macaque SARS-CoV-2 infection model so that this question can be addressed.
Interestingly, the response kinetics in pigtail macaques showed a prolonged increase in serum antibody titers through day 42 and plateauing through at least day 70 postvaccination. This is most likely due to the sustained antigen expression mediated by repRNA, which may result in plasma cell activation akin to what happens following traditional vaccine booster immunizations. This potential mechanism is First release: 20 July 2020 stm.sciencemag.org (Page numbers not final at time of first release) 5 currently being investigated. However, because the 250 μg dose was administered concurrently to five injection sites in the pigtail macaques, it cannot be ruled out that immunogenicity and safety measures would be different to those following administration of this dose via a single bolus injection.
Studies are ongoing to address this question. Additionally, we demonstrated that LION/repRNA-CoV2S induces S-specific T cell responses in mice and macaques. Given the relatively recent emergence of SARS-CoV-2, we can only speculate based on limited knowledge from previous and recent reports of coronavirus infection as to how T cell responses may contribute to protection from infection and disease. Following natural infection of humans with the related SARS-CoV, neutralizing antibody and memory B cell responses in some individuals are reported to be short lived (~ 3 years), whereas memory T cells persist for at least 6 years (54-56), suggesting a potential role for T cells in long-term responses especially in those who lack robust memory B cell responses. Additionally, anti-S T-cell responses to the related SARS-and MERS-CoVs contribute toward viral clearance in normal as well as aged mice infected with SARS-or MERS-CoV, respectively (43)(44)(45). A recent study indicates SARS-CoV-2-specific T cell responses may be associated with better recovery from COVID-19 (57). Our findings demonstrate that LION/repRNA-CoV2S elicited modest but detectable T cell responses in macaques (Fig. 5), an outcome that is consistent with previous studies of SARS-CoV-2 S-targeted DNA and live viral vector vaccines tested in nonhuman primates (53,58). Additionally, the ability of our repRNA vaccine to induce anti-SARS-CoV-2 S-specific memory T cell responses in macaques (Fig. S6) has implications for contributions to longterm protection and recovery from SARS-CoV-2 infection, however, testing of later time points is needed to determine the longevity of this response.
Together, our results demonstrate potential for LION/repRNA-CoV2S, which will enter clinical development under the name HDT-301, to induce rapid immune protection from SARS-CoV-2 infection. A scalable and widely-distributed vaccine capable of inducing robust immunity in both young and aged populations against SARS-CoV-2 infection in a single shot would provide immediate and effective containment of the pandemic. Although our results demonstrate the potential for a prime-only LION/repRNA-CoV2S vaccine (administered over 5 injection sites), only through clinical evaluation of HDT-301 will we know if a single or a primeboost regimen is required to induce protective immunity in the most optimal and cost-effective way. It is possible that certain demographics, such as the elderly, may require either two doses or a single but higher dose. Critically, the vaccine induced Th1-biased antibody and T cell responses in both young and aged mice as well as pigtail macaques, an attribute that may be important for preventing vaccine-induced immune enhancement of disease. However, these conclusions are drawn in the absence of challenge data and from a limited number of animals and at time points after vaccination where observed responses, especially T-cell responses, exhibited great variability. Additional larger studies will be needed to evaluate safety, kinetics and durability of immune responses, and protection from disease including the absence of immune-enhancement of disease. Together, our results support further development of LION/repRNA-CoV2S as a vaccine candidate for protection against COVID-19.

MATERIALS AND METHODS Study design
The primary objective of this study was to determine and characterize the immunogenicity and safety of a LION/repRNA-CoV2S vaccine in mice and pigtail macaques. Mouse group sizes were based on power analyses using data from previous experiments using a similar repRNA platform and mice were randomly distributed between groups. Group sizes in the macaque study were limited by availability of animals. Five pigtail macaques were stratified based on age and weight. No blinding was used throughout all studies. Endpoints were selected prior to the start of each study and were selected based on the primary objective of characterizing the safety and immune responses to vaccination with a LION/repRNA-CoV2S vaccine. Replication of experiments and the number of biological and technical replicates varied between experiments as described in the figure legends.

SARS-CoV-2 repRNA vaccine production and qualification
Codon optimized gene sequences for SARS-CoV-2 full S corresponding to positions 21,536 to 25,384 in SARS-CoV-2 isolate Wuhan-Hu-1 (GenBank: MN908947.3) fused to a c-terminal v5 epitope tag was synthesized as double stranded DNA fragments (IDT) and cloned into a plasmid vector encoding the 5′ and 3′ untranslated regions as well as the nonstructural open reading frame of Venezuelan equine encephalitis virus, strain TC-83, between PflFI and SacII sites by Gibson assembly (SGI-DNA). The complete antigen sequence used in repRNA-CoV2S is included in data file S1. Clones were then sanger sequenced and prepped for RNA production as follows. Template DNA was linearized by enzymatic digestion with NotI followed by phenol chloroform treatment and ethanol precipitation. Linearized template was transcribed using MEGAscript® T7 Transcription Kit (Invitrogen) followed by capping with NEB Vaccinia Capping System as previously described (59). To qualify the vaccine candidate in vitro, Baby Hamster Kidney (BHK) cells (ATCC) were transfected with repRNA or mock transfected using TransIT-mRNA transfection kit (Mirus Bio) and cells analyzed 24 hours later by immunofluorescence using a mouse anti-v5 AF488 secondary antibody (Invitrogen). Additionally, BHK cells were transfected with repRNA-CoV2S and repRNA-GFP and cell lysates were collected 24 hours later for analysis by SDS-PAGE and by Western blot using recombinant SARS-CoV-2 S protein as a positive control. To detect repRNA-mediated protein expression following transfer to nitrocellulose membrane, anti-v5-HRP or convalescent human serum collected 29 days after onset of PCR-confirmed COVID-19 followed by anti-human Ig-HRP secondary antibody (Southern Biotech) was used.

LION formulation of the repRNA vaccine
To protect the RNA replicons from degradation, we partnered them with a Lipid InOrganic Nanoparticle (LION) formulation that consists of inorganic superparamagnetic iron oxide (SPIO) nanoparticles within a hydrophobic squalene core to enhance formulation stability. LIONs comprise 37.5 mg/ml squalene (Millipore Sigma), 37 mg/ml Span® 60 (Millipore Sigma), 37 mg/ml Tween® 80 (Fisher Chemical), 30 mg/ml DOTAP chloride (Corden Pharma), 0.2 mg/ml 15 nm oleic acid-coated iron oxide nanoparticles (Ocean Nanotech) and 10 mM sodium citrate dihydrate (Fisher Chemical). LION particles were manufactured by combining the iron oxide nanoparticles with the oil phase (Squalene, Span 60, and DOTAP) and sonicating for 30 min in a 65°C water bath. Separately, the aqueous phase, containing Tween 80 and sodium citrate dihydrate solution in water, was prepared with continuous stirring until all components were dissolved. The oil and aqueous phases were then mixed and emulsified using a VWR 200 homogenizer (VWR International) and the crude colloid was subsequently processed by passaging through a microfluidizer at 20,000 psi with a LM10 microfluidizer equipped with a H10Z-100 μm ceramic interaction chamber (Microfluidics) until the z-average hydrodynamic diametermeasured by dynamic light scattering (Malvern Zetasizer Nano S) -reached 50 ±5 nm with a 0.2 polydispersity index. The microfluidized LION was terminally filtered with a 200 nm pore-size polyethersulfone (PES) filter and stored at 2-8°C.

RNase protection assays
Replicon RNA was complexed with LION formulations and placed on ice for 30 min. After diluting the complex using nuclease-free water, complexes containing 1 μg of repRNA at 20 μg/mL were treated with 50 ng of RNase A (Thermo Scientific) for 30 min at room temperature, followed by an incubation with 5 μg of recombinant Proteinase K (Thermo Scientific) for 10 min at 55°C. RNA was then extracted using an equal volume of 25:24:1 phenol:chloroform:isoamyl alcohol (Invitrogen). After vortexing, samples were centrifuged at 17,000 × g for 15 min. The supernatant was collected and mixed 1:1 with Glyoxal load dye (Invitrogen) and heated at 50°C for 15 min. The equivalent of 200 ng of RNA was loaded and run on a denatured 150 mL 1% agarose gel in Northern Max Gly running buffer (Invitrogen) at 120 V for 45 min. Gels were imaged using a ChemiDoc MP imaging system (BioRad). The intensity of the intact RNA band was compared to phenol:chloroform:isoamyl extracted RNA from complexes that were not subjected to RNase and Proteinase K treatment.

Mouse immunizations
All mouse experiments were conducted in accordance with procedures approved by the institutional animal care and use committee. Female C57BL/6 or BALB/C mice (Charles River) were maintained in specific pathogen-free conditions and entered experiments at 6-12 weeks of age unless otherwise indicated. LION and repRNA-CoV2S were complexed at a nitrogen-to-phosphate molar ratio of 15 in 10mM sodium citrate and 20% sucrose buffer and incubated on ice. Animals were dosed within 30 min after complexing LION and repRNA-CoV2S. All described doses indicate the total quantity of RNA that is formulated and administered to the animals. Mice were immunized by intramuscular injection of vaccine delivered in a total volume of 50 μl in the thigh.

Pigtail macaque immunizations
Five adult male pigtail macaques were used in these studies (aged 3-6 years, weight 5-13 kg) due to immediate availability of this species of macaque and prior to availability of rhesus macaque SARS-CoV-2 data. All animals received a previous Hepatitis B virus (HBV) DNA and protein vaccine regimen, comprised of HBV core and surface antigens and anti-CD180 (60), and were re-enrolled in this study in response to the SARS-CoV-2 pandemic. All animals were housed at the Washington National Primate Research Center (WaNPRC), accredited by the American Association for the Accreditation of Laboratory Animal Care International (AAALAC), and as previously described (61). All procedures performed on the animals were with the approval of the University of Washington's Institutional Animal Care and Use Committee (IACUC).
Blood was collected at baseline (week -2 or -1), and at days 10, 14, 28, and 42 post-prime vaccination (Fig. 5A). Blood was also collected 10 days post-boost (38 days post-prime) in the 50μg vaccinated animals. Serum and plasma were collected and PBMCs were isolated from whole blood as previously described (62). Animals were sedated with an intramuscular injection (10 mg/kg) of ketamine (Ketaset®; Henry Schein) prior to blood collection or vaccination. Animals were observed daily for general health (activity, appetite) and for evidence of reactogenicity at the vaccine inoculation site (swelling, redness). They also received full physical exams including temperature and weights measurements at each study timepoint. None of the animals became severely ill during the course of the study and none required euthanasia.
Vaccine was prepared as described above. The 50 μg vaccine was delivered intramuscularly into the quadriceps muscle with one 250 μl injection on weeks 0 and 4. To maintain consistency in the vaccine formulation and concentration, First release: 20 July 2020 stm.sciencemag.org (Page numbers not final at time of first release) 7 the 250μg vaccine was delivered intramuscularly by inoculating 250 μl injections into 5 intramuscular injections sites, 2 in the right quadriceps, 1 in the left quadricep, and 1 each in the left and right deltoids on week 0. All injection sites were shaved prior to injection and monitored post-injection for any signs of local reactogenicity.

Serum chemistries and complete blood counts
Serum chemistries were run on a Beckman Coulter AU 680/5812 system and complete blood counts were determined on a Sysmex XN9000 analyzer by the University of Washington Department of Laboratory Medicine.

Antigen-specific antibody responses
Blood was collected from the retro-orbital sinus of immunized mice, or venipuncture of anesthetized macaques, and serum prepared. Antigen-specific IgG, IgG1, IgG2a, and IgG2c responses were detected by enzyme linked immunosorbent assay (ELISA) using a previously described recombinant SARS-CoV-2 S as the capture antigen (63). ELISA plates (Nunc) were coated with 1 μg/ml antigen or with serial dilutions of purified polyclonal IgG from mouse or monkeys to generate a standard curve in 0.1 M PBS buffer and blocked with 0.2% BSA-PBS. Then, in consecutive order, washes in PBS/Tween, serially diluted serum samples, anti-mouse ormonkey IgG, IgG1, IgG2a, or IgG2c-HRP (Southern Biotech) and TMB then HCL were added to the plates. Plates were analyzed at 405nm (ELX808, Bio-Tek Instruments Inc). Absorbance values from the linear segment of each serum dilution curve was used to interpolate the standard curve and calculate the IgG concentration present in each sample. For IgG isotype-specific ELISAs, endpoint titers were determined using 3 times the standard deviation above the mean of triplicate negative-control sera as the cut-off absorbance value. IgG2c isotype was selected for C57BL/6 sera due to their lack of IgG2a isotype (64), while IgG2a is utilized for BALB/C sera (65).

SARS-CoV-2 pseudovirus neutralization assay
Murine leukemia virus (MLV)-based SARS-CoV-2 Spseudotyped viruses were prepared as previously described (63,66). In brief, HEK293T cells were co-transfected with a SARS-CoV-2 (based on Wuhan-Hu-1 isolate) S-encoding plasmid, an MLV Gag-Pol packaging construct, and the MLV transfer vector encoding a luciferase reporter using the Lipofectamine 2000 transfection reagent (Life Technologies) according to the manufacturer's instructions. Cells were incubated for 5 hours at 37°C with 8% CO2 with DNA, lipofectamine, and OPTIMEM transfection medium. Following incubation DMEM containing 10% FBS was added for 72 hours. Pseudovirus was then concentrated using a 30kDa Amicon concentrator for 10 min at 3,000 × g and frozen at -80C.
BHK cells were plated in 96 well plates for 16-24 hours prior to being transfected with human ACE2 using standard lipofectamine 2000 protocol and incubated for 5 hours at 37°C with 8% CO2 with DNA, lipofectamine, and OPTIMEM transfection medium. Following incubation, DMEM containing 20% FBS was added in equal volume to the OPTIMEM transfection media for 16-24 hours. Concentrated pseudovirus with or without serial dilution of antibodies as incubated for 1 hour at room temperature and then added to the wells after washing 3X with DMEM and removing all media. After 2-3 hours, equal volumes of DMEM containing 20% FBS and 2% PenStrep was added to the cells for 48 hours. Following 48 hours of infection, equal volume of One-Glo-EX (Promega) was added to the cells and incubated in the dark for 5-10 min prior to reading on a Varioskan LUX plate reader (ThermoFisher). Measurements were done in duplicate and relative luciferase units (RLU) were recorded.

Human convalescent sera
De-identified remnant diagnostic samples from individuals with COVID-19 in Washington state were obtained via Northwest Biospecimens (Seattle, WA) following registration for exemption from IRB approval with the University of Washington Human Subjects Division.

SARS-CoV-2 neutralization assay
Three-fold (pigtail macaque) or four-fold (human) serial dilutions of heat inactivated serum and 600 plaque-forming units (PFU)/ml solution of SARS-CoV-2/WA/20 (BEI resources) were mixed 1:1 in DPBS (Fisher Scientific) + 0.3% gelatin (Sigma G7041) and incubated for 30 min at 37°C. Serum/virus mixtures were added in duplicate, along with virus only and mock controls, to Vero E6 cells (ATCC) in a 12-well plate and incubated for 1hr at 37°C. Following adsorption, plates were washed once with DPBS and overlayed with a 1:1 mixture of Avicel RC-591 (FMC) + 2 x MEM (ThermoFisher) supplemented with 4% heat-inactivated FBS and Penicillin/Streptomycin (Fisher Scientific). Plates were then incubated for 2 days at 37°C. Following incubation, overlay was removed and plates were washed once with DPBS and then 10% formaldehyde (Sigma-Aldrich) in DPBS was added to cells and incubated for 30 min at room temp. Plates were washed again with DPBS and stained with 1% crystal violet (Sigma-Aldrich) in 20% EtOH (Fisher Scientific). Plaques were enumerated and percent neutralization was calculated relative to the virus-only control.

Mouse IFN-γ ELISpot assay
Spleen and lung lymphocytes were isolated from mice 12 days after the second vaccination. MIAPS4510-Multiscreen plates (Millipore) were coated with rat anti mouse IFNgamma capture antibody (BD) in PBS and incubated overnight at 4°C. The plates were washed in PBS and then blocked (2h, RT) with RPMI medium (Invitrogen) containing 10% heat inactivated fetal calf serum (Gibco). Lung and spleen cells were plated at 5x10 5  15 mer peptides from Genscript) at 1.5ug/ml/peptide and cultured for 20 hours (37°C, 5% CO2). Biotinylated anti-mouse IFN-gamma antibody (BD) and streptavidin-Alkaline Phosphotase-substrate (Biolegend) were used to detect IFNgamma secreting cells. Spot forming cells were enumerated using an Immunospot Analyzer from CTL Immunospot profession software (Cellular Technology Ltd).

Macaque IFN-γ ELISpot assay
PBMCs were stimulated for 24 hours with SAR-CoV-2 Spike (Genscript) as described above including 17-or 18-mers with 11 amino acid overlap, peptide pools at a final concentration of 1μg/mL. Concanavalin A (Thermofisher) 2.5 μg/mL was used as a positive control, and DMSO at a concentration equal to peptide stimulations was used as a negative control. Antigen-specific cells secreting IFN-γ were detected using an Immunospot Human IFN-γ Single-Color Enzymatic ELISPOT Assay, per manufacturer's protocol. Spot forming cells (SFC) were enumerated as described above and results were considered positive if the number of SFC was greater than that of the negative control and for 1x10 5 cells was ≥ 1 per well.

Statistical analyses
Statistical analyses were conducted in Prism (Graphpad) using one-way analysis of variance and Tukey's multiple comparison test to compare more than two groups. Student's t test, Mann Whitney U test, or Wilcoxon rank sum test were used to compare two groups. Statistical significance was considered when the p-values were < 0.05.

SUPPLEMENTARY MATERIALS
stm.sciencemag.org/cgi/content/full/scitranslmed.abc9396/DC1 Fig. S1. Breadth of T cell responses in C57BL/6 and BALB/c mice. Fig. S2. Vaccination did not induce adverse reactions in pigtail macaques. Fig. S3. Raw ELISA absorbance values for pigtail macaques. Fig. S4. Neutralization curves for pigtail macaque and human samples against SARS-CoV-2/WA/2020 or pseudotyped virus. Fig. S5. Gating strategies used for evaluation of peptide-specific intracellular cytokine staining. Fig. S6. LION/repRNA-CoV2S induces memory T-cell responses in pigtail macaques. Table S1. Convalescent sera from individuals with COVID-19. Data file S1. Source data for all figures. . This construct was fused to a C-terminal v5 epitope tag and then was cloned into an alphavirus replicon encoding the 4 nonstructural protein (nsP1-4) genes of Venezuelan equine encephalitis virus, strain TC-83. Following RNA transcription and capping, repRNA-CoV2S was transfected into BHK cells. 24 hours later, the transfected BHK cells were analyzed by (B) anti-v5 immunofluorescence and (C) Western blot using either convalescent human serum or anti-v5 serum for immunodetection. Recombinant SARS-CoV2 spike protein (rCoV2-Spike) and repRNA-GFP were used as positive and negative controls, respectively. (D) Shown is a graphical representation of LION and formation of the vaccine complex after mixing with repRNA. (E) Shown is the evolution of LION particle size over 15 weeks measured by dynamic light scattering during storage at 4°C, 25°C or 42°C. (F) After mixing LION particles and repRNA, complex formation was confirmed by a shift in size distribution. (G) Gel electrophoresis analysis of triplicate preparations of repRNA extracted from LION particles after a concentrated RNase challenge showed substantial protection relative to a triplicate preparation of a dose-matched naked RNA following RNAse challenge. The formulated vaccine was stable for at least one week after mixing and storage at 4°C or 25°C as determined by (H) gel electrophoresis of repRNA extracted by phenol-chloroform treatment and (I) particle size of the complex. Data in B and C are representative of 2 independent experiments. Data in E, H, and I are from a single experiment, whereas data in F and G are representative of 3 independent experiments. Data in E, G, and I are shown as mean ± s.d. of 3 technical replicates. Scale bar in B is 100 μm.    to determine mean 50% inhibitory concentrations (IC50) of each sample. Additionally, (D) 80% plaquereduction neutralizing antibody titers (PRNT80) against the SARS-CoV2/WA/2020 isolate were measured at days 28 and 42 alongside 7 human convalescent serum samples collected from confirmed COVID-19 patients (see Table S1). The experiment was performed once. Each line in B and C represents each individual animal. Data in D are reported as individual values as well as mean ± s.d. *p<0.05 as determined by students t test comparing 250 μg dose groups at days 14 and 28. There was no significant difference (ns) between mean PRNT80 titers in all 5 animals at day 42 and titers in sera from 7 convalescent humans, as measured by Mann-Whitney U test.
First release: 20 July 2020 stm.sciencemag.org (Page numbers not final at time of first release) 16