DDO-adjuvanted influenza A virus nucleoprotein mRNA vaccine induces robust humoral and cellular type 1 immune responses and protects mice from challenge

We have demonstrated that DDO268 induces expression of type I IFN at the site of inoculation and not systemically (12). To further assess the safety of DDO268, we inoculated mice subcutaneously in the footpad with 50 µg of DDO268, a dose 10 times higher than that used in previous published studies (11, 12), and we evaluated complete blood counts (CBCs), serum chemistry, and systemic cytokine levels up to 72 h post-inoculation. As shown in Fig. 1 and Table S1, DDO268 did not induce significant changes in any of the measured parameters. Analysis of the leukogram, erythrogram, and thrombocyte counts (Fig. 1A through G) revealed no significant alterations in blood cell populations. Similarly, chemical parameters such as blood urea nitrogen, bilirubin, and aspartate transferase levels (Fig. 1H through J) were comparable to those in mice inoculated with PBS. Additionally, transcripts of pro-inflammatory cytokines, including Il6 and TNFα, and of IFN-stimulated genes, such as Mx1, were not detected in the liver at any timepoint (Fig. 1K through M). These findings confirm the ability of DDO268 to safely elicit a localized immune response without detectable systemic effects.

To first evaluate the impact of DDO268 in the context of the mRNA vaccine platform, we tested its impact on the induction of type 1 immune responses during vaccination with the original severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) mRNA vaccine, BNT162b2. We vaccinated mice subcutaneously in the footpad with a simple mix of 0.125 µg BNT162b2 vaccine and 5 µg DDO268, a dose that showed no local visible inflammation, and boosted the mice 28 days later (Fig. 2A). Migration of conventional dendritic cells type 1 (cDC1) to the draining lymph node is an essential early step in the generation of type 1 immune responses upon immunization in the presence of DDO268 (12), so we first assessed cDC1 migration to the popliteal lymph node upon vaccination with either DDO268 alone, BNT162b2 alone, BNT162b2 combined with an inert RNA (a synthetic 30 nucleotide-long and non-immunostimulatory RNA), or BNT162b2 combined with DDO268. All combinations were used at equivalent molar amounts. DDO268, but not the non-immunostimulatory RNA, significantly increased the number of cDC1 in the draining lymph nodes at 12 h post vaccination (Fig. 2B; Fig. S1), a previously determined peak timepoint for cDC1 accumulation in the lymph node upon vaccination (12).

To confirm the induction of type I IFNs at the site of immunization, we measured expression of Ifnb1 and the IFN-stimulated gene Mx1 mRNA in the footpad after vaccine administration. DDO268 transiently and significantly enhanced the expression of these genes (Fig. 2C). In addition, in mice lacking the type I IFN receptor (Ifnar^−/−), cDC1 migration was abolished, underscoring the role of type I IFN in the response to DDO268 (Fig. 2D). Moreover, DDO268 increased titers of circulating IgG2c antibodies (Fig. 2E) and promoted the generation of a larger population of activated SARS-CoV-2 specific CD8⁺ IFNγ⁺ T-cells in the spleen compared to BNT162b2 alone (Fig. 2F). The adaptive immune response enhancement by DDO268 depended on type I IFN signaling, similar to our previous findings with inactivated virus vaccines (12). Overall, these proof-of-principle experiments suggest that DDO268 can enhance the immune response elicited by an mRNA vaccine and bias it toward type 1 immunity in a type I IFN-dependent manner.

We next investigated if DDO268 promotes type 1 immunity when co-packaged with the mRNA. To this end, we developed an mRNA vaccine using IAV NP as target antigen. To eliminate unspecific immunostimulatory dsRNA, we cellulose-purified the in vitro transcribed unmodified IAV NP mRNA followed by stringent quality control (Fig. 3A and B). Codon-optimized influenza A/Puerto Rico/8/1934 H1N1 NP mRNA was generated from the pJB201.1 vector (Fig. 3A) and the mRNA functionality was tested in A549 cells. For the purification process, mRNA was incubated with cellulose under optimized binding conditions to ensure selective removal of double-stranded RNA while maintaining mRNA stability and functionality, followed by column purification. NP expression levels were comparable between purified and unpurified mRNAs (Fig. 3C). However, purified mRNA resulted in significantly reduced Ifnb1 expression (Fig. 3D), confirming the effectiveness of cellulose purification in mitigating immune activation. In subsequent experiments, we used unmodified dNTPs for mRNA production followed by cellulose-purification.

We formulated mRNA vaccines into LNPs using the Gen-Voy ILM reagent and co-packaged the IAV NP mRNA with a series of immunostimulatory and non-immunostimulatory RNAs. The formulations included IAV NP mRNA alone, IAV NP mRNA/DDO268, IAV NP mRNA/DDO268B (a modified DDO268 molecule, with the immunostimulatory motif encoded in reverse direction), IAV NP mRNA/X region (a non-immunostimulatory small RNA derived from the hepatitis C virus genome [19, 20]), and empty LNPs (Fig. 4A). Each formulation contained a 1:1 molar ratio of mRNA to RNA, ensuring consistent total number of RNA molecules. The RNA concentration and the encapsulation efficiency were assessed using RiboGreen RNA assay (21), and nanoparticle sizes (ranging from 120 to 150 nm) were measured by dynamic light scattering (DLS) (Fig. 4B).

To assess protein expression and identify the cell types expressing the viral protein in vivo, we inoculated mice subcutaneously and evaluated NP expression at the inoculation site and in the draining lymph nodes at 24 h post-vaccination. NP was expressed in CD45⁺ cells in the footpad (Fig. 4C) and in myeloid cells within draining lymph nodes (Fig. 4D). These results agree with previous reports of mRNA vaccines being primarily taken up by immune cells at the site of immunization (22–24).

We have reported that DDO delivered subcutaneously in its naked form triggers signaling by the cell-type-restricted endosomal RNA sensor TLR3 (12). However, when transfected into cells, DDO is sensed by cytoplasmic RIG-I-like receptors (20). We hypothesized that DDO delivered within LNPs would be detected by the ubiquitously expressed intracellular RNA sensors, providing the advantage of adjuvant functionality in a much larger population of target cells. To test this hypothesis, we compared LNPs containing IAV NP mRNA alone or LNPs containing IAV NP mRNA/DDO268 for their ability to trigger type I IFN expression in cells lacking RIG-I signaling (MAVS KO) or control cells. As expected, DDO induced high levels of MAVS-dependent transcription of Ifnb1 and IL29 (Fig. 5A and B).

To determine if encapsulated DDO triggers RIG-I -like receptors in vivo, we inoculated control, Mavs^−/−, or Tlr3^−/− mice and tested type I IFN expression at the injection site, as well as cDC1 cell accumulation in the draining lymph node 12 h later (Fig. 5C). In WT mice, the DDO-adjuvanted vaccine induced high levels of Ifnb1 while significantly lower levels of Ifnb1 were detected in Mavs^−/− mice (Fig. 5D). Tlr3^−/− mice showed a moderate decrease in Ifnb1 expression in both formulations, suggesting that TLR3 is not sufficient nor necessary for DDO adjuvancy in this context. These findings provide two critical insights: (i) cells detect and respond to DDO268 when encapsulated in LNPs; and (ii) DDO268 induces type I IFN expression necessary to initiate a type 1 immune response in a RIG-I-dependent manner.

To confirm that our vaccine induced local expression of type I IFN, cytokines, and chemokines, we analyzed footpads and popliteal lymph nodes 12 h after subcutaneous inoculation, as well as spleens at 12 and 24 h after subcutaneous inoculation (Fig. 6A). As expected, empty LNPs induced Il1b, Il6, and Ccl2 expression (7) but showed minimal Ifnb1 or Cxcl10 expression at the inoculation site (Fig. 6B). LNPs containing IAV NP mRNA or IAV NP mRNA/X region induced low Ifnb1 levels, while the LNPs containing IAV NP mRNA/DDO268B showed higher levels than controls but lower than the DDO268 formulation (Fig. 6C). All vaccine formulations induced Il6, Il1b, Ccl2, and Cxcl10 transcription (Fig. 6D through G). Transcripts of Il13, a type 2 immunity-associated cytokine were undetected at the inoculation site for all formulations (Fig. 6H). Importantly, DDO268-containing formulations did not induce significant Ifnb1 in the spleen at 12 h after vaccination (Fig. 6I) or Il6, Mx1, and Tnfα expression in the spleen at 24 h after vaccination (Fig. 6J through L), confirming that DDO268 adjuvancy induces a localized immune response (12). Additionally, cDC1 recruitment to the draining lymph nodes increased significantly with DDO268 (Fig. 6M). These findings confirm that the DDO268 enhances type I IFN expression without triggering systemic inflammation.

We next evaluated the adaptive immune responses elicited by the IAV NP mRNA/DDO268 vaccine after primary vaccination and a booster immunization administered 4 weeks later (Fig. 7A). We also tested DDO268B, which triggers less robust responses (Fig. 6; Fig. S2). Three weeks post-booster, anti-NP IgG1 and IgG2c antibodies were measured from sera to determine Th2- and Th1-biased responses, respectively. The IAV NP mRNA vaccine induced higher IgG1 than IgG2c levels, while the IAV NP mRNA/DDO268 or DDO268B vaccine induced higher IgG2c levels, indicating a Th1-biased humoral response (Fig. 7B; Fig. S2B). These results align with our findings of enhanced IgG2c antibodies to an inactivated vaccine when DDO268 was added as an adjuvant (11, 12).

To assess T-cell responses, mice were euthanized 32 days post-booster immunization and their splenocytes were in vitro restimulated with ionomycin/phorbol myristate acetate (PMA). Both mRNA vaccines induced CD4⁺IFNγ⁺ and CD8⁺IFNγ⁺ T-cells in the spleen (Fig. 7C and D). However, the DDO268-adjuvanted NP mRNA vaccine significantly increased numbers of double-positive (TNFα⁺IFNγ⁺) CD4⁺ and CD8⁺T-cells (Fig. 7C and E), suggesting that DDO268 facilitates T-cell activation. As IAV NP contains T-cells epitopes (25, 26), we also evaluated the induction of IAV-specific CD4⁺ T-cells and CD8⁺ T-cells. Both IAV NP mRNA and IAV NP mRNA/DDO268 or DDO268B formulations induced NP-specific CD8⁺ and CD4⁺ T-cells in the spleen. After in vitro IAV NP restimulation, the DDO268-adjuvanted mRNA vaccine induced significantly higher populations of IAV-specific CD8⁺ Tetramer (NP336-374)⁺ and CD4⁺ Tetramer (NP311-325)⁺ double-positive for TNFα⁺IFNγ⁺ (Fig. 7F and G). Additionally, a two-color (IFNγ and IL2) ELISpot assay showed higher double-positive spots for IFNγ and IL2 (Fig. 7I; Fig. S2F) in the DDO268 (or DDO268B)-containing formulation after in vitro restimulation. Overall, DDO268-adjuvanted formulations induced CD8⁺ and CD4⁺ T-cells characterized by both IFNγ and TNFα or IL2 production.

To evaluate the protective efficacy of our mRNA vaccines against challenge with a lethal dose of IAV PR8 (H1N1), mice were challenged intranasally with 40 TCID₅₀ of IAV PR8 39 days after booster immunization and monitored for weight loss for 20 days (Fig. 8A). In two independent experiments, mice inoculated with the IAV NP mRNA/DDO268 vaccine had a higher survival probability after a lethal dose of virus than mice vaccinated with LNPs containing IAV NP mRNA (Fig. 8B). Although all mice experienced bodyweight loss (Fig. 8C), the DDO268-containing formulation induced more robust specific type 1 immune response, effectively protecting against IAV infection. Viral RNA transcripts were assessed 7 days post challenge (Fig. 8D) with IAV NP transcripts detected in both groups but reduced in mice vaccinated with IAV NP mRNA/DDO268. At the conclusion of the experiments, mice were sacrificed, and the presence of IAV NP-specific memory effectors CD8⁺ and CD103⁺ CD8⁺ T-cells in the lungs was evaluated. As shown in Fig. 8E and F, both vaccines generated a population of effector memory CD8⁺ Tetramer⁺ (Fig. 8E) and CD8⁺ Tetramer⁺ CD103⁺ (Fig. 8F). However, the percentage of these cells was higher in mice vaccinated with IAV NP mRNA/DDO268, indicating stronger protective immune response elicited by the DDO-adjuvanted vaccine. Moreover, the DDO-containing formulation reduced lung pathology in challenged mice compared with vaccine lacking DDO268 (Fig. 8G).

C57BL/6 mice, Mavs^−/−, Tlr3^−/− (The Jackson Laboratory), and Ifnar1^−/− (provided by Dr. Thomas Moran) (37) were bred in-house. All mice were sex and age matched. Both male and female mice were included in the experiments.

IAV/Puerto Rico/8/1934 H1N1 was used as challenge strain. IAV was grown in 10-day-old embryonated chicken eggs as described in dx.doi.org/10.17504/protocols.io.5jyl8dm79g2w/v1↗.

The mRNA construct encodes the NP of IAV/Puerto Rico/8/1934 H1N1 (GenBank: ACV49549.1). The codon-optimized sequence (Twist Bioscience) was cloned into the mRNA production plasmid (pJB201.1) that contains a T7 type II promoter, the beta globin 5′ UTR, a Kozak sequence with an AG mutation allowing for 5′ capping, and the alpha globin 3′ UTR followed by a 142 base pair poly(A) tail. mRNA was produced as described in doi.org/10.17504/protocols.io.e6nvw11d7lmk/v1↗. mRNA size and integrity were assessed in a Agilent Bioanalyzer 2100. dsRNA contaminants were removed following the Baiersdorfer protocol (9) as described in dx.doi.org/10.17504/protocols.io.n2bvjnn3wgk5/v1↗.

IVT mRNA was tested in HEK-293t and A549 cells using jetMESSENGER (Polyplus) as transfection reagent. IAV NP production from the IVT mRNA was measured 24 and 48 h after transfection by intracellular staining using NP IAV antibody NBP3-12741AF647 (Novus Biological) followed by flow cytometry.

DDO268, DDO268B, and the control X region expressing plasmid (with a T7 promoter) were linearized and in vitro transcribed with Hi Scribe T7 (NEB). RNA products were DNase treated and precipitated with LiCl. The OD260/OD280 ratios were between 2.00 and 2.25, and the OD260/OD230 ratios between 2.20 and 2.60. RNA purity and integrity were confirmed using an Agilent Bioanalyzer 2100. Endotoxin level was below 0.1 EU/mL/300 µg.

NP mRNA, NP mRNA/X region, NP mRNA/DDO268, and NP mRNA/DDO268B were encapsulated in the GenVoy Ionizable Lipid Mix (ILM) using a NanoAssemblr Ignite machine (Precision Nanosystems) following the manufacturer's instructions. Encapsulation efficiency and RNA concentration were tested by Ribogreen Assay (Thermo Fisher). Final particle size was measured by DLS. mRNA and adjuvant RNA co-packaging was performed at a 1:1 molar ratio. All formulations contained the same total amount of RNA (about 0.5 µg total RNA per dose).

Mice were anesthetized with isoflurane and injected subcutaneously into a rear footpad. For toxicity studies, mice were injected with 50 µg of DDO268 or phosphate-buffered saline (PBS) at a final volume of 30 µL per dose. For immune response studies, mice were inoculated with BNT162b2 (0.125 µg); BNT162b2 (0.125 µg) + DDO268 (5 µg); BNT162b2 (0.125 µg) + RNA (5 µg); LNPs containing 1 ug IAV NP mRNA; LNPs containing 1 µg IAV NP mRNA/DDO268; LNPs containing 0.5 µg IAV NP mRNA; LNPs containing 0.5 µg IAV NP mRNA/X region; LNPs containing 0.5 µg IAV NP mRNA/DDO268; LNPs containing 0.5 µg IAV NP mRNA/DDO268B; empty LNPs or PBS at a final volume of 30 µL per dose. Leftover BNT162b2 SARS-CoV-2 vaccine was stored at −80°C prior to use in this study. Mice were primed and boosted 28 days later with the same vaccine formulations for adaptive immunity studies.

Mice were challenged intranasally with 40 TCID₅₀ of IAV-PR8 39 days after boost. All mice were weighed daily post-challenge, and survival probability was recorded based on an increase or decrease in mortality.

Peripheral blood samples were collected from mice via cardiac puncture and blood was collected into EDTA-coated tubes to prevent clotting. Samples were processed immediately and analyzed using an automatic hematology analyzer HEMAVET. The results were generated automatically by the analyzer software and recorded.

Peripheral blood samples were collected from mice via cardiac puncture and collected into serum-separating tubes. Samples were processed immediately following standard protocols for each parameter using ACE Axcel Clinical Chemistry System.

RNA was extracted from cells and tissue using TRIzol (Ambion Inc.) cDNA synthesis was performed with the High-Capacity cDNA Reverse Transcription kit (Applied Biosystems). For quantitative analysis by RT-PCR (qPCR), 10 ng/µL of cDNA was amplified using SYBR Green Mastermix (Thermo Fisher) in a Bio-Rad C1000 Touch thermal cycler (Bio-Rad). The primers used are listed in Table 1.

Footpads were processed following Fangzhou et al. (38) with modifications. Lymph nodes were digested using DNase (1 µg/mL) and Liberase (5 µg/mL) in RPMI 1640 Medium (Life Technologies) for 20 min at 37°C. Spleens were collected in RPMI 1640 Medium and dissociated. Lungs were inflated with 0.7 mL of digestion mix containing collagenase A (Sigma), dispase (Thermo Fisher), Liberase TL (Sigma), and DNAse I (Sigma) and incubated at 37°C for 30 min with agitation. Digested samples were filtered through a 70 µm mesh to obtain single-cell suspensions. Cells were washed with PBS containing 5% fetal bovine serum (FBS), treated with red blood cells lysis buffer (Sigma), and total viable cells were quantified with trypan blue staining using an automated cell counter (TC-20 Automated Cell Counter; Bio-Rad).

Flow cytometry experiments were performed using a Cytek Aurora spectral flow cytometer (Cytek Biosciences), with at least 1 × 10⁶ events acquired. Data were analyzed using FlowJo V12 software (Tree Star Inc.). Single-cell suspensions were stained with fluorochrome-labeled antibodies. Fixable Viability Dye eFluor506 and monoclonal antibodies specific for mouse IFNγ (XMG1.2), Ly6c (HK1.4), CD3 (17A2), CD19 (eBio1D3), B220 (RA3-6B2), NK1.1 (PK136), CD11b (M1/70), and TNFα (MP6-XT22) were obtained from eBioscience. Monoclonal antibodies specific for mouse CD3 (17A2), CD4 (GK1.5), CD11a (H11578), XCR1 (ZET), PDCA1 (129 cl), CD11c (cN418), SIRPα (P84), CD64 (x54/7.1), CD44 (NIM-R8), CD62L (cMEL-14), CCR7 (4B12), CD103 (QA17A24) and MHC-II (M5/114.15.2) were obtained from BioLegend. Monoclonal antibodies for mouse CD86 (GL1), and CD8 (53-6.7) were obtained from BD BioSciences. For IAV-specific tetramers, H-2D^b tetramers bearing NP366-374 (ASNENMETM) and I-A^b tetramers bearing NP311-325 (QVYSLIRPNENPAHK) were obtained from NIH Tetramer Core Facility at Emory University. APC-labeled SARS-CoV-2 S-specific tetramer was MHC class I tetramer, residues 539–546, VNFNFNGL, H-2Kb. For intracellular staining, cells were fixed/permeabilized with the FoxP3/Transcription Factor Staining Buffer Set (eBioscience) and incubated with anti-IAV NP antibody conjugated with AlexaFluor-647 (Invitrogen).

For each sample, 1 × 10⁶ cells were restimulated using PMA (0.1 mg/mL) and ionomycin (1 mg/mL) or purified IAV NP protein (Sino Biological) for 4 h at 37°C followed by brefeldin treatment. Cells were resuspended in PBS with 1% FBS (Sigma) and blocked with anti-mouse FcψRIII/II (CD16/32; BD Biosciences) for 20 min on ice. T-cell subpopulations were defined using a surface marker panel, with activation determined by intracellular staining for IFNγ and TNFα.

Sera from 21 post-boost immunized mice were analyzed for anti-SARS-CoV-2 Spike or anti-IAV NP IgG1 and IgG2c antibodies. ELISA plates (Nunc, MaxiSorp) were coated with 2 µg of purified SARS-CoV-2 Spike (kindly provided by Dr. Ellebedy, Washington University in St. Louis) or IAV-NP (Sino Biological) and treated with pre-diluted sera in triplicate, followed by HRP-conjugated anti-mouse IgG1 or IgG2c (Southern Biotech) and TMB substrate (Sera Care).

Mice lungs were perfused and inflated with 0.7 mL of OCT compound (Tissue-Tek) mixed 1:1 (vol/vol) with 4% paraformaldehyde (Electron Microscopy Sciences) diluted in PBS. Inflated lungs were snap-frozen and stored at −80°C until sectioning. Tissue sections (4 µm) were stained with hematoxylin and eosin, and chronic lung disease was scored for peribroncholitis and goblet cell metaplasia. Area affected was determined, multiplied by the intensity scores previously defined, and the resulting weighted scores were graphed.

Double-color fluorescent ImmunoSpot (CTL) was used for the detection of primed IFN-γ-IL2-producing Th1-type memory T-cells. In brief, IAV NP or ionomycin/PMA (mitogen) were plated at the specified concentrations into capture antibody-precoated ELISpot assay plates. Then, 200,000 viable cells/well in 100 µL CTL-medium and cultured with the NP or mitogen for 24 h at 37°C and 9% CO₂. After addition of detection antibody, the plates were scanned and spot-forming units were counted and analyzed using an S6 Ultra M2 Fluorospot reader. Spot numbers were automatically calculated using the Autogate function of the ImmunoSpot Software.

Statistical significance was inferred using GraphPad Prism software version 10.0 (GraphPad Software, San Diego, CA). Unpaired t-test, one-way and two-way ANOVA with Bonferroni's multiple comparison test were used. Log-rank Mantel-Cox test was used for survival. For animal experiments, group size consisted of 3–5 mice per group.