Antibodies used in this study were purchased from BioLegend (San Diego, CA, USA) or Southern Biotechnology (Birmingham, AL, USA), and listed in Table S1 . Specific pathogen-free C57BL/6 or BALB/c female mice at 6 weeks of age were purchased from Samtako Bio Korea (Kyounggi, Korea) and maintained at the NA Vaccine Institute (NAVI) animal facility (Seoul, Korea). The mice were fed a sterile, commercial mouse diet and provided with water ad libitum. The experimental protocols used in this study were reviewed and approved by the Ethics Committee and Institutional Animal Care and Use Committee (Permit Number: NAVI-2019-0002) of the NAVI.

As a template for IVT, the nucleotide segment (1,701-3,360; 1,660 nucleotides) from CSBV genome (GenBank accession number KF960044.1) was cloned into the pUC-GW-Amp vector and the template was amplified by PCR using primer pairs containing the T7 RNA promoter sequence. PCR cycling conditions were as follows: 95°C for 5 min, followed by 35 cycles of 95°C for 30 s, 60°C for 30 s, and 72°C for 1 min. A final elongation step was carried out for 5 min at 72°C. Synthesis and purification of dsRNA were performed using a MEGAscript RNAi Kit (Thermo Fisher Scientific, Waltham, MA, USA) according to the manufacturer’s instructions. Briefly, IVT reactions were carried out by incubating the PCR products with 10× T7 Reaction Buffer, 4 ribonucleotide solutions, and T7 Enzyme Mix at 37°C for 2-4 h. The IVT products were incubated at 75°C for 5 min and then the complementary RNAs were allowed to anneal to each other at room temperature to form dsRNA. The template DNAs and unannealed ssRNAs were removed by treatment with DNase I and RNase A. Lastly, each dsRNA sample was purified with 100% ethanol and eluted with a solution containing 10 mM Tris-HCl (pH 7) and 1 mM EDTA. For the synthesis of NVT used as a vaccine adjuvant, partial nucleotide segment (1,701-2,112; 412 nucleotides) was selectively cloned, and ssRNAs were digested by treatment with RNase T1 instead of RNase A. Template and primer details for NVT synthesis are shown in Table S2 .

The purified NVT sample was analyzed by RP-HPLC on the Waters Alliance HPLC System (Waters, Milford, MA, USA) using a Waters XBridge OST C18 Column with 130 Å, 2.5 μm, and 4.6 × 50 mm (Waters, Milford, MA, USA). Two eluent-buffers were used to perform the chromatography. Buffer A is an aqueous solution containing 0.1 M triethylammonium acetate (TEAA), pH 7.0, and buffer B is an aqueous solution of 0.1 M TEAA, pH 7.0, containing 25% (v/v) acetonitrile. The whole procedure was undertaken at 50°C which is a non-denaturing temperature. Ion-pair RP-HPLC analysis was performed using the following linear gradient condition: flow rate 0.9 ml/min, 40–70% buffer B over 15 min. Peak quantification was performed by recording chromatograms at 260 nm and integrating peak areas.

Human TLR3-expressing HEK293 cells (InvivoGen, San Diego, CA, USA) were cultured in Dulbecco’s modified Eagle medium (DMEM) supplemented with 4.5 g/l glucose, 2 mM L-glutamine, 10% heat-inactivated fetal bovine serum (FBS), 1% penicillin/streptomycin in the presence of blasticidin (Sigma-Aldrich, St Louis, MO, USA) (30 μg/ml), zeocin (InvivoGen) (100 μg/ml), and normocin (InvivoGen) (100 μg/ml). For stimulation, the cells were seeded on a 96-well culture plate at 5 × 10⁴ cells/well and incubated overnight at 37°C. The attached cells were treated with various concentrations of dsRNA, NVT, or poly (I:C) (InvivoGen) and CpG (InvivoGen) for 24 h, and 20 μl of supernatant and 180 μl of Quanti-Blue were reacted in a 96-well enzyme-linked immunosorbent assay (ELISA) plate. While incubating at 37°C for 1-3 h, the absorbance was measured at 655 nm when the most color development occurred.

C57BL/6 mice were injected intramuscularly with optimal dose (10 μg) of NVT, poly(I:C), or CpG (25). Five hours later, total RNA was isolated from the local iLN of the immunized mice using TRIzol reagent (Thermo Fisher Scientific). Briefly, the cells from iLN were lysed by treatment with 1 ml of TRIzol, and 0.2 ml of chloroform was added to the lysate. After centrifugation at 12,000 × g for 15 min at 4°C, the aqueous phase containing RNA was transferred to a new tube. Then, the same amount of isopropanol was added and incubated for 10 min at room temperature, followed by centrifugation for 10 min at 12,000 × g at 4°C. The RNA pellets were washed with 75% ethanol, air-dried, and resuspended in RNase-free water. The amount of total RNA was measured at 260 nm with NanoDrop (Molecular Devices, San Jose, CA, USA). The purity was confirmed at 260/280 nm (a ratio between 1.8 - 2.0 is considered pure) and 260/230 nm (a ratio between 2.0 - 2.2 is considered pure).

One Step TB Green PrimeScript RT-PCR kit (Takara Bio, Shiga, Japan) was used according to the manufacturer’s specifications. Briefly, RNA samples were mixed in a 96-well reaction plate with 2× One Step TB Green RT-PCR Buffer 4, PrimeScript 1 step Enzyme Mix 2, forward and reverse primers (10 μM), and RNase-free dH₂O. Then, PCR reactions were performed with pre-incubation at 42°C for 5 min, followed by 40 thermal cycles at 95°C for 10 s and 55°C for 30 s using a CFX real-time PCR system (Bio-Rad, Hercules, CA, USA). The expression level of target genes was normalized to the reference gene HPRT and calculated using the 2 ^−ΔΔCt method as previously described with minor modifications (26). Primers were selected from previous study (27) and synthesized by Cosmogenetech (Seoul, Korea). The specific primer sequences are shown in Table S2 .

To test NVT for long-term stability, the substance was stored in an accelerated condition at 25 ± 2°C and relative humidity of 60 ± 5% under the ICH guideline (https://www.ich.org/page/quality-guidelines) on “Stability testing of new drug substances and drug products”. Endotoxin test was performed using the Pierce™ LAL Chromogenic Endotoxin Quantitation Kit (Thermo Fisher Scientific, Waltham, MA, USA) according to the manufacturer’s instructions. Their appearance and concentration were determined for each period, and the stability was monitored by electrophoresis on an 1% agarose gel.

Preclinical GLP toxicity profiling experiments of the NVT administered through intramuscular and subcutaneous routes were undertaken at a GLP-certified organization (Biotoxtech, Cheongju, Korea) according to the Organisation for Economic Co-operation and Development (OECD) protocol (https://www.oecd.org/chemicalsafety/testing/good-laboratory-practiceglp.htm). The test profiles include single dose, dose range finding test, repeated dose toxicity tests (two doses, four doses, and recovery test) for Sprague Dawley rats and New Zealand white rabbits, antigenicity assays [passive cutaneous anaphylaxis (PCA) and active systemic anaphylaxis (ASA)] for guinea pigs, and in vivo micronucleus assay for ICR mice. Also, in vitro chromosomal aberration assay using a Chinese hamster lung cell line, and in vitro bacterial reverse mutation test (Ames test) for Salmonella typhimurium and Escherichia coli were performed ( Table 1 ).

C57BL/6 female mice (n=4 per group) were administered either with 2 μg of ovalbumin (OVA; InvivoGen) alone or formulated with 10 μg of NVT or10 μg of poly(I:C) as an adjuvant via the intramuscular route. At 0, 6 or 24 h, the draining iLN was harvested to dissociate into a single cell suspension. The cells were labeled with a fixable viability dye to distinguish live cells from dead cells followed by surface staining for macrophage (CD11b⁺F4/80⁺), neutrophil (CD11b⁺Ly6G⁺), and DCs (CD11c^hiMHC-II^hi) ( Supplementary Figure 1 ). The specificity of staining was confirmed by fluorescence minus one and the migration of innate cells was expressed as an absolute number after flow cytometric analysis. To analyze the activation of DCs, CD11c^hiMHC-II^hi cells were gated and then the intensity of the surface expression of CD40, CD80, CD86, and MHC class II was determined by NovoCyte flow cytometry (ACEA Biosciences, San Diego, CA, USA). IFN-β was measured by sandwich ELISA according to the manufacturer’s instructions (BioLegend, San Diego, CA, USA).

To explore the efficacy of the adjuvants, C57BL/6 mice (n=5 per group, female, 6-8 weeks old) were immunized intramuscularly twice at two-week intervals with 2 μg of OVA with either alone or adjuvanted with 10 μg of NVT or 10 μg of poly (I:C). No severe clinical signs including weight loss were observed after immunization. Two weeks later, blood was collected to determine OVA-specific antibodies and the spleen was isolated to analyze OVA-specific T cell responses.

The 96-well microplates were coated with either 2 μg/ml of goat anti-mouse IgG antibody for standard curve construction or 5 μg/ml of OVA for OVA-specific antibody titer overnight at 4°C and then non-specific sites were blocked with 1% skim milk. Serially diluted mouse IgG antibodies or serum samples were incubated for 2 h at room temperature. For the OVA-specific Ig isotypes, horseradish peroxidase-conjugated anti-mouse IgG-, IgG1-, and IgG2c-antibodies were used as secondary antibodies and developed by using 3,3′,5,5′-tetramethylbenzidine substrate solution. The reaction was stopped with 0.5 M HCl and the absorbance was measured at 450-590 nm using a Multiskan Sky microplate reader (Thermo Fisher Scientific). A standard curve using mouse IgG was generated and antibody titers of serum samples were calculated as arbitrary units relative to the standard curve.

Two weeks after the last vaccination, spleens were obtained from immunized mice and dissociated into single cells. To examine OVA-specific T cell responses, the cell suspensions were restimulated with 20 μg/ml of OVA peptides (OVA_257-264 and OVA_323-339) (28) and incubated overnight at 37°C. Then, 1 μg/ml of a protein transport inhibitor GolgiPlug (BD Bioscience, Franklin Lakes, NJ, USA) was added and further cultured for 4-6 h. After washing with phosphate-buffered saline (PBS), the cells were stained with a viability dye, anti-CD4, anti-CD8, and anti-CD44 antibodies for 20 min at 4°C. For intracellular staining, the cells were permeabilized and fixed using a Cytofix/Cytoperm solution (BD Bioscience) and stained with anti-IFN-γ and anti-IL-4 antibodies for 20 min at 4°C. The cells were washed and resuspended with FACS buffer (PBS containing 1% FBS and 0.1% NaN₃) and analyzed by flow cytometry.

Experimental results were expressed as the mean value ± standard deviation (S.D.). The Kruskal-Wallis test followed by Dunn’s post-test was used for multiple comparison of more than three groups by using GraphPad Prism version 6.00 for Windows (GraphPad Software, La Jolla, CA, USA, www.graphpad.com). *p<0.05 was considered statistically significant.

The nucleotide segment (1,701-3,360; 1,660 nucleotides) from CSBV genome (GenBank accession number KF960044.1), which does not match with a human DNA sequence, was cloned into a vector. Then, PCR products of various sizes were produced by 3′-serial deletion and used to synthesize dsRNAs of 319, 397, 466, 508, 664, 733, 822, 885, 1,032, 1,153, or 1,648 bp length by IVT. When each dsRNA was electrophoresed on an 1% agarose gel, dsRNA longer than 508 bp showed polymeric impurities similar to poly(I:C), indicating the need to further select dsRNA of less than 500 bp (data not shown). Since dsRNA tend to activate TLR3 signaling, we tested whether the synthetic dsRNA could act as a TLR3 agonist. To this end, the TLR3-expressing reporter cell line was stimulated with dsRNA of 319 bp to 508 bp in length. The induction of TLR3 activation was similar regardless of the dsRNA size ( Supplementary Figure 2A ). To select a specific length of dsRNA that was the most effective as an adjuvant, mice were intramuscularly injected with OVA formulated with each of the dsRNAs twice at intervals of 2 weeks, and OVA-specific IgG levels in each serum were measured 2 weeks after boosting. Unlike the results of in vitro assays, dsRNAs shorter than 300 bp were found to be less effective in inducing antigen-specific antibodies as vaccine adjuvants in vivo, suggesting that candidate molecules were further selected from dsRNAs with lengths of 400 to 500 bp ( Supplementary Figure 2B ). Next, to determine the dsRNA size for optimal production efficiency, dsRNAs of various lengths were synthesized under the same conditions and their yields were compared on an 1% agarose gel. The yield of dsRNA synthesis was highest at 424 bp to 460 bp, about 120 μg per 20 μl reaction ( Supplementary Figure 2C ). In consideration of adjuvanticity, production efficiency, and manufacturing cost, dsRNA with a length of 424 bp (approximately 275 kDa) was selected as a final candidate for product analyses, and the partial genomic sequence (1,701-2,112; 412 nucleotides) of CSBV was used for cloning to synthesize it ( Figure 1 ).

When dsRNA is synthesized with a commercial kit, ssRNA and DNA template are removed by treating the product with RNase A and DNase I. However, since RNase A can partially or completely degrade dsRNA (29), there is a possibility that it may affect the stability of the product after synthesis or purification. Considering that RNase T1 specifically degrades ssRNA at G residues and cannot cleave dsRNA, we used RNase T1 instead of RNase A for dsRNA purification. This step generated 424 bp dsRNA with UAUAG-3′ at both ends ( Figure 1 ). We named this RNase T1-treated final product NexaVant (NVT). Treatment of NVT with RNase III alone or in combination with RNase T1 generated approximately 25 bp dsRNAs, indicating that NVT consists of dsRNA that forms a completely complementary pairing ( Supplementary Figure 3 ). We also found the importance of RNase T1 use for the synthesis and purification of NVT by confirming that NVT is completely degraded by RNase A ( Supplementary Figure 3 ).

NVT was tested for stability under accelerated storage conditions (25 ± 2°C). The result showed that the appearance, concentration, and molecular size of NVT were unaffected for 6 months ( Figure 2A and Table 2 ). Next, an RP-HPLC assay was used to confirm that the finally-produced NVT is a single compound. The RP-HPLC study confirmed that NVT was eluted as a single peak at a retention time of 17.601 min and accordingly exhibited almost 100% purity, demonstrating that NVT exhibits an extreme homogeneity ( Figure 2B ). To determine whether NVT acts as a TLR3 ligand, human TLR3-expressing HEK 293 cells were stimulated with 10, 50, or 100 μg/ml of NVT, or 100 μg/ml of CpG (negative control), or poly(I:C) (positive control), and NF-κB/AP-1 activation was measured. NVT induced TLR3 activation in a dose-dependent manner in vitro ( Figure 2C ). To determine whether the expression of viral nucleic acid sensors in vivo is increased by NVT, mice were injected intramuscularly with 10 μg of NVT, CpG, or poly(I:C), then local draining iLN were isolated and the expression of TLR3, MDA5, and RIG-I was measured by RT-qPCR analysis. We found that NVT promoted the expression of MDA5 and RIG-I as well as TLR3 to the same level as poly(I:C) ( Supplementary Figure 4 ). For biomedical product development, the in vivo kinetics of the candidate molecule is essential. Notably, pharmacokinetic data of poly(I:C) were difficult to obtain due to the lack of analytical methods for such a heterogeneous mixture (15). In contrast, RT-qPCR reactions can be used to determine the exact copy number of NVT in blood. Indeed, we were able to quantify the content of NVT in blood by collecting blood from rats injected subcutaneously with NVT, extracting RNA, and performing RT-qPCR. The pharmacokinetic analysis showed that the amount of NVT present in the blood decreased rapidly ( Figure 2D ), indicating that unpredictable responses by NVT caused by long-term retention in the body are unlikely.

To evaluate the role of NVT in the induction of an immune response as an adjuvant, C57BL/6 mice were intramuscularly injected with OVA either alone (control) or formulated with NVT or poly(I:C) as an adjuvant, and draining iLNs were collected 0, 6 and 24 h after injection to analyze innate immune cell migration and DC activation ( Figure 3A ). Compared to the data at 0 h, NVT efficiently promoted the migration of macrophages (CD11b⁺F4/80⁺) and neutrophils (CD11b⁺Ly6G⁺) as well as DCs (CD11c^hiMHC-II^hi) into draining iLN soon after the injection as poly(I:C) did ( Figures 3B-D ). Next, to measure the expression of DC maturation, the mean fluorescence intensity of CD40, CD80, CD86, and MHC-II in CD11c^hiMHC-II^hi DCs was analyzed by flow cytometry, and we found both NVT and poly(I:C) promoted the activation of DCs. NVT significantly increased CD40 at 6 h compared to the control (OVA alone). After that, the expression of CD40 decreased and became similar to control at 24 h ( Figure 3E ). NVT enhanced the expression of the costimulatory molecules CD80 and CD86 at 24 h after injection, significantly higher than the control and poly(I:C) groups ( Figure 3F, G ). NVT also significantly promoted an increase in MHC class II expression compared to control at 24 h ( Figure 3H ). In addition to DC activation, NVT substantially promoted the production of IFN-β, the major effector cytokine induced by TLR3 agonists, compared to poly(I:C) at 6 h after injection, indicating that NVT provides strong in vivo activation of type I IFN signaling cascades ( Figure 3I ). Taken together, these results suggest NVT works as a potential adjuvant that can activate adaptive immunity through migration of APCs to LN, induction of DC activation, and promotion of IFN-β production.

Next, we examined the ability of NVT to induce antigen-specific antibody responses. C57BL/6 mice were immunized intramuscularly twice at two-week intervals with either OVA alone or adjuvanted with NVT or poly(I:C) according to the schedule ( Figure 4A ). Then, OVA-specific total IgG, IgG1, and IgG2c were measured in sera from mice two weeks after the last immunization. The groups in which OVA was formulated with an adjuvant such as NVT or poly(I:C) showed significant increases in antigen-specific IgG. Interestingly, NVT had higher levels of Th1-type IgG2c antibody but lower levels of Th2-type IgG1 antibody compared to poly(I:C) ( Figures 4B-D ). These results suggest that NVT is a strong effective vaccine adjuvant that can induce a potent Th1-biased antibody response.

To further explore whether NVT could promote the antigen-specific T cell response, C57BL/6 mice were vaccinated twice with either OVA alone or together with NVT or poly(I:C). Two weeks after boosting, splenocytes isolated from the mice were restimulated with epitopes of the OVA protein, OVA_257-264 and OVA_323-339, to analyze OVA-specific T cell responses (28). We found that OVA alone is insufficient to induce OVA-specific CD4⁺ T cell responses. NVT enhanced OVA-specific IFN-γ-producing CD4⁺ T cells to a level similar to that of poly(I:C) ( Figure 4E ) whereas OVA-specific IL-4-producing CD4⁺ T cells responses were not observed in any group (data not shown). Moreover, NVT promoted an OVA-specific cytotoxic T lymphocyte (CTL) response in the spleen to a level comparable to that of poly(I:C) ( Figure 4F ). Taken together, NVT would be an effective vaccine adjuvant that promotes potent Th1 and CTL responses, especially for vaccines against intracellular pathogens or cancers.

Pharmaceutical development includes GLP preclinical toxicology studies as required by regulatory agencies (30). Thus, we performed a GLP toxicity test to see if NVT is a safe substance without causing severe toxicity, and to obtain the clinical starting dose and safety margin we determined the maximum tolerated dose (MTD) and no observed adverse effect level (NOAEL). We first determined the MTD and lethal dose of NVT in rats, which is a standard toxicological test animal model (30), by subcutaneously administering doses of 0, 10, 20, and 40 mg/kg. No severe adverse effects were observed even at the highest dose and therefore lethal dose and MTD were considered as >40 mg/kg. In addition, when a repeated dose toxicity test (subcutaneous route in rats, 4 doses and recovery for 2 weeks) was performed, there was no mortality or serious symptoms, and NOAEL and MTD were confirmed to be 3 mg/kg for males and 6 mg/kg for females. Next, according to the standard protocol of the OECD, NVT (1.6, 0.8, 0.4, and 0.2 mg) was injected into ICR mice through the intramuscular route, and the effect on micronuclei induction was confirmed. There was no abnormality in the generation of micronuclear polychromatic red blood cells (MNPCE), indicating that NVT did not cause chromosomal damage, consistent with the in vitro chromosomal aberration assay. Lastly, the results of antigenicity (guinea pig) and Ames test (Salmonella) confirmed that NVT was a non-antigenic and non-mutagenic substance, respectively. Therefore, NVT is expected to be a safe vaccine adjuvant without serious toxicity in the range below the MTD and NOAEL ( Table 1 ).