INTRANASAL ADMINISTRATION OF THERMOSTABLE RNA VACCINES CROSS-REFERENCE TO RELATED APPLICATIONS [0001] This application claims the benefit of U.S. Provisional Patent Application Serial No.63/345,345, filed on May 24, 2022, the entirety of which is hereby incorporated herein by reference. BACKGROUND Field of the Invention [0002] The present disclosure relates to intranasal vaccine delivery systems. Description of the Related Art [0003] The ongoing severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic has led to the rapid advancement and development of novel vaccines for coronavirus disease 2019 (COVID-19). Since their introduction in late 2020, two now FDA- approved mRNA vaccines (Comirnaty and Spikevax) have had a remarkable impact on the trajectory of the pandemic and were developed at unprecedented speeds. However, while these vaccines have displayed remarkable efficacy at preventing severe COVID-19 and slowing the worldwide pandemic, the lack of vaccine-induced local immunity in the respiratory tract allows for mild infection of vaccinated individuals, resulting in transmission of the virus, despite prior vaccination. Between lack of vaccine induction of mucosal immunity, newly emerging variants, waning efficacy of currently approved mRNA vaccines, the need for frequent booster doses, vaccine hesitancy in part fueled by a fear of needles, and limited access to cold chain storage, global susceptibility to SARS-CoV-2 remains high. Improved vaccines and associated technologies are therefore necessary to fully combat the epidemic. [0004] Although the World Health Organization recently declared an end to the global COVID-19 health emergency, the pandemic is nonetheless ongoing and second-generation SARS-CoV-2 vaccines are urgently needed. While several vaccines have been developed and approved since the pandemic began—including many using novel mRNA vaccine technologies—that have admirably contributed to the waning of disease mortality globally, emerging trends highlight the insufficiencies of current vaccine approaches. Vaccine efficacy has waned rapidly and frequent boosters have been recommended for individuals that received the Pfizer and/or Moderna mRNA vaccines. Increasing evidence of post- infection and post-vaccination reinfection, particularly with the Omicron variant of SARS- CoV-2 and its numerous sub-variants, and ongoing viral transmission post-vaccination, has negatively impacted vaccine credibility. Moreover, new variants will continually place pressure on the efficacy of first-generation SARS-CoV-2 vaccines. [0005] While more research is necessary as to why current vaccines are waning, emerging research is pointing to insufficient nasal and mucosal immunity. Increasingly, data are demonstrating that intramuscular vaccination alone may not generate potent mucosal immunity that is predicted to be critical for early control and clearance of SARS-CoV-2. Vaccinees still show evidence of viral RNA in the upper respiratory tract after viral exposure, suggesting ongoing viral replication and transmission even by vaccinated populations. However, breakthrough infection with SARS-CoV-2 allows vaccinees to elicit potent mucosal and nasal immunity. [0006] In addition, although it is impossible to predict the origin of the next global pandemic, over the past century influenza has been a recurrent threat that threatens global health and security. Despite the availability of seasonal influenza vaccines, vaccine effectiveness is only 10-60%, and vaccines provide no protection against new pandemic strains. Historically, as influenza variants spread, they have the potential to cause three to four pandemics per century, most recently the 2009 H1N1 influenza pandemic that stressed medical and military operations. Avian influenza A H5N1 viruses have recently spread throughout domestic and wild animal populations and have been termed “pre-pandemic.” There is a strong concern that H5N1 will mutate and become readily transmissible between humans and drive the next pandemic. Recent cases have shown H5N1 influenza is highly pathogenic and causes human disease with high mortality in Africa, Asia, and the Middle East. [0007] mRNA vaccines have shown exceptional utility during the COVID-19 pandemic as a tool for pandemic response, enabling rapid adaptability to new pathogen targets. For influenza response, the use of RNA vaccine technology would also avoid the need for tedious and resource-limited vaccine production in embryonated chicken eggs. However, current mRNA vaccine technologies have several key limitations that need to be overcome to be optimized for future influenza pandemics. First, all mRNA vaccines to date are administered through intramuscular (i.m.) injection, which induces serum antibody titers and cellular immunity that protect against viremia and serious disease but not against asymptomatic infection and viral shedding. Second, the currently authorized or approved mRNA vaccines are all monovalent, which limits their flexibility in rapidly evolving pandemic situations. Third, current mRNA vaccines’ use of lipid nanoparticles (LNPs) requires complex manufacturing processes, difficult-to-source ionizable lipids, and encapsulation of the RNA during LNP manufacture—complicating technology transfer and supply chain management. Fourth, the use of LNPs as the delivery vehicle restricts the ability to meaningfully stockpile drug substances because LNPs must be manufactured with the target RNA, thus delaying the speed at which new vaccine constructs to emerging variants or pathogens can be introduced, manufactured, and distributed. Finally, current mRNA vaccines require cold chain storage (-20°C or -80°C), significantly impacting vaccine distribution particularly in low-resource areas or in situations of active military deployment. [0008] Improving the induction of nasal and mucosal immunity following vaccination is a critical component for second-generation SARS-CoV-2 vaccines to finally achieve durable and flexible protection. However, data also show that generating this immunity is generally not possible via standard intramuscular vaccination and likely requires administration of an intranasal vaccine. [0009] Intranasal vaccines to date have primarily been developed as live-attenuated influenza vaccines. Recently, four new intranasal or inhalable vaccines have been approved for SARS-CoV-2 in China, India, Iran, and Russia, including two adenovirus-vectored vaccines. Further development of nasal or airway administration of vaccines for influenza and other respiratory viruses has remained an attractive target given the potential for development of critical airway resident memory T cells (T _RM ) at this first line of defense. Several other intranasal vaccines are in development for SARS-CoV-2, but most rely on standard protein or adenovirus vector vaccine technologies. An RNA-based vaccine, in contrast, can be developed more rapidly, and allows cross-over intramuscular to intranasal vaccination without developing a new vaccine, allowing for significantly greater utility in pandemic situations. Validating an RNA-based, intranasally administered vaccine technology would be a significant development not only for the SARS-CoV-2 pandemic but for other respiratory pathogens such as influenza, combining the rapidity and reliability of RNA vaccine adaptation to new targets with the key mucosal immune responses induced by mucosal-specific dosing. [0010] While the expected benefits of RNA vaccines have been proven over the course of the COVID-19 pandemic, no intranasally-administered RNA vaccine for COVID-19 has yet been developed. This is on account of significant challenges of intranasally administering RNA vaccines, including ensuring that the vaccine is delivered via an appropriate particle size and charge to allow efficient crossing of the mucosa and uptake by airway epithelial cells, ensuring that it is protected from enzymatic degradation while traversing the mucosa, and ensuring that it is sufficiently stable to allow delivery via a liquid or powdered nasal spray. [0011] Thus, there remains a need for an effective RNA vaccine delivery systems for intranasal delivery of RNA vaccines. This disclosure is made with respect to these and other considerations. SUMMARY [0012] A novel thermostable nanostructured lipid carrier (NLC)-based RNA vaccine delivery system for intranasally delivery of self-amplifying RNA vaccines is disclosed herein. [0013] NLC-based vaccine delivery systems for intramuscular delivery of RNA vaccines have been disclosed previously. See, e.g., PCT Patent Application Publication No. WO 2022/051022; U.S. Patent Application Publication No. 2020/0230056. These delivery systems are generally composed of NLC particles comprising (a) an oil core comprising a liquid phase lipid and a solid phase lipid, (b) a cationic lipid, (c) a hydrophobic surfactant, and (d) an additional surfactant. These delivery systems were previously known to be effective for intramuscular and subcutaneous vaccine administration but were not known to be suitable for intranasal vaccine administration. [0014] Intranasal vaccine delivery promises many benefits over typical intramuscular or subcutaneous vaccine administration, particularly in the induction of the highly specialized mucosal immune system that confers early and effective protection against infectious respiratory diseases. In addition, intranasal vaccines offer greater ease of administration, eliminating the need for medically-trained personnel as well as potentially improving patient acceptance and compliance. Thus, it is extremely desirable to be able to administer vaccines intranasally. [0015] The challenges of intranasally administering RNA vaccines are significant. As the mucosal system is designed to trap and remove large particles from the nasal passages, the particle size and charge of an intranasally-administered vaccine affect crossing of the mucosa and uptake by airway epithelial cells. Further, as RNA is a molecule particularly prone to degradation, it is desirable for an intranasal RNA vaccine delivery system to effectively protect vaccine RNA from enzymatic degradation while traversing the mucosa. Finally, the stability of an intranasal vaccine delivery system affects the ability to deliver the vaccine via a liquid or powdered nasal spray. Due to these and other challenges, it is inherently difficult to adapt a vaccine platform to a new route of administration. [0016] Surprisingly, it was determined that the previously disclosed RNA/NLC vaccine delivery systems could be modified to be suitable for use in intranasal vaccine administration. It was determined that adjusting the ratio of NLC-contained-amine-group to RNA-phosphate (N/P) is one parameter governing the size, charge, and ability of the NLC formulation to effectively protect RNA from enzymatic degradation. The N/P ratio is the ratio of delivery NLC formulation to RNA in the complexed vaccine. It was found that enzymatic degradation occurs when the zeta potential of the vaccine is a positive value, which occurs at N/P ratios of 3 or greater. Based on particle size analysis using DLS and NTA, it is likely that at least a small excess of NLC (i.e., NLC is not saturated with bound RNA) is necessary to confer protection to the RNA—a negative zeta potential reflects the fact that NLCs are approaching complete saturation with RNA. Therefore, the zeta potential, and thus by necessity the N/P ratio, is a parameter to take into consideration when optimizing the vaccine to maximize efficacy in a specific application while minimizing cytotoxicity risk. [0017] In addition, the identity of the liquid phase lipid in the oil core of the NLC is also a factor in minimizing reactogenicity and optimizing immunogenicity, as discussed below. [0018] Using the results of experiments on NLC formulations and the composition of the RNA/NLC complexes, it was possible to adjust the composition of the NLC and the N/P ratio for the RNA vaccine and NLC delivery system to develop an RNA vaccine delivery system for effective intranasal vaccination. This delivery system yielded vaccines that showed prophylactic effects and were also surprisingly effective at preventing transmission from infected subjects to uninfected subjects. This benefit of intranasal vaccine delivery was successfully achieved using the intranasal RNA vaccine delivery systems of this disclosure. Thus, vaccines may be delivered intranasally to inoculate subjects against respiratory viruses such as, but not limited to, SARS-CoV-2 and influenza while benefitting from the advantages in vaccine administration and efficacy exhibited by intranasal vaccine delivery. [0019] It is to be understood that one, some, or all of the properties of the various implementations described herein may be combined to form other implementations of the present invention. These and other aspects of the present invention will become evident upon reference to the following detailed description and attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS [0020] FIGS. 1A-B illustrate the design of a self-amplifying RNA/NLC vaccine delivery system for SARS-CoV-2. FIG. 1A shows a diagram of the self-amplifying RNA/NLC construct. FIG.1B shows a diagram of the self-amplifying RNA/NLC complex. [0021] FIG.2 shows SARS-CoV-2 protein binding antibody titers 14 days post-boost. [0022] FIG. 3 shows SARS-CoV-2 pseudovirus-neutralizing antibody titers 14 days post-boost. [0023] FIGS. 4A-I show results for systemic immunogenicity of an intranasally administered SARS-CoV-2 self-amplifying RNA/NLC vaccine assessed three weeks post- prime and three weeks post-boost. FIG. 4A shows serum SARS-CoV-2 spike-specific IgG responses measured post-prime and post-boost, where LOD indicates the limit of detection. FIG.4B shows serum SARS-CoV-2 spike Wuhan pseudovirus neutralizing titers measured post-prime and post-boost. FIG.4C shows serum spike-specific IgG1 and IgG2a responses measured post-boost. FIG.4D shows induction of bone marrow-resident spike-specific IgG- secreting ASCs as measured by ELISpot 3 weeks post-boost. FIGS. 4E-G show SARS- CoV-2 spike-responsive secretion of IFNγ, IL-5, or IL-17a by splenocytes post-boost as measured by ELISpot in mice dosed with 10 µg of saRNA. FIGS. H-I show splenic spike- responsive polyfunctional T cells (IFNγ+, IL-2+, and TNFα+ CD4+ or CD8+ T cells) post- boost as measured by intracellular cytokine staining and flow cytometry. [0024] FIGS. 5A-C show mucosal T cell populations following intranasal vaccination with SARS-CoV-2 self-amplifying RNA/NLC vaccine assessed three weeks post-prime and three weeks post-boost. FIGS. 5A-B show CD69+ lung-resident memory for CD8+ and CD103+ CD8+ T cells, respectively. FIG.5C shows CD40L+ (CD154+)-expressing CD4+ T cells. [0025] FIGS.6A-E show mucosal SARS-CoV-2 spike polyfunctionally reactive T cell populations following immunization with the intranasal or intramuscular SARS-CoV-2 self- amplifying RNA/NLC vaccine. FIGS. 6A-B show polyfunctional CD69+ lung-resident memory for CD8+ and CD103+ CD8+ T cells, respectively. FIG.6C shows polyfunctional CD40L-expressing (CD154+) CD4+ T cells. FIGS. 6D-E show polyfunctional CCR4- expressing (CD194+) lung-homing CD4+ and CD8+ T cells, respectively. [0026] FIGS. 7A-F show systemic and mucosal immune responses following boost immunization with the intranasal or intramuscular SARS-CoV-2 self-amplifying RNA/NLC vaccines seven weeks post-prime. FIG.7A shows SARS-CoV-2 spike-specific IgG titers as measured by ELISA three weeks post-boost. FIG. 7B shows serum SARS-CoV-2 pseudovirus neutralizing titers three weeks post-boost. FIGS. 7C-D show spleen polyfunctional CD4+ and CD8+ T cell responses, respectively. FIG. 7E shows polyfunctional CD40L-expressing (CD154+) CD4+ T cells. FIG.7F shows polyfunctional lung-resident memory CD69+ CD8+ T cell responses. [0027] FIGS. 8A-H show protection from SARS-CoV-2 infection-induced morbidity, viral load, and lung damage in vaccinated hamsters. FIG.8A illustrates the study design for experiments in vaccinated and sham-vaccinated co-housed hamsters. FIGS. 8B-C show vaccine antibody responses in sham-vaccinated, i.n.-prime/i.n.-boost vaccinated, i.m.- prime/i.m.-boost vaccinated, and i.m.-prime/i.n.-boost vaccinated animals. FIG. 8B shows mean fluorescence intensity (MFI) of serum anti-spike IgG on day 21, and FIG.8C shows mean fluorescence intensity (MFI) of serum anti-spike IgG on day 33. FIGS. 8D-H show experimental infection in sham-vaccinated and vaccinated hamsters. FIG. 8D shows body weights of vaccinated-infected, sham-vaccinated-infected, and sham-vaccinated-uninfected hamsters. FIGS. 8E-F show total N RNA and subgenomic N RNA copy number, respectively, in nasal swipes from vaccinated-infected and sham-vaccinated-infected hamsters. FIG.8G shows anti-N IgG in serum of vaccinated-infected and sham-vaccinated- infected animals. FIG. 8H shows lung pathology scores calculated for sham-vaccinated- infected and vaccinated-infected animals. [0028] FIGS.9A-F show prevention of transmission of SARS-CoV-2 to naive hamster cagemates that results from vaccination. FIG. 9A illustrates the study design for transmission experiments in naive cagemates. FIG. 9B shows body weights of the naive cagemates that were re-paired for 24 hours with vaccinated-infected or sham-vaccinated- infected hamsters. FIGS.9C-D show total N RNA and subgenomic N RNA copy numbers, respectively, in nasal swipes from naive cagemates that were re-paired for 24 hours with vaccinated-infected and sham-vaccinated-infected hamsters. FIG. 9E shows anti-N IgG in serum of naive cagemates that were re-paired for 24 hours with vaccinated-infected and sham-vaccinated-infected groups. FIG.9F shows lung pathology scores calculated for naive cagemates that were re-paired for 24 hours with sham-vaccinated-infected and vaccinated- infected animals. [0029] FIG. 10 shows a non-human primate study diagram for a SARS-CoV-2 self- amplifying RNA/NLC vaccine, indicating timepoints where samples were taken. [0030] FIGS.11A-D show serum antibody titers induced by self-amplifying RNA/NLC vaccination against SARS-CoV-2. FIGS.11A-B show ELISA evaluations of IgG levels in vaccinated non-human primates. FIG. 11A shows IgG titers to the WA 1/2020 strain (vaccine strain), and FIG.11B shows IgG titers to the B.1.617.2 strain (the challenge strain). FIGS.11C-D show the increase from challenge to necropsy in titers for all three groups to WA 1/2020 and B.1.617.2, respectively. [0031] FIGS. 12A-D show antibody levels in the BAL and nasal mucosa for SARS- CoV-2 vaccination. FIG.12A shows IgG levels in the BAL. FIG.12B shows IgA levels in the BAL. FIG. 12C shows IgG levels in the nasal mucosa. FIG. 12D shows IgA levels in the nasal mucosa. [0032] FIGS. 13A-C show serum pseudovirus neutralization titers for SARS-CoV-2 vaccination. [0033] FIGS.14A-D shows T cell immunogenicity in non-human primates for SARS- CoV-2 vaccination. [0034] FIGS.15 A-D show IgG antibody secreting cells post-challenge in non-human primates for SARS-CoV-2 vaccination. [0035] FIG.16 shows viral load in non-human primates post-challenge for SARS-CoV- 2 vaccination. [0036] FIGS. 17A-C show dynamic light scattering (DLS) size and polydispersity analyses of SARS-CoV-2 self-amplifying RNA/NLC complexes of varying N/P ratios. FIG. 17A shows intensity size distribution of complexes formed at increasing N/P ratios. FIG. 17B shows a plot of the Z-average diameters of complexes formed at increasing N/P ratios. FIG.17C shows the polydispersity index (PDI) of complexes at increasing N/P ratios. [0037] FIGS.18A-B show a nanoparticle tracking analysis (NTA) of SARS-CoV-2 self- amplifying RNA/NLC complexes formed at varying N/P ratios. FIG.18A shows NTA size distribution of complexes formed at different N/P ratios. FIG.18B shows mean particle size of complexes as a function of N/P ratio, determined by NTA. [0038] FIGS. 19A-C show a correlation between zeta potential and RNase protection for SARS-CoV-2 self-amplifying RNA/NLC vaccines. FIG. 19A shows zeta potential of self-amplifying RNA/NLC complexes as a function of N/P ratio. FIG. 19B shows percent protection of RNA from degradation by RNase A as a function of N/P ratio, as determined by agarose gel electrophoresis. FIG. 19C shows agarose gel electrophoresis of self- amplifying RNA extracted directly from complexes or “challenged” with RNase A prior to extraction from complexes. [0039] FIGS.20A-D show cryo-TEM imaging of NLCs alone or in complex with self- amplifying RNA at various N/P ratios. FIG.20A shows the NLC alone. FIG.20B shows the saRNA/NLC complex at N/P = 15. FIG.20C shows the saRNA/NLC complex at N/P = 5. FIG.20D shows the saRNA/NLC complex at N/P = 0.6. [0040] FIG. 21 shows illustrates the design of a self-amplifying RNA/NLC vaccine delivery system for influenza. FIG.1A shows a diagram of the self-amplifying RNA/NLC construct. FIG.1B shows a diagram of the self-amplifying RNA/NLC complex. [0041] FIG.22 shows a timeline of sample collection and assays performed for an initial N/P dosing comparison for influenza vaccination. [0042] FIGS. 23A-B show H5 vaccine-induced serum antibody responses. FIG. 23A shows serum H5-specific IgG responses measured post-prime and post-boost. FIG. 23B shows serum H5 pseudovirus neutralizing antibody titers measured post-prime and post- boost. [0043] FIG. 24 shows ELISpot measurements of the levels of H5 IgG-secreting antibody secreting cells (ASCs) in bone marrow. [0044] FIGS.25A-F show H5-reactive vaccine-induced T cell populations assessed via intracellular cytokine staining and flow cytometry. [0045] FIG.26 shows a timeline of sample collection and assays performed for in-depth N/P ratio comparison for influenza vaccination. [0046] FIGS. 27A-B show post-immunization body weights of mice for influenza vaccination at various N/P ratios. FIG. 27A shows post-prime body weights. FIG. 27B shows post-boost body weights. [0047] FIG. 28 shows vaccine-induced H5-binding serum IgG levels at various N/P ratios. [0048] FIG. 29 shows serum pseudovirus neutralizing antibody titers for intranasal or intramuscular influenza vaccination at various N/P ratios.. [0049] FIGS.30A-B show the presence of IgG and IgA antibody-secreting cells (ASCs) in bone marrow at various N/P ratios as assessed by ELISpot. FIG.30A shows the presence of IgG antibody-secreting cells (ASCs) in bone marrow. FIG. 30B shows the presence of IgA antibody-secreting cells (ASCs) in bone marrow. [0050] FIGS. 31A-F show vaccine-induced antigen-responsive T cell populations for influenza vaccination. [0051] FIG. 32 shows a timeline of sample collection and assays performed for liquid oil comparison studies for influenza vaccination. [0052] FIGS.33A-B show body weight changes post-immunization with alternative-oil self-amplifying RNA/NLC vaccine formulations for influenza vaccination. FIG.33A shows post-prime body weight change. FIG.33B shows post-boost body weight change. [0053] FIG. 34 shows H5-binding serum IgG titers for influenza vaccination assessed by ELISA. [0054] FIG. 35 shows serum pseudovirus neutralizing antibody titers induced by alternate oil NLC formulated self-amplifying RNAs for influenza vaccination. [0055] FIGS. 36A-F show antigen-specific T cell responses to alternate oil NLC formulated H5 self-amplifying RNA vaccines, as assessed by intracellular cytokine staining with flow cytometry by intracellular cytokine staining with flow cytometry. DETAILED DESCRIPTION [0056] A novel thermostable nanostructured lipid carrier (NLC)-based RNA vaccine delivery system for intranasally delivery of self-amplifying RNA (saRNA) vaccines is disclosed herein. In some implementations, the vaccine delivery system is composed of a self-amplifying RNA vaccine complexed with a nanostructured lipid carrier. [0057] The disclosed delivery systems extend previously disclosed development of RNA/NLC vaccine delivery systems that have been used for intramuscular delivery. The disclosed delivery systems allow delivery of an intranasal RNA vaccine that induces respiratory mucosal immune responses resulting in protective efficacy equal to intramuscular administration and benefits unique to intranasal administration. [0058] These delivery systems are generally composed of NLC particles comprising (a) an oil core comprising a liquid phase lipid and a solid phase lipid, (b) a cationic lipid, (c) a hydrophobic surfactant, and (d) an additional surfactant. These delivery systems were previously known to be effective for intramuscular and subcutaneous vaccine administration but were not known to be suitable for intranasal vaccine administration. [0059] As discussed above, it was determined that the previously disclosed RNA/NLC vaccine delivery systems could be modified to be suitable for use in intranasal vaccine administration. It was determined that adjusting the ratio of NLC-contained-amine-group to RNA-phosphate (N/P) is an important parameter governing the size, charge, and ability of the NLC formulation to effectively protect RNA from enzymatic degradation. It was determined that the N/P ratio is one parameter to take into consideration when optimizing the vaccine to maximize efficacy in a specific application while minimizing cytotoxicity risk. In addition, it was determined that the identity of the liquid phase lipid in the oil core of the NLC is also a factor in minimizing reactogenicity and optimizing immunogenicity, as discussed below. [0060] Using the results of experiments on NLC formulations and the composition of the RNA/NLC complexes, it was possible to adjust the composition of the NLC and the N/P ratio for the RNA vaccine and NLC delivery system to develop an RNA vaccine delivery system for effective intranasal vaccination. This delivery system yielded vaccines that showed prophylactic effects and were also surprisingly effective at preventing transmission from infected subjects to uninfected subjects. Thus, vaccines may be delivered intranasally to inoculate subjects against respiratory viruses such as, but not limited to, SARS-CoV-2 and influenza while benefitting from the known advantages in vaccine administration and efficacy exhibited by intranasal vaccine delivery. [0061] The disclosed delivery system has been shown to be effective for intranasal delivery of saRNA vaccines for COVID-19 and influenza. It is expected that the disclosed delivery system will similarly be effective for intranasal delivery of other saRNA vaccines. The disclosed NLC-based saRNA vaccine delivery system has enhanced thermostability relative to current clinical RNA vaccine formulations and allows characteristics of the vaccine nanoparticles—such as particle size, charge, and immunostimulatory characteristics—to be easily tuned to achieve effective delivery of an RNA vaccine through an intranasal route. [0062] In some implementations, the self-amplifying RNA vaccine expresses the SARS-CoV-2 S spike protein. In some other implementations, the self-amplifying RNA vaccine expresses antigens associated with the H5N1 strain of influenza or another strain of influenza. [0063] The disclosed delivery system allows intranasal administration of an alphavirus- based self-amplifying RNA vaccine using an NLC delivery system. An optimized NLC delivery system allows for safe and effective intranasal delivery of an RNA vaccine that elicits a robust mucosal immune response. Tuning the size and charge of the vaccine nanoparticle makes it possible to achieve uptake at mucosal surfaces. Further, by tuning the NLC composition for intranasal delivery, the level of immune stimulation at the site of administration may be optimized to allow for maximum vaccine immunogenicity. The vaccine nanoparticle may be delivered as a liquid or lyophilized and delivered in a powdered form. The vaccine may be delivered to subjects using a metered spray. A subject may be a human or non-human mammal. [0064] The nanoparticle size for the RNA/NLC vaccine complex may be optimized to maximize effective delivery of RNA through the nasal mucosa. Without being bound by any specific theory, it is believed that smaller particles will be better able to cross the mucosa and thus generate a more robust mucosal immune response. Baseline immunogenicity is established by administering the RNA/NLC vaccine at the optimized size to test subjects intranasally at various doses. Additionally, immunogenicity of an intranasally-delivered RNA vaccine may be optimized by varying particle composition through different liquid oils and effective surface charge. For example, the immunostimulatory properties of the NLC vaccine carrier may be optimized by incorporation of different liquid oils and assessing their effect on vaccine immunogenicity. Without being bound by any specific theory, it is believed that for a SARS-CoV-2 RNA vaccine, using Miglyol ^® 810 as a liquid oil may provide superior immunostimulatory properties relative to squalene, because intranasal delivery of squalene may cause an overly strong immune response. Additionally, the effective surface charge of the vaccine particles may be varied to optimize RNA delivery. [0065] With respect to vaccination against SARS-CoV-2, the intranasally administered RNA vaccine induced systemic immunity in both the humoral and cellular compartments that rivaled intramuscular administration. Additionally, intranasal administration elicited not only robust populations of spike-responsive polyfunctional T cells in the spleen but also in the lung, demonstrating induction of key respiratory mucosal T cell populations by an RNA vaccine. Further, it was demonstrated that the disclosed system may be used as an intranasal booster of prior intramuscular vaccine-induced immunity, a common real-life scenario given the largely COVID-experienced population to whom the delivery system is expected to be used to deliver vaccines. This intranasal boost of prior intramuscular vaccine- induced immunity significantly boosted systemic responses in addition to inducing mucosal T cell populations not stimulated by intramuscular vaccine boosting. In addition, it was demonstrated that both intranasal-intranasal, intramuscular-intramuscular, and heterologous intramuscular-intranasal vaccination regimens prevented virus-associated morbidity in a hamster model and, most importantly, prevented viral transmission and development of disease in naive cagemates. [0066] A SARS-CoV-2 vaccine delivered via the disclosed delivery system elicits both robust systemic humoral immunity and spike-responsive systemic polyfunctional T cells as an intramuscular or intranasal vaccine. Delivered intranasally, it also elicits tissue-resident T cell populations within the lungs, suggesting broad T cell presence across the respiratory tract—an effect that intramuscularly delivered mRNA vaccines have been unable to induce to date. While CD4 and CD8 T cell populations are now being recognized as a major determinant of disease progression and critical to long-lasting immunity, the main vaccines on the market elicit minimal peripheral T cell responses and negligible mucosal/respiratory T cell responses. The observed thermostability and ability to administer intranasally may make vaccines delivered using the disclosed delivery system substantially more desirable in the fight against respiratory viruses. Self-amplifying RNA/NLC vaccines delivered using the disclosed intranasal delivery system may be effective boosters to pre-existing SARS- CoV-2 immunity. RNA/NLC vaccines delivered using the disclosed intranasal delivery system outperformed an intramuscular booster, showing that the additive effect of delivering a vaccine intranasally to pre-existing systemic immune responses generated the greatest benefit of all vaccination regimens tested. [0067] The intranasal self-amplifying RNA vaccine delivered using the disclosed delivery system may be most effective as a booster dose following an intramuscular prime dose. This aligns with the results of other studies where priming an immune response with a strong intramuscular vaccine followed by intranasal vaccination is able to optimally stimulate the TRM compartment while still boosting the extant intramuscular induced immunity. It has been demonstrated that use of an identical or similar RNA vaccine product for both intramuscular and intranasal vaccination may be appropriate. This potentially eliminates many manufacturing hurdles and relies upon an already clinically tested vaccine platform technology for multiple routes of administration. It was observed that intranasal boosting outperformed intramuscular boosting, particularly in mucosal immune measures, hinting at the potential for intranasal dosing to access novel cell compartments and mechanisms. [0068] In some implementations, intranasal vaccines may be delivered using the disclosed delivery system via use of nasal spray devices. This will promote induction of immunity while avoiding toxicity and reducing heterogeneity in nasal delivery. [0069] The thermostability and concomitant stockpiling ability of the disclosed RNA/NLC vaccine platform additionally provide key attributes for rapid and widespread pandemic response. [0070] In addition to inducing potent systemic serum neutralizing antibodies, bone marrow-resident IgG-secreting cells, and robust lymphoid tissue T cell immune responses, the disclosed intranasal vaccines additionally induce robust respiratory mucosal immune responses, including SARS-CoV-2-reactive lung-resident memory and lung-homing T cell populations. As a booster following previous intramuscular vaccination, the intranasal vaccine also elicits the development of mucosal virus-specific T cells. The intranasally administered vaccines protect hamsters from infection-associated morbidity upon viral challenge, significantly reducing viral loads and preventing challenged hamsters from transmitting virus to naive cagemates. The disclosed self-amplifying RNA vaccine’s potent systemic immunogenicity, and additional mucosal immunogenicity when delivered intranasally, may be key for combating SARS-CoV-2 and other respiratory pathogens. [0071] Intranasal SARS-CoV-2 self-amplifying RNA vaccination induces systemic and mucosal immunity in mice and prevents morbidity and blocks viral transmission in hamsters. Similar results are observed in non-human primates. [0072] The extensive biophysical characterization of differently formulated self- amplifying RNA/NLC vaccine candidates described herein demonstrates that a parameter governing the size, charge, and ability of the NLC formulation to effectively protect RNA from enzymatic degradation is the N/P ratio, or ratio of delivery NLC formulation to RNA in the complexed vaccine. It was found that enzymatic degradation occurs when the zeta potential of the vaccine is a positive value, which occurs at N/P ratios of about 3 or greater. Based on particle size analysis using DLS and NTA, it is likely that at least a small excess of NLC (i.e., NLC is not saturated with bound RNA) is necessary to confer protection to the RNA—a negative zeta potential reflects the fact that NLCs are approaching complete saturation with RNA. Therefore, the zeta potential, and thus by necessity the N/P ratio, is a parameter to take into consideration when optimizing the vaccine to maximize efficacy in a specific application while minimizing cytotoxicity risk. [0073] With respect to H5N1, the optimized self-amplifying RNA/NLC vaccine for intranasal delivery was shown to effectively induce serum neutralizing antibodies at high titer, as well as bone marrow antibody-secreting cells and both systemic and lung-resident antigen-specific polyfunctional T cell responses. It was observed that certain T-cell responses, such as CD69+ lung cells, were enhanced specifically by i.n. vaccine delivery relative to standard i.m. vaccination. It was shown that minimal reactogenicity and optimal immunogenicity are induced by an intranasal vaccine formulated at about N/P=8-10, and that use of Miglyol ^® 810 instead of squalene as the liquid oil component of the NLC delivery formulation both further minimizes intranasal reactogenicity and maximizes vaccine immunogenicity. [0074] The disclosed self-amplifying RNA/NLC vaccine platform represents a promising technology to combat both the existing SARS-CoV-2 pandemic and other respiratory illnesses, such as seasonal and pandemic influenza. For respiratory pathogens, intranasal vaccination or intranasal vaccine boosting of intramuscular vaccination may help raise RNA vaccine technologies to a new level, through stimulating respiratory T cell and B cell populations and generating effector subsets within the lungs and nasal mucosa primed to respond rapidly at the first point of infection. Beyond its efficacy at stimulating T _RM subsets and bolstering systemic immunity, intranasal administration may also be desirable as a needle-less delivery method with potentially higher uptake, particularly in pediatric and vaccine-hesitant populations. Altogether, the disclosed unique intranasally delivered RNA vaccine that elicits strong systemic and mucosal immunity is a promising new technology ripe for further development and application. N/P Ratio for RNA/NLC Complexes [0075] In some implementations, the RNA/NLC complex has an NLC-contained- amine-group to RNA-phosphate (N/P) ratio of at least 15. [0076] In some implementations, the RNA/NLC complex has an NLC-contained- amine-group to RNA-phosphate (N/P) ratio of at least 3. This will guarantee protection of the RNA, maintain the positive zeta potential necessary for cellular uptake, and limit the formation of very large aggregate species. [0077] In some implementations, the RNA/NLC complex has an NLC-contained- amine-group to RNA-phosphate (N/P) ratio of between about 3 and about 15. [0078] In some implementations, the RNA/NLC complex has an NLC-contained- amine-group to RNA-phosphate (N/P) ratio of between about 4 and about 14. In some implementations, the RNA/NLC complex has an NLC-contained-amine-group to RNA- phosphate (N/P) ratio of between about 5 and about 13. In some implementations, the RNA/NLC complex has an NLC-contained-amine-group to RNA-phosphate (N/P) ratio of between about 5 and about 12. In some implementations, the RNA/NLC complex has an NLC-contained-amine-group to RNA-phosphate (N/P) ratio of between about 5 and about 11. In some implementations, the RNA/NLC complex has an NLC-contained-amine-group to RNA-phosphate (N/P) ratio of between about 6 and about 11. In some implementations, the RNA/NLC complex has an NLC-contained-amine-group to RNA-phosphate (N/P) ratio of between about 6 and about 10. In some implementations, the RNA/NLC complex has an NLC-contained-amine-group to RNA-phosphate (N/P) ratio of between about 7 and about 10. [0079] In some implementations, the RNA/NLC complex has an NLC-contained- amine-group to RNA-phosphate (N/P) ratio of between about 5 and about 10. In some implementations, the RNA/NLC complex has an NLC-contained-amine-group to RNA- phosphate (N/P) ratio of between about 5 and about 9. In some implementations, the RNA/NLC complex has an NLC-contained-amine-group to RNA-phosphate (N/P) ratio of between about 6 and about 9. In some implementations, the RNA/NLC complex has an NLC-contained-amine-group to RNA-phosphate (N/P) ratio of between about 7 and about 9. In some implementations, the RNA/NLC complex has an NLC-contained-amine-group to RNA-phosphate (N/P) ratio of between about 8 and about 9. In some implementations, the RNA/NLC complex has an NLC-contained-amine-group to RNA-phosphate (N/P) ratio of between about 3 and about 6. In some implementations, the RNA/NLC complex has an NLC-contained-amine-group to RNA-phosphate (N/P) ratio of between about 4 and about 7. [0080] In some implementations, the RNA/NLC complex has an NLC-contained- amine-group to RNA-phosphate (N/P) ratio of about 3, about 4, about 5, about 6, about 7, about 8, about 9, about 10, about 11, about 12, about 13, about 14, or about 15. [0081] In some preferred implementations, the RNA/NLC complex has an NLC- contained-amine-group to RNA-phosphate (N/P) ratio of between about 8 and about 10. It was observed that vaccine reactogenicity is reduced for vaccines formulated at N/P of 10 and below. Effective immunogenicity was observed at N/P ratios of 8 and above. Composition of Nanostructured Lipid Carriers [0082] The composition of the NLCs may be according to the NLCs described in PCT Patent Application Publication No. WO 2022/051022, U.S. Patent Application Publication No.2020/0230056, or combinations or modifications thereof. These references are hereby incorporated herein in their entireties by reference. [0083] The NLC compositions are composed of NLC particles comprising (a) an oil core comprising a liquid phase lipid and a solid phase lipid, (b) a cationic lipid, (c) a hydrophobic surfactant, preferably a sorbitan ester (e.g., sorbitan monoester, diester, or triester), and (d) a surfactant (preferably, a hydrophilic surfactant). The NLCs typically comprise an unstructured or amorphous solid lipid matrix composed of a mixture of blended solid and liquid lipids dispersed in an aqueous phase. One or more of the surfactants may be present in the oil phase, the aqueous phase, or at the interface between the oil and aqueous phase. In certain aspects the sorbitan ester and the cationic lipid are present at the interface between the oil and aqueous phase. Solid-Phase and Liquid-Phase Lipids [0084] The NLCs are composed of a blend of solid and liquid lipids. The liquid and solid lipids may be any lipid capable of forming an unstructured or amorphous solid lipid matrix and forming a stable composition. The weight ratio of solid to liquid may vary widely, for example from 0.1:99.9 to 99.9:0.1. In some illustrative implementations, the solid lipids are mixed with liquid lipids in a solid/liquid lipid weight ratio of from about 70:30 to about 99.9:0.1 or from about 1:10 to about 1:30. In some aspects, the solid lipids are mixed with liquid lipids in a solid/liquid lipid weight of about 1:16. [0085] The total oil core component (solid lipid + liquid oil) of the NLC-based composition or formulation is typically present in an amount from about 0.2% to about 50% (w/v). For example, the NLC may comprise from about 0.2% to about 50% (w/v) oil core component, 0.2% to about 40% (w/v) oil core component, from about 0.2% to about 30% (w/v) oil core component, from about 0.2% to about 20% (w/v) oil core component, from about 0.2% to about 15% (w/v) oil core component, from about 0.2% to about 10% (w/v) oil core component, from about 0.2% to about 9% (w/v) oil core component, from about 0.2% to about 8% (w/v) oil core component, from about 0.2% to about 7% (w/v) oil core component, from about 0.2% to about 6% (w/v) oil core component, from about 0.2% to about 5% (w/v) oil core component, from about 0.2% to about 4.3% (w/v) oil core component, from about 0.3% to about 20% (w/v) oil core component, from about 0.4% to about 20% (w/v) oil core component, from about 0.5% to about 20% (w/v) oil core component, from about 1% to about 20% (w/v) oil core component, from about 2% to about 20% (w/v) oil core component, from about 3% to about 20% (w/v) oil core component, from about 4% to about 20% (w/v) oil core component, from about 5% to about 20% (w/v) oil core component, about 0.5% (w/v) oil core component, about 1% (w/v) oil core component, about 1.5% (w/v) oil core component, about 2% (w/v) oil core component, about 2.5% (w/v) oil core component, about 3% (w/v) oil core component, about 3.5% (w/v) oil core component, about 4% (w/v) oil core component, about 4.3% (w/v) oil core component, about 5% (w/v) oil core component, or about 10% (w/v) oil core component or any other amount or range described herein for the oil core component. Higher or lower w/v percentages are contemplated herein, particularly when considering diluted or concentrated formulations. [0086] The oil core of the NLC comprises a liquid phase lipid. Preferably, although not necessarily, the liquid phase lipid is a metabolizable, non-toxic oil; more preferably one of about 6 to about 30 carbon atoms including, but not limited to, alkanes, alkenes, alkynes, and their corresponding acids and alcohols, ethers and esters thereof, and mixtures thereof. The oil may be, for example, any vegetable oil, fish oil, animal oil, or synthetically prepared oil that can be administered to a subject. In some aspects, the liquid phase lipid will be non- metabolizable. [0087] The oil may be, for example, any long chain alkane, alkene, or alkyne, or an acid or alcohol derivative thereof either as the free acid, its salt, or an ester such as a mono-, or di- or triester, such as the triglycerides and esters of 1 ,2-propanediol or similar poly- hydroxy alcohols. Alcohols may be acylated employing a mono- or poly-functional acid, for example acetic acid, propanoic acid, citric acid, or the like. Ethers derived from long chain alcohols which are oils and meet the other criteria set forth herein may also be used. [0088] The individual alkane, alkene, or alkyne moiety and its acid or alcohol derivatives will generally have from about 6 to about 40 or from 6 to about 30 carbon atoms. The moiety may have a straight or branched chain structure. It may be fully saturated or have one or more double or triple bonds. Where mono or poly ester- or ether-based oils are employed, the limitation of about 6 to about 40 carbons applies to the individual fatty acid or fatty alcohol moieties, not the total carbon count. [0089] Any suitable oils from an animal, fish or vegetable source may be used. Sources for vegetable oils include nuts, seeds, and grains, and suitable oils include, for example, peanut oil, soybean oil, coconut oil, olive oil, and the like. Other suitable seed oils include safflower oil, cottonseed oil, sunflower seed oil, sesame seed oil, and the like. In the grain group, com oil, and the oil of other cereal grains such as wheat, oats, rye, rice, teff, triticale, and the like may also be used. The technology for obtaining vegetable oils is well developed and well known. The compositions of these and other similar oils may be found in, for example, the Merck Index, and source materials on foods, nutrition, and food technology. [0090] Most fish contain metabolizable oils which may be readily recovered. For example, cod liver oil, shark liver oils, and whale oil such as spermaceti exemplify several of the fish oils which may be used herein. A number of branched chain oils are synthesized biochemically in 5-carbon isoprene units and are generally referred to as terpenoids. Naturally occurring or synthetic terpenoids, also referred to as isoprenoids, may be used herein as a liquid phase lipid. Squalene is a branched, unsaturated terpenoid. A major source of squalene is shark liver oil, although plant oils (primarily vegetable oils), including amaranth seed, rice bran, wheat germ, and olive oils, are also suitable sources. Squalane is the saturated analog to squalene. Oils, including fish oils such as squalene and squalane, are readily available from commercial sources or may be obtained by methods known in the art. Oils to be used herein may also be made using synthetic means, including genetic engineering (e.g., oils made from bioengineered yeast, including squalene). Synthetic squalene has been successfully produced from bioengineered yeast and exhibits immunomodulating characteristics equal to squalene obtained from sharks. See Tateno, M., et al. “Synthetic Biology-Derived Triterpenes as Efficacious Immunomodulating Adjuvants,” Sci. Rep.2020, 10, 17090. Squalene has also been synthesized by the controlled oligomerization of isoprene. See Adlington, K., et al. “Molecular Design of Squalene/Squalane Countertypes via the Controlled Oligomerization of Isoprene and Evaluation of Vaccine Adjuvant Applications,” Biomacromolecules, 2016, 17(1), 165-72. [0091] Illustrative liquid phase lipids that may be used include, for example, castor oil, coconut oil, corn oil, cottonseed oil, evening primrose oil, fish oil, grapeseed oil, jojoba oil, lard oil, linseed oil, olive oil, peanut oil, safflower oil, sesame oil, soybean oil, squalene, squalane, sunflower oil, wheatgerm oil, mineral oil, capric/caprylic triglyceride (e.g., Miglyol ^® 810, Miglyol ^® 812, Labrafac™), vitamin E (e.g., TOS, TPGS), lauroyl polyoxylglycerides (e.g., Gelucire ^® 44/14), monoacylglycerols (e.g., Myverol 18-99K), soy lecithin (e.g., Epikuron™ 200), famesene, or a combination thereof. [0092] The liquid phase lipid may include for example, squalene, sunflower oil, soybean oil, olive oil, grapeseed oil, squalane, capric/caprylic triglyceride, or a combination thereof. [0093] The liquid phase lipid may include for example, squalene, squalene, capric/caprylic triglyceride, or a combination thereof. [0094] The liquid phase lipid may include for example, capric/caprylic triglyceride, vitamin E, lauroyl polyoxylglycerides, monoacylglycerols, soy lecithin, squalene, squalene, or a combination thereof. [0095] The liquid phase lipid can include for example, squalene, squalene, famesene, or a combination thereof. [0096] The oil core of the NLC comprises a solid phase lipid. A wide variety of solid phase lipids can be used, including for example, glycerolipids. Glycerolipids are fatty molecules composed of glycerol linked esterically to a fatty acid. Glycerolipids include triglycerides and diglycerides. [0097] Illustrative solid phase lipids include, for example, glyceryl palmitostearate (Precitol ^® ATO 5), glycerylmonostearate, glyceryl dibehenate (Compritol ^® 888 ATO), cetyl palmitate (Crodamol™ CP), stearic acid, tripalmitin, or a microcrystalline triglyceride. Illustrative microcrystalline triglycerides include those sold under the trade name Dynasan ^® (e.g., trimyristin (Dynasan ^® 114) or tristearin (Dynasan ^® 118) or tripalmitin (Dynasan ^® 116)). [0098] The solid phase lipid may be, for example, a microcrystalline triglyceride, for example, one selected from trimyristin (Dynasan ^® 114) or tristearin (Dynasan ^® 118). [0099] Preferably, the solid phase lipid of the oil core is solid at ambient temperature. When indoors, ambient temperature is typically between 15°C and 25°C. [0100] In any of the implementations provided herein, the solid phase lipid may be a glycerolipid, for example, a microcrystalline triglyceride. [0101] In any of the implementations provided herein, the liquid phase lipid may be synthetic or naturally-occurring squalene. [0102] In some implementations, the NLC may comprise an oil selected from the group consisting of Miglyol ^® 810, cottonseed oil, castor oil, soybean oil, and squalene. Cationic Lipids [0103] The NLCs described herein comprise a cationic lipid. The cationic lipid is useful for interacting with negatively charged bioactive agents on the surface of the NLC. Any cationic lipid capable of interacting with negatively charged bioactive agents that will not disturb the stability of the NLC and can be administered to a subject may be used. Generally, the cationic lipid contains a nitrogen atom that is positively charged under physiological conditions. Suitable cationic lipids include, benzalkonium chloride (BAK), benzethonium chloride, cetrimide (which contains tetradecyltrimethylammonium bromide and possibly small amounts of dodecyltrimethylammonium bromide and hexadecyltrimethyl ammonium bromide), cetylpyridinium chloride (CPC), cetyl trimethylammonium chloride (CTAC), primary amines, secondary amines, tertiary amines, including but not limited to N,N’,N’- polyoxyethylene (10)-N-tallow-l,3-diaminopropane, other quaternary amine salts, including but not limited to dodecyltrimethylammonium bromide, hexadecyltrimethylammonium bromide, mixed alkyl-trimethyl-ammonium bromide, benzyldimethyldodecylammonium chloride, benzyldimethylhexadecylammonium chloride, benzyltrimethylammonium methoxide, cetyldimethylethylammonium bromide, dimethyldioctadecyl ammonium bromide (DDAB), methylbenzethonium chloride, decamethonium chloride, methyl mixed trialkyl ammonium chloride, methyl trioctylammonium chloride, N,N-dimethyl-N-[2-(2- methyl-4-(l,1,3,3-tetramethylbutyl)-phenoxy]-ethoxy)ethyl]-b enzenemetha-naminium chloride (DEBDA), dialkyldimethylammonium salts, [1-(2,3-dioleyloxy)-propyl]-N,N,N- trimethylammonium chloride, l,2-diacyl-3-(trimethylammonio)propane (acyl group=dimyristoyl, dipalmitoyl, distearoyl, dioleoyl), l,2-diacyl- 3(dimethylammonio)propane (acyl group=dimyristoyl, dipalmitoyl, distearoyl, dioleoyl), l,2-dioleoyl-3-(4’-trimethyl-ammonio)butanoyl-sn-glycerol, 1,2-dioleoyl 3-succinyl-sn- glycerol choline ester, cholesteryl (4’-trimethylammonio) butanoate), N-alkyl pyridinium salts (e.g., cetylpyridinium bromide and cetylpyridinium chloride), N-alkylpiperidinium salts, dicationic bolaform electrolytes (C12Me6; C12Bu6), dialkylglycetylphosphorylcholine, lysolecithin, L-α-dioleoylphosphatidylethanolamine, cholesterol hemisuccinate choline ester, lipopolyamines, including but not limited to dioctadecylamidoglycylspermine (DOGS), dipalmitoyl phosphatidylethanol- amidospermine (DPPES), lipopoly-L (or D)-lysine (LPLL, LPDL), poly (L (or D)-lysine conjugated to N-glutarylphosphatidylethanolamine, didodecyl glutamate ester with pendant amino group (C12GluPhCnN+), ditetradecyl glutamate ester with pendant amino group (C14GluCnN+), cationic derivatives of cholesterol, including but not limited to cholesteryl- 3-β-oxysuccinamidoethylenetrimethylammonium salt, cholesteryl-3-β- oxysuccinamidoethylenedimethylamine, cholesteryl-3-β-carboxy amidoethylenetrimethylammonium salt, cholesteryl-3-β-carboxy amidoethylenedimethylamine, and 3-γ-[N-(N’,N- dimethylaminoethanecarbomoyl]cholesterol) (DC Cholesterol), l,2-dioleoyloxy-3- (trimethylammonio)propane (DOTAP), dimethyldioctadecylammonium (DDA), 1,2- dimyristoyl-3-trimethylammonium propane (DMTAP), dipalmitoyl(C16:0)trimethyl ammonium propane (DPTAP), distearoyltrimethylammonium propane (DSTAP), and combinations thereof. [0104] Other cationic lipids suitable for use in the invention include, e.g., the cationic lipids described in U.S. Patent Application Publication Nos. 2008/0085870 and 2008/0057080. [0105] Other cationic lipids suitable for use in the invention include, e.g., Lipids E0001- E0118 or E0119-E0180 as disclosed in Table 6 (pp. 112-139) of PCT Patent Application Publication No. WO 2011/076807 (which also discloses methods of making and methods of using these cationic lipids). Additional suitable cationic lipids include N-[l-(2,3- dioleyloxy)propyl]-N,N,N-trimethylammonium chloride (DOTMA), N,N-dioleoyl-N,N- dimethylammonium chloride (DODAC), 1,2-dioleoyl-sn-glycero-3-ethylphosphocholine (DOEPC), 1,2-dioleoyl-3-dimethylammonium propane (DODAP), l,2-dilinoleyloxy-3- dimethylamino propane (DLinDMA). [0106] The NLCs may comprise one or any combination of two or more of the cationic lipids described herein. [0107] In illustrative implementations, the cationic lipid is selected from the group consisting of l,2-dioleoyloxy-3-(trimethylammonio)propane (DOTAP), 313-[N-(N’,N’- dimethylaminoethane)-carbamoyl]cholesterol (DC Cholesterol), dimethyldioctadecylammonium (DDA), 1,2-dimyristoyl-3-trimethylammonium propane (DMTAP), dipalmitoyl(C16:0)trimethyl ammonium propane (DPTAP), distearoyltrimethylammonium propane (DSTAP), Lipids E0001-E0118 or E0119-E0180 as disclosed in Table 6 (pp. 112-139) of PCT Patent Application Publication No. WO 2011/076807, and combinations thereof. [0108] In other illustrative implementations, the cationic lipid is selected from the group consisting of l,2-dioleoyloxy-3-(trimethylammonio)propane (DOTAP), 313-[N-(N’,N’- dimethylaminoethane)-carbamoyl] cholesterol (DC Cholesterol), dimethyldioctadecylammonium (DDA), 1,2-dimyristoyl-3-trimethylammonium propane (DMTAP), dipalmitoyl(C16:0)trimethyl ammonium propane (DPTAP), distearoyltrimethylammonium propane (DSTAP), N-[l-(2,3-dioleyloxy)propyl]-N,N,N- trimethylammonium chloride (DOTMA), N,N-dioleoyl-N,N-dimethylammonium chloride (DODAC), l,2-dioleoyl-sn-glycero-3-ethylphosphocholine (DOEPC), l,2-dioleoyl-3- dimethylammonium-propane (DODAP), 1,2-dilinoleyloxy-3-dimethylaminopropane (DLinDMA), Lipids E0001-E0118 or E0119-E0180 as disclosed in Table 6 (pp. 112-139) of PCT Patent Application Publication No. WO 2011/076807, and combinations thereof. [0109] Illustrative cationic lipids are selected from the following: l,2-dioleoyloxy-3- (trimethylammonio)propane (DOTAP), 3[3-[N-(N’,N’-dimethylaminoethane)-carbamoyl] cholesterol (DC Cholesterol), dimethyldioctadecylammonium (DDA), 1,2-dimyristoyl-3- trimethylammonium propane (DMTAP), dipalmitoyl(C16:0)trimethylammonium propane (DPTAP), distearoyltrimethylammonium propane (DSTAP), N-[1-(2,3-dioleyloxy)propyl]- N,N,N-trimethylammonium chloride (DOTMA), N,N-dioleoyl-N,N-dimethylammonium chloride (DODAC), l,2-dioleoyl-sn-glycero-3-ethylphosphocholine (DOEPC), l,2-dioleoyl- 3-dimethylammonium-propane (DODAP), l,2-dilinoleyloxy-3-dimethylaminopropane (DLinDMA), or combinations thereof. Additional suitable cationic lipids may be known by one of skill in the art. [0110] In certain implementations, the NLC-based composition or formulation comprises from about 0.5 mg/ml to about 50 mg/ml of the cationic lipid. In certain implementations, the cationic lipid is DOTAP. The NLC may comprise, for example, from about 0.5 mg/ml to about 25 mg/ml or 30 mg/ml DOTAP or any other amount or range described herein for DOTAP. [0111] In certain implementations, the cationic lipid is DC Cholesterol. In certain aspects, the NLC may comprise DC Cholesterol at from about 0.1 mg/ml to about 5 mg/ml DC Cholesterol. In certain implementations, the cationic lipid is DDA. The NLC may comprise, for example, from about 0.1 mg/ml to about 5 mg/ml DDA. In certain implementations, the cationic lipid is DOTMA. The NLC may comprise, for example, from about 0.5 mg/ml to about 25 or 30 mg/ml DOTMA. In certain implementations, the cationic lipid is DOEPC. The NLC may comprise, for example, from about 0.5 mg/ml to about 25 mg/ml DOEPC. In certain implementations, the cationic lipid is DSTAP. The NLC may comprise, for example, from about 0.5 mg/ml to about 50 mg/ml DSTAP. In certain implementations, the cationic lipid is DODAC. The NLC may comprise, for example, from about 0.5 mg/ml to about 50 mg/ml DODAC. In certain implementations, the cationic lipid is DODAP. The NLC may comprise, for example, from about 0.5 mg/ml to about 50 mg/ml DODAP. [0112] With respect to weight per volume, an illustrative NLC-based composition or formulation may comprise, for example, from about 0.05 % to about 5% or to about 10% w/v cationic lipid such as DOTAP, from about 0.2% to about 10% w/v cationic lipid such as DOTAP, from about 0.2% to about 5% w/v cationic lipid such as DOTAP, from about 0.2% to about 2% w/v cationic lipid such as DOTAP, from about 2% to 10% w/v cationic lipid such as DOTAP, from about 2% to about 5% w/v cationic lipid such as DOTAP, from about 1% to about 5% w/v cationic lipid such as DOTAP, from about 3% to about 5% w/v cationic lipid such as DOTAP, or from about 3% to about 4% w/v cationic lipid such as DOTAP or any other amount or range described herein for the cationic lipid. Higher or lower w/v percentages are contemplated herein, particularly when considering diluted or concentrated formulations. [0113] In some cases, it may be desirable to use a cationic lipid that is soluble in the oil core. For example, DOTAP DOEPC, DODAC, and DOTMA are soluble in squalene or squalane. In other cases, it may be desirable to use a cationic lipid that is not soluble in the oil core. For example, DDA and DSTAP are not soluble in squalene. It is within the knowledge in the art to determine whether a particular lipid is soluble or insoluble in the oil and choose an appropriate oil and lipid combination accordingly. For example, solubility can be predicted based on the structures of the lipid and oil (e.g., the solubility of a lipid may be determined by the structure of its tail). For example, lipids having one or two unsaturated fatty acid chains (e.g., oleoyl tails), such as DOTAP, DOEPC, DODAC, DOTMA, are soluble in squalene or squalene, whereas lipids having saturated fatty acid chains (e.g., stearoyl tails) are not soluble in squalene. Alternatively, solubility can be determined according to the quantity of the lipid that dissolves in a given quantity of the oil to form a saturated solution). [0114] The NLC may comprise additional lipids (i.e., neutral and anionic lipids) in combination with the cationic lipid so long as the net surface charge of the NLC prior to mixing with the bioactive agent is positive. Methods of measuring surface charge of a NLC are known in the art and include for example, measurement by Dynamic Light Scattering (DLS), Photon Correlation Spectroscopy (PCS), or gel electrophoresis. Hydrophobic Surfactants [0115] The NLC further includes a hydrophobic surfactant, preferably a sorbitan ester. The term “sorbitan ester” as used herein refers to an ester of sorbitan. Sorbitan is as shown in Formula A: Formula A [0116] Suitable sorbitan esters are sorbitan alkyl esters, wherein the alkyl is a C1-C30 alkyl group, preferably a saturated or unsaturated C1-C20 alkyl group, more preferably a saturated or unsaturated C10-C20 alkyl group. [0117] It was previously discovered that the immune response to encoded proteins in a bioactive nucleic acid may be modulated by selection of sorbitan ester used in the NLC. It was surprisingly discovered that use of a sorbitan monoester was particularly effective at enhancing the effectiveness of the NLC. In some aspects, the acyl chain of the sorbitan monoester is saturated. In addition, without being bound by any specific theory, it was surprisingly discovered that the sorbitan ester, and in particular, sorbitan monoester, acts in combination with the solid lipid (e.g., microcrystalline triglycerides) to enhance the effectiveness of the adjuvant activity of the NLC (e.g., in eliciting antibodies to an antigen in a subject where the bioactive agent is an antigen or encodes antigen and the composition is administered to a subject). [0118] Illustrative sorbitan monoesters are commercially available under the tradenames Span ^® or Arlacel ^® . An illustrative sorbitan monoester for use herein may be represented as a compound of Formula I or a stereoisomer thereof (including, but not limited to, Formula Ia, Ib, Ic, or Id) wherein R is a saturated or unsaturated C1-C30 alkyl group, preferably a saturated or unsaturated C1-C20 alkyl group, and more preferably a saturated or unsaturated C10-C20 alkyl group. In illustrative implementations, the alkyl group is non- cyclic. Illustrative sorbitan monoesters also include positional isomers of Formulas I, Ia, Ib, Ic or Id (e.g., one of the hydroxy functional groups is replaced by an ester functional group (e.g., an alkyl ester wherein the alkyl is a saturated or unsaturated C1-C30 alkyl group, preferably a saturated or unsaturated C1-C20 alkyl group, and more preferably a saturated or unsaturated C10-C20 alkyl group, and R is OH). The skilled artisan will appreciate that illustrative sorbitan monoesters may be salt forms (e.g., pharmaceutically acceptable salts) of Formulas I, Ia, Ib, Ic, Id, and stereoisomers or positional isomers thereof. Formula Ic Formula Id [0119] Suitable sorbitan monoesters in this regard are sorbitan monostearate (also known as Span ^® 60 and shown below) and sorbitan monooleate (also known as Span ^® 80 and shown below), although other sorbitan monoesters can be used (including, but not limited to, sorbitan monolaurate (Span ^® 20) and sorbitan monopalmitate (Span ^® 40)). Illustrative sorbitan monostearate is represented by Formula II or IIa or a salt form thereof and illustrative sorbitan monooleate is represented by Formula III or IIIa or a salt form thereof.

[0120] In addition to providing sorbitan monoesters as a component of the NLC, also contemplated is the substitution of the sorbitan monoester for an alternative hydrophobic surfactant, including alternative sorbitan-based non-ionic surfactants. Accordingly, also provided herein are NLC particles comprising an oil core comprising a liquid phase lipid and a solid phase lipid, a cationic lipid, a hydrophobic surfactant (e.g., non-ionic surfactants including sorbitan-based non-ionic surfactants) and a hydrophilic surfactant. Sorbitan-based non-ionic surfactants include sorbitan esters other than sorbitan monoesters, for example sorbitan diesters and sorbitan triesters, such as for example, sorbitan trioleate (Span 85™) and sorbitan tristearate (Span 65™). Generally, the non-ionic surfactant (including sorbitan- based non-ionic surfactants) will have a hydrophilic-lipophilic balance (HLB) number between 1.8 to 8.6. All of the implementations provided herein for the NLCs comprising a sorbitan monoester are applicable and contemplated for the NLCs comprising an alternative hydrophobic surfactant in place of the sorbitan monoester, e.g., NLCs comprising a sorbitan diester or triester in place of the sorbitan monoester. The sorbitan diester and triester or other hydrophobic surfactant may be present in the same concentrations as the sorbitan monoester. In some aspects, the acyl chains of the sorbitan diester or triester will be saturated. [0121] Generally, the sorbitan esters (e.g., sorbitan monoesters) have a hydrophilic- lipophilic balance (HLB) value from 1 to 9. In some implementations, the sorbitan esters (e.g., sorbitan monoesters) have an HLB value from 1 to 5. In some implementations, the hydrophobic surfactant has an HLB value from about 4 to 5. [0122] An illustrative sorbitan diester for use herein may be represented as a compound of Formula IV below or a stereoisomer thereof (e.g., wherein R is a saturated or unsaturated C1-C30 alkyl group, preferably a saturated or unsaturated C1-C20 alkyl group, and more preferably a saturated or unsaturated C10-C20 alkyl group, and at least one of R1 is H while the other is -C(=O)Y wherein Y is a saturated or unsaturated C1-C30 alkyl group, preferably a saturated or unsaturated C1-C20 alkyl group, and more preferably a saturated or unsaturated C10-C20 alkyl group). In illustrative implementations, the alkyl group is non- cyclic. Illustrative sorbitan diesters also include positional isomers of Formulas IV. The skilled artisan will appreciate that illustrative sorbitan diesters may be salt forms (e.g., pharmaceutically acceptable salts) of Formula IV and stereoisomers or positional isomers thereof. [0123] An illustrative sorbitan triester for use herein may be represented as a compound of Formula V below or a stereoisomer thereof (including, but not limited to, Formula Va, Vb, or Vc) wherein R is a saturated or unsaturated C1-C30 alkyl group, preferably a saturated or unsaturated C1-C20 alkyl group, and more preferably a saturated or unsaturated C10-C20 alkyl group, and R1 is -C(=O)Y wherein Y may be the same or different in each instance and is a saturated or unsaturated C1-C30 alkyl group, preferably a saturated or unsaturated Cl- C20 alkyl group, and more preferably a saturated or unsaturated C10-C20 alkyl group. In illustrative implementations, the alkyl group is non-cyclic. Illustrative sorbitan triesters also include positional isomers of Formulas V, Va, Vb, or Vc (e.g., the hydroxy functional group is replaced by an ester functional group (e.g., an alkyl ester wherein the alkyl is a saturated or unsaturated C1-C30 alkyl group, preferably a saturated or unsaturated C1-C20 alkyl group, and more preferably a saturated or unsaturated C10-C20 alkyl group) and one of the alkyl esters (e.g., a ring alkyl ester or non-ring alkyl ester) is replaced by a hydroxy functional group). The skilled artisan will appreciate that illustrative sorbitan triesters may be salt forms (e.g., pharmaceutically acceptable salts) of Formulas V, Va, Vb, or Vc or stereoisomers or positional isomers thereof. [0124] With respect to stereoisomers, the skilled artisan will understand that the sorbitan esters may have chiral centers and may occur, for example, as racemates, racemic mixtures, or as individual enantiomers and diastereomers. [0125] In implementations wherein the sorbitan-based non-ionic surfactant is a sorbitan ester, typically, the NLC-based composition or formulation typically contains, for example, from about 0.1% to about 15% sorbitan ester (w/v), 0.1% to about 10% sorbitan ester (w/v), from 0.1% to about 5% sorbitan ester (w/v), about 0.1% to about 4 % sorbitan ester (w/v), about 0.1% to about 4% sorbitan ester (w/v), about 0.1% to about 2.5% sorbitan ester (w/v), about 0.1% to about 2% sorbitan ester (w/v), 0.1% to about 1.5% sorbitan ester (w/v), 0.1% to about 1% sorbitan ester (w/v), 0.1% to about 0.5% sorbitan ester (w/v), 0.3% to about 2.5% sorbitan ester (w/v), about 0.3% to about 2% sorbitan ester (w/v), 0.3% to about 1.5% sorbitan ester (w/v), 0.3% to about 1% sorbitan ester (w/v), 0.3% to about 0.5% sorbitan ester (w/v), or any other amount or range described herein for a sorbitan ester, including from about 0.25 % to about 15% sorbitan ester. In some aspects, the NLC-based compositions contain about 0.1%, about 0.2%, about 0.3%, about 0.4%, about 0.5%, about 0.6%, about 0.7%, about 0.8%, about 0.9%, about 1%, about 2%, about 3%, or about 4% (w/v) sorbitan ester. Higher or lower w/v percentages are contemplated herein, particularly when considering diluted or concentrated formulations. [0126] Accordingly, when the sorbitan ester is a sorbitan monoester (e.g., Span 60™, Span 80™), the NLC-based composition or formulation typically contains, for example, from about 0.1% to about 15% sorbitan monoester (w/v), 0.1% to about 10% sorbitan monoester (w/v), from 0.1% to about 5% sorbitan monoester (w/v), about 0.1% to about 4 % sorbitan monoester (w/v), about 0. 1% to about 4% sorbitan monoester (w/v), about 0. 1% to about 2.5% sorbitan monoester (w/v), about 0. 1% to about 2% sorbitan monoester (w/v), 0.1% to about 1.5% sorbitan monoester (w/v), 0.1% to about 1% sorbitan monoester (w/v), 0. 1% to about 0.5% sorbitan monoester (w/v), 0.3% to about 2.5% sorbitan monoester (w/v), about 0.3% to about 2% sorbitan monoester (w/v), 0.3% to about 1.5% sorbitan monoester (w/v), 0.3% to about 1% sorbitan monoester (w/v), 0.3% to about 0.5% sorbitan monoester (w/v), or any other amount or range described herein for sorbitan monoester, including from about 0.25 % to about 15% sorbitan monoester. In some aspects, the NLC-based composition or formulation contains about 0.1%, about 0.2%, about 0.3%, about 0.4%, about 0.5%, about 0.6%, about 0.7%, about 0.8%, about 0.9%, or about 1%, about 2%, about 3% or about 4% (w/v) sorbitan monoester. Higher or lower w/v percentages are contemplated herein, particularly when considering diluted or concentrated formulations. [0127] Accordingly, when the sorbitan ester is a sorbitan diester, the NLC-based composition or formulation typically contains, for example, from about 0.1% to about 15% sorbitan diester (w/v), 0.1% to about 10% sorbitan diester (w/v), from 0.1% to about 5% sorbitan diester (w/v), about 0.1% to about 4 % sorbitan diester (w/v), about 0.1% to about 4% sorbitan diester (w/v), about 0.1% to about 2.5% sorbitan diester (w/v), about 0.1% to about 2% sorbitan diester (w/v), 0.1% to about 1.5% sorbitan diester (w/v), 0.1% to about 1% sorbitan diester (w/v), 0.1% to about 0.5% sorbitan diester (w/v), 0.3% to about 2.5% sorbitan diester (w/v), about 0.3% to about 2% sorbitan diester (w/v), 0.3% to about 1.5% sorbitan diester (w/v), 0.3% to about 1% sorbitan diester (w/v), 0.3% to about 0.5% sorbitan diester (w/v), or any other amount or range described herein for sorbitan diester, including from about 0.25 % to about 15% sorbitan diester. In some aspects, the NLC-based composition or formulation contains about 0.1%, about 0.2%, about 0.3%, about 0.4%, about 0.5%, about 0.6%, about 0.7%, about 0.8%, about 0.9%, about 1%, about 2%, about 3%, or about 4% (w/v) sorbitan diester. Higher or lower w/v percentages are contemplated herein, particularly when considering diluted or concentrated formulations. [0128] Accordingly, when the sorbitan ester is a sorbitan triester (e.g., Span 85™ or Span 65™), the NLC-based composition or formulation typically contains, for example, from about 0.1% to about 15% sorbitan triester (w/v), 0.1% to about 10% sorbitan triester (w/v), from 0.1% to about 5% sorbitan triester (w/v), about 0. 1% to about 4 % sorbitan triester (w/v), about 0. 1% to about 4% sorbitan triester (w/v), about 0. 1% to about 2.5% sorbitan triester (w/v), about 0.1% to about 2% sorbitan triester (w/v), 0.1% to about 1.5% sorbitan triester (w/v), 0.1% to about 1% sorbitan triester (w/v), 0.1% to about 0.5% sorbitan triester (w/v), 0.3% to about 2.5% sorbitan triester (w/v), about 0.3% to about 2% sorbitan triester (w/v), 0.3% to about 1.5% sorbitan triester (w/v), 0.3% to about 1% sorbitan triester (w/v), 0.3% to about 0.5% sorbitan triester (w/v), or any other amount or range described herein for sorbitan triester, including from about 0.25 % to about 15% sorbitan triester. In some aspects, the NLC-based composition or formulation contains about 0.1%, about 0.2%, about 0.3%, about 0.4%, about 0.5%, about 0.6%, about 0.7%, about 0.8%, about 0.9%, about 1%, about 2%, about 3%, or about 4% (w/v) sorbitan tri ester. Higher or lower w/v percentages are contemplated herein, particularly when considering diluted or concentrated formulations. [0129] In illustrative implementations, the sorbitan ester (e.g., sorbitan monoester, diester, or triester) is present in an amount sufficient to increase the ability of the composition to facilitate delivery and/or expression of RNA as compared to a comparable composition lacking the sorbitan ester (e.g., sorbitan monoester, diester, or triester respectively). In implementations where the composition is administered to the subject in an effective amount, the composition may elicit antibody titers to the antigen equal to or greater than the antibody titers elicited when a comparable composition lacking the sorbitan ester is administered to the subject or when the RNA is administered to the subject without the NLC. In some implementations, the composition induces an immune response (e.g., neutralizing antibody titers) in the subject at a higher level than the immune response induced in the subject by a comparable composition lacking the sorbitan ester. Surfactants [0130] The NLCs comprise a surfactant, in addition to the sorbitan-based non-ionic surfactants (e.g., sorbitan ester). There are a number of surfactants specifically designed for and commonly used in biological applications. Such surfactants are divided into four basic types: anionic, cationic, zwitterionic, and non-ionic. A particularly useful group of surfactants are the hydrophilic non-ionic surfactants and, in particular, polyoxyethylene sorbitan monoesters, and polyoxyethylene sorbitan triesters. These materials are referred to as polysorbates and are commercially available under the mark Tween ^® and are useful for preparing the NLCs. Tween ^® surfactants generally have an HLB value falling between 9.6 to 16.7. Tween ^® surfactants are commercially available. Other non-ionic surfactants which may be used include, for example, polyoxyethylene fatty acid ethers derived from lauryl, acetyl, stearyl, and oleyl alcohols, polyoxyethylene fatty acids made by the reaction of ethylene oxide with a long-chain fatty acid, polyoxyethylene, polyol fatty acid esters, polyoxyethylene ether, polyoxypropylene fatty ethers, bee’s wax derivatives containing polyoxyethylene, polyoxyethylene lanolin derivatives, polyoxyethylene fatty glycerides, glycerol fatty acid esters or other polyoxyethylene fatty acids, alcohol or ether derivatives of C12-C22 long-chain fatty acids. [0131] In some implementations, it is preferable to choose a non-ionic surfactant which has an HLB value in the range of about 7 to 16. This value may be obtained through the use of a single non-ionic surfactant such as a Tween ^® surfactant or may be achieved by the use of a blend of surfactants. In certain implementations, the NLC comprises a single non-ionic surfactant, most particularly a Tween ^® surfactant, as the emulsion stabilizing non-ionic surfactant. In an illustrative implementation, the emulsion comprises Tween ^® 80, otherwise known as polysorbate 80. [0132] The NLC-based composition or formulation may contain, for example, from about 0.01% to about 15% surfactant (w/v), from about 0.01% to about 10% surfactant (w/v), from about 0.01% to about 5% surfactant (w/v), about 0.01% to about 2.5% surfactant, about 0.01% to about 2% surfactant, 0.01% to about 1.5% surfactant, 0.01% to about 1% surfactant, 0.01% to about 0.5% surfactant, 0.05% to about 0.5% surfactant, 0.08% to about 0.5% surfactant, about 0.08% surfactant, about 0.5% surfactant, about 0.6% surfactant, about 0.7% surfactant, about 0.8% surfactant, about 0.9% surfactant, about 1% surfactant, about 2%, about 3%, about 4 % surfactant, or any other amount or range described herein for the surfactant. Higher or lower w/v percentages are contemplated herein, particularly when considering diluted or concentrated formulations. Optional Components [0133] Additional components may be included in the NLCs including, for example, components that promote NLC formation, improve the complex formation between the negatively charged molecules and the cationic particles, facilitate appropriate release of negatively charged RNA molecules and/or increase the stability of negatively charged RNA molecules (e.g., to prevent degradation of the RNA). [0134] The aqueous phase (continuous phase) of the NLCs is typically a salt solution (e.g., saline) or water. The salt solution is typically an aqueous solution that comprises a salt (e.g., sodium citrate), and may further comprise, for example, a buffer (e.g., a citrate buffer), an osmolality adjusting agent (e.g., a saccharide), a polymer, a surfactant, or a combination thereof. It may be preferable to maintain a pH compatible with normal physiological conditions. Also, in certain instances, it may be desirable to maintain the pH at a particular level to ensure the stability of certain components of the NLC. For example, it may be desirable to prepare a NLC that is isotonic (i.e., the same permeable solute (e.g., salt) concentration as the normal cells of the body and the blood) and isosmotic. To control tonicity, the NLC may comprise a physiological salt, such as a sodium salt. In some aspects, sodium chloride (NaCl), for example, may be used at about 0.9% (w/v) (physiological saline). Other salts that may be present include, for example, potassium chloride, potassium dihydrogen phosphate, disodium phosphate, magnesium chloride, calcium chloride, and the like. Non-ionic tonicifying agents may also be used to control tonicity. Monosaccharides classified as aldoses such as glucose, mannose, arabinose, and ribose, as well as those classified as ketoses such as fructose, sorbose, and xylulose may be used as non-ionic tonicifying agents. Disaccharides such a sucrose, maltose, trehalose, and lactose may also be used. In addition, alditols (acyclic polyhydroxy alcohols, also referred to as sugar alcohols) such as glycerol, mannitol, xylitol, and sorbitol are non-ionic tonicifying agents that may be used. Non-ionic tonicity modifying agents may be present, for example, at a concentration of from about 0.1% to about 10% or about 1% to about 10%, depending upon the agent that is used. [0135] The aqueous phase may be, but is not necessarily, buffered. Any physiologically acceptable buffer that provides adequate protection for the RNA may be used herein, such as water, citrate buffers, phosphate buffers, acetate buffers, tris buffers, bicarbonate buffers, carbonate buffers, succinate buffer, or the like. The pH of the aqueous component will preferably be between 4.0-8.0 or from about 4.5 to about 6.8. In another illustrative implementation, the aqueous phase is, or the buffer is prepared using, RNase-free water or DEPC treated water. In some cases, high salt in the buffer may interfere with complexation of the negatively RNA charged molecule to the emulsion particle and therefore is avoided. In other cases, a certain amount of salt in the buffer may be included. [0136] In an illustrative implementation, the aqueous solution is sodium citrate with a pH between about 5.0 and 8.0. The sodium citrate solution may have a concentration of between 1-20 mM such as, 5 mM, 10 mM, 15 mM, or 20 mM. In another illustrative implementation, the aqueous phase is, or the buffer is prepared using, RNase-free water or DEPC treated water. [0137] The aqueous phase may also comprise additional components such as molecules that change the osmolarity of the aqueous phase or molecules that stabilize the negatively charged molecule after complexation. Preferably, the osmolarity of the aqueous phase is adjusting using a non-ionic tonicifying agent, such as a sugar (e.g., trehalose, sucrose, dextrose, fructose, reduced palatinose, or the like), a sugar alcohol (e.g., mannitol, sorbitol, xylitol, erythritol, lactitol, maltitol, glycerol, or the like), or combinations thereof. If desired, a non-ionic polymer (e.g., a poly(alkyl glycol) such as polyethylene glycol, polypropylene glycol, or polybutylene glycol) or nonionic surfactant may be used. [0138] Excipients may be used singly or in combination with other excipients which include, but are not limited to, cake-forming excipients, cake-forming bulking agents, bulking agents, buffering agents, chelating agents, solubilizing agents, isotonicity agents, tonicifying agents, surfactants, emulsifiers, antimicrobial agents, and/or collapse temperature modifiers. [0139] The excipients are substances other than a bioactive agent, which are included in the manufacturing process, or fill-finish process for storage or shipment of the composition including, without limitation, lyophilization, and are contained in a finished vaccine platform or vaccine. An excipient is a substance added to a liquid stable oil-in-water emulsion formulation prior to lyophilization which yields a cake following lyophilization. [0140] Excipients suitable for vaccine formulations and/or lyophilization are known in the art. See, e.g., Bahetia, A., et al. “Excipients Used in Lyophilization of Small Molecules,” J. Excip. Food Chem. 2010, 1(1), 41-54; Grabenstein, J. D. ImmunoFacts: Vaccines and Immunologic Drugs, 2012 (8th ed.), Wolters Kluwer Health. These include cake-forming excipients, cake- forming bulking agents, chelating agents, bulking agents, buffering agents, solubilizing agents, isotonicity agents, tonicifying agents, surfactants, emulsifiers, antimicrobial agents, and/or collapse temperature modifiers. Excipients in approved vaccines include without limitation sucrose, D-mannose, D-fructose, dextrose, potassium phosphate, plasdone C, anhydrous lactose, micro crystalline cellulose, polacrilin potassium, magnesium stearate, cellulose acetate phthalate, alcohol, acetone, castor oil, FD&C Yellow aluminum lake dye, human serum albumin, fetal bovine serum, sodium bicarbonate, human-diploid fibroblast cell cultures (WI-38), Dulbecco’s Modified Eagle’s Medium, aluminum hydroxide, benzethonium chloride, formaldehyde, gluteraldehyde, amino acids, vitamins, inorganic salts, sugars, glycerin, asparagine, citric acid, potassium phosphate, magnesium sulfate, iron ammonium citrate, lactose, aluminum potassium sulfate, aluminum hydroxyphosphate, potassium aluminum sulfate, peptone, bovine extract, thimerosal (trace), modified Mueller and Miller medium, beta-propiolactone, thimerosol (multi-dose vials only), monobasic sodium phosphate, dibasic sodium phosphate, monobasic potassium phosphate, potassium chloride, potassium glutamate, calcium chloride, sodium taurodeoxy cholate, neomycin sulfate, polymyxin B, egg protein, lactalbumin hydrolysate, and neomycin sulfate. [0141] A cake-forming excipient is a substance added to a liquid stable oil-in-water emulsion formulation prior to lyophilization which yields a cake following lyophilization. Upon reconstitution of the lyophilized cake, an oil-in-water stable emulsion forms which is suitable for delivery of a pharmacologically active drug. In some implementations, cake- forming excipients are those substances which do not disrupt an emulsion upon reconstitution of the cake. In some implementations the agents useful as cake-forming excipients, also referred to as bulking agents, include sugars/saccharides or sugars/saccharides in combination with sugar alcohols. In some implementations disclosed herein, the sugars/saccharides or sugars/saccharides in combination with sugar alcohols are useful as bulking agents or cake-forming excipients include, but are not limited to, trehalose, dextrose, lactose, maltose, sucrose, raffinose, mannose, stachyose, fructose, lactulose, glucose, glycerol, sorbitol, and/or mannitol. In one implementation, the cake-forming excipient is sucrose. In one implementation, the cake-forming excipient is trehalose. [0142] In some implementations, the cake-forming excipient is a saccharide and the saccharide is present in the NLC formulation prior to lyophilization at a concentration range of about 5% w/v to about 22% w/v, about 5% to about 20%, about 5% w/v to about 18% w/v, about 8% w/v to about 15% w/v, or about 9% w/v to about 11% w/v. In some implementations, the saccharide is present in the NLC formulation prior to lyophilization a concentration of about 5%, 6%, 7%, 8%, 9%, 10%, 11%, 12%, 13%, 14%, 15%, 16%, 17%, 18%, 19%, or 20%. [0143] In some implementations, the compositions comprise a buffering agent. Buffering agents useful as excipients include tris acetate, tris base, tris-HCl, ammonium phosphate, citric acid, sodium citrate, potassium citrate, tartic acid, sodium phosphate, zinc chloride, arginine, and histidine. Concentration of the buffering agents may range between 1-20 mM such as, for example 5 mM, 10 mM, or 20 mM. In some implementations buffering agents include pH adjusting agents such as hydrochloric acid, sodium hydroxide, and meglumine. [0144] Chelating agents such as ethylenediaminetetraacetic acid (EDTA) may be present at concentrations of between about 0.1-1 mM. Oil/Surfactant Ratios [0145] Illustrative NLCs are composed of a hydrophobic core containing the liquid oil and solid lipid, and surfactants (also known as emulsifiers or emulsifying agents) that make up the interface separating the hydrophobic phase—liquid oil and solid lipid, collectively referred to here as oil—from the aqueous phase. Since surfactants typically reside on the surface of NLC nanoparticles, their amount dictates the total available surface area. On the other hand, the oil resides in the core and primarily contributes to the total available volume. Increasing the surfactant to oil ratio consequently increases the surface area (SA) to volume ratio (V); thus, for a fixed volume of material, increasing the SA/V ratio translates to reducing NLC particle diameter. Instead of, or in addition to, describing illustrative NLC compositions in terms of the w/v percentages of various components, they can be described by the molar ratios of various components. In some aspects, illustrative NLCs have an oil to surfactant molar ratio of from about 0.05 to about 12 or from about 0.05 to about 9 or from about 0.05 to about 8 or from about 0.05 to about 1 or from about 0.1 to about 1. By reducing the oil to surfactant molar ratio, smaller NLCs may be synthesized. In addition, by reducing the amount of oil in the NLCs, potential toxicity of the formulations may be reduced. In other aspects, illustrative NLCs have an oil to surfactant molar ratio of from about 0.5 to about 12, from about 0.5 to about 9, from 1 to about 9, from about 2 to about 9, from about 3 to about 9, from about 4 to about 9, from about 4.5 to about 9, or from about 4.5 or about 5 to about 7. Illustrative formulations have an oil to surfactant molar ratio of about 0.5, about 1, about 1.5, about 2, about 2.5, about 3, about 3.5, about 4, about 4.5, about 5, about 5.5, about 6, about 7, about 8, about 9, about 10, about 11, or about 12. As used herein, the oil to surfactant molar ratio is determined by (i) adding the moles of lipid that make up the oil core (solid phase lipid and liquid phase lipid) to arrive at a value for moles of oil core lipid (ii) adding the moles of the cationic lipid (e.g., DOTAP), hydrophobic surfactant (e.g., sorbitan ester) and hydrophilic surfactant (e.g., Tween ^® 80) to arrive at a value for moles surfactant, and (iii) dividing moles of oil core lipid by moles of surfactant. Hydrophilic Surfactant/Cationic Lipid Ratios [0146] The ratio of hydrophilic surfactant to cationic lipid may impact the ability of the NLC to have a protective effect from RNase degradation and may impact the immunogenicity of the formulations. In particular, hydrophilic surfactant/cationic lipid ratios at about 0.6 are beneficial for obtaining consistent results for delivery and expression of RNA bioactive agents whereas hydrophilic surfactant/cationic lipid ratios at about 2.0 and higher are not as beneficial for obtaining such consistency. Accordingly, illustrative NLCs have a hydrophilic surfactant/cationic lipid ratio of from about 0.2 to about 1.5, from about 0.2 to about 1, or from about 0.5 to about 1. When Tween and DOTAP are in the composition, illustrative NLCs have a Tween/DOTAP ratio of from about 0.2 to about 1.5, from about 0.2 to about 1, or from about 0.5 to about 1. As used herein, the hydrophilic surfactant: cationic lipid ratio is determined by (i) adding the moles of hydrophilic surfactant to arrive at a value for moles of hydrophilic surfactant, (ii) adding the moles of the cationic lipid to arrive at a value for moles of cationic lipid, and (iii) dividing moles of hydrophilic surfactant by moles of cationic lipid. Loading Capacities [0147] The loading capacity of the NLC formulations may be manipulated by modulating the ratio of hydrophilic surfactant to cationic lipid and the amount of oil present in the formulations, thereby reducing the average NLC particle size. Illustrative NLC formulations have loading capacity for RNA of at least about 10 pg/ml RNA, at least about 20 pg/ml RNA, at least about 50 pg/ml RNA, at least about 100 pg/ml RNA, at least about 200 pg/ml RNA, at least about 300 pg/ml, or at least about 400 pg/ml RNA. NLC formulations having an average particle size of from 20 nm to about 110 nm, from about 20 nm to about 80 nm, from about 20 nm to about 70 nm, and from about 20 nm to about 60 nm typically have increased loading capacity. Persons of ordinary skill in the art will appreciate how to adjust the NLC formulation to achieve a desired loading capacity. Examples of NLC Formulations [0148] In some implementations, the NLC may comprise about 3.75% w/v GMP shark squalene, about 0.24% w/v Dynasan ^® 114, about 3.00% w/v DOTAP (1,2-dioleoyl-3- trimethylammonium propane), about 3.70% w/v Span ^® 60, about 3.70% Tween ^® 80, and about 10 mM sodium citrate. [0149] In some alternate preferred implementations, the NLC may preferably comprise Miglyol ^® 810. [0150] In some other alternate implementations, the NLC may comprise any of the NLC formulations described in U.S. Patent Application Publication No. 2020/0230056, filed on December 13, 2019 and entitled “Nanostructured Lipid Carriers and Stable Emulsions and Uses Thereof.” In other alternate implementations, the NLC may comprise any of the NLC formulations described in U.S. Patent Application Serial No.63/144169, filed on February 1, 2021 and entitled “A Thermostable, Flexible RNA Vaccine Delivery Platform For Pandemic Response.” Both of these references are incorporated herein in their entireties by reference. EXPERIMENTAL RESULTS AND METHODS SARS-CoV-2 Studies [0151] Without being bound by any specific theory, it is believed that mucosal immune responses are important to improving sustained, potent immunity to SARS-CoV-2 and other respiratory pathogens. Unlike i.m. administered vaccines, i.n. vaccines induce critical immune cells within the mucosa and respiratory tract, generating IgA-secreting cells as well as specific effector cell subsets, including T _RM cells that are maintained within the lung. Such mucosal immunity, mediated by IgA antibodies and lung-resident memory T and B cells, can prevent early-stage infection and viral replication and shedding, minimizing chances for viral transmission. Effective i.n. vaccines can also induce cross-reactive antibodies and thus provide expanded protection to multiple strains. Intranasal vaccination also presents many other advantages, such as easier needle-free administration, higher patient compliance, simpler pediatric dosing, and easier storage and transport, and thus potentially increased vaccine uptake. However, while several i.n. vaccines are currently under development for SARS-CoV-2, most of these rely on older vaccine technologies, i.e., live-attenuated viruses and viral vectors, creating a slow pathway to manufacturing, testing, and licensing. Fortunately, bridging RNA vaccine technologies to i.n. delivery is possible, combining the production speed of RNA vaccines with the mucosal immune stimulation of i.n. vaccines. [0152] It was observed that both i.n. and i.m. administrations of self-amplifying RNA SARS-CoV-2 vaccine delivered via the disclosed vaccine delivery system strongly induce systemic immunity in preclinical models and prevent viral transmission by vaccinated, virally-challenged hamsters. Intranasal dosing uniquely induces strong mucosal T cell immunity, and a heterologous dosing strategy—i.m. prime followed by i.n. boosting— appears to show the greatest potential to maximize mucosal and systemic immunity, leading to prevention of viral transmission. The self-amplifying RNA/NLC vaccine delivered via the disclosed delivery system combines the flexibility and manufacturability of RNA vaccines with potent induction of mucosal immunity using a thermostable formulation that represents a significant new tool for current and future pandemic responses. RESULTS [0153] The previously developed long-term thermostable SARS-CoV-2 self-amplifying RNA/NLC vaccine, see Voigt, E. A., et al. “A Self-Amplifying RNA Vaccine against COVID-19 with Long-Term Room-Temperature Stability,” npj Vaccines, 2022, 7, 1-13, was used to develop an intranasally delivered self-amplifying RNA vaccine against SARS- CoV-2. This vaccine contains a codon-optimized Wuhan-strain D614G SARS-CoV-2 spike protein sequence driven by a Venezuelan equine encephalitis virus (VEEV)-based replicon, as shown in FIG. 1A. This self-amplifying RNA is complexed to the exterior of an NLC nanoparticle to create the final vaccine product, as shown in FIG.1B. Intranasal Self-Amplifying RNA/NLC Vaccine Induces Antibody Generation [0154] C57Bl/6 mice were immunized intranasally with the optimized SARS-CoV-2 self-amplifying RNA/NLC vaccine and immunogenicity was evaluated relative to a traditional intramuscularly-administered SARS-CoV-2 vaccine. [0155] A self-amplifying RNA vaccine expressing the SARS-CoV-2 S spike protein, as illustrated in FIG. 1, was complexed with a nanostructured lipid carrier comprising 3.75% w/v GMP shark squalene, 0.24% w/v Dynasan 114, 3.00% w/v DOTAP (1,2-dioleoyl-3- trimethylammonium-propane), 3.70% w/v Span 60, 3.70% Tween 80, and 10 mM sodium citrate. The complex had an NLC-contained-amine-group to RNA-phosphate (N/P) ratio of 15, creating vaccine nanoparticles in a liquid solution of 5 mM citrate, 10% sucrose for isotonicity. [0156] A prime dose was administered intranasally to C57Bl/6 mice (n=6) with a prime dose of 5 µg RNA per mouse delivered in a 20 µl volume (10 µl/nare, 6 mice total) or i.m. (20 µl 1x i.m. injection, 2 control mice). Dosing was performed using standard techniques. The mice were lightly anesthetized using isoflurane, breathing was monitored, and the liquid vaccine droplet was pipetted in front of a nostril just before the mouse was about to breathe—at which point the droplet was inhaled by the mouse via its nose. This was then repeated for the second nostril. [0157] The mice were given a boost dose of 12 µg RNA/mouse administered intranasally in 50 µl (25 µl per nare, 6 mice total) or i.m. (50 µl 1x i.m. injection, 2 control mice) 21 days post-prime. [0158] Mouse serum samples were taken from all eight mice 14 days post-boost and tested for SARS-CoV-2 binding and neutralizing antibodies, as shown in FIGS. 2 and 3 respectively. Intranasal Self-Amplifying RNA/NLC Vaccine Stimulates Robust Systemic Immune Responses [0159] To investigate the ability of the self-amplifying RNA/NLC vaccine to generate effective immunogenicity when delivered intranasally, 6–8-week-old C57BL/6J mice were given prime and boost vaccine doses of 1, 5, or 10 µg, 3 weeks apart. Control mice were i.n. dosed with 10 µg of a non-immunogenic self-amplifying RNA/NLC expressing the reporter protein secreted alkaline phosphatase 2 (SEAP). Systemic immunogenicity was compared between groups. Serum, bone marrow, lung, and spleen tissues were harvested after vaccination and tested for humoral (serum IgG and pseudovirus neutralization titers; bone marrow antibody-secreting cells [ASCs] by enzyme-linked immunospot [ELISpot]) and cellular (systemic T cell responses by intracellular cytokine staining and flow cytometry [ICS-flow] and T cell ELISpot of splenocytes; mucosal T cell responses by ICS-flow of lungs) vaccine-induced SARS-CoV-2-directed immune responses. [0160] As shown in FIG. 4, systemic immunogenicity of intranasally administered SARS-CoV-2 saRNA/NLC vaccine was assessed. [0161] The i.n. self-amplifying RNA/NLC vaccine elicited high serum spike-specific IgG and SARS-CoV-2 Wuhan-strain neutralization titers that would be predicted to be protective (>1000 IC50) at all i.n. doses after prime and boost, as shown in FIGS. 4A-B. Strikingly, at the 5 and 10-µg doses, no significant differences were observed between i.n. and i.m. vaccines post-boost, suggesting equivalent induction of serum responses—the key correlates of protection for SARS-CoV-2—by the i.n. vaccine. IgG1 and IgG2a serum antibody titers indicated that i.n. vaccination with the self-amplifying RNA/NLC vaccine maintains the desired Th1 bias seen with the corresponding i.m. vaccine, as shown in FIG. 4C. Next, bone marrow-resident ASCs were assessed post-boost by ELISpot to determine if there was establishment of a memory niche for IgG-secreting ASCs. No differences were seen between the i.n. and i.m. vaccines at either the 5- or 10-µg dose i.n. as compared to the 10-µg dose i.m., as shown in FIG.4D. This further suggests effective systemic induction by both delivery routes and suggesting durability of response established by both vaccines. [0162] Cell-mediated immunity, particularly T cell immunity, is also thought to be important for sustained, long-term protection against SARS-CoV-2. Thus, systemic cellular responses were investigated in i.n. and i.m. vaccinated mouse spleens after boost, both by ELISpot and ICS. Significant populations of IFNγ-secreting T cells were observed in both the i.n. and i.m. dosed (10-µg dose) mouse groups as measured by ELISpot, as shown in FIG.4E. This indicated effective induction of effector T cells primed to respond to SARS- CoV-2. While the cell numbers were significantly higher in the i.m. dosed group, both groups showed robust T cell responses. IL-5- and IL-17-secreting cells were seen at negligible populations in either group, as shown in FIGS. 4F-G. This suggested a strong Th1-biased response induced by both the i.n. and i.m. vaccines with very little induction of a potentially harmful Th2 response associated with immunopathology and natural infection. [0163] These robust, strongly Th1-skewed systemic CD4 ⁺ and CD8 ⁺ T cell responses were verified using ICS-flow performed on mouse splenocytes collected post-vaccination. After prime-boost immunization, significant populations of spike-reactive polyfunctional (IFNγ ⁺ , IL-2 ⁺ , and TNFα ⁺ ) CD8 ⁺ and CD4 ⁺ T cells were observed in mice immunized with the 5- and 10-μg i.n. vaccine doses and the 10-μg i.m. dose, as shown in FIGS.4H-I. This, demonstrated strong T cell-mediated responses induced by both i.m. and i.n. vaccines. The spike-reactive polyfunctional T cell populations of CD4 ⁺ T cells induced by the i.n. vaccine were lower than those induced by the i.m. vaccine (20-30% reduction between the i.n. and i.m. vaccines at a 10-µg dose, p = 0.0191); however, spike-reactive polyfunctional CD8 ⁺ T cell populations were not significantly different between the i.n. and i.m. vaccinated mice, and CD4 ⁺ T cell responses induced by the i.n. vaccine were still high, demonstrating effective seeding of systemic T cell immunity across all i.n. and i.m. vaccinated mice. As seen with vaccine-induced antibody titers, a stronger induction of a Th1-specific response (secretion of IFNγ ⁺ , IL-2 ⁺ , or TNFα ⁺ ) compared to a Th2 response (secretion of IL-5 ⁺ or IL- 10 ⁺ ) or a Th17 response (secretion of IL-17 ⁺ ) occurred within the T cell compartment in both i.n. and i.m. vaccinated mice and across all doses. This suggested an overall systemic functional response to the virus. These data therefore indicate that the i.n. self-amplifying RNA/NLC vaccine elicits humoral and cellular responses at levels predicted to be protective. Intranasal Vaccine Administration Induces Specialized Lung T Cell Populations [0164] While both i.m. and i.n. vaccines induce excellent systemic immunity, a key objective to assessing i.n. immunization was to determine whether effective mucosal responses could be generated with i.n. administration of an RNA vaccine. To that end, the lung T cell compartment was analyzed from C57BL/6J mice vaccinated i.n. or i.m. with 10 µg of AAHI-SC23 weeks post-prime and 3 weeks post-boost; the lung T cell compartment was analyzed by ICS-flow for the presence of specialized T cell subsets that would indicate the development of long-lasting respiratory-specific immunity. [0165] FIG. 5 shows mucosal T cell populations following intranasal vaccination with a SARS-CoV-2 self-amplifying RNA/NLC vaccine. [0166] First, post-vaccination lung infiltration of antigen-experienced memory T cells was examined. These TRM cells are characterized by CD69 expression as well as CD103 expression in the CD8 compartment. The percentage of T _RM cells were significantly increased in the CD8 compartment with i.n. vaccination (median of CD69 ⁺ cells >20% of CD8 ⁺ ) compared to i.m. vaccination (median of CD69 ⁺ cells <10% of CD8 ⁺ ) following boost (CD69 ⁺ CD8 ⁺ cells, p = 0.0029; CD69 ⁺ CD103 ⁺ CD8 ⁺ cells, p < 0.0001), as shown in FIGS. 5A-B. While the fractions of CD8 ⁺ T _RM cells were equivalent post-prime, lung expansion of TRM cells followed boost administration in the i.n. dosed group only, suggesting seeding of these durable populations followed by recall and expansion upon boost dosing. [0167] CD154 ⁺ CD4 ⁺ T cells were then assessed. CD154 (CD40 ligand [CD40L]) is often expressed on activated T cells and T follicular helper cells and is critical for licensing dendritic cells and engaging with B cells to enhance mucosal antibody production. Here, it was found that similar patterns to those seen with T _RM cells; significantly increased populations of CD154 ⁺ CD4 ⁺ cells were observed in the lungs after i.n. self-amplifying RNA vaccination post-boost compared to i.m. vaccination (p = 0.0029), as shown in FIG. 5C. Collectively, these data show that i.n. administration appears to elicit an expansion of TRM and CD40L ⁺ activated CD4 ⁺ T cells within the lung. In contrast, i.m. vaccination elicits minimal levels of these cell subsets, suggesting that these specialized respiratory cells are not inducible by i.m. vaccination. Intranasal Administration Elicits Polyfunctional SARS-CoV-2 Reactive Mucosal T Cell Responses [0168] While the lung T cell subsets were present in large populations post-i.n. administration, it was desirable to verify that these cell subsets were functional effector cells and confirm that they responded to SARS-CoV-2. Here, mice that received the vector control or had been immunized and boosted with AAHI-SC2 i.m. (10 µg) or i.n. (1 µg, 5 µg, or 10 µg) were used, and ICS-flow cytometry was used to determine the frequency of polyfunctional lung IFNγ ⁺ IL-2 ⁺ TNFα ⁺ CD4 ⁺ and CD8 ⁺ T cells following re-stimulation for 6 hours with overlapping spike peptides. Polyfunctional IFNγ ⁺ IL-2 ⁺ TNFα ⁺ -secreting SARS-CoV-2 spike-responsive T cells composed a large fraction of the key T _RM , CD69 ⁺ population in the 5- and 10-µg i.n. immunized mice (12.8% and 9.11%, respectively), which was significant in the CD69 ⁺ CD8 ⁺ subset of these T _RM cells compared to the i.m. immunized mice (p = 0.0005 for the 5-µg dose and p = 0.0258 for the 10-µg dose), as well as in the CD69 ⁺ CD103 ⁺ CD8 ⁺ subset (p = 0.0181) from 5-µg i.n. immunized mice, as shown in FIGS. 6A-B. Intramuscularly immunized groups again induced consistently negligible levels of polyfunctional CD8 ⁺ cells. The activated CD154 ⁺ /CD40L ⁺ CD4 ⁺ T cells and the lung-homing CD194 ⁺ /CCR4 ⁺ CD4 ⁺ and CD8 ⁺ cells were also examined. Similar to the TRM cells, CD154 ⁺ CD4 ⁺ cells showed significantly higher percentages of polyfunctionally responding cells at the 5-µg i.n. dose compared to the i.m. dose (p = 0.0006), as shown in FIG.6C. Likewise, CD194 ⁺ CD4 ⁺ and CD8 ⁺ cells were also induced by the i.n. vaccine at significantly higher percentages at the 5-µg dose (p = 0.0441 and p = 0.0071, respectively) compared to the i.m. vaccine, as shown in FIGS.6D-E. [0169] Finally, the quality of the response by lung T cells was examined to confirm that i.n. administration elicited Th1-biased lung T cell responses, as was observed in the spleen. The majority of the responses to both the i.n. and i.m. vaccines were Th1-specific responses, with negligible Th2 and Th17 responses, similar to what was observed in the spleen. Th2 responses were not significantly different between i.m. vaccinated mice and mice receiving the vaccine via the i.n. route, did not scale with increasing vaccine dose, and were not significantly increased from buffer-only groups. Collectively, these data demonstrate that i.n. administration seeds specialized polyfunctional CD4 ⁺ and CD8 ⁺ T cell populations in the lung – populations which are not elicited following i.m. administration. Intranasal Boosting of Previous Intramuscular Vaccination Induces Both Systemic and Mucosal Immune Responses [0170] To understand whether i.n. administration could be viable as a boosting strategy for individuals previously vaccinated with an i.m. SARS-CoV-2 vaccine, mice that were primed with an i.m. vaccine and then boosted with a medium dose (5 µg) either i.m. or i.n. were examined. Intranasally boosted mice showed equivalent serum anti-spike IgG levels and similar systemic neutralizing antibody responses to the i.m. boosted mice, as shown in FIGS. 7A-B. Within the splenic CD4 ⁺ and CD8 ⁺ T cell compartment, i.n. boost immunization induced equivalent or increased spleen-resident polyfunctional T cell responses in CD4 ⁺ or CD8 ⁺ (p = 0.0454) cells, respectively, compared to an i.m. boost immunization, as shown in FIGS.7C-D. This indicated that an i.n. booster may effectively boost the systemic immunity developed in response to an original i.m. vaccination. With respect to mucosal immunity, the i.n. booster induced a higher but not significant percentage of CD40L ⁺ (CD154 ⁺ ) activated CD4 ⁺ T cells within the lungs than the i.m. booster, as shown in FIG. 7E. Most importantly, the i.n. booster elicited significantly more polyfunctional lung-resident CD8 ⁺ CD69 ⁺ T _RM cells (p = 0.0236) than the i.m. booster, as shown in FIG. 7F. This indicated that just boosting with an i.n. dose may be sufficient to elicit the full mucosal immune-stimulating effect of i.n. vaccination. These data are consistent with other i.n. vaccines that suggest i.n. boosting can have enhanced efficacy systemically as compared to i.m. vaccines. This suggests that an i.n. boost vaccination strategy may be ideal to boost systemic and induce respiratory immunity in previously vaccinated individuals, reflecting the real-world scenario of second-generation SARS-CoV-2 vaccine rollouts. Self-Amplifying RNA Vaccination Protects Against SARS-CoV-2-Induced Disease in Hamsters [0171] Given the data showing that i.m. prime and i.n. boost with the self-amplifying RNA/NLC SARS-CoV-2 vaccine described herein elicited strong systemic and pulmonary immunity in mice, it was hypothesized that vaccination, particularly via the i.m./i.n. route, would be sufficient to protect animals from morbidity due to SARS-CoV-2 infection and would prevent these animals from transmitting the virus to naive bystander animals. To test this hypothesis, hamster pairs were co-housed (1:1) for several weeks to acclimate them to one another. Then, paired animals were separated and one animal in each pair was either sham-vaccinated or vaccinated with 5 µg of the self-amplifying RNA/NLC SARS-CoV-2 vaccine given either i.m. or i.n., as shown in FIG.8A. After 24 hours, the sham-vaccinated and vaccinated animals were re-paired with their naive (non-vaccinated) cagemates. On day 21, peripheral blood was collected from all animals. On day 28, all pairs of animals were separated, and the sham-vaccinated and vaccinated animals were boosted via the i.m. or i.n. route. All animals were re-paired with their naive cagemates after 24 hours and blood was collected from all animals five days later. As expected, vaccinated hamsters, particularly i.m.-dosed hamsters, demonstrated strong induction of spike-specific serum IgG post-prime, which was absent in the sham-vaccinated group, as shown in FIG. 8B, and in all the naive cagemates (not shown). Intranasally primed animals also showed development of significant serum IgG titers, albeit at titers significantly lower than those observed in i.m.-primed animals (p < 0.0001). Boosting on day 28 did not result in a significant increase in systemic anti-spike IgG responses in any of the vaccinated groups, as shown in FIG.8C. This may have been due to the still ongoing robust primary response. [0172] At 9 days post-boost (day 37), the vaccinated and sham-vaccinated hamsters were separated from their naive cagemates, as shown in FIG. 8A. Some sham-vaccinated hamsters were set aside to serve as uninfected negative controls (sham-vaccinated- uninfected), and all other vaccinated or sham-vaccinated hamsters were experimentally infected with 1 x 10 ⁶ pfu SARS-CoV-2 USA/WA-1 strain delivered intranasally. Morbidity of the vaccinated-infected and sham-vaccinated-infected animals was then compared to the sham-vaccinated-uninfected controls. As expected, the sham-vaccinated-uninfected hamsters exhibited a modest increase in weight over the next 14 days, as shown in FIG.8D. By contrast, the sham-vaccinated-infected animals had increased morbidity with peak body weight loss of 9-10% by day 6 post-infection and minimal body weight recovery by termination of the study. The vaccinated-infected animals (i.n.-i.n., i.m.-i.n., and i.m.-i.m. vaccination), on the other hand, showed minimal weight loss and exhibited rapid recovery in weight by day 3-4 post-infection, with statistically significantly reduced virus infection- induced morbidity when compared to the sham-vaccinated-infected group (p = 0.0091, p = 0.0074, and p = 0.0206 for i.n.-i.n., i.m.-i.n., and i.m.-i.m. vaccinated-infected animals, respectively). Body weight loss after challenge did not correlate with systemic anti-spike IgG titers, suggesting that either all three vaccination regimens induced sufficient systemic IgG to protect against SARS-CoV-2-induced weight loss and/or cellular or mucosal responses played a role in protection (not shown). [0173] To further understand the dynamics of infection and transmission, total and subgenomic viral nucleocapsid (N) RNA was measured in nasal swipes from all infected hamsters across the first 10 days after infection. Total viral N RNA was detected in all infected hamsters as early as 1 hour post infection, as shown in FIG. 8E. This confirmed successful intranasal instillation of SARS-CoV-2. Total viral copies peaked between 24-48 hours in all infected groups. However, total viral N RNA copies were significantly lower in the vaccinated-infected groups compared to the sham-vaccinated-infected group (p < 0.0001, p < 0.0001, and p = 0.0003 for i.n.-i.n., i.m.-i.n., and i.m.-i.m., respectively). Next, subgenomic viral N RNA was quantified in the nasal discharge to compare infectious, actively replicating viral load between the sham-vaccinated-infected and vaccinated- infected hamsters. Subgenomic viral N RNA was detected in the nasal swipes from all infected hamsters. However, significantly diminished subgenomic viral N RNA copies were observed in all vaccinated-infected animals relative to the sham-vaccinated-infected animals, as shown in FIG.8F (p < 0.0001, p = 0.0002, and p < 0.0001 for i.n.-i.n., i.m.-i.n., and i.m.-i.m. vaccine regimens, respectively). Moreover, the peak copies of subgenomic N RNA in nasal swipes were observed within 24 hours of infection in the vaccinated-infected hamsters, while subgenomic N RNA peaked in the sham-vaccinated-infected hamsters between days 1-2, as shown in FIG. 8F. Importantly, the vaccinated-infected hamsters had increased clearance of viral subgenomic N RNA that declined to undetectable levels by day 6 (i.m.-i.n. and i.m.-i.m. groups) and day 8 (i.n.-i.n. group) post-infection. In contrast, the sham-vaccinated-infected control animals had detectable viral subgenomic N RNA until day 10, indicating that vaccination promoted more rapid reduction in replicating virus. [0174] The viral RNA measurements from nasal swipes suggested that none of the self- amplifying RNA/NLC vaccination regimens induced sterilizing immunity following direct intranasal instillation with 1x10 ⁶ pfu of SARS-CoV-2. To address whether the vaccinated animals made an immune response to experimental infection, systemic anti-N IgG antibodies were measured on day 35 post-infection. Indeed, anti-N IgG antibody was detected in the serum from all three groups of vaccinated-infected animals, as shown in FIG. 8G. This indicated that the amount of N protein made by the infectious virus was sufficient to elicit a de novo N-specific antibody response in these hamsters. However, compared to the sham-vaccinated-infected animals, the amount of anti-N IgG antibody was significantly lower in the i.m.-i.m. and i.m.-i.n. vaccinated-infected hamsters (p < 0.0001 and p = 0.0470, respectively) and modestly lower in the i.n.-i.n. vaccinated-infected hamsters, again consistent with a decreased viral load in all vaccinated animals. [0175] Next, to assess SARS-CoV-2-induced pulmonary pathology, cross sections of the hamster lungs were examined and scored for histology by a board-certified pathologist. As expected, the sham-vaccinated-infected hamsters had mild to moderate histology scores confirming that the SARS-CoV-2 viral inoculum used to infect the hamsters caused detectable pulmonary immunopathology as late as day 35 post-infection, as shown in FIG. 8H. These sham-vaccinated-infected hamsters exhibited signs of inflammation that consisted primarily of accumulation of mononuclear and polymorphonuclear infiltrates in the alveolar interstitium and lumen, and adjacent inflammation around blood vessels and bronchioles (not shown). Conversely, all three groups of vaccinated-infected hamsters had significantly decreased total histology scores, as shown in FIG. 8H, and exhibited only sporadic interstitial, perivascular, and peribronchiolar inflammation that were comparable to sham-vaccinated-uninfected negative control animals (not shown). This suggested that vaccination reduced the extent and severity of lung pathology following infection with a high dose of SARS-CoV-2. Therefore, while vaccination did not prevent infection following a robust viral challenge, all vaccine regimens effectively prevented disease-associated morbidity as measured by weight loss, decreased peak viral copies in the nasal discharge, suppression of viral replication, increased viral clearance, and protection against SARS- CoV-2-induced lung damage. Self-Amplifying RNA Vaccination Prevents Viral Transmission by Virally Challenged Hamsters [0176] Although all three groups of vaccinated-infected hamsters appeared to have been productively infected following experimental SARS-CoV-2 exposure, the amount of subgenomic viral N RNA in the nasal secretions of these hamsters had already declined significantly by day 2 post-infection relative to the sham-vaccinated-infected group, as shown in FIG.8F. To assess whether this reduction in infectious viral load was sufficient to prevent transmission, the sham-vaccinated-infected and vaccinated-infected hamsters were re-paired with their naive cagemates on day 2 post-infection (day 39 on timeline in FIG.8A; day 0 on timeline in FIG.9A) for a period of 24 hours, as shown in FIG.9A, and assessed morbidity, nasal swipe viral RNA, serum IgG antibodies, and lung immunopathology in the naturally virally challenged cagemates. As expected, naive cagemates of the sham- vaccinated-infected animals began losing weight within 24 hours of re-pairing, reaching maximum weight loss of 6-7% by day 7 post-re-pairing and exhibiting minimal body weight recovery by the termination of the study, as shown in FIG.9B. In striking contrast, the naive cagemates of the vaccinated-infected hamsters exhibited no body weight loss following re- pairing (p = 0.0412, p = 0.0470, and p = 0.0478 for i.n.-i.n., i.m.-i.n., and i.m.-i.m., respectively). [0177] Given the little morbidity observed in the naive cagemates co-housed with the vaccinated-infected hamsters, it was hypothesized that viral transmission between the vaccinated-infected animals and their naive cagemates would also be diminished. To test this, total and subgenomic viral N RNA was measured in the nasal swipes, as shown in FIGS.9C and 9D respectively, from the naive cagemates that were paired for 24 hours with the sham-vaccinated-infected and vaccinated-infected groups. Total viral N RNA, which reflects both replicating and non-replicating virus, was present in nasal swipes from all the naive cagemates as early as 24 hours post-re-pairing with infected hamsters, as shown in FIG. 9C. Interestingly, the naive cagemates paired with vaccinated-infected hamsters had on average 70-fold lower total viral N RNA at 24 hours post-re-pairing compared to naive hamsters that were co-housed with sham-vaccinated-infected animals, as shown in FIG.9C. Hence, immediately following re-pairing, the viral load transmitted to naive cagemates by the vaccinated-infected animals was lower than that transmitted by the sham-vaccinated- infected animals. Importantly, total viral N RNA in nasal secretions from the naive cagemates paired with vaccinated-infected hamsters never surpassed 10 ⁴ copies/mL, as shown in FIG.9C, and remained just above the limit of detection starting at day 4 post-re- pairing (p < 0.0001 for all groups). Conversely, the naive cagemates paired with sham- vaccinated-infected hamsters had total viral N RNA nasal discharge titers that peaked at 10 ⁷ copies/mL by day 2 post-re-pairing, with minimal clearance by end of study at day 8 post- re-pairing, as shown in FIG.9C. Similarly, subgenomic viral N RNA in the nasal discharge, which is indicative of viral replication/infection, was detected at 24 hours post-re-pairing in the naive cagemates co-housed with sham-vaccinated-infected hamsters, peaked at 10 ⁵ copies/mL by day 2-3 post-re-pairing and remained detectable even out to day 8 post-re- pairing, as shown in FIG. 9D. This was in striking contrast to the naive cagemates of the vaccinated-infected animals, which had minimal to non-detectable levels of subgenomic viral N RNA in the nasal discharge, as shown in FIG.9D (p < 0.0001 for all three groups). [0178] To assess whether the naive cagemates exposed to the vaccinated-infected animals were protected from active infection, de novo anti-N serum IgG responses were measured. As expected, the naive cagemates co-housed with sham-vaccinated-infected hamsters developed high serum N-specific IgG responses, as shown in FIG.9E. The naive cagemates co-housed with vaccinated-infected hamsters, on the other hand, exhibited negligible responses (p < 0.0001 for all vaccine regimen groups). Indeed, no anti-N IgG response was detected in the naive cagemates of the i.n.-i.n. vaccinated-infected group, and only one naive cagemate from the i.m.-i.n and the i.m.-i.m. vaccinated-infected groups developed a measurable anti-N IgG serum antibody response. Thus, all three vaccine regimens were effective at preventing significant transmission of infectious virus capable of inducing an immune response to naive cagemates. [0179] Next, the total histology score of each group of naive cagemates was compared, as shown in FIG.9F. Naive hamsters co-housed with the sham-vaccinated-infected hamsters exhibited mild to moderate histology scores, primarily due to mild inflammation in the alveolar interstitium and lumen that was similar in severity to that observed in the sham- vaccinated-infected hamsters (not shown). Naive hamsters co-housed with the vaccinated- infected hamsters, regardless of vaccination regimen, exhibited a significantly decreased histology score (p < 0.0001 for all vaccine regimen groups), with minimal lung pathology that was very similar to that observed in the vaccinated-infected cagemates (not shown). Taken together, the data demonstrate that i.m.-i.m., i.n.-i.n., or i.m.-i.n. dosing of the self- amplifying RNA/NLC spike vaccine was not only effective in preventing disease-associated morbidity in hamsters but also prevented the vaccinated and subsequently experimentally infected hamsters from transmitting infectious virus to naive cagemates, even at a time when viral load peaked in the nasal discharge. Non-Human Primate Studies [0180] To establish feasibility of use of the disclosed SARS-CoV-2 intranasal vaccine delivery system in large primates, an i.n. vaccine arm was added to an existing nonhuman primate (NHP) SARS-CoV-2 i.m. self-amplifying RNA/NLC vaccine immunogenicity and efficacy study conducted by Dr. Sudhir Kasturi at the Emory National Primate Research Center using rhesus macaques. This study sought to establish safety, tolerability, and immunogenicity of the proof-of-concept self-amplifying RNA/NLC vaccine. As indicated in Table 1, the study consisted of 15 animals total, 5 each in the i.m.-vaccinated, unvaccinated, and i.n.-vaccinated arms. The i.m. vaccinated animals were dosed using our base self-amplifying RNA/NLC SARS-CoV-2 vaccine formulation (N/P=15), and the i.n. vaccinated animals dosed with a similar self-amplifying RNA/NLC SARS-CoV-2 vaccine formation with a reduced RNA loading ratio for i.n. delivery (N/P=5). After prime-boost vaccination, animals were then challenged with the heterologous Delta strain of SARS- CoV-2 to investigate ability of the vaccine to protect against robust challenge of a further- evolved viral strain. Over the course of the study, as shown in FIG.10, blood, nasal swabs, bronchoalveolar lavage (BAL) and bone marrow samples were taken to assess vaccine immunogenicity and protection against viral infection. Upon conclusion of the study one week post-challenge, study animals were euthanized and tissue samples taken to investigate immune responses and viral loads. Table 1 [0181] The purpose of this study was to assess the immunogenicity and efficacy of the vaccines in NHPs, assess whether i.m. and/or i.n. vaccine preparation protected NHPs from virus-induced morbidity, and to compare the efficacy of the intranasal and intramuscular preparations of the vaccine. Collectively these data demonstrate that i.n. and i.m. vaccination elicit qualitatively different immune responses, but both show protection from challenge in the NHPs. [0182] First, serum immunogenicity was assessed across the study, as illustrated in FIG.11. Both vaccinated groups generated IgG antibodies to the vaccine strain (WA1/2020) as well as the challenge strain (B.1.617.2). Vaccinated animals showed increased titers after prime and boost vaccination, with boost particularly beneficial to boosting the i.n.- vaccinated animals. The i.m. vaccinated animals showed generally higher serum IgG titers than the i.n. vaccinated animals. Titers had dipped slightly by the day of challenge (5 weeks post-boost) but increased by the day of necropsy (7/8 days post challenge, in response to challenge), with both i.m. and i.n. vaccinated animals demonstrating high titers at necropsy. [0183] Antibody levels were also assessed in the mucosal tract, specifically in the BAL and nasal mucosa, as shown in FIG. 12. Only the i.m. vaccinated primates elicited IgG antibodies in either of these compartments prior to infection. However, upon infection the i.n. vaccinated animals peaked to the same BAL IgG antibody levels as seen in the i.m. vaccinated animals. Furthermore, only the i.n. vaccinated animals were capable of eliciting BAL anti-RBD IgA after infection, indicating seeding of a mucosal antibody response in this compartment. Similarly in the nasal mucosa there was no response initially noted in the i.n. vaccinated animals. However, there was indication of a higher IgA response in the i.n. vaccinated animals as well. [0184] Next, virus neutralization potential of animal sera was measured for the same two viral strains for the vaccinated animals, as shown in FIG.13. The intramuscular animals responded clearly with neutralization potential in all animals against the vaccine and challenge strain. However, this was not the case in the i.n. group: just 2/5 animals post- vaccination showed measurable levels of neutralizing antibodies in serum, and only one of those had demonstrated cross-protective neutralizing antibodies to the challenge strain. Any protective responses induced by the intranasal vaccine, therefore, are unlikely due to circulating serum neutralizing antibodies. [0185] However, post-challenge neutralizing antibody data clearly show that the while the i.n. vaccinated animals show lower titers of serum neutralizing antibody prior to challenge than the i.m. vaccinated animals, they are able to robustly respond with antibody post-challenge, as shown in FIG. 13. The control animals were unable to produce serum neutralizing antibodies in this timeframe, despite being challenged with virus, suggesting that the serum response in the i.n. vaccinated animals was a recall response to vaccine- induced memory rather than any de novo response to viral challenge. This suggests further that humoral memory B-cells established in response to the intranasal vaccine exist in a compartment unable to be easily sampled pre-necropsy. [0186] T cell responses to vaccination were also assessed across the study, as shown in FIG.14. During the study, the i.m. vaccinated group showed a dramatic induction in CD4+ and CD8+ T cells found in peripheral blood. No circulating CD4+ T cells were seen in i.n. vaccinated animals, and only transient populations of circulating CD8+ T cells were detected prior to challenge. However, similar to the pattern seen with humoral immune responses, upon viral challenge both CD4+ and CD8+ T cells were detected at high levels in PBMCs in the i.n.-vaccinated animals, at levels that rivaled or exceeded the levels seen in i.m.-vaccinated animals, representing a robust recall response to infection. Unvaccinated animals were again unable to produce detectable T cell responses to viral challenge within this timeframe. These data thus repeat the humoral immunogenicity patterns—while vaccine-induced T cell populations were not seen systemically in i.n.-vaccinated animals post-vaccination, cellular memory responses were clearly established by vaccination, allowing for rapid T cell expansion and response to viral infection. [0187] Lastly, the induction of SARS-CoV-2 spike-binding IgG secreting cells was assessed. Lymph nodes, spleen, lung and bone marrow were all sampled at necropsy, one week after viral challenge, and the presence of IgG+ RBD or spike specific plasma cells was measured by ELISpot, as shown in FIG.15. The unvaccinated animals had essentially no reactive plasma cells, as expected. However, both vaccine strategies seeded significant populations of reactive plasma cells across tissues. While the i.m. vaccinated animals showed significantly higher levels of these plasma cells than i.n.-vaccinated animals in the lymph nodes and bone marrow, i.n. animals still saw significant induction of these cell subsets across all tissues measured indicating robust establishment of vaccine-induced immune memory prior to and capable of responding to infection. [0188] With all of this immunogenicity data on the disparity in the response following i.n. or i.m. vaccination, it was also desired to observe how different vaccine strategies affected susceptibility to infection. Thus the viral load in nasal swabs, throat swabs, and BAL was measured across 7 or 8 days of infection, as shown in FIG.16. All animals showed evidence of productive infection, with rising subgenomic RNA in all sites by day 1 post challenge and peaking viral load on day 2. However, both the i.m. and i.n. vaccinated animals showed clear evidence of faster resolution of infection, with most animals having viral loads return to baseline by day 7/8, unlike the naïve controls which continued. Although a complete statistical analysis is ongoing, it is hypothesized that these data will demonstrate that the i.m. vaccine is reducing peak viral load in the upper and lower respiratory tract and i.n. vaccination is reducing peak viral load in the upper respiratory tract only. MATERIALS AND METHODS Study Design [0189] Research subjects. Research was conducted on C57BL/6J mice obtained from The Jackson Laboratory. Lakeview Golden (LVG) Syrian hamsters were obtained from Charles River Laboratories. [0190] Experimental design. Several mouse vaccine immunogenicity experiments were conducted, involving vaccinating mice i.m and/or i.n. with the above-described SARS-CoV- 2 self-amplifying RNA vaccine. Mice were vaccinated using different doses and dosing strategies (homologous vs. heterologous prime-boost vaccine regimens), and immunogenicity was assessed by ELISA and pseudoneutralization assay on serum samples, T cell intracellular cytokine staining (ICS) on spleen and lung samples, bone marrow B cell ELISpot, and T cell ELISpot to assess systemic antibody responses and induction of specific vaccine-induced T cell and B cell populations. [0191] Randomization. Rodents were obtained commercially and randomized into study groups prior to study onset. [0192] Blinding. Investigators were not blinded to the study groups during data collection and analysis as this is typically unnecessary for this type of inbred animal study and introduces more systemic and human error than it prevents. Quantitative, unbiased assays were developed with highly regulated SOPs to minimize data biases. [0193] Sample size. Statistical analyses were the guiding strategy to maximize statistical power while reducing the number of animals used to a minimum. It was previously determined that 6-8 animals per group (3-4 male, 3-4 female) is the lowest estimate to achieve meaningful comparisons between vaccine candidates with subtle differences in immunogenicity. 10 mice/group, 5 male and 5 female, is required to provide statistical power of 90% with an alpha (p-value) of 0.05 to detect a statistically significant difference in pseudovirus neutralizing antibody titers between vaccinated groups with two-fold differences in mean antibody titers with expected statistical variance. Both male and female mice were used in most study groups to avoid sex-driven data biases and allow for identification of any sex-specific differences in vaccine immunogenicity. For one select study where mouse numbers were limited, as shown in FIG.5), n = 5 all-female mice were used per group to provide statistical power of 80% to detect two-fold differences in serum antibody titers. The use of all female mice in this study was conducted to reduce data variability sufficiently to minimize animal use without compromising the main study immunogenicity readouts. [0194] Data inclusion/exclusion criteria. No data were excluded from any plots or analyses except two cases: (1) for flow cytometry after ICS where <50,000 cells were collected in the well (this was established as a guideline prospectively) and (2) one i.n. vaccinated hamster was removed from FIG. 8 where the complete lack of anti-spike IgG antibody established that it had not been effectively vaccinated. [0195] Outliers. Outliers were not excluded. Assays were repeated on any suspected outliers to verify data prior to inclusion. [0196] Replicates. Replication of data was conducted whenever possible. All experiments were conducted with a minimum of biological triplicates. ELISAs and pseudoneutralization assays were conducted with technical duplicates for each study sample. T cell immunogenicity was only assessed in technical singlicate due to limiting tissue quantity and cell number. Bridging groups were used between independent animal studies to provide replication of key study groups across multiple studies offset by months. All attempts at replication verified the robustness of the scientific approach. Self-Amplifying RNA Expression Plasmid Design, Cloning, and Production [0197] The self-amplifying RNA plasmid was designed and selected based on immunogenicity measures from i.m. immunization. Briefly, the self-amplifying RNA encodes a SARS-CoV-2 spike sequence based on a GenBank sequence (MT246667.1) containing the D614G mutation, a diproline at sites 987-988, and a QQAQ substitution for RRAR in the furin cleavage site (683-686). The sequence was codon-optimized for expression in humans, synthesized by BioXp, and inserted via Gibson cloning into AAHI’s backbone self-amplifying RNA expression vector. A SEAP-expressing plasmid was also used as a control, created similarly with the SEAP sequence in the place of the vaccine antigen. Sanger sequencing was used to confirm plasmid sequences. These were then amplified in Escherichia coli and extracted using Qiagen maxi- or gigaprep kits, linearized with NotI (New England Biolabs), and purified using a Qiagen DNA purification kit. RNA Manufacture [0198] NotI-linearized DNA plasmids containing T7 promoters were used as templates for in vitro transcription (IVT) of self-amplifying RNA for vaccination. An in-house optimized IVT protocol with T7 polymerase, RNase inhibitor, and pyrophosphatase (Aldevron) was used. Next, a Dnase step was used to digest the templates, and Cap0 structures were added to the RNA transcripts with guanylyltransferase (Aldevron), GTP, and S-adenosylmethionine (New England Biolabs). CaptoCore 700 resin (GE Healthcare) was used to chromatographically purify the RNA followed by diafiltration and concentration. self-amplifying RNA material was filtered through a 0.22-µm polyethersulfone filter and stored at -80°C until further use. Self-amplifying RNA size and integrity was characterized by gel electrophoresis, and RNA concentration was quantified by UV absorbance (NanoDrop 1000) and RiboGreen assay (Thermo Fisher).  NLC Manufacture [0199] A mixture of trimyristin (IOI Oleochemical), squalene (Sigma-Aldrich), sorbitan monostearate (Sigma-Aldrich), and the cationic lipid 1,2-dioleoyl-3-trimethylammonium- propane (DOTAP; Corden) was heated at 70°C in a bath sonicator. Separately, polysorbate 80 (Fisher Scientific) was diluted in 10 mM sodium citrate trihydrate and heated to 70°C in a bath sonicator. After all components were in suspension, a high-speed laboratory emulsifier (Silverson Machines) running at 7000 rpm was used to blend the oil and aqueous phases. Particle size of the mixture was further decreased through high-shear homogenization. The colloid mixture was then processed at 30,000 psi for 11 discrete microfluidization passes using an M-110P microfluidizer (Microfluidics). The NLC product was then filtered through a 0.22-µm polyethersulfone filter and stored at 2-8°C until use. Vaccine Complexing and Characterization [0200] Vaccine complexes were produced by combining aqueous RNA at a 1:1 volume ratio with NLC that was pre-diluted in a 10 mM sodium citrate and 20% w/v sucrose buffer. Vaccines were prepared at a nitrogen/phosphate (N/P) ratio of 5-15, representing the ratio of amine groups on the NLC DOTAP to the RNA’s backbone phosphate groups. This complexing resulted in vaccine containing the intended dose of complexed self-amplifying RNA/NLC in an isotonic 10% w/v sucrose, 5 mM sodium citrate solution. The vaccine solution was incubated on ice for 30 minutes after mixing to ensure complete complexing. Mouse Studies [0201] All mouse studies were conducted according to the Bloodworks Northwest Research Institute’s Institutional Animal Care and Use Committee. All mouse work was in compliance with all applicable sections of the Final Rules of the Animal Welfare Act regulations, see 9 C.F.R. Parts 1, 2, and 3, and the Guide for the Care and Use of Laboratory Animals. [0202] C57BL/6J mice purchased from The Jackson Laboratory were used for all mouse studies. Mice were between 6 and 8 weeks of age at study onset and evenly split between male and female. At arrival, mice were simultaneously randomized into study groups, half male and half female, housed separately 3-5 mice per cage with irradiated bedding in Allentown individually ventilated cages. Sterile water gel packs were given to all newly arrived animals during the acclimation period, and animals had access to fresh potable water ad libitum via an Edstrom automatic reverse-osmosis and chlorinated water system. Cages were changed every 2 weeks inside a laminar flow cage change station. A spot check for dirty cages was conducted on the non-change weeks. [0203] Mice were observed daily for 3 days post-injection after both prime and boost injections by trained personnel. Isoflurane delivered by inhalation was used for temporary anesthesia immediately prior to retro-orbital blood collection and i.n. immunization to mitigate animal pain and distress. At study completion, animals were humanely euthanized by CO2 overdose. [0204] Animal handling technicians were not blinded to study groups. After 1 week of acclimatization to the facility, mice were then primed with vaccine. Mice were then boosted 3 weeks later and harvested 3 weeks after boost. Mouse weights were monitored for 4 days following prime and boost vaccinations. Mice were immunized by i.m. injection in both rear quadriceps muscles (50 µL/leg, 100 µL total) or by i.n. inoculation in two 25-µL doses. Intranasal inoculation was done under full anesthesia, and mice were allowed to recover under a heat lamp. During ongoing studies, serum samples were taken by retro-orbital bleed. Three-weeks post-boost, mice were harvested, terminal bleeds were taken, spleens and lungs were dissected, and femurs were harvested for bone marrow purification. Mouse Serum IgG, IgG1, and IgG2a Titers by ELISA [0205] SARS-CoV-2 spike-specific IgG was measured by ELISA. Plates (Corning 384- well high-binding #CLS3700) were coated with Recombinant SARS-Cov-2 Spike His Protein, Carrier Free (R&D Systems #10549-CV), at 1 µg/mL and incubated overnight at 4°C. Plates were incubated in blocking buffer (2% dry milk, 0.05% Tween 20, 1% goat serum) for over 1 hour. Samples were plated on low-binding plates, starting at a dilution of 1:40 and then serially diluting 1:2 across the plate. High-binding plates were washed, and samples were then transferred from the low-binding plates to the coated, high-binding plates. Naive mice serum at 1:40 was used as a negative control. A SARS-CoV-2 neutralizing monoclonal antibody (GenScript #A02057) was used as a positive control at a starting concentration of 3.2 ng/µL. Following a 1 hour incubation, plates were washed, and a secondary antibody, Anti-Mouse IgG (Fc Specific)-Alkaline Phosphatase antibody (Sigma-Aldrich #A2429) at a 1:4000 dilution was added. Following the secondary antibody and another wash step, phosphatase substrate tablets (Sigma-Aldrich #S0942) were dissolved in diethanolamine substrate buffer (Fisher Scientific #PI34064) at a final concentration of 1 mg/mL and added to the plates. Plates were incubated for 30 minutes and then read spectrophotometrically at 405 nm. A 4-point logistic (4PL) curve was used to fit the antibody standard titration curve, and sample concentrations were interpolated off the linear region of the curve. [0206] For IgG1 and IgG2a isotype-specific ELISAs, plates were coated and blocked as described above. For IgG1 and IgG2a, the standard curve was run using SARS-CoV-2 neutralizing antibodies (GenScript #A02055 and #BS-M0220, respectively). Sample dilution and incubation were identical to the total IgG curve, and plates were probed with IgG1- and IgG2a-specific secondary alkaline phosphatase (AP)-conjugated detection antibodies (Sigma-Aldrich #SAB3701172 and #SAB3701179, respectively) prior to development, reading, and quantification as described above. Pseudovirus Neutralization Assay [0207] SARS-CoV-2 pseudovirus neutralizing antibody titers in mouse sera were measured via a pseudoneutralization assay. Lentiviral SARS-CoV-2 spike protein pseudotyped particles were prepared by co-transfecting HEK-293 cells (American Type Culture Collection CRL #11268) with plasmids containing a lentiviral backbone-expressing luciferase and ZsGreen (BEI Resources #NR-52516), lentiviral helper genes (BEI Resources #NR-52517, NR-52518, and NR-52519), or a delta19 cytoplasmic tail-truncated SARS-CoV-2 spike protein (Wuhan strain plasmids from Jesse Bloom of Fred Hutchinson Cancer Center). Cells were incubated for 72 hours at standard cell culture conditions (37°C, 5% CO ₂ ), and cell culture media were harvested and filtered through a 0.2-μm filter. Pseudovirus-containing supernatant was frozen until titering and use. [0208] To perform the assay, plates were seeded with Human Angiotensin-Converting Enzyme 2 (hACE2)-expressing HEK-293 cells (BEI Resources #NR52511) and incubated overnight. Serum samples were pre-diluted 1:10 in media (Gibco Dulbecco’s Modified Eagle Medium + GlutaMAX + 10% fetal bovine serum [FBS]) and then serially diluted 1:2 for 11 total dilutions. These were then incubated with polybrene (Sigma-Aldrich #TR-1003- G) and pseudovirus for 1 hour. Then, the serum samples were added onto the hACE2 cells in duplicate and incubated at 37°C and 5% CO ₂ for 72 hours. To determine 50% inhibitory concentration (IC50) values, plates were scanned on a fluorescent imager (Molecular Devices ImageXpress Pico Automated Cell Imaging System) for ZsGreen expression. Total integrated intensity per well was used to calculate the percent of pseudovirus inhibition per well. Neutralization data for each sample were fit with a 4-parameter sigmoidal curve, which was used to interpolate IC ₅₀ values. Spleen, Lung, and Bone Marrow Cell Harvest [0209] Spleens were prepared in 4 mL of RPMI medium by manual maceration through a filter using the back of a syringe. Dissociated splenocyte samples were centrifuged briefly at 400 x g to pellet fat cells, and the supernatants containing lymphocytes were either transferred to 5-mL mesh-cap tubes to strain out any remining tissue debris or lysed with ammonium-chloride-potassium (ACK) lysing buffer. Cells were counted on a Guava easyCyte cytometer (Luminex) and plated in round-bottom 96-well plates at 1-2 x 10 ⁶ cells per well in complete RPMI (cRPMI) medium containing CD28 costimulatory antibody (BD Biosciences #553294) and brefeldin A. One of three stimulation treatments was added to each well: 0.0475% dimethyl sulphoxide [DMSO], 2 μg/μL spike peptide pool (JPT peptides #PM-WCPV-S-1), or phorbol myristate acetate (PMA)-ionomycin. Plates were incubated for 6 hours at 37 C with 5% CO2. [0210] Lung cells were isolated via enzymatic digestion using a gentleMACS Dissociator (Miltenyi Biotec). Lungs were dissociated in 4 mL of Hanks’ Balanced Salt Solution (HBSS) supplemented with 10% Liberase (MilliporeSigma), 10% aminoguanidine, 0.1% KN-62, and 1.25% Dnase. Lungs and enzymatic mix were added to a gentleMACS M tube (Miltneyi Biotec), and the m_lung_01.02 program was run. Samples were then incubated at 37°C and 5% CO ₂ for 30 minutes. Directly after, samples were run again on the gentleMACS Dissociator using the m_lung_02.01 program. The resulting slurry was added to 10 mL of RPMI medium and centrifuged for 5 minutes. The supernatant was discarded, and the cells were either filtered through a 5-mL snap-top tube or washed again in RPMI medium prior to counting on a Guava easyCyte cytometer and plated in round-bottom 96-well plates at 1-2 x 10 ⁶ cells per well in RPMI medium containing 10% FBS, 50 μM beta-mercaptoethanol, CD28 costimulatory antibody, brefeldin A, and one of three stimulation treatments: 0.0475% DMSO, 2 μg/μL spike peptide pool (JPT peptides #PM-WCPV-S-1), or PMA-ionomycin. Plates were incubated for 6 hours at 37 C with 5% CO2. Intracellular Cytokine Staining and Flow Cytometry [0211] Following incubation, plates were centrifuged at 400 x g for 3 minutes, the supernatants were removed, and cells were resuspended in phosphate-buffered saline (PBS). Splenocytes were stained for viability with Zombie Green (Biolegend) in 50 μL of PBS. Cells were washed twice, and then spleen samples were incubated with CD16/CD32 antibody (Invitrogen #14-0161-86) to block Fc receptors. Next, the cells were surface stained in an additional 50 μL of staining buffer (PBS with 0.5% bovine serum albumin and 0.1% sodium azide). Spleen samples were stained with fluorochrome-labeled monoclonal antibody (mAb) specific for mouse CD4 (eBioscience #45-0042-82), CD8 (BD Biosciences #563068), CD44 (BD Biosciences #560568), and CD107a (BioLegend #121614), while lung samples were stained for CD4 (BioLegend #100526), CD8 (BD Biosciences #563068), CD44 (BD Biosciences #562464), CD69 (BioLegend #104508), CD103 (BioLegend #121435), CD194 (BioLegend #131220), and CD154 (BD Biosciences #745242). Cells were washed twice, permeabilized using the Fixation/Permeabilization Kit (BD Biosciences #554714), and stained intracellularly. Spleens were stained for TNFα (BioLegend #506327), IL-2 (BioLegend #503824), IFNγ (Invitrogen #25731182), IL-5 (eBioscience #12-7052-82), IL-10 (BD Biosciences #564081), and IL-17a (BD Biosciences #560820), while lungs were stained for TNFα (BioLegend #506327), IL-2 (BioLegend #503824), IFNγ (Invitrogen #25731182), IL-5 (BioLegend #504306), and IL-17a (BD Biosciences #560820). After two washes in staining buffer, cells were resuspended in 100 μL of staining buffer and analyzed using an LSRFortessa flow cytometer (BD Biosciences). After initial gating for live CD4 ⁺  or CD8 ⁺  lymphocytes, cells gating positive for all three activation markers IL-2, IFNγ, and TNFα were counted as activated polyfunctional T cells. Bone Marrow Harvest and ASC ELISpot [0212] Antibody-secreting bone marrow-resident cell counts were measured by ELISpot. MultiScreenHTS IP Filter plates (0.45 µm, MilliporeSigma) were treated with 15 µL of 35% ethanol for 30 seconds. Recombinant SARS-CoV-2 Spike His Protein, Carrier Free (R&D Systems #10549-CV-100), was diluted at 2 µg/mL in ELISpot coating buffer (eBioscience), and plates were coated with 100 µL. Plates were incubated overnight at 4°C. The next day, plates were washed with PBS with 0.1% Tween 20 and blocked with cRPMI medium for 2 hours. [0213] Femurs were removed from mice and inserted into a snipped-end 0.6-mL Eppendorf tube and then inserted into a 1.5-mL Eppendorf tube with 1 mL of cRPMI medium. Femurs were centrifuged for 15 seconds at 10,000 rpm, and supernatant was discarded. Cell pellets were vortexed, resuspended in 200 µL of RBC Lysis Buffer (Invitrogen), and then incubated on ice for 30 seconds. An additional 800 μL of RPMI medium was added, and cells were centrifuged for 5 minutes at 400 x g before the supernatant was decanted. Cells were resuspended in 1 mL of cRPMI medium, counted, and transferred to prepared filter plates at 1 million cells per well followed by a 3-fold dilution across five adjacent wells. [0214] Plates were incubated for 3 hours, then washed three times with PBS plus 0.1% Tween 20. Secondary antibody (Goat Anti-Mouse IgG-HRP or IgA-HRP [SouthernBiotech #1030-05 and #1040-05, respectively]) was added at a 1:1000 dilution in PBS with 0.1% Tween and 5% FBS overnight at 4°C. Plates were then washed three times in PBS with 0.1% Tween 20 and two times in PBS. 100 µL of Vector NovaRED Substrate Peroxidase (Vector Laboratories #SK-4800) was applied for 7 minutes to develop the plates. The reaction was quenched by rinsing plates with distilled water for 2 minutes, and plates were dried in the dark. Spots were counted and data were analyzed using ImmunoSpot v7 software (Cellular Technology Limited). T Cell ELISpot Assay [0215] IFNγ (BD Biosciences #51-2525KZ), IL-17A (Invitrogen #88-7371-88), or IL- 5 (BD Biosciences #51-1805KZ) at a 1:200 dilution in Dulbecco’s PBS (DPBS; Gibco) was used to coat ELISpot plates (MilliporeSigma). Plates were incubated overnight at 4°C and then washed and blocked with cRPMI medium for at least 2 hours. Previously harvested splenocytes (see above) were added at 2 x 10 ⁵ cells per well. Samples were then stimulated with 1 µg/mL PepMix SARS-CoV-2 (JPT Peptide Technologies #PM-WCPV-S-1). Plates were incubated at 37°C and 5% CO2 for 48 hours. Then plates were washed with PBS with 0.1% Tween 20. Next, detection antibody (IFNγ, BD Biosciences #51-1818KA; IL-17A, Invitrogen #88-7371-88; and IL-5, BD Biosciences #51-1806KZ) was diluted in ELISpot diluent (eBiosciences) at 1:250 and added overnight at 4°C. Plates were washed and then developed using Vector NovaRED Substrate Peroxidase for 15 minutes. The reaction was quenched with deionized water. Plates were then dried in the dark. Spots were counted and data were analyzed using ImmunoSpot. Hamster Studies [0216] Nine-week-old male Lakeview Golden (LVG) Syrian hamsters were purchased from Charles River Laboratories and maintained in the University of Alabama at Birmingham (UAB) animal facilities, which include the UAB Southeastern Biosafety Lab (SEBLAB) ABSL3 facilities. All hamster procedures were approved by the UAB Institutional Animal Care and Use Committee (IACUC protocol 22628) and the UAB Biosafety Committee (Protocol 22-160 and 22-161). All procedures were performed in accordance with National Resource Council guidelines. Hamster Vaccination and Infection [0217] Adult male hamsters were housed in pairs and separated at the time of vaccination. One hamster from each pair was vaccinated with 5 mg of self-amplifying RNA/NLC delivered i.m. (50 mL per rear quadriceps muscle, 100 mL total) or i.n. (25 mL/nare, 50 mL total). Sham-vaccinated hamsters received dPBS i.m. Vaccinated and sham-vaccinated animals were re-paired with naive cagemates after 24 hours. Cagemates were separated on day 28, and vaccinated animals were boosted with the same vaccine given via the i.m. or i.n. route. Animals were re-paired after 24 hours. Nine days after the boost, animals were separated, and the vaccinated and sham-vaccinated hamsters were infected i.n. with 1 x 10 ⁶ pfu SARS-CoV-2 USA/WA-1 variant (50 mL/nare, 100 mL total). Uninfected controls received 100 mL (50 mL/nare) of dPBS (vehicle). The infected animals were re- paired with the uninfected, unvaccinated naive cagemates at 48 hours post-infection. Cagemates were separated 24 hours later and remained apart until the end of the study. Hamster weights were measured before infection and daily until day 15 post-infection. Hamster Blood Collection and Processing [0218] Blood samples from sedated hamsters were collected from the lateral saphenous vein on day 21 post-prime and 5 days post-boost vaccinations. Blood samples following euthanasia were collected from terminal bleed via the posterior vena cava. Blood samples were transferred into BD Microtainer blood collection tubes (BD Biosciences), and serum was separated by centrifugation at 10,000 rpm at RT for 10 minutes. Serum samples collected post-infection were heated at 60°C for 20 minutes to inactivate SARS-CoV-2 virus. All samples were aliquoted and frozen at -80°C until analyzed by cytometric bead array (CBA). Hamster Nasal Viral Load Samples [0219] Nasal swipes were collected immediately prior to i.n. infection to establish baseline, at 1 hour post-infection, and then daily until day 10 post-infection. Awake hamsters were transferred into an open container to minimize movement, and the exterior surface of their nose was swiped for 5-10 seconds with a polyester-tipped swab (Medical Packaging #SP-7D) that had been dipped in a screw-cap tube containing 300 mL of viral transport medium (VTM; HBSS (+Ca ²⁺ +Mg ²⁺ ) containing 2% FBS, 100 mg/mL gentamicin, and 0.5 mg/mL amphotericin B). The tip of the swab was cut and placed inside the VTM-containing tube. The tubes were vortexed, and swab tips were discarded. Samples were then processed for viral RNA quantitation. Hamster Nasal Wash Samples [0220] Hamsters were sedated with isoflurane and euthanized with an intraperitoneal injection of tribromoethanol (500 mg/kg, 3 mL/animal). The trachea was exposed, an incision was made, and a blunt 19-gauge needle attached to a 1-mL insulin syringe was inserted into the trachea. While holding the mouth of the hamster shut with one hand, 400 mL of VTM was squeezed through the nasal passages, out of the nose, and into a screw-cap tube. Samples were heated at 60°C for 20 minutes to inactivate SARS-CoV-2 virus, then aliquoted, and stored at -80°C until analyzed by CBA. Recombinant SARS-CoV-2 Spike, RBD, and N Protein Production [0221] Recombinant SARS-CoV-2 spike ectodomain trimeric proteins were prepared as previously described. See King, R. G., et al. “Single-Dose Intranasal Administration of AdCOVID Elicits Systemic and Mucosal Immunity against SARS-CoV-2 and Fully Protects Mice from Lethal Challenge,” Vaccines (Basel), 2021, 9, 881. Briefly, two human codon-optimized constructs were generated with a human IgG leader sequence, the SARS- CoV-2 spike ectodomain (amino acids 14-1211), a GGSG linker, T4 fibritin foldon sequence, a GS linker, and a 15 amino acid biotinylation consensus site (AviTag, construct 1) or 6X-HisTag (construct 2). Each construct was engineered with two sets of mutations to stabilize the protein in a pre-fusion conformation. Recombinant SARS-CoV-2 spike RBD domain monomers (human codon-optimized) were generated with a human IgG leader sequence, the spike RBD (amino acids 319-541 from SARS-CoV-2 Wuhan-Hu-1 strain), a GS linker, an AviTag, and a 6X-HisTag. The coding sequence of N from SARS-CoV-2 Wuhan-Hu-1 strain was synthesized in frame with the coding sequence for an AviTag and was cloned in frame to the 6X-HisTag in the pTrcHis2C expression vector (Invitrogen). Biotinylated recombinant N protein was produced by co-transforming Rosetta cells with the N expression plasmid and an inducible BirA expression plasmid. Bacteria were grown in the presence of chloramphenicol, ampicillin, and streptomycin, induced with IPTG, and supplemented with biotin. Recombinant SARS-CoV-2 Avi/His-tagged spike trimeric proteins were produced by co-transfecting each plasmid construct (1AVI:2HIS construct ratio) into FreeStyle 293-F cells. The Avi-His-tagged spike RBD protein was produced by transfecting the single dual-tagged RBD construct into FreeStyle 293-F cells. All recombinant proteins were purified by FPLC using a nickel-affinity column and subsequent size exclusion chromatography. After buffer exchange, purified spike ectodomain trimers and RBD domain protein were biotinylated by the addition of biotin-protein ligase (Avidity). Biotinylated proteins were buffer exchanged into PBS, sterile filtered, aliquoted, and stored at -80ºC until used. SARS-CoV-2 Spike Cytometric Bead Array [0222] CBAs with recombinant SARS-CoV-2 proteins were prepared as previously described. See King, et al., infra. Briefly, recombinant proteins were passively absorbed onto streptavidin functionalized 4-µm fluorescent microparticles (Carboxy Blue Particle Array Kit, Spherotech).500 µg of biotinylated recombinant protein was incubated with 2 x 10 ⁷ streptavidin functionalized fluorescent microparticles in 400 µL of 1% BSA in PBS. Protein coupled beads were washed and resuspended at 1 x 10 ⁸ beads/mL and stored at 4°C. Cytometric Bead Array Measurement of Spike-Specific IgG Responses [0223] 50 µL of serum (diluted to 1/4000 in PBS) or nasal wash (diluted 1/4 in PBS) was arrayed in 96-well u-bottom polystyrene plates.5 µL of a suspension containing 5 x 10 ⁵ spike beads, RBD beads, and N beads was added to the samples. Suspensions were mixed, incubated for 15 minutes at RT, washed in PBS, stained with FITC-conjugated polyclonal anti-hamster IgG (Southern Biotech) for 15 minutes at RT, washed in PBS, and then resuspended in 100 µL of 1% paraformaldehyde in PBS. Samples were collected on a BD CytoFLEX flow cytometer in plate mode at a sample rate of 100 µL/minute for 1 minute. Following acquisition, the FCS files were analyzed in FlowJo (TreeStar). Briefly, the beads were identified by gating on singlet 4-µm particles in log scale in the forward scatter and side scatter parameters. APC-Cy7 channel fluorescence gates were used to segregate the particles by bead identity. Geometric mean fluorescent intensity was calculated in the FITC channel. Propagation and Titer Determination of SARS-CoV-2 [0224] The original SARS-CoV-2 isolate USA-WA1/2020 was obtained from BEI Resources (#NR-52281), propagated in Vero E6 cells (ATCC #C1008), and titered as previously described (84). Viral RNA Quantitation by qRT-PCR [0225] SARS-CoV-2 viral N RNA and viral subgenomic N RNA was measured by PCR in nasal swipe (300 µL) samples as previously described. See Schultz, M. D., et al. “A Single Intranasal Administration of AdCOVID Protects Against SARS-CoV-2 Infection in the Upper and Lower Respiratory Tracts,” Human Vaccines & Immunotherapeutics, 2022, 0, 2127292. AccuPlex SARS-CoV-2 Reference Material (SeraCare Life Sciences, Inc., # 0505-0126) was extracted and amplified in parallel to generate a standard curve enabling viral N quantitation. The subgenomic RNA standard was constructed and validated in- house. Histopathology [0226] Hamster lungs were inflated with 10% neutral-buffered formalin (Thermo Fisher) using a blunt 19-gauge needle. Each lung was excised en bloc and fixed for 7 days at RT in 15-fold volume of 10% neutral-buffered formalin. The fixed tissues were then embedded dorsal side down in paraffin using standard procedures. Samples were sectioned at 5 µm, and resulting slides were stained with hematoxylin and eosin (H&E). All tissue slides were evaluated by light microscopy by a board-certified veterinary pathologist blinded to study group allocations. Representative photo images were collected using a Nikon Eclipse Ci microscope (Nikon Inc) and analyzed with NIS-Elements software (Nikon Inc). Scoring was performed following a published algorithm. See Carroll, T., et al. “The B.1.427/1.429 (epsilon) SARS-CoV-2 Variants Are More Virulent Than Ancestral B.1 (614G) in Syrian Hamsters,” PLOS Pathogens, 2022, 18, e1009914; Gruber, A.D., et al. “Standardization of Reporting Criteria for Lung Pathology in SARS-CoV-2-Infected Hamsters: What Matters?” Am. J. Respir. Cell. Mol Biol. 2020, 63, 856-859. Briefly, the airway, alveolar, and vascular pathology in affected areas were assessed and scored on a scale of 0-4, see Schultz, et al., infra, and then added together to generate a total histopathological score of the whole lung. Statistical Analysis [0227] The effect of dosing strategy on CD4 and CD8 cell polyfunctionality was assessed by one-way ANOVA with Tukey’s multiple comparisons test. The effect of dosing strategy on IgG titer and pseudovirus neutralizing antibody titers was assessed in log- transformed data using a mixed-effects model with Tukey’s multiple comparisons test. The effect of immunization strategy on IgG titer was assessed on log-transformed data using unpaired t tests or two-way ANOVA with Tukey’s or Sidak’s (if any sample data were missing) multiple comparisons test. The effect of immunization strategy on pseudovirus neutralizing antibody titers was assessed on log-transformed data using unpaired t tests or a mixed-effects model with Sidak’s multiple comparisons test. The effect of immunization strategy on polyfunctional CD4 and CD8 cell responses was assessed in untransformed data using unpaired t tests or one-way ANOVA with Holm-Sidak’s multiple comparisons test. Lung, bone marrow, and splenocyte ELISpot data were analyzed using log-transformed data by one-way ANOVA with Tukey’s multiple comparisons test. Hamster body weight data was analyzed using area under the curve measures assessed statistically by one-way ANOVA with Tukey’s multiple comparison test. Hamster viral load measures were analyzed using log-transformed area under the curve measures, assessed with one-way ANOVA with Tukey’s multiple comparison test. All statistical analyses were conducted using Prism version 9 (GraphPad Software). Optimization of Vaccine Particle Size and Charge for Intranasal Delivery [0228] Particle size and surface charge are optimized for intranasal delivery of the self- amplifying RNA/NLC vaccine complex described herein. The size and charge of the complex is dictated by the ratio of the number of positively charged amine groups on the DOTAP lipid contained in the NLC to the number of negatively charged phosphate groups from the self-amplifying RNA backbone (N/P ratio). To understand how particle size and surface charge affect the delivery efficiency of the drug product, a range of N/P ratios were tested, and the particle size, polydispersity, zeta potential, particle concentration, and protection of the self-amplifying RNA from degradation by RNase A were characterized. These experiments were performed in conjunction with the H5N1 vaccine immunogenicity testing experiments described below. Effect of N/P Ratio on Particle Size and Polydispersity [0229] To investigate the effect of the N/P ratio on particle size, complexes were formed at N/P ratios ranging from N/P = 0.6 to N/P = 15. N/P = 15 was the highest N/P ratio selected as a historical starting point due to previous intramuscular dosing studies conducted by the Applicant. An N/P = 0.6 was previously determined to be the point at which the NLCs are saturated with self-amplifying RNA. Below this point, significant levels of free (non- complexed) self-amplifying RNA are detectable by agarose gel electrophoresis. [0230] Particle size and polydispersity was first assessed using dynamic light scattering (DLS). Immediately apparent in the intensity size distribution is the presence of two major species: a smaller species (~40 nm) corresponding to free NLCs, and a larger species (~130 nm) corresponding to the final self-amplifying RNA/NLC complex, as shown in FIG.17A. As the N/P ratio is decreased from 15, the particle size distribution profile shifts towards larger species, reflecting a decrease in the amount of free NLC in solution as it becomes complexed with the increasing amount of RNA. At N/P = 1, multiple large peaks greater than 1 µm are observed, and the two peaks corresponding to free NLC and the self- amplifying RNA/NLC complex are largely absent. This is likely due to the formation of large insoluble aggregates, as will be discussed further below. When the N/P ratio is further lowered below N/P = 1, the size distribution shifts towards a smaller size and the large aggregate peak greater than 1 µm is no longer present. This indicates that the previously aggregated self-amplifying RNA/NLC complexes are now fully solubilized by the RNA – visually the turbidity of the formulation is substantially decreased below N/P = 1. This trend in particle size becomes readily apparent when plotting the Z-average diameter as a function of the N/P ratio, as shown in FIG.17B. Sample polydispersity remained constant across all N/P ratios at a value of approximately 0.4 apart from N/P = 1, which increased to a value of nearly 0.8, reflecting the increased degree of nanoparticle aggregation at this ratio, as shown in FIG.17C. Effect of N/P Ratio on Apparent Particle Size and Concentration using Nanoparticle Tracking Analysis [0231] Particle size and concentration as a function of N/P ratio was further evaluated using nanoparticle tracking analysis (NTA). In contrast to DLS, which records an ensemble measurement of the solution, NTA tracks individual nanoparticles, providing increased resolution of the self-amplifying RNA/NLC complex size distribution. This affords us greater insight as to the true state of the complexes once formed. In contrast to DLS, the size distribution profiles remained relatively similar in shape at all N/P ratios tested, the only difference being the height of the peak at approximately 50 – 70 nm in diameter, as shown in FIG. 18A. This peak can be attributed to non-complexed NLC. Like DLS, the particle concentration of this peak decreases with decreasing N/P ratio, reflecting the fact that less NLCs are free in solution as they are used to form complexes with the increasing amount of RNA. This is also reflected in the mean particle size - as the N/P ratio is decreased, mean particle size begins to increase, as shown in FIG. 18B. However, unlike with DLS, no apparent increase in particle size is observed at N/P = 1 by NTA. [0232] Particle concentration also decreases with decreasing N/P ratio, as shown in FIG. 18C. This is unsurprising considering that to achieve a lower N/P ratio at a constant RNA concentration, the concentration of NLCs used must be decreased. Interestingly, at N/P = 1 very few particles are detected, such that the particle concentration nearly appears to be zero, however, very large aggregate species are clearly detectable by DLS at this ratio. A possible explanation for this is that at N/P = 1, RNA and NLC molecules are forming very large and complex networks of crosslinked material beyond the range of detection of the instrument, leaving only a few particles that are small enough to be detected by NTA. Alternatively, at N/P = 1 the net surface charge of the complex is approaching neutral, leading to physical precipitation of the nanoparticles out of solution. This appears to be the more likely explanation since of the few particles that were observed by NTA, a very small fraction of those were approaching the 1 µm range. At N/P = 0.6 the particle concentration is increased compared to N/P = 1, indicating that some threshold has been surpassed whereby the RNA now sufficiently saturates the NLC surface to allow for charge repulsion and solubilization of the once aggregated self-amplifying RNA/NLC complex. However, complete saturation of the NLC surface may have potential implications on the efficiency of RNA protection from enzymatic degradation by RNase A, as will be discussed further below. Relationship Between Zeta Potential and Protection of RNA from Enzymatic Degradation [0233] To assess the effect of the N/P ratio on complex zeta potential, self-amplifying RNA/NLC complexes were formed at N/P ratios ranging from N/P = 0.6 to N/P = 15. The zeta potential of the complexes was then measured after diluting 100-fold with water, as shown in FIG. 19A. The complexes maintained a constant zeta potential of approximately +12 mV from N/P = 15 to N/P = 5. At N/P = 1 the zeta potential dropped substantially to a value of nearly -40 mV, and at N/P = 0.6 where the RNA is fully saturating the NLCs, the zeta potential dropped further to nearly -60 mV. As discussed above, it is likely that at around N/P = 1 the RNA fully masks the surface of the NLCs and thus the dominant factor determining the surface charge of the complex. A highly cooperative effect of RNA binding to NLC begins to take place, which could result in a significant fraction of the particles having a net neutral charge at N/P = 1, contributing to the large increase in particle aggregation and decreased solubility at this ratio. Below N/P = 1, more RNA is present on the NLC surface, further decreasing the charge. This is consistent with the RNA effectively solubilizing the NLCs, which would simultaneously have the effect of increasing the apparent particle concentration below N/P = 1, as discussed above. Based on the zeta potential, a decrease in delivery efficiency of the RNA is expected at N/P = 1 and below due to charge repulsion between the complexes and cell membranes, which are largely anionic in nature. [0234] Protection of the complexed RNA at different N/P ratios was assessed by treating complexes with a small amount of RNase A at room temperature for 30 minutes, extraction of RNA from the complexes, followed by agarose gel electrophoresis and visualization with ethidium bromide, as shown in FIG. 19B-C. Protection is equally maintained from N/P = 15 down to N/P = 5, however protection is completely lost at N/P = 1 and below. This rapid loss in protection occurs simultaneously with the rapid switch in zeta potential from a positive to a negative value when the N/P ratio is decreased to a value of 1 and below, indicating that RNA protection depends on the particle charge. Taken together with the information gained from size and particle concentration analysis, this is consistent with a mechanism by which the RNA overcrowds the NLC surface in the final complex, resulting in an increase the amount of RNA that is extending away from the NLC surface and left exposed to degradation by RNases in solution. Overall, these data suggest that an N/P ratio of about 3 or greater is beneficial, as this will protect the RNA, maintain the positive zeta potential necessary for cellular uptake, and limit the formation of very large aggregate species. Cryo-TEM Imaging of self-amplifying RNA/NLC Complexes [0235] Cryo-TEM was used to visualize the appearance of NLCs as well as the self- amplifying RNA NLC complex, as shown in FIG. 20. Non-complexed NLCs show a size of approximately 40-50 nm in diameter, with some heterogeneity in the particle morphology that is to be expected for NLCs, as shown in FIG. 20A. NLCs complexed with self- amplifying RNA at N/P = 15, as shown in FIG.20B, visually appear indistinguishable from non-complexed NLCs. However, the appearance of a second, larger peak of approximately 130 nm in the DLS intensity size distribution which does not appear in the size distribution of the bulk NLC material indicates that the self-amplifying RNA is forming a complex with the NLC at this ratio. This would suggest that the bulk of the NLCs at N/P = 15 are not in complex with the self-amplifying RNA; that is, there is a large excess of free NLC in solution. It should be noted that DLS measurements are easily skewed by the presence of large particles, which makes it appear as though there are more large particles present in the media than there are in reality. When the N/P ratio is decreased to N/P = 5, dark string-like material may be observed protruding from the surfaces of the NLCs—this is the self- amplifying RNA becoming visible as it is no longer saturated by the NLCs, as shown in FIG.20C. Further decreasing to N/P = 0.6 allows the self-amplifying RNA to fully saturate the NLCs, an expansive RNA “shell” around the NLC surface, as shown in FIG.20D. [0236] At N/P = 5 the RNA ais protruding away from the RNA surface to a high degree and visually appears largely similar to complexes formed at N/P = 0.6. However, it is known that at N/P = 5 RNA protection is maintained and the zeta potential remains positive, whereas at N/P = 0.6 RNA protection is completely lost and the zeta potential has now decreased to nearly -60 mV. A likely explanation for this is that at N/P = 5, there is still a significant population of non-complexed (free) NLCs that are not readily apparent by TEM imaging, but still serve to confer protection of the RNA from enzymatic degradation, possibly by denaturing RNase A through a hydrophobic and/or electrostatic means. At N/P = 0.6, no NLCs remain free in solution. As a result, protection of the RNA from enzymatic degradation is completely lost, as well as the contribution of the excess DOTAP from the free NLCs to the zeta potential. This may also partially explain the decreased potency of the vaccine at low N/P ratios, as there is no longer sufficient positive charge present in the final complexed product to assist in efficient transfection of cells with self-amplifying RNA for antigen expression. This will be further discussed below. H5N1 Studies [0237] The H5N1 influenza construct was produced using the HA sequence from H5N1, A/Vietnam/1203/2004 (Genbank: EU122404) inserted into an optimized VEEV-based self- amplifying RNA replicon, as shown in FIG. 21. The sequence was produced using an optimized in vitro transcription protocol. The resulting RNA construct is complexed to the exterior of a nanostructured lipid carrier (NLC) described herein to create the final vaccine product. Initial H5 Vaccine Immunogenicity Confirmation [0238] To confirm whether an H5 influenza vaccine induces similar immune profiles to the intranasal SARS-CoV-2 spike vaccine described herein, initial vaccine immunogenicity was tested at the previously-used N/P ratio of 15, with a 5 µg dose/mouse as previously demonstrated to be optimal for the SARS-CoV-2 i.n.-delivered self-amplifying RNA/NLC vaccine. While this delivery formulation is known to show some signs of reactogenicity, as indicated by transient mouse weight loss, attributed to the NLC component, it is clearly functional as an intranasal RNA delivery formulation as described herein. [0239] Each study group included ten 6- to 8-week-old C57BL/6 mice, evenly split between males and females, as shown in Table 2. Each study group was administered the vaccine at doses indicated in FIG. 22 at day 0 (prime vaccination) and day 22 (boost vaccination) of the study. Blood was collected by the retro-orbital route every 7 days. Each mouse was individually ear-tagged and tracked across the entire study. Post-prime and post- boost serum was analyzed for IgG titer by ELISA and pseudovirus neutralization titers. Bone marrow was collected at harvest to measure B cell IgG secretion, while spleens and lungs were isolated for T cell flow cytometry. Experimental Groups for Initial N/P Dosing Comparison Table 2 [0240] To assess the induction of serum immunogenicity, H5-binding serum IgG titers were measured post-prime (day 21) and post-boost (day 37), as shown in FIG.23A. Serum IgG was detectable at high levels in animals vaccinated with the H5 self-amplifying RNA/NLC vaccine, both intramuscularly and intranasally, with serum IgG at >10 ⁵ ng/mL post-boost, with no significant differences between the two dosing strategies. Pseudovirus neutralization assays indicated that post-prime, i.m. vaccine dosing was able to generate IC50s greater than 1:100, while minimal responses were seen post-prime in the i.n. vaccinated group with a minimal detectable response was seen in a single animal, as shown in FIG. 23B. This is generally consistent with the SARS-CoV-2 vaccine responses, where i.n. priming induced significantly lower serum antibody titers than i.m. dosing, differences that disappeared after a boost dose. Similarly, for this H5 vaccine, both dosing strategies induced post-boost serum IC50 averages > 1:1000. This indicates the i.n. dosing of this vaccine construct can establish comparable serum antibody responses to i.m. dosing after a prime-boost regimen. [0241] Cellular immunogenicity with bone marrow ELISpots was then assessed to detect H5-specific IgG-secreting antibody secreting cells in bone marrow, as shown in FIG. 24. High levels of these ASCs, >32 spots/million bone marrow cells, were detected in both immunized mouse groups after the prime-boost regimens, and no significant differences between the two groups were observed. [0242] To assess the induction of antigen-responding polyfunctional CD4+ and CD8+ T cell populations, both spleens (systemic immunogenicity) and lungs (respiratory immunogenicity) were isolated and stimulated with H5 overlapping peptides. ICS and flow cytometry were used to evaluate the levels of polyfunctional T cells. Substantial levels of polyfunctional CD4 and CD8 T cells were detected in spleens of both i.m. and i.n. vaccinated animals, with no significant differences seen between the two groups, as shown in FIGS.25A-B. [0243] Both i.m. and i.n. vaccination also induced antigen-reactive CD4 and CD8 T cells in the lungs. While differences were not always statistically significant due to the natural variance observed in T cell data, there is a consistent trend towards higher responses with intranasal administration, as shown in FIGS.25C-D. Looking at the specialized CD69+ TRM subset, there were significantly higher responses in CD4+ T cells induced by the i.n. vaccine relative to the i.m. vaccine, and a trend in the same direction with CD8+ cells, as shown in FIG.25E-F. These results are consistent with the data described above using the intranasally-delivered SARS-CoV-2 self-amplifying RNA/NLC vaccine candidate. [0244] These data demonstrate the in vitro immunogenicity of this vaccine construct. Further optimization of the vaccine size and charge parameters, controlled predominantly by the formulation N/P ratio, was carried out by screening intranasally-dosed vaccines at various N/P ratios, as described below. In-Depth N/P Ratio Comparison [0245] To further optimize the vaccine construct described herein for intranasal delivery, a range of vaccine N/P ratios were assessed to determine which was optimal. The optimal ratio of N/P for this vaccine formulation minimizes reactogenicity while maximizing immunogenicity. [0246] Study groups spanned a range from N/P=0.6-15, and included a control of mice i.m. dosed with the H5 self-amplifying RNA/NLC vaccine at N/P=15, as well as a vector control group as before. Each study group included five 6- to 8-week-old C57BL/6 mice female mice, as shown in Table 3. [0247] Each mouse was administered vaccine at 5 µg per mouse as indicated in FIG.26 at day 0 (prime vaccination) and day 21 (boost vaccination) of the study. Blood was collected by the retro-orbital route every 7 days. Each mouse was individually ear-tagged and tracked across the entire study. Vaccine reactogenicity was assessed by tracking the body weight of each mouse for the 5 days post-dosing. Post-prime and post-boost serum was analyzed for IgG titer by ELISA and pseudovirus neutralization assays. Bone marrow was collected at harvest to measure B cell IgG secretion, and spleens and lungs were isolated for T cell flow cytometry. The resulting data were analyzed to address whether the N/P ratio affects vaccine reactogenicity as assessed by weight loss, whether the N/P ratio affects vaccine immunogenicity, and what the optimal N/P ratio is to balance reactogenicity and immunogenicity. Experimental Groups for In-Depth N/P Ratio Comparison Table 3 [0248] The first measure of N/P ratio effect was vaccine reactogenicity. Mice were weighed daily for 7 days following prime and boost vaccination and data are shown in FIG. 27. As shown, vaccines with higher N/P ratios (10 and above) led to greater magnitudes of transient bodyweight loss immediately after immunization in the post-prime days, that recovered by day 5. Vaccine formulated at N/P=8 resulted in significantly reduced bodyweight drop, >5% weight loss at day 1 that recovered by day 3. All other study groups (vaccines at N/P<8) showed no effect on bodyweight. This effect on bodyweight was more evident after the prime vaccination dose. These dynamics suggest that N/P of 8 or lower show reduced reactogenicity for intranasal dosing than the previous N/P=15 vaccine. [0249] Serum antibody responses were assessed next. IgG ELISA titers demonstrated a steep drop-off in vaccine immunogenicity in vaccines formulated at N/P ratios below 5, with these low N/P groups showing no detectable H5-binding serum IgG response. However, H5-binding IgG titers were induced at significant levels at N/P=5 and above. IgG levels were highly similar across all groups formulated at N/P ratios of 8 and above, as shown in FIG.28. These data demonstrate that vaccine formulated at lower N/P clearly is unable to effectively deliver vaccine self-amplifying RNA to cells, resulting in a total lack of immune response. Vaccine formulated at N/P=5 shows significant induction of responses, and vaccines at N/P above 5 stimulate the highest and equivalent levels of serum IgG. These results were confirmed by a full repeat of this study, with very similar data collected (not shown). [0250] Likewise, the serum pseudovirus neutralizing antibody titer data demonstrated a similar response cliff at N/P 5 and below, as shown in FIG.29. Here, there was no response below N/P=8 post-prime, though the N/P=5 group did recover a response upon boost vaccination. Boost immunization overall had a greater effect on serum pseudovirus neutralizing antibody titers in boosting final titers. After boost, vaccines formulated at N/P=5 and above had induced demonstrated serum neutralizing antibody titers in mice, in patterns similar to those seen with the serum IgG titers measured by ELISA. [0251] Next, bone marrow was assessed for IgG and IgA antibody-secreting cells, as shown in FIG 30. Similar results were evident across both measures, with no responses seen induced by vaccines formulated at N/P ratios below 5, and significant responses induced by vaccines formulated at N/P=5 and above. Again, this confirms that while there is a cutoff below N/P=5, below which no vaccine-induced immune responses are detected and presumably self-amplifying RNA is not effectively delivered to cells, vaccine formulation at N/P ratios above 5 appear generally equivalent in their ability to induce bone marrow- resident immune responses. [0252] Lastly, T cell immunogenicity was assessed in splenocytes and lung T cells, as shown in FIG 31. The splenocytes showed increases in responses as N/P ratio increased, with vaccine formulated at N/P=8 and above inducing noticeably higher CD4+ and C8+ T cell responses, as shown in FIGS.31A-B, than seen induced by vaccine formulated at lower N/P ratios. Similar patterns were observed in lung CD4+ cells, though the CD8+ cells showed low polyfunctional responses to antigen stimulation overall relative to CD4+ cells, across vaccines formulated at all effective N/P ratios, as has been seen in previous experiments, as shown in FIGS.25C-D. Finally, the specialized T resident memory cells in the lungs, CD69+ T cells, also showed enhanced polyfunctionality in both CD4+ T cells and CD8+ T cells at N/P=8 and above, as shown in FIGS.25E-F. [0253] Based on the experimental results discussed above, it was determined that an N/P ratio of 8-10 optimizes both vaccine reactogenicity and immunogenicity. Testing Alternate Oils [0254] Further optimization of the NLC formulation for intranasal vaccination was achieved through testing of alternative liquid oils for suitability. The oil used in the NLC formulation may further improve vaccine immunogenicity while reducing signs of vaccine reactogenicity. The base self-amplifying RNA delivery NLC formulation contains squalene, which has known tissue-irritating properties and may not be optimal for use in intranasal vaccine delivery to sensitive mucosal tissues. Additionally, moving away from a squalene- containing formulation may allow for an improved supply chain, as shark-based squalene is a non-renewable material. Alternate oils were studied with the H5 vaccine self-amplifying RNA formulated at an N/P ratio of 10, to allow gauging of any reduced reactogenicity of the alternate-oil NLC formulations relative to the standard squalene-based NLC. Using an N/P that showed some evidence of reactogenicity with the squalene NLC formulation allowed for detection of reduction in reactogenicity by any of these alternate-oil NLC formulations. [0255] New alternate oil NLCs were created with Miglyol ^® 810, cottonseed oil, castor oil, and soybean oil, in addition to the standard squalene-based NLCs. H5-expressing self- amplifying RNA formulated with each of these NLCs comprised the study groups, all complexed at N/P=10. The vaccine was administered intranasally (i.n.) at 5 µg per mouse. A non-immunogenic vector control self-amplifying RNA/NLC construct (SEAP self- amplifying RNA) was used as a negative control and administered i.n. formulated with squalene-containing NLCs at N/P=10. Each study group included 6 (controls) or 8 (study groups) 6- to 8-week-old C57BL/6 mice, evenly split between males and females, as shown in Table 4. [0256] Each study group was administered vaccine as indicated in FIG. 32 at day 0 (prime vaccination) and day 22 (boost vaccination) of the study. Blood was collected by the retro-orbital route every 7 days. Each mouse was individually ear-tagged and tracked across the entire study. Post-prime and post-boost serum was analyzed for IgG titer by ELISA and pseudovirus neutralizing antibody titers. Spleens and lungs were isolated for T cell flow cytometry at the end of the study. The resulting data were statistically analyzed to address whether any of the alternate oil NLCs are less apparently reactogenic than the standard squalene NLC and whether any of the alternate oil NLCs exhibit equivalent or enhanced immunogenicity relative to the standard squalene NLC.

Experimental Groups for Testing Alternate Oil NLC-Delivered Self-Amplifying RNA Vaccines Table 4 [0257] To address a primary goal of reducing apparent intranasal vaccine reactogenicity, mice were weighed after prime and boost vaccinations. Several of the new preparations induced reduced weight loss relative to the squalene NLC formulation, as shown in FIG 33. In particular, vaccines formulated with the Miglyol ^® 810, cottonseed, and castor oil NLCs showed improved profiles relative to vaccine formulated with the standard NLC post-prime, as shown in FIG 33A. Post-boost there was far less variability in the mouse weights, as previously seen, as shown in FIG 33B., with all vaccine formulations showing generally the same slight reduction immediately post-vaccination with a rapid return to baseline over the course of a few days, as may be observed in response to mouse handling for dosing. [0258] Serum immunogenicity was then assessed to determine how these alternate oil NLCs affected the vaccines’ induction of systemic antibody responses. The ELISA measures demonstrated that none of the oils markedly reduced the antibody response induced by the various vaccine formulations relative to the squalene NLC formulated vaccine, as shown in FIG 34. However, the Miglyol ^® 810 NLC was significantly higher than the castor and soybean oil NLCs post boost. And the cottonseed oil based NLCs appeared to have induced enhanced serum antibody titers relative to the standard preparation both post-prime and post-boost. [0259] This was further confirmed by assessing the capacity of these vaccine preparations to induce serum pseudovirus neutralizing antibodies in mice, as shown in FIG 35. It was observed that the Miglyol ^® 810 oil and cottonseed oil containing NLCs all induced improved serum pseudovirus neutralization in mice relative to the standard squalene NLCs, though with some greater spread in the responses (cottonseed oil NLC formulated vaccine). [0260] T cell responses in these different vaccine groups were then assessed via ICS/flow cytometry, as shown in FIG 36. Compared to all other formulations and the standard squalene-based formulation, the Miglyol ^® 810-based NLC showed a significant enhancement of antigen-responsive polyfunctional CD4 and CD8 T cell populations. This was demonstrated not only in spleens, but also in lung T _RM cells (CD69+). [0261] Collectively, these data indicate that an optimal vaccine formulation is based on Miglyol ^® 810 containing NLCs, reducing reactogenicity and enhancing vaccine immunogenicity as compared to squalene containing NLCs. [0262] DEFINITIONS [0263] The following terms have the following meanings unless otherwise indicated. Any undefined terms have their art recognized meanings. [0264] In the present description, the terms “about,” “around,” “approximately,” and similar referents mean ± 20% of the indicated range, value, or structure, unless otherwise indicated. [0265] The use of the alternative (e.g., “or”) should be understood to mean either one, both, or any combination thereof of the alternatives. [0266] As used herein, the terms “include,” “have” and “comprise” are used synonymously, which terms and variants thereof are intended to be construed as non- limiting. [0267] As used herein and in the appended claims, the singular forms “a,” “an,” and “the” include plural reference unless the context clearly indicates otherwise. [0268] The terms thermostable lyophilized vaccine composition, lyophilized vaccine composition, lyophilized thermostable cake, and lyophilized cake are used interchangeably herein. These terms generally refer to a lyophilized oil-in-water stable emulsion comprising a biodegradable oil or metabolizable oil, cake-forming excipients used to produce the cake, and optionally one or more bioactive agents. [0269] The term “alkyl” means a straight chain or branched, noncyclic or cyclic, unsaturated or saturated aliphatic hydrocarbon containing the indicated number of carbon atoms. Unsaturated alkyls contain at least one double or triple bond between adjacent carbon atoms. [0270] The terms “polypeptide,” “peptide,” and “protein” are used interchangeably herein to refer to polymers of amino acids of any length. The polymer may be linear or branched, it may comprise modified nucleotides or amino acids, and it may be interrupted by non- nucleotides or non-amino acids. The terms also encompass a nucleotide or amino acid polymer that has been modified naturally or by intervention; for example, disulfide bond formation, glycosylation, lipidation, acetylation, phosphorylation, or any other manipulation or modification, such as conjugation with a labeling component. Also included within the definition are, for example, polynucleotides or polypeptides containing one or more analogs of a nucleotide or an amino acid (including, for example, unnatural amino acids, etc.), as well as other modifications known in the art. [0271] The term “isolated” means the molecule has been removed from its natural environment. [0272] “Purified” means that the molecule has been increased in purity, such that it exists in a form that is more pure than it exists in its natural environment and/or when initially synthesized and/or amplified under laboratory conditions. Purity is a relative term and does not necessarily mean absolute purity. [0273] A “polynucleotide” or “nucleic acid,” as used interchangeably herein, refer to polymers of nucleotides of any length, include DNA and RNA. The nucleotides can be, for example, deoxyribonucleotides, ribonucleotides, modified nucleotides or bases, and/or their analogs, or any substrate that can be incorporated into a polymer by DNA or RNA polymerase, or by a synthetic reaction. A polynucleotide may comprise modified nucleotides, such as methylated nucleotides and their analogs. If present, modification to the nucleotide structure may be imparted before or after assembly of the polymer. [0274] The term “RNA integrity” as used herein means the quantity of intact RNA remaining after an event or passage of time. For example, RNA integrity may be evaluated following freezing, lyophilization, or storage. RNA integrity may be evaluated by both the size and strength of bands shown in agarose gel electrophoresis. [0275] An “individual” or a “subject” is any vertebrate. Vertebrates include, but are not limited to humans, primates, farm animals (such as cows, pigs, sheep, chickens), sport animals, pets (such as cats, dogs, birds, horses), and rodents. [0276] A “replicon” as used herein includes any genetic element, for example, a plasmid, cosmid, bacmid, phage or virus that is capable of replication largely under its own control. A replicon may be either RNA or DNA and may be single or double stranded. [0277] The term liquid phase lipid refers to a lipid that, prior to mixing with any other component, is liquid at ambient temperature. [0278] The term solid phase lipid refers to a lipid that, prior to mixing with any other component, is solid at ambient temperature. [0279] Ambient temperature is between 15°C and 25°C. [0280] Cake-forming excipient and lyoprotectant are used herein interchangeably. A cake-forming excipient refers to a substance added to a liquid stable oil-in-water emulsion formulation prior to lyophilization which yields a cake following lyophilization. Upon reconstitution of the lyophilized cake, a stable emulsion forms, that is suitable for delivery of a bioactive agent including vaccine antigens or polynucleotides encoding vaccine antigens. As used herein, cake-forming excipients are those substances which do not disrupt an emulsion upon reconstitution of the lyophilized cake. [0281] Excipients as used herein refers to substances other than the pharmacologically active drugs, which are included in the manufacturing process, or fill-finish process for storage or shipment of the pharmacologically active drug including, without limitation, lyophilization, and are contained in a finished pharmaceutical process. [0282] The practice of the present disclosure will employ, unless otherwise indicated, conventional techniques of molecular biology, recombinant DNA, biochemistry, and chemistry, which are within the skill of the art. Such techniques are explained fully in the literature. See, e.g., Molecular Cloning A Laboratory Manual, 2nd Ed., Sambrook et al., ed., Cold Spring Harbor Laboratory Press: (1989); DNA Cloning, Volumes I and II (D. N. Glover ed., 1985); Oligonucleotide Synthesis (M. J. Gait ed., 1984); Mullis et al., U.S. Pat. No: 4,683,195; Nucleic Acid Hybridization (B. D. Hames & S. J. Higgins eds. 1984); B. Perbal, A Practical Guide to Molecular Cloning (1984); the treatise, Methods in Enzymology (Academic Press, Inc., N.Y.); and in Ausubel et al., Current Protocols in Molecular Biology, John Wiley and Sons, Baltimore, Maryland (1989). CONCLUSION [0283] The previous description of the disclosed implementations is provided to enable any person skilled in the art to make or use the invention disclosed herein. Although the various inventive aspects are disclosed in the context of one or more illustrated embodiments, implementations, and examples, it should be understood by those skilled in the art that the invention extends beyond the specifically disclosed implementations to other alternative implementations and/or uses of the invention and obvious modifications and equivalents thereof. It should be also understood that the scope of this disclosure includes the various combinations or sub-combinations of the specific features and aspects of the implementations disclosed herein, such that the various features, modes of implementation, and aspects of the disclosed subject matter may be combined with or substituted for one another. The generic principles defined herein may be applied to other implementations without departing from the spirit or scope of the disclosure. Thus, the present disclosure is not intended to be limited to the implementations shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein. [0284] All references listed herein, including patent applications and patent publications are herein incorporated by reference in their entirety, as if each individual reference is specifically and individually indicated to be incorporated by reference.