Combination of a Sindbis-SARS-CoV-2 Spike Vaccine and αOX40 Antibody Elicits Protective Immunity Against SARS-CoV-2 Induced Disease and Potentiates Long-Term SARS-CoV-2-Specific Humoral and T-Cell Immunity

The COVID-19 pandemic caused by the coronavirus SARS-CoV-2 is a major global public threat. Currently, a worldwide effort has been mounted to generate billions of effective SARS-CoV-2 vaccine doses to immunize the world’s population at record speeds. However, there is still a demand for alternative effective vaccines that rapidly confer long-term protection and rely upon cost-effective, easily scaled-up manufacturing. Here, we present a Sindbis alphavirus vector (SV), transiently expressing the SARS-CoV-2 spike protein (SV.Spike), combined with the OX40 immunostimulatory antibody (αOX40) as a novel, highly effective vaccine approach. We show that SV.Spike plus αOX40 elicits long-lasting neutralizing antibodies and a vigorous T-cell response in mice. Protein binding, immunohistochemical, and cellular infection assays all show that vaccinated mice sera inhibits spike functions. Immunophenotyping, RNA Seq transcriptome profiles, and metabolic analysis indicate a reprogramming of T cells in vaccinated mice. Activated T cells were found to mobilize to lung tissue. Most importantly, SV.Spike plus αOX40 provided robust immune protection against infection with authentic coronavirus in transgenic mice expressing the human ACE2 receptor (hACE2-Tg). Finally, our immunization strategy induced strong effector memory response, potentiating protective immunity against re-exposure to SARS-CoV-2 spike protein. Our results show the potential of a new Sindbis virus-based vaccine platform to counteract waning immune response, which can be used as a new candidate to combat SARS-CoV-2. Given the T-cell responses elicited, our vaccine is likely to be effective against variants that are proving challenging, as well as serve as a platform to develop a broader spectrum pancoronavirus vaccine. Similarly, the vaccine approach is likely to be applicable to other pathogens.


INTRODUCTION
In the ongoing COVID19 pandemic, vaccines play a key role in the strategy to bring SARS-CoV-2 transmission under control. Safety and eliciting a broad-spectrum immune response are paramount for coronavirus vaccine development. Data from vaccine clinical trials and real-world evidence show that available coronavirus vaccines are able to cut the risk of severe COVID19 disease and transmission. However, even with first-generation vaccines currently being globally administered to reduce transmission and severity of the disease, the emergence of circulating variants has raised major concerns that challenge sustained vaccine efficacy, particularly in the face of waning immunity following vaccination (1)(2)(3)(4)(5)(6)(7). Recent data have indicated that escape (appearance and spread of viral variants that can infect and cause illness in vaccinated hosts) from protection by vaccines designed against the Wuhan-1 strain is inevitable (4).
The global COVID19 pandemic is unlikely to end until there is an efficient pan-global roll-out of SARS-CoV-2 vaccines. Although multiple vaccines are currently available, vaccine rollout and distribution at the time of writing this paper is quite incomplete. The three largest countries in the western hemisphere-United States, Brazil, and Mexico-have vaccinated 46%, 13%, and 15% of their populations, respectively, compared with only 4.2% in India (8). Vaccine distribution to date has been highly non-uniform among these and other countries around the globe, encountering many challenges. Unequal vaccine roll-out and the new B.1.617 variant are highly concerning. Major challenges have been supply shortages, logistic problems, complex storage conditions, priced affordably, and safety (9). Consequently, the pandemic is currently sweeping through India at a pace faster than ever before. The countries' second wave became the worst COVID19 surge in the world, despite previous high infection rates in megacities that should have resulted in some immunity. More cost-effective and facilitated delivery of broad-spectrum SARS-CoV-2 vaccines would help improve wide and rapid distribution that would in turn minimize vaccine escape.
Traditionally, vaccines have been designed to induce antibody responses and have been licensed on their capacity to induce high titers of circulating antibody to the pathogen (10). With increased knowledge of host-virus interactions, it has become clear that the cellular arm of the immune response is also crucial for the efficacy of vaccines against pathogens and to provide appropriate help for antibody induction. Various strategies have emerged that specialize in developing candidate vaccines that solely induce either cellular or humoral responses (10). However, as most viruses and pathogens reside at some point during their infectious cycle in the extracellular, as well as intracellular space, vaccines need to promptly elicit a strong T-cell memory response against intracellular pathogens, so that, at the earliest stages of the infective process, prevention can be addressed in coordination with antibodies.
It has been reported that recovered COVID19 patients consistently generate a substantial CD4+ T (OX40+CD137+) cell response against SARS-CoV-2 spike (11). SARS-CoV-2 −specific CD4+ T cells produced IL-2 and substantial amounts of IFNg, hallmarks of Th-1 type effector T-cell polarization. Th-1 type effector T cells provide critical help for CD8 T-cell priming and conferring cytotoxic T cell-mediated immune protection. The costimulatory molecule OX40 is a member of TNF receptor superfamily (TNFRSF) that is upregulated on activated T cells shortly after T-cell receptor recognition of specific antigen (12,13). It is mainly expressed on CD4+ T cells, although activated CD8 T also express OX40, albeit at lower levels (14). Once activated, OX40 receptor is the key molecule for clonal expansion, differentiation, and survival of Th1-effector cells and cytokine production (12,(15)(16)(17)(18)(19). Although OX40 does not directly initiate T-cell memory formation, it contributes to homeostasis of memory T cells and enhances effector memory T-cell function (20). In addition to its role in direct T cellmediated viral clearance (T-cell immunity), OX40 stimulation is found to cooperate with the inducible costimulating (ICOS) molecule on follicular T helper (Tfh) cells augmenting their amplification and development to coordinate humoral immune response (21). Antigen-specific activated Tfh cells help B cells produce high-affinity antibodies against pathogens and are indispensable for vaccine-induced long-lasting humoral immunity by facilitating differentiation of memory B cells and long-lived plasma cells from Germinal Centers (GC) (22)(23)(24). Therefore, designing a vaccine that could stimulate OX40 would provide a powerful platform for T cell-mediated immunity.
Alphaviruses have demonstrated strong attributes as a development-and-manufacturing platform for vaccines (25)(26)(27)(28)(29)(30)(31)(32). Particularly, studies with SARS-CoV strains bearing epidemic and zoonotic spike variants are promising (31). The strength of the use of alphavirus vaccine utilization is the generation of rapid, high-level, and transient nature of transgene expression (33). Importantly, we have shown in our earlier preclinical work (34)(35)(36) that alphavirus vaccine platforms have the advantage to directly deliver antigens and immune modulatory molecules to lymph nodes, where they are expressed transiently to elicit diversified CD4+ and CD8+ T-cell immunity effective at controlling tumors throughout the body.
These vectors represent a highly effective self-amplifying mRNA vaccine that can be engineered to express multiple antigens and stimulatory molecules. Within 3 h after infection, the vector generates hundreds of thousands of mRNA copies within the infected cells and high levels transgene expression (e.g., the spike antigen and anti-OX40 antibody) (37). At the same time, the transient nature and cytosolic location of RNA improve the safety profile of SV vector-based vaccines. The replication defective nature of our vectors ensures no further transmission of the virus beyond the infected cells (38). Replication-deficient alphavirus-based vaccines are immunogenic, safe, well tolerated, and can be cost-effectively stored and transported using conventional 2°C to 8°C storage as well as lyophilization (39,40).
Here we describe a new Sindbis virus (SV) vaccine transiently expressing the SARS-CoV-2 spike protein (SV.Spike). Adaptive immunity induced by SV.Spike fully protects transgenic mice that express the SARS-CoV receptor (human angiotensin-converting enzyme 2 [hACE2]), hACE2-Tg, against authentic SARS-CoV-2 virus infection. In addition, we demonstrate that combination of our vaccine with aOX40 agonistic antibody significantly enhances the induction of immunity by the SV.spike vector. Specifically, seroconversion and abundance of IgG neutralizing antibodies and T-cell immunity through early initiation of Th1-type T-cell polarization are markedly augmented to potentiate long-term immunity protective against SARS-CoV-2 infection in mice. Together these studies develop a safe and effective vaccine platform that provides humoral and cellular immunity to the SARS-CoV-2 spike. This platform has the potential application to other emerging pathogens.

Cell Lines
Baby hamster kidney (BHK) and 293T-cell lines were obtained from the American Type Culture Collection (ATCC). 293T/ ACE2 cell line was obtained from BEI Resources.

SV Production
SV.Spike expressing vector was produced as previously described (34,35,41,42). Briefly, plasmids carrying the replicon (pT7-SV-Spike) or the helper T7-DMHelper RNAs were linearized with XhoI. In vitro transcription was performed using the mMessage mMachine RNA transcription kit (Invitrogen Life Sciences). Helper and replicon RNAs were then electroporated into BHK cells and incubated at 37°C in aMEM supplemented with 10% FBS. After 12 h, the media was replaced with OPTI-MEM (GIBCO-BRL) supplemented with CaCl 2 (100 mg/l) and cells were incubated at 37°C. After 24 h, the supernatant was collected, centrifuged to remove cellular debris, and frozen at −80°C. Vectors were titrated as previously described (43) and titer expressed as transducing units per milliliter (TU/ml).

Detection of SARS-CoV-2 Spike Pseudotyped Lentivirus Infectivity
Luciferase-and nLacZ-encoding SARS CoV-2 Spike or VSV-G pseudotyped lentivirus titers were determined making serial dilutions of the vectors in DMEM and by infecting 293T/ACE2 cells pre-plated in 96-well culture plates (10 4 cells/well). 24 h later, fresh media was added. For Luciferase-encoding pseudotype, cells were lysed 72 h later using cell lysis buffer and lysates were transferred into fresh 96-well luminometer plates, where luciferase substrate was added (Thermo Fisher), and relative luciferase activity was determined (Supplementary Figure 4C). For nLacZ-encoding pseudotypes, cells were washed with PBS and stained for 16 h at 37°C with X-Gal Solution [1 mg/ml X-Gal in PBS (pH 7. 0) containing 20 mM potassium ferricyanide, 20 mM potassium ferrocyanide, and 1mM MgCl2] (Supplementary Figure 4D). Vector titers refer to the number of infectious particles (transducing units per milliliter of supernatant [TU/mL] and were estimated as the last dilution having detectable reporter activity. Correct assembling of pseudotypes was assessed by western blot following standard protocol, to detect the expression of SARS-CoV-2-spike and p24 proteins. SARS-CoV-2 spike (BPS Bioscience) and p24 (Abcam) recombinant proteins were used as positive controls (Supplementary Figures 4A, B).

In Vivo Experiments
All experiments were performed in accordance with protocols approved by the Institutional Animal Care and Use Committee of New York University Grossman School of Medicine. Six-to 12-week-old female C57BL/6J albino mice (B6(Cg)-Tyr<c-2J>/J, Cat 000058) and hemizygous (B6(Cg)-Tg(K18-ACE2)2Prlmn/J; Cat 034860) (hACE2-Tg) mice expressing the human ACE2 receptor or non-carrier controls were purchased from Jackson Laboratory.

ABSL3 Experiments Using SARS-CoV-2 Coronavirus
Three weeks after prime and boost vaccination doses, hACE2-Tg and non-carrier control mice were challenged with 10 4 pfu particles of SARS-CoV-2 Coronavirus via the intranasal (i.n.) route ( Figure 4F). We recorded daily the body weight of each mouse after infection for a total of 14 days. The New York University Grossman School of Medicine (NYUSOM) Animal Biosafety Level 3 (ABSL3) Facility, located on the third floor of the Alexandria Center for Life Science West Tower, is a 3,000 sq. ft. high-containment research facility under the responsibility of the Office of Science & Research and its Director of High-Containment Laboratories. It has been designed, and it is operated in compliance with the guidelines of the Centers for Disease Control and Prevention (CDC) and the National Institutes of Health (NIH). All research and non-research operations are governed by institutional standard operating procedures (SOPs). As per those SOPs, all users undergo specific training and require medical and respiratory protection clearance. The facility and its SOPs are re-certified by an outside consultant on a yearly basis. The NYUSOM ABSL3 has also been registered with the Department of Health and Mental Hygiene of the city of New York since March 2017.
Exposure of the animals to pseudotyped ( Figures 4D, E), or authentic SARS-CoV-2 ( Figures 4F-H) or SV.Spike only ( Figure 11) subsequent to prime and boost vaccination doses is defined in the manuscript as challenge or recall. Both terms will be mostly interchangeable, where the term challenge describes the condition of the animals after exposure to the authentic coronavirus, whereas recall or "challenge" will be used for reference to the ability of the antigen from pseudotyped SARS-CoV-2 and SV-Spike to recall the immune response.

Mouse Vaccination and Serum Collection
Mice were intraperitoneally (i.p.) immunized with SV.Spike (10 7 TU/ml) in a total volume of 500 ml was injected i.p. into the left side of the animal. The immunostimulatory aOX40 antibody (clone OX-86, BioXCell) was injected i.p. into the left side of the animal at a dose of 250 mg per injection. Mice were boosted once at 2 weeks. Sera were collected at 7 days post-2nd vaccination and used to detect neutralizing antibody (NAbs) activity.
Therapeutic efficacy of vaccines was monitored in two ways: vaccinated hACE2-Tg mice that were challenged with SARS-CoV-2 coronavirus in BSL3 were tested for survival compared to their non immunized control group. Survival and weight was monitored and recorded daily.
In Vivo Delivery of nLacZ-SARS-CoV-2 Pseudotyped Lentivirus and X-Gal Histochemistry Isoflurane-anesthetized 4-week-old young adult hACE2-Tg mice were dosed intranasally with a 70-µl volume of nLacZ-encoding lentiviral vector (titer 5.18 × 10 3 TU/ml). Isoflurane anesthesia (2.5% isoflurane/1.5 L oxygen per minute), and dosing of animals was carried out in a vented BSL-2 biological safety cabinet. For processing of mouse lungs for X-Gal staining of intact tissue, lungs were inflated through the trachea with OCT embedding as described previously (44). Intact airways were submerged in 0.5% glutaraldehyde for 2 h at 4°C, washed in PBS/1 mM MgCl 2 and stained for 16 h at 37°C with X-Gal Solution [1 mg/ml X-Gal in PBS (pH 7. 0) containing 20 mM potassium ferricyanide, 20 mM potassium ferrocyanide, and 1mM MgCl 2 ].

SARS-CoV-2 Spike-hACE2 Blocking Assay
To measure protective NAbs, COVID-19 convalescent plasma from C57BL/6J mice was diluted (1:10) and incubated with recombinant SARS-CoV-2 full-length Spike (BPS Bioscience) for 1 h at 37°C prior to adding to an hACE2 pre-coated ELISA plates. The NAb levels were calculated based on their inhibition extents of Spike and hACE2 interactions according to the following equation: [(1-OD value of samples/OD value of negative control) × 100%]. A neutralizing antibody against SARS-CoV-2 spike (Bio Legend) was used as a positive control.

SARS-CoV-2 Spike Pseudotyped Lentivirus Inhibition Assay
Pseudotyped lentivirus inhibition assay was established to detect neutralizing activity of vaccinated mouse sera and inhibitory ability of antiviral agents against infection of SARS-CoV-2 spike pseudotyped lentivirus in target cells. Briefly, pseudotyped expressing luciferase virus containing supernatants were respectively incubated with serially diluted mouse sera at 37°C for 1 h before adding to target cells pre-plated in 96-well culture plates (10 4 cells/well). 24 h later, fresh media was added, and cells were lysed 72 h later using cell lysis buffer. Lysates were transferred into fresh 96-well luminometer plates. Luciferase substrate was added (Promega), and relative luciferase activity was determined. The inhibition of SARS-COV-2 Spike pseudotype lentivirus was presented as % inhibition.

Cell-Cell Fusion Assay
The establishment and detection of several cell-cell fusion assays are as previously described (45). In brief, 293T/ACE2 cells were used as target cells. For preparing effector cells expressing SARS-CoV-2 spike, 293T cells were transiently co-transfected with pCDNA3.1-Spike and pMAX-GFP or with pMAX-GFP only as control, and applied onto 293T/ACE2 cells after 48 h from transfection. Effector and target cells were cocultured in DMEM plus 10% FBS for 6 h. After incubation, five fields were randomly selected in each well to count the number of fused and unfused cells under an inverted fluorescence microscope (Nikon Eclipse Ti-S).

Inhibition of SARS-CoV-2-Spike-Mediated Cell-Cell Fusion
The inhibitory activity of neutralizing antibodies from immunized mice sera on a SARS-CoV-2-spike-mediated cellcell fusion was assessed as previously described (46,47).
Briefly, a total of 2 × 10 4 target cells/well (293T/ACE2) were incubated for 5 h. Afterward, medium was removed, and 10 4 effector cells/well (293T/Spike/GFP) were added in the presence of serum from C57BL/6J immunized mice at 1:100 dilution in medium at 37°C for 2 h. The fusion rate was calculated by observing the fused and unfused cells using fluorescence microscopy.

Immunocytochemistry
Cell immunocytochemistry was performed as described previously (48). Briefl y, cells were fi xed with 4% paraformaldehyde (PFA) for 20 min at room temperature, and then the membrane was permeabilized with 0.1% (vol/vol) Triton X-100 (Fisher Scientific). Incubation with blocking solution (5% normal goat serum) was performed at room temperature for 45 min. Anti-mouse SARS-CoV-2-spike (GTX, 1:100) and anti-rabbit hACE2 (Thermo Fisher,1:100) were applied overnight at 4°C followed by incubation of appropriate secondary antibodies conjugated with fluorophores. Confocal images were captured using the Zeiss LSM-800 system.

Flow Cytometry
For flow cytometry analysis, spleens were harvested from mice and processed as previously described (35). Extracted lungs were chopped in small pieces and incubated with a digestive mix containing RPMI with collagenase IV (50 mg/ml) and DNAseI (20 U/ml) for 30 min at 37°C. Spleens and lungs were mashed through a 70-mm strainer before red blood cells were lysed using ammonium-chloride-potassium (ACK) lysis buffer (Gibco). Cells were washed with PBS containing 1% FBS and surface receptors were stained using various antibodies. Fluorochromeconjugated antibodies against mouse CD3, CD4, CD44, CD38, ICOS, OX40, CD62L, Perforin, Granzyme B, and Tbet, CXCR5 were purchased from Biolegend. Fluorochrome-conjugated antibodies against mouse CD8a were purchased from BD Biosciences. Fluorochrome-conjugated antibodies against CXCR3 and Ki67 were purchased from Thermofisher. Stained cells were fixed with PBS containing 4% Formaldehyde. For intracellular staining, the forkhead box P3 (FOXP3) staining buffer set was used (eBioscience). Flow cytometry analysis was performed on a LSR II machine (BD Bioscience) and data were analyzed using FlowJo (Tree Star).

T-and B-Cell Isolation
Total T cells were freshly isolated with the EasySep ™ mouse T Cell Isolation Kit. Total B cells were freshly isolated with the EasySep ™ mouse B-cell Isolation Kit. Isolation of T and B cells were performed according to the manufacturer's protocols (Stemcell Technologies).

Enzyme-Linked Immunospot
Enzyme-linked immunospot was performed as previously described (35). Mouse IFNg ELISPOT was performed according to the manufacturer's protocol (BD Bioscience). Freshly isolated T cells (1 × 10 5 /well) were directly plated per well overnight in RPMI supplemented with 10% FBS. No in vitro activation step was included. As positive control, cells were stimulated with 5ng/ml PMA+1mg/ml Ionomycin.

Ex Vivo Cytotoxic Assay
T cells (8 × 10 5 /ml) from C57BL/6J immunized splenocytes were co-cultured with 293T/ACE2 cells (2 × 10 4 /ml), previously infected with 3 × 10 5 TU of SARS-CoV-2 Luc-SARS-CoV-2 spike pseudotyped lentivirus. Cells were co-cultured in a 24-well plate for 2 days in 1 ml of RPMI 1640 supplemented with 10% FBS, washed with PBS, and lysed with 100 ml of M-PER mammalian protein extraction reagent (Thermo Fisher) per well. Cytotoxicity was assessed based on the viability of 293T/ ACE2 cells, which was determined by measuring the luciferase activity in each well. Luciferase activity was measured by adding 100 ml of Steady-Glo reagent (Promega) to each cell lysate and measuring the luminescence using a GloMax portable luminometer (Promega).

Transcriptome Analysis of T Cells
Total RNA was extracted from freshly isolated T cells on day 7 of treatment from spleens using RNeasy Kit (Qiagen). For each group, five C57BL/6J mice were used for biological repeats. RNA-seq was done by NYUMC Genome Center. RNA quality and quantity were analyzed. RNAseq libraries were prepared and loaded on the automated Illumina Novae 6000 Sequencing System (Illumina). 1× S1 100 Cycle Flow Cell v1.5, 30 automated stranded RNA-seq library prep polyA selection, per sample.

RNA-Seq Data Analysis
RNA-seq data were analyzed by sns rna-star pipeline (https:// github.com/igordot/sns/blob/master/routes/rna-star.md). Sequencing reads were mapped to the reference genome (mm10) using the STAR aligner (v2.6.1d) (49). Alignments were guided by a Gene Transfer Format (GTF) file. The mean read insert sizes and their standard deviations were calculated using Picard tools (v.2.18.20) (http://broadinstitute.github.io/picard). The read count tables were generated using subread (v1.6.3) (50), (normalized based on their library size factors using DEseq2 (51), and differential expression analysis was performed. To compare the level of similarity among the samples and their replicates, we used principal-component. All the downstream statistical analyses and generating plots were performed in R environment (v4.0.3) (https://www.r-project.org/). The results of gene set enrichment analysis were generated by GSEA software (52,53). The network of Gene Ontology terms was generated by Enrichment Map in Cytoscape. Additional protein-protein functional associations used in this study for bar graphs were retrieved from STRING (http://www.string-db.org/, version 11) (54), a well-known public database on several collected associations between proteins from various organisms.

Measurement of Oxygen Consumption and Extracellular Acidification Rates of T and B Cells
T-and B-cell metabolic output was measured by Seahorse technology as previously described (55). Purified T cells from C57BL/6J immunized or control mice were plated at 6 × 10 5 cells/well in a Seahorse XF24 cell culture microplate. Oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) were measured using an Agilent Seahorse XFe24 metabolic analyzer following the procedure recommended by the manufacturer (Agilent). For the mitochondrial stress test: (1) oligomycin (1 mM), (2) FCCP (1.5 mM), and (3) rotenone (100 nM) and antimycin A (1 mM) were injected sequentially through ports A, B, and C.

Immunoblot Analysis
Western blot was performed to detect SARS-CoV-2 spike protein in 293T cells infected with SV.Spike and in the generated pseudotyped lentivirus. Cells were lysed in M-PER ® Mammalian Protein Extraction Reagent (Thermo Fisher) according to the manufacturer's protocol. Lysates were separated by SDS-PAGE on 4-12% Bio-Rad gels, transferred to polyvinylidene difluoride (PVDF) membranes, blocked in 5% milk in TBS buffer with 0.1% Tween-20 (TBST). Primary antibodies to SARS-CoV-2 spike (GTX, 1:1000) and p24 (Abcam, 1:1000) were added overnight at 4°C. HRPconjugated secondary antibodies were added in 5% milk in TBST for 1 h at room temperature. BioRad Imaging System was used for visualization.

Statistical Analysis
Statistical analysis was performed using GraphPad Prism 7.0 as described in figure legends. All data are shown as mean ± SEM. Figures were prepared using GraphPad Prism 7, Adobe Photoshop, and ImageJ Software. Treated groups were compared using a one-way or two-way ANOVA analysis with Bonferroni correction) to naïve mice. Immunophenotyping statistical analysis was determined with the Kruskal-Wallis test followed by Dunns' test. Correlation coefficients (R) were calculated by using the Pearson's rank matrix. Differences with a P value of <0.05 were considered significant: *P<0.05; **P<0.005; ***P<0.001,****P<0.0001.

Data Availability Statement
All sequencing data that support the findings of this study have been deposited in the National Center for Biotechnology Information Gene Expression Omnibus (GEO) and are accessible through the GEO Series accession number GSE176203. All other relevant data are included in the article/ Supplementary Material, further inquiries can be directed to the corresponding authors.

Construction and Characterization of Sindbis Carrying the SARS-CoV-2-Spike
We designed and generated a Sindbis alphavirus replicon carrying the SARS-CoV-2 spike mRNA. SV vectors are generated from two plasmids: a replicon and helper ( Figure 1 and Supplementary Figure 1). Genes of interest (GOI) can be substituted for the 5kb structural genes that were removed to generate the helper plasmid. The plasmid encoding the structural genes does not contain a packaging signal, preventing further virus assembly beyond the initial preparation of the vectors in BHK-21 cells. Plasmids are transcribed from the T7 promoter and the RNA transcripts are electroporated into BHK-21 cells to produce viral vectors.
The combination of SV vectors encoding a selected antigen with immunomodulatory antibodies makes them far more effective than they are alone (35,36,56). In particular we have found that combining SV vectors expressing specific antigens with aOX40 generates very potent immune responses capable of eradicating tumors in multiple murine models and conferring long-term protection against tumor recurrences or rechallenges (36,56).
The overall design in the production of Sindbis SARS-CoV-2 spike (SV.Spike) is illustrated in Figure 1 and Supplementary  Figure 1. We determined the expression of the full-length SARS-CoV-2 spike from infected cells by western blot in Figure 1B.
The immune responses induced by the Sindbis SARS-CoV-2 spike (SV.Spike) vaccine candidate were analyzed in C57BL/6J mice. Groups of mice (n = 5) were immunized by intraperitoneal (i.p.) or intramuscular (i.m) route, by prime-boost vaccine strategy with SV.Spike and/or aOX40, with 14 days difference between the two doses ( Figure 1C). Immune response analysis from splenocytes, T cells, or mice sera was carried at different time points, as indicated in Figure 1C.

Sindbis Vaccine-Elicited Antibodies to SARS-CoV-2 Spike
Serum IgM, IgG, and IgA responses to SV.Spike, SV.Spike +aOX40, injections were measured on days 21, 75, and 100 days after vaccination by enzyme-linked immunosorbent assay (ELISA) against recombinant SARS CoV2 spike protein (11,57). Sera from all of mice tested showed reactivity to recombinant SARS-CoV-2 spike protein and, as might be expected, levels of antibodies varied based on the experimental group and time point ( Figure 2). Consistent with previous reports (58-60), levels of IgM and IgG measured at day 21 and 75 post injection (p.i.) were significantly higher in the mice vaccinated with SV.Spike and combination of SV.Spike+aOX40 than in the mice who had received aOX40 alone or the naïve group ( Figure 2A). Moreover, the SV.Spike+aOX40 group showed higher titers of IgG compared with only SV.Spike treatment, for which IgM was the predominant isotype and did not show seroconversion to IgG over the different time points. Specifically, both SARS CoV2specific IgG and IgM antibodies demonstrated the highest expression on day 21 post immunization for the indicated groups (IgG-OD450 of 2.3 for SV.Spike+aOX40 serum, and IgM-OD450 of 1.9 for SV.Spike serum). At days 75 p.i., IgG were still significantly predominant in the sera of the mice immunized with the SV.Spike+aOX40 combination (IgG-OD450 = 1.3), whereas IgM levels did not significantly vary from day 21 to day 100 compared with the control groups ( Figure 2B). Instead, IgM levels in the SV.Spike mice showed a more significant decrease and less lasting reactivity from days 21 to 75 days p.i. (IgM-OD450 of 1.2) compared to the control group, whereas the IgG trend demonstrated significant high levels only at day 21 p.i.
Conversely, IgA levels did not show any significant difference in any of the groups and time points tested (Figures 2A, B). These data support the evidence that immunization of mice with SV.Spike combined with aOX40 elicits a strong and specific immune response compared to SV.Spike alone, which is p r e d o m i n a n t l y r e p r e s e n t e d b y S A R S -C o V -2 I g Gspecific antibodies.

Anti-SARS-CoV-2 Spike Neutralizing Antibodies Induced in Sindbis Vaccinated Mice Block the SARS-CoV-2 Spike Protein From Binding to hACE2 Receptor Proteins
Immediately after SARS-CoV-2 was identified as the causative agent of the COVID-19 outbreak, it was shown that human ACE2 (hACE2) is the main functional receptor for viral entry (61). We hypothesized that the virus-receptor binding can be mimicked in vitro via a protein-protein interaction using purified recombinant hACE2 and the SARS-CoV-2 spike protein. This interaction can be blocked by virus naturalizing antibodies (NAbs) present in the test serum of vaccinated mice.
A competition ELISA assay was developed to detect whether SARS-CoV-2 spike-specific antisera from mice immunized with aOX40, SV.Spike, and SV.Spike+aOX40 could block the interaction between SARS-CoV-2 spike and hACE2. Our assay demonstrated that the specific Spike-hACE2 binding can be neutralized by SV.Spike or SV.Spike+aOX40 sera in a dose-dependent manner, but not by sera from aOX40 alone or naïve groups ( Supplementary Figures 2A, B). Similar results were obtained by the intramuscular route (Supplementary Figure 2C). As shown in Figure 3A, antibodies in the antisera from mice immunized with SV.Spike and combination of SV.Spike and aOX40 at day 21 post-immunization significantly inhibited the binding of SARS-CoV-2 spike to hACE2 compared to the sera from naïve mice, indicating that SV.Spike-induced antibodies could strongly neutralize SARS-CoV-2 infection by blocking the binding of Spike protein on the surface of SARS-CoV-2 to hACE2.
To investigate whether the neutralizing antibody response in immunized mice could maintain a high level for a longer period of time, we tested the neutralization activity of mice sera at 75 days post-immunization. The results showed that, although the overall antibody neutralizing capacity decreased compared to day 21, antibodies from SV.Spike and SV.Spike+aOX40 groups still significantly competed for the binding of the SARS-CoV-2 spike and hACE2 ( Figure 3B), indicating that our SV.Spike vaccine is able to induce relative long-term neutralizing antibody responses.
Next, we investigated if the serum from mice immunized with SV.Spike could inhibit the cell membrane fusion process for viral entry (45,46,62) occurring upon the binding of SARS-CoV-2 spike Receptor Binding Domain (RBD) fragment to the ACE2 receptor on target cells. To establish an assay for measuring SARS-CoV-2-spike-mediated cell-cell fusion, we employed 293T with EGFP only did not elicit such an effect, confirming that CoV-2 Spike-hACE2 engagement is essential for viral fusion and entry.
To determine whether the serum of mice immunized with SV.Spike can block Spike protein-mediated cell-cell fusion, we incubated the effector cells with serum from Naïve, SV.Spike, and/or aOX40 mice (diluted 1:100) at 37°C for 1 h and then we co-cultured them with the 293T/ACE2 target cells. We found that not only were fewer fusing cells observed, but also the size of fused cells were visually smaller in the groups of SARS-CoV-2spike/293T effector cells pre-incubated SV.Spike with or without aOX40 sera compared to controls ( Figure 3C). Quantification of fused cells per field in at least four randomly selected fields revealed a remarkably lower number of cell-cell fusions in both SV.Spike and SV.Spike+aOX40 groups compared to all the other groups. Moreover, SARS-CoV-2 spike-mediated cell-cell fusions were significantly inhibited by serum derived from SV.Spike +aOX40 vaccinated mice, indicating that addition of aOX40 to the vaccination protocol elicits antibodies with enhanced interference of syncytium formation mediated by SARS-CoV-2 infection (Figures 3C, D).
The interference of immunized sera NABs with SARS-CoV-2-hACE2 binding was also determined by immunofluorescence experiments performed by culturing 293T/ACE2 cells with recombinant SARS-CoV-2 spike previously incubated with serum from naïve and SV.Spike and aOX40 immunized mice. The binding between Spike and hACE2 expressed on the cell surface was subsequently visualized via confocal fluorescence microscopy ( Figure 3E). As expected, Spike incubated with SV.Spike+aOX40 serum was incapable of binding to hACE2, while the control group showed evident co-localization with hACE2 on the cell surface.
Taken together, these data demonstrate that SV.Spike sera alone and to a greater extent SV.Spike+aOX40 sera can neutralize SARS-CoV-2 spike-hACE2 interaction and in turn counteract virus entry mediated by cell-membrane fusion.

SV.Spike Vaccine Prevents Infection of SARS-CoV-2 in Transgenic hACE2-Tg Mice
The neutralizing activity of serum from vaccinated mice was determined using Luciferase-encoding SARS-CoV-2 spike pseudotyped lentivirus (63)(64)(65) (Supplementary Figures 5A,  B), by testing the impact of the serum on the lentivirus transduction. Serial dilutions (1:300, 1:600, 1:900: 1:1800, 1:3200, and 1:6400) of mice sera harvested at day 21 and 75 p.i. were incubated with equal amounts of lentivirus for 1 h at 37°C, then plated on 293T/ACE2 cells. We then measured the amount of blocked pseudotyped viral particles in infected cells by determining the amount of luminescence reduction, which reflects the level of neutralizing antibody or molecular inhibitors in the sample. The results showed that the antisera could inhibit SARS-CoV-2 pseudotype infection in a dosedependent manner (Supplementary Figure 5), consistent with the result from the antibody neutralization assay (Supplementary Figure 3). Our results demonstrate that sera from SV.Spike with or without aOX40 immunized mice groups resulted in significantly high levels of neutralizing antibodies both at day 21 and 75, since they overcame the pseudotyped lentivirus infectivity inhibition threshold of 30% ( Figures 4A, B). Moreover, serum from these mice receiving combination of SV.Spike and aOX40 gave the highest levels of neutralization at day 21 after vaccination (95.3% of inhibition), with a slight decrease at day 75 (79% of inhibition). Naïve and aOX40 groups did not develop a neutralizing antibody response (% inhibition < 30%) at the timepoints tested, consistent with their lack of SARS-CoV-2 spike binding antibodies.
Recently, hACE2 transgenic (B6(Cg)-Tg(K18-ACE2)2Prlmn/ J or hACE2-Tg) mice were used for the development of an animal model of SARS-CoV-2 infection (66). In order to test pseudotyped lentivirus infectivity rate in vivo, we produced a nLacZ-encoding lentivirus expressing SARS-CoV-2 spike protein (Supplementary Figures 4B, D) and we evaluated the vector expression following delivery to hACE-Tg mice airways, by administrating a single dose of nLacZ-pseudotype to 4-weekold hACE2-Tg mice by intranasal inhalation. After 7 days, the airways were harvested and intact glutaraldehyde-fixed tissues were processed for staining with X-Gal for detection of bgalactosidase activity expressed from the nuclear-localized lacZ reporter gene (nlacZ; Figure 4C). Positive X-Gal staining observed in airways upon lentivirus intranasal administration indicated the successful SARS-CoV-2-spike pseudotyped lentiviral vector expression and delivery in mice airways.
In order to investigate the protective effects of SV.Spike vaccination in vivo, we subsequently immunized hACE2-Tg mice with the same strategy as used for the C57BL/6J mice ( Figure 1D). The hACE2-Tg mice were vaccinated at 0 and 2 weeks and then "challenged" with pseudotyped SARS-CoV-2 intranasally at day 21 and 75 post-immunization ( Figure 4D). The lungs were collected at 7 days "post-challenge" and pseudotype delivery was tested by X-Gal staining. As shown in Figure 4E, the nLacZ-SARS-CoV-2-spike lentivirus could not be detected in the lungs from SV.Spike+aOX40 immunized mice, while substantially reduced infectious virus burden was still detected in the lungs from SV.Spike treated mice compared with the naïve group at the indicated time points. As expected, lungs from animals treated with aOX40 showed high amount of pseudotyped particles, as indicated from the very high signal of X-Gal staining ( Figure 4E). Finally, protective immunity was also assessed in young adult vaccinated Tg-ACE2 mice challenged with live SARS-CoV-2 coronavirus. Three weeks after prime and boost vaccination doses, all mice were challenged with 10 4 particles of SARS-CoV-2 via the intranasal (i.n.) route ( Figure 4F). We recorded the daily the body weight of each mouse after infection for a total of 14 days and found that the body weights of both SV.Spike and SV.Spike+aOX40 mice showed a slow decrease at 3-5 days post infection (dpi), with a progressive stabilization, and increase of their weight at day 8-9 post infection. The naïve unvaccinated group showed a faster decrease during 3-5 dpi ( Figure 4G), which led to early mortality around day 8 dpi ( Figure 4H). Vaccinated mice did not evidence any signs of disease at the time the experiment was terminated but were culled on day 14 as required by the protocol, which was performed in an ABSL3 facility. Together, these data suggest that combination of SV.Spike and aOX40 vaccine in mice conferred long-term protection against SARS-CoV-2 infection by eliciting a durable humoral response in mice.

SV.Spike in Combination With aOX40 Metabolically Reprograms and Activates T Cells Shortly After Prime Vaccine Doses
Analysis of SARS-COV-2 specific adaptive immune responses during acute COVID-19 identified coordination between SARS-COV-2-specific CD4+ T cells and CD8+ T cells in limiting disease severity (67). We analyzed vaccine elicited T-cell responses in the spleen 7 days after mice received prime doses of SV.Spike and/or aOX40 and compared the initial T-cell response to naïve mice ( Figure 5). Spleens of mice were excised and a single cell suspension was stained and analyzed by flow cytometry. For a successful vaccine-elicited immune response, differentiation of virus-specific T cells from the naïve to the effector state requires a change in the metabolic pathways utilized for energy production (68). Therefore, metabolic profiles of vaccine-induced T cells are of interest and correlate to vaccinemediated immunity (69).
We performed metabolic analysis of isolated T cells from spleens in an Extracellular Flux Analyzer XFe24 (Seahorse Bioscience) to investigate metabolic changes of T cells. We found, that combining our SV.Spike vaccine with agonistic aOX40 antibody metabolically rewires T cells in vivo shortly after initial vaccine doses (Figures 5A-D). T cells freshly isolated from mice on day 7 after first doses with SV.Spike+aOX40 combination displayed a metabolic shift to a highly bioenergetic state compared with single agent treatment or naïve mice that show a quiescent metabolism ( Figures 5A, B). Naïve T cells are quiescent and characterized by a metabolic program that favors energy production over biosynthesis. Upon T-cell receptor (TCR)-mediated stimulation, T cells become activated and metabolically reprogrammed. The bioenergetic state of metabolically reprogrammed T cells is characterized by a strong increase of oxygen consumption rate (OCR), which is a parameter for mitochondrial respiration (Figure 5A), and a strong increase of baseline extracellular acidification rate (ECAR) ( Figure 5C), which is measured as a parameter for glycolysis. It has been shown that TCR signaling is directly tied to glycolysis (70). We found that T cells isolated from mice vaccinated with SV.Spike+aOX40 displayed a 3-fold increase of OCR and a 10fold increase of ECAR compared to naïve and single agent vaccinated mice. T cells switched to the energetic state ramped up their ATP production ( Figure 5D). A metabolic rapid adaptation is further required for effector T cells cytokine production and signaling. Rapid switch to type-1 cytokine production, such as IFNg and granzyme B (GrB) in antiviral CD8+ T cells is more reliant on oxidative phosphorylation (71). Indeed, immunophenotyping of CD4+ and CD8+ T cells by flow cytometry revealed rapid clonal expansion of CD4+ T and CD8+ T subsets within one week after prime vaccine doses indicated by Ki67 expression on gated CD4+ and CD8+ T cells. As KI67 can be considered an indirect marker for T cell expansion, CD4+ T  Figures 5E, F). Both T-cell subsets were highly activated, indicated by CD38 and CD44 expression (Figures 5G-J) underlining successful vaccine elicited effector T-cell engagement by our vaccine shortly after initial vaccine doses. Similar results were obtained by the intramuscular route (Supplementary Figure 6).

SV.Spike+aOX40 Vaccinated Mice Are Characterized by a Unique T-Cell Transcriptome Signature Profile After Prime Vaccine Doses
To reveal the molecular profile of SV.Spike vaccine induced Tcell responses, we isolated T cells 7 days after prime vaccine doses from spleens of mice from SV.Spike and/or aOX40 vaccinated groups and naïve group. We then performed mRNA deep sequencing (RNAseq) and network analysis ( Figure 6).

Principal-component analysis (PCA) showed a distinct segregation between combined SV.Spike and aOX40
vaccination and all other groups ( Figure 6A). These data suggest, that SV.Spike and aOX40 induces a distinct T-cell response. Indeed, we next looked at gene expression profiles of naïve versus SV.Spike and/or aOX40 and we found that naïve versus SV.Spike+aOX40 markedly showed the highest amount of uniquely upregulated and downregulated total genes with 1,126 upregulated (left) and 328 downregulated transcripts ( Figure 6B). Overall, in all groups more genes were significantly upregulated than downregulated ( Figures 6B, C). These data suggest that SV.Spike+aOX40 changes the transcriptome signature of T cells. We performed Gene Ontology (GO) functional enrichment analysis (also Gene Set Enrichment Analysis, GSEA) and network analysis from naïve mice versus SV.Spike+aOX40 ( Figure 6D) and naïve versus SV.Spike only ( Figure 6E) immunized mice to determine key pathways and intersections of these pathways. The majority of pathways were upregulated in T cells isolated from mice immunized with SV.Spike+aOX40 with the exception of one cluster downregulated (ribosomal biogenesis). The upregulated pathways in the combination immunized mice were dominated by immune response, T-cell activation, chemokine/cytokine signaling, immune cell migration, DNA replication, chromosomal organization, cell cycle regulation, and chromatin modification that formed the central nodes of this network ( Figure 6D). SV.Spike single agent immunized mice showed a smaller network of seven upregulated pathways including a main cluster of immune response closely connected to a cluster for to B-cell engagement, a small cluster of cytokine production, chemotaxis, cell cycle, DNA replication, regulation of ROS ( Figure 6E). We next identified the top 10 hub GO terms by employing the Maximal Clique Centrality (MCC) for SV.Spike ( Figure 6F) and SV.Spike+aOX40 ( Figure 6G) immunized mice. We found that top 10 hub GO terms in SV.Spike only immunized mice were a selected network cluster of B-cell stimulation and Immunoglobulin regulating pathways compared to the combination that represents a cluster of lymphocyte activation and differentiation regulating pathways. Additionally, we performed Protein Association Network Analysis using STRING to identify differentially expressed genes (DEGs)encoded protein-protein interactions (PPIs). Significantly upregulated DEGs (log2FC>2, p<0.05) in T cells of SV.Spike and/or aOX40 vaccinated mice compared to naïve were analyzed to assess overrepresentation of Gene Ontology (GO) categories in Biological Processes in all groups (Supplementary Figure 7). GO Biological Processes (Strength ≥1; p<0.05) identified by STRING for each group were assigned to one of 7 clusters (apoptosis, light green; cell cycle, red; cellular signaling, dark blue; chemokines/chemotaxis, yellow; cytokines, pink; immune response, light blue; mitochondrial ATP production, dark green). Each GO Biological Process term is defined by one gene set. The amount of contributing DEGs from mice immunized with SV.Spike and/or aOX40 in each gene set is shown as percentage. We identified fourteen biological processes for aOX40, thirteen for SV.Spike, and forty-five for the combination vaccine strategy. We found cell cycle-related processes solely in the SV.Spike+aOX40 combination. The highest amount of chemokines/chemotaxis-related processes was observed in the combination (eleven) compared to aOX40 (four) and SV.Spike (four) alone. Six cytokine-related pathways were upregulated in the combination versus SV.Spike (one) and aOX40 (two) and 14 immune response-related terms were upregulated in the combination versus SV.Spike (four) and aOX40 (three). Overall, the percentage of DEGs that contribute to each biological process was highest in the combination vaccinated group compared to SV.Spike and aOX40 alone. Top 20 ranking of selectively enriched GO terms in the GSEA (FDA<0.05) revealed (GO) immunoglobulin production in the SV.Spike group ( Figure 6H) and (GO) response to chemokine in the combination immunized mice group (Figures 6I, J). We analyzed expression of single signature gene transcripts for each immunized mouse group. We found the highest upregulation of DEGs (p<0.05) indicating T-celldependent B-cell stimulation for building up humoral immunity against SARS-CoV-2 (ICos, Cxcr5, Il21, Cxcl13), differentiation of Th-1 type effector T cells associated with vaccine effectiveness (Tnfrsf4, Cd44, ICos, Cxcr3, Ccr5, Il2, Ifng, Tbx21, Ccl3, Ccl4, Ccl9) and antiviral cytotoxic T-cell stimulation for T-cell immunity (Gzma, Gzmb, Gzmk) in the SV.Spike +aOX40 immunized mice compared to single agent treated groups ( Figure 6H).
In conclusion, these findings indicate that synergistic SV.Spike+aOX40 vaccine combination successfully changes the transcriptome profile of T cells that is indispensable for building up humoral and T-cell immunity. memory T cells, and T-cell clustering around activated antigen presenting cells (APCs). For example, virus-specific cytotoxic T lymphocytes (CTLs) are quickly recruited to influenzainfected lungs by a Th1 response, specifically due to the production of IFNg (72) Two of these Th1-type effector T-cell chemokine receptors are CXCR3 and CX3CR1. We performed flow cytometry analysis to investigate CD4+ T helper differentiation, formation, and antiviral cytotoxic effector T-cell differentiation in T cells from SV.Spike and/or aOX40 immunized animals (Figure 7). Vaccines mimicking an infection can help to build up immunity to the site of inflammation. Two of these Tfh-like effector T-cell chemokine receptors are CXCR3 and CX3CR1. We found a significant increase of CXCR3 and CX3CR1 positive expressing CD4+ T cells (Figures 7A, B) from spleens 7 days after administration of prime vaccine doses in the SV.Spike+aOX40 immunized mice group indicating mobility of generated Tfh-like effector T cells. Immunophenotyping by flow cytometry revealed a 2-fold increase of the transcription factor Tbet and immune costimulatory molecule ICOS-double-positive Tfh-like effector CD4+ T cells compared with single agent vaccinated mice. Tbet+ ICOS+ are hallmarks of Tfh-like T-cell polarization ( Figures 7C, D).

CD4+ T-Cell Help Promotes Effector Differentiation of Cytotoxic T Cells
The predominant pathway used by human and murine CD8+ T cells to kill virus-infected cells is granule exocytosis, involving the release of perforin and GrB. It is known from influenza vaccine research that GrB correlates with protection and enhanced CTL response to influenza vaccination in older adults (73). We looked at CTLs after day 7 of prime doses and found that combination immunization significantly increased differentiation of CTLs as indicated by GrB+ expression (Figures 7E-H) and perforin (Figures 7I, J) upregulation within one week after initial vaccine doses. Seven days after mice groups received booster doses that were administered on day 14, we found a robust 10-fold upregulation of GrB+ positive CD8+ T cells indicating successful vaccine elicited differentiation of cytotoxic T cells (Supplementary Figure 8).
Interestingly, it has been reported that cytotoxic CD4+ T cells can compensate for age-related decline of immune cell protection such as B-cell loss and a less robust antibody response (74). Strikingly, we found in SV.Spike+aOX40 immunized mice showed a significant increase of cytotoxic CD4+ T cells indicating that our vaccine not only induced Tfh-like CD4+ T helper functions but has the potential to improve direct CD4+ T cell-mediated virus-killing, thus, adding an extra layer to immune protection against SARS-CoV-2 in more vulnerable older populations (Figures 7E-J). One important early feature of response to the SV.Spike+aOX40 immunization is a strong interferon-gamma (IFNg) secretion ( Figure 7K), which is associated with polarization to Tfh-like effector cells and cytotoxic T cells. To investigate the recruitment and specificity in CTLs to prevent SARS-CoV-2 cell entry, we analyzed the potential of T cells isolated from SV.Spike and/or aOX40 immunized and naïve mice on day 7 after prime doses to block the infection of 293T cells with SARS-CoV-2-spike expressing, luciferase-encoding pseudovirus. VSVG expressing, luciferase-encoding pseudovirus was used as control. We found that splenic T cells from SV.Spike and SV.Spike+aOX40 mice potently inhibited infection with SARS-CoV-2 pseudotyped lentivirus ( Figure 7L) compared with VSV-G pseudotyped control T cells ( Figure 7M). In conclusion, SV.Spike+aOX40 activated T cells display a Th-1 effector phenotype that promotes effector differentiation and direct T cell-mediated cytotoxicity against SARS-CoV-2 spike within one week after prime vaccine doses.

SV.Spike in Combination With aOX40 Drives Metabolic Activation of B-Cells and T Cell-Dependent B-Cell Support
Almost all durable neutralizing antibody responses as well as affinity matured B-cell memory depend on CD4+ T-cell helper. GSEA of RNAseq data between T cells from the SV.Spike +aOX40 vaccinated and naive group one week after prime vaccine doses revealed selective enrichment of the gene set characteristic for activation of B-cells ( Figure 8A) (p<0.05). To test if SV.Spike combination with aOX40 selectively regulates T cell-dependent B-cell activation, we investigated CD4+ T cell activation and differentiation in mice vaccinated with SV.Spike and/or aOX40 one week after booster vaccine doses by flow cytometry analysis. We found that SV.Spike+aOX40 immunized mice had a 3-fold significant increase of overall CD44+-positive splenic CD4+ T cells compared to naïve mice (Supplementary Figure 9). We next analyzed follicular CD4+ T helper (Tfh) cells that are a subset of CD4+ T cells required for most IgG responses promoting high-quality neutralizing antibodies and we found a 3-fold increase of ICOS+CXCR5+ (Figures 8B, C) and a 2 fold increase CD44+CXCR5+ (Figures 8D, E) positive CD4+ T cells in splenocytes from the SV.Spike+aOX40 group indicating Tfh cell differentiation. We isolated B-cells from spleens and performed a metabolic flux analysis on day 21 after initial vaccine doses and we found that isolated B-cells from SV.Spike +aOX40 immunized mice were metabolically reprogrammed indicating successful vaccine elicited B-cell activation ( Figures 8F-H). Activated B-cells in the combination immunized group experienced a 2.5-fold increase in mitochondrial respiration ( Figures 8F, G) and glycolysis ( Figures 8G, H) when compared to B-cells isolated from mice spleens that were vaccinated with a single agent or compared to naïve mice. Association analysis of the frequencies of Tfh cells with SARS-CoV-2 spike IgG antibody titers revealed that Tfh cells positively correlated with the SARS-CoV-2 spike IgG serum levels in the SV.Spike (R 2 = 0.9722, P=0.002) and SV.Spike+aOX40 group (R 2 = 0.83, P = 0.0290) with the highest amounts of IgG antibodies and Tfh cells in the combination ( Figure 8I). Taken together, these results indicate SV.Spike+aOX40 vaccine induced the most potent T celldependent B-cell response.

Combination of SV.Spike and aOX40 Promotes Robust T Cell Immune Response in Lungs
Most vaccines for airborne infectious diseases are designed for delivery via the muscle or skin for enhanced protection in the lung. We investigated if SV.Spike vaccine-induced T cells can readily home most efficiently to the lungs prior to and shortly after pathogen exposure. To address the immune responses in the lungs, we immunized mice with SV.Spike and/or aOX40 and excised PBS-perfused lungs one week after booster doses for single cell suspensions and performed flow cytometry staining ( Figure 9 and Supplementary Figure 9). We found an increase of ICOS+ CXCR5+ double-positive T helper cells indicating presence of B-cell supporting Tfh cells in the SV.Spike single agent and combination immunized group. We further found an increase of Th-1 type effector CD4+ T cells in lungs from combination treated mice indicated by expression of ICOS +Tbet+ double-positive effector CD4+ T cells (Figures 9C, D). We next investigated if effector CTLs were successfully recruited into the lungs after 3 weeks of initial vaccine administration. While we found the highest increase of differentiated cytotoxic CD4+ T and CD8+ T cells in lungs from the combination treated group ( Figures 9E-H and Supplementary Figure 8), we observed a significant increase of differentiated cytotoxic CD8+ T cells homing in the lungs of the SV.Spike single agent immunized group, although this increase was less pronounced compared to the combination group. These data indicate a successful recruitment of vaccine mediated antiviral Tfh-like effector T cells to the lungs.

SV.Spike and aOX40 Promotes CD4+ T-Cell Memory Formation and Long-Term Protection Upon Recall With SARS-CoV-2 Spike Antigen
Boosting both, local and systemic memory T-cell response is a useful strategy to achieve long term immunity. We analyzed development of T-cell memory in spleens fourteen weeks after initial prime vaccine doses of SV.Spike and/or aOX40 primeboost immunized mice by flow cytometry. We found that mice in the SV.Spike+aOX40 combination group developed significant effector CD4+ T memory indicated by CD44+ CD62L− CD4+ T cells ( Figures 10A-C) compared with naïve mice, reiterating the importance of the combination vaccination in generating strong immune responses memory protection from infection and/or disease against SARS-CoV-2.
To further explore the long-term protection efficacy of our SV.Spike vaccine against SARS-CoV-2 virus exposure, C57BL/6J mice (n = 5 each group) received prime and boost immunizations of SV.Spike and/or aOX40 and placebo (naïve group) via the i.p. route. At day 100 post-immunization, we additionally administered one dose of SV.Spike, to recapitulate spike antigen endogenous entry through SV vector injection ( Figure 11A). Spleens or sera from "challenged" mice were collected 3 days after SARS-CoV-2 spike antigen injection and processed for T-cell response analysis ( Figures 11B-F and Supplementary Figure 10) and detection of specific anti-spike protein IgA, IgM, and IgG isotypes by ELISA ( Figure 11G). The SARS-CoV-2 pseudotyped lentivirus infectivity assay revealed that mice immunized with SV.Spike or SV.Spike and aOX40 are effective in reactivating circulating cytotoxic T cells (CTLs) upon "challenge" with spike antigen ( Figure 11B). CTLs reactivation was also observed by flow cytometry as indicated by granzyme B upregulation in mice receiving combination vaccination ( Figures 11C, D). Moreover, immunophenotyping analysis showed that CXCR5-ICOS-double-positive Tfh like effector CD4+ T cells were increased in "challenged" mice receiving SV.Spike combination vaccination compared to the same group of "unchallenged" mice ( Figures 11E, F).
Antibody response analysis showed that immunization with SV.Spike or SV.Spike+aOX40 followed by Spike antigen injection induced strong production of IgM antibodies compared with the mice that did not received the antigen and the naïve groups. This was particularly evident in mice vaccinated with SV.Spike ( Figure 11G). Strikingly, we noticed that combination of SV.Spike and aOX40 followed by "challenge" with antigen stimulated a high peak of spikespecific IgG antibodies levels that were about 4 times higher than the IgG levels of "unchallenged" mice and control group mice. No significant difference in the spike-specific IgG response was detected in SV.Spike or single aOX40 "challenged" mice compared with the respective "unchallenged" mice and the control groups, whereas no SARS-CoV-2 spike-specific IgA was not detected in any of the groups ( Figure 11G). To corroborate and validate our previous data, C57BL/6J mice (n = 5 each group) were "challenged" with SV.Spike at 180 days after receiving prime and boost immunization via the intramuscular (i.m.) route (Supplementary Figure 11).  Figure 11B). Moreover, immunization with SV.Spike and aOX40 resulted in activation of effector CD4+ memory T cells, as indicated by CD44+ CD62L − CD4+ T cells upon "challenge" (Supplementary Figure 11C). Additionally, IgG seroconversion (Supplementary Figure 11D) and neutralizing antibodies (Supplementary Figure 11E) levels 180 days after immunization with vaccine combination were found significantly higher in mice exposed to spike antigen compared to the "unchallenged" and naïve control groups. Together, these data suggest that combination vaccination with SV.Spike and aOX40 conferred remarkably long-term and specific protection against SARS-CoV-2 infection by eliciting a durable humoral and T-cell response, regardless of the vaccine administration route.

DISCUSSION
An ideal vaccine against SARS-CoV-2 would be effective after one or two immunizations, conferring long-term protection to target populations such as the elderly or immunocompromised individuals, and reducing onward transmission of the virus to contacts (75). It would protect against a broad range of coronaviruses and evolving variants, i.e., offer pancoronavirus protection. The benefit of developing such a vaccine would be Potentiates Sindbis-Spike SARS-CoV-2-Elicited Immunity even greater if it were available to be rapidly deployed in time to prevent repeated or continuous epidemics, economical and readily distributable worldwide without temperature constraints that limit access. This supports the use of alphavirus vaccine platforms that are rapid and straightforward to produce inexpensively, with less challenging temperature requirements, and with previously proven safety and efficacy (76)(77)(78)(79). The alphavirus-based replicon platform technology has been developed as vaccine candidates for many different infectious diseases, including influenza A virus (IAV), respiratory syncytial virus (RSV) (80, 81) Ebola (EBOV), hepatitis C virus (HCV), chikungunya (CHIKV, now in phase III) (82,83) HIV (now in phase I), human papilloma virus (HPV, now in therapeutic phase II) (84). Given the generic design of these platform and that new constructs can be made rapidly with synthetic design of the insert, it can be readily adapted to SARS-CoV-2 as we have demonstrated here. Moreover, when new virus species emerge, a vaccine platform that can be rapidly adapted to emerging viruses is highly desirable.
Sindbis and other alphaviruses have a natural tropism for lymphatic tissues and dendritic cells, relative resistance to interferon, high expression levels, lack of pre-existing antivector immunity in most human and animal populations, and efficient production in a broad range of cell lines, with an accepted regulatory pedigree (85). These observations indicate that a vaccine platform based on Sindbis virus vectors could contribute significantly to dealing with current and future vaccine needs. In nature SV has the safest profile among alphaviruses. SV is an RNA virus without replicative DNA intermediates and poses no risk of chromosomal integration or insertional mutagenesis. Hence, its presence is transitory. To avoid even transient adverse effects, our vectors have been attenuated by splitting the SV genome and by removing the packaging signal from the genomic strand that encodes the structural genes. Moreover, the combination of SV vectors with immunomodulatory antibodies like aOX40 makes them extremely effective compared to the single vector treatments (86,87).
Neutralizing antibodies (NAbs) have conventionally been the desired outcome of vaccination, as they are capable of intercepting and neutralizing microbes and their components as well as eliciting destructive anti-microbial innate immune responses (88). Nonetheless, humoral immunity can decline over time and, as seen with influenza, can only last as short as one season. Many newer vaccines and vaccines in development are also designed to generate T-cell responses that have the potential to help the antibody response, promote long-term immune memory, have direct effector functions themselves, or activate innate effector cells such as macrophages and neutrophils (60,89).
Here, we developed a Sindbis-based Spike-encoding RNA vaccine against SARS-CoV-2 and demonstrated that immunization with SV vector expressing SARS-CoV-2 spike along with a costimulatory agonistic aOX40 antibody induced a synergistic T cell and antibody response. Our adaptable approach has the potential to boost tissue specific immunity and immune memory against respiratory viruses and aims to develop vaccines with long-term protection. As a viral vector, we found that a Sindbis vector expressing SARS-CoV-2 spike antigen in combination with aOX40 markedly improves the initial T-cell priming, compared with the viral vector alone, which results in a robust CD4+ and CD8+ T-cell response and stable SARS-CoV-2 specific neutralizing antibodies, compared with the viral vector alone. The vaccine efficiently elicits effector T-cell memory in respiratory tissues with a potential for long lasting protection against COVID19 through multiple beneficial mechanisms. Importantly, it provided complete protection against authentic SARS-CoV-2 challenge in hACE2 transgenic mice preventing morbidity and mortality. It has been shown that OX40 controls survival of primed CD8+ T cells and confers CTL-mediated protection (19,90). CTLs are a critical component of the adaptive immune response but during aging, uncoordinated adaptive responses have been identified as potential risk factors that are linked to disease severity for the outcome of COVID19 patients. It is known from influenza vaccine research that granzyme B correlates with protection and enhanced CTL response to influenza vaccination in older adults. We looked at cytotoxic T cells (CTLs) and found that combination vaccination significantly increased CD8+ cytotoxic T cells indicated by granzyme B and perforin upregulation. A limitation of the present study is the lack of demonstration of antigen specific T cells. However, the nature of the T-cell response has been investigated through impact from the pseudotyped and authentic SARS-CoV-2 experiments. Almost all durable neutralizing antibody responses as well as affinity matured B-cell memory depend on CD4+ T helper cells. Our data indicates that agonistic OX40 antibody facilitates class switching from SARS-CoV-2 IgM and IgG antibodies in combination with SV.Spike. Moreover, in combination vaccinated mice, we found a significant increase of cytotoxic CD4+ T cells indicating that our vaccine not only induced CD4+ T helper functions but has the potential to improve direct CD4+ T cell−mediated virus-killing adding an extra layer to long-term immunity/protection in more vulnerable older populations.
Virus-specific CTL are quickly recruited to influenza-infected lungs by a Th1 response, specifically due to the production of IFNg (72). IFNg regulates various immune responses that are critical for vaccine-induced protection and has been well studied (91,92). In a clinical trial of the now approved BNT162b1 COVID19 vaccine, IFNg secreting T cells increased in participants 7 days after boost (60). In this regard, one important early feature of the response to the SV.Spike +aOX40 immunization is a strong interferon-gamma (IFNg) secretion compared to naïve and single agent immunized mice. We found a significant increase of CXCR3 and CX3CR1 positive expressing CD4+ T cells, indicating effective recruitment and mobility of generated effector Th1 type T cells in mice. The presence of Th1-type effector T cells in the lungs indicates mobility of vaccine induced activated T cells/tissue surveillance, which positively correlates with vaccine induced long-term immune protection and generation of neutralizing antibodies against SARS-CoV-2.
Both humoral and cell-mediated immune responses have been associated with vaccine-induced protection against challenge or subsequent re-challenge after live SARS-CoV-2 infection in recent rhesus macaque studies (93,94) and there is mounting evidence that T-cell responses play an important role in COVID-19 mitigation (11,95,96). We demonstrated that two doses of SV.Spike with or without aOX40 candidate vaccines induced neutralizing antibody titers in all immunized mice, with a strong IgG response in the mice receiving combination vaccination. Moreover, our data show that SV.Spike+aOX40 skewed Tfh cells toward CXCR5 + Tfh differentiation, which positively correlated with the magnitude of IgG isotype response. These findings indicate that the induction of CXCR5 + Tfh cell differentiation through vaccination may be beneficial for eliciting broad and specific NAb responses. Importantly, the synergistic activity of combination vaccination elicited antibodies that were able to efficiently neutralize SARS-CoV-2 pseudotyped lentivirus in all the mice tested. In addition, we show SV-Spike-based recall in mice immunized with combination vaccination led to an increase in cytotoxic T cells and IgG seroconversion and response 3 days only after exposure to recall antigen compared to all other and "unchallenged" groups, reiterating the importance of the involvement of both humoral and cellular immune responses in SARS-CoV-2-mediated immunity.
The SV.Spike platform evaluated in this study has the advantage that it is inexpensive, stable, and easy to produce. Cost projections based on using our upstream and downstream processes for production of a SV based vaccine are in line with or below costs per dose for other vaccines in use today. Moreover, unlike other mRNA vaccine candidates this viral platform does not require a cold-chain during transportation and storage. It can be easily reconstitute after lyophilization process and is suitable for rapid adaptation such that potential new viruses/threats in an emerging outbreak can be rapidly targeted (97). Thus, for emerging pathogens like SARS-CoV-2, the SV platform can be an efficient and cost-effective alternative to the traditional large-scale antigen production or technology platforms that require extended time for implementation. Development of a successful SV vector vaccine is readily translatable into human vaccination efforts.
As shown in this study, SV.Spike can be applied alone or can be combined with immunomodulatory reagents like aOX40 in a remarkably efficient prime-boost regimen.
Our goal is to exploit the combined SV.Spike+aOX40 formulation and integrate the two components into a single vector, to further facilitate administration and immunomodulatory response. Our lab has recently demonstrated that the expression of full-length antibodies from SV vectors is feasible and effective and that we can also integrate a third gene of interest such as an antigen or a cytokine (unpublished). Taken together, the overall data presented provide an insight into antigen design and preclinical evaluation of immunogenicity of SV-based vaccine immunogenicity, and support further development of SV.Spike as a vaccine candidate for protection against COVID-19 and pancoronavirus vaccine generation.

DATA AVAILABILITY STATEMENT
The RNAseq data has been uploaded to NCBI -accession number is GSE176203. Data and materials availability: Correspondence should be addressed to DM.

ETHICS STATEMENT
The animal study was reviewed and approved by the Institutional Animal Care and Use Committee (IACUC) of New York University Grossman School of Medicine. Centers for Disease Control and Prevention (CDC), and the National Institutes of Health (NIH) for the authentic coronavirus experiments.

AUTHOR CONTRIBUTIONS
AS, SO, and DM conceived the study. AS, SO, AH, CP designed experiments. AS and SO performed mouse experiments and related data analysis. AH and CP designed the Sindbis vector. ZL processed and analyzed the RNAseq data. MN, ST, and KS performed BSL3 experiments with live coronavirus and related data analysis. AS, SO, and DM wrote the manuscript. All authors contributed to the article and approved the submitted version.
Licensing agreement with NYU. AS, SO, AH, ZL, CP, and DM are inventors on one or several issued patents and/or patent applications held by NYU that cover Sindbis treatment of neoplasia and COVID19. As part of the Research and Licensing agreement authors who are inventors on patents are entitled to a portion of NYU Langone's royalties received, should Sindbis vectors be approved by the FDA for the therapeutic or vaccination use.
Publisher's Note: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Copyright © 2021 Scaglione, Opp, Hurtado, Lin, Pampeno, Noval, Thannickal, Stapleford and Meruelo. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.