Broadly Active Intranasal Influenza Vaccine with a Nanocomplex Particulate Adjuvant Targeting Mast Cells and Toll-like Receptor 9

Abstract

Flumist is the only FDA-approved intranasal influenza vaccine. Although it has recently been approved for at-home use, it has significant limitations. These include reduced effectiveness in generating a protective immune response in patients with extensive influenza exposure, safety concerns due to its live attenuated virus formulation, and reduced efficacy due to viral drift/shift. To address this limitation, we have developed a nanocomplex adjuvant comprised of a mast cell (MC) agonist and toll-like receptor 9 (TLR9) ligand to adjuvant a broadly acting influenza vaccine. The newly reported MC agonist was identified by screening mastoparan-7 analogs for MC degranulation activity, which led to a more active peptide analog, MP12W. Positively charged MP12W spontaneously forms nanoparticulate complexes (NPs) with CpG 1826 that were then used to intranasally vaccinate mice with a computationally optimized broadly reactive antigen (COBRA) hemagglutinin (HA) protein. The NPs were further optimized by substituting CpG 1826 with CpG 55.2, a TLR-9 agonist identified by machine learning to be more active in humans. MP12W-CpG 1826 NPs showed increased pro-inflammatory response and decreased cytotoxicity in vitro compared to M7 complexes, translating into a safer profile in CC027 mice. Intranasal vaccination with this complex and broadly reactive HA resulted in higher mucosal antibody concentration and increased cytokine production with antigen recall. These responses were enhanced with MP12W-CpG 55.2 NP vaccination. MP12W-CpG NPs provided similar protection in an influenza challenge model. This study demonstrates the potential of this novel intranasal nanocomplex for vaccination.

Graphical Abstract

Introduction

New mucosal adjuvant systems are needed to develop safe and effective mucosally-administered vaccines that better protect against respiratory viruses like influenza and SARS-CoV-2 (COVID), when compared to traditional parenterally-administered vaccines. Infections from Influenza virus alone represent a worldwide public health problem, with an estimated 3–5 million cases of severe illness and up to 650,000 respiratory deaths globally each year.[1] Adding to that, COVID has led to ~7M deaths worldwide (WHO).[2] Despite the availability of antiviral medications, vaccination continues to be the most effective strategy for preventing respiratory infections like influenza and COVID.

Developing a vaccine that provides mucosal protection is highly desirable for combating respiratory infections. An intranasal (i.n.) vaccine is particularly advantageous because it eliminates the need for needles and can be easily administered in resource-limited settings. It can also provide stronger mucosal immune responses by locally targeting the mucosal surfaces to stop the virus at the infection site and decrease the viral spread.[3–5] Following the COVID pandemic, the US government launched the “American Pandemic Preparedness: Transforming Our Capabilities” initiative, which calls for the development of i.n. vaccines. These measures aim to streamline vaccine administration during pandemics.

New approaches for i.n. and thereby mucosal vaccination are needed. mRNA lipid nanoparticles (LNPs), like those used for the COVID vaccine, have been shown to have limitations when delivered i.n.[6] The influenza vaccine FluMist is the only FDA approved i.n. vaccine, a Live Attenuated Influenza Vaccine (LAIV) adjuvanted through the natural pathogen-associated molecular patterns (PAMPs) inherent to the virus. FluMist cannot be administered to immunocompromised patients (e.g., individuals with HIV or those undergoing treatment for cancer or autoimmune diseases) due to safety concerns. Furthermore, FluMist is mainly administered to children, who have less extensive exposure to influenza compared to adults (i.e., reduced antigenic imprinting).[7] FluMist sales reached $206 million during the 2015–16 flu season, marking a 20% increase from the previous year and demonstrating a strong market potential for an i.n. influenza vaccine.[8] However, despite this success, the CDC withdrew FluMist for the 2016–17 and 2017–18 flu seasons due to concerns over low vaccine efficacy. In September 2024, FluMist was approved for at-home use, which will likely increase sales further.

Compared to LAIV, protein or subunit vaccines offer a safer alternative with broader applicability across the population. However, subunit vaccines must often be adjuvanted to achieve a protective immune response. A major challenge in developing mucosal subunit vaccines is the scarcity of safe and effective adjuvants that can be incorporated into the vaccine formulation. For instance, while cholera toxin is regarded as the gold standard mucosal adjuvant in animal models, it is not safe for human use.[9, 10] To overcome key limitations of i.n. vaccination—such as poor durability and increased patient variability compared to intramuscular administration—an effective i.n. adjuvant is essential.[11]

Adjuvants are important components that increases the immunogenicity of vaccines through the activation of the innate immune system. Intranasal adjuvant activity can be mediated through mast cells (MCs), which are highly prevalent in the dermal region of the skin and in the mucosa of the respiratory, gastrointestinal, and urinary tracts.[12] MCs store vast quantities of inflammatory mediators (e.g., TNFα, histamine, tryptase) that are released immediately upon activation. MCs are implicated in rapid, robust, and sustained inflammatory responses (e.g., allergic hypersensitivity) and play roles in immunity against infectious pathogens.[13, 14] They regulate the recruitment of immune cells to regional lymph nodes following bacterial exposure, a process dependent on MC production of TNF-α.[15] This recruitment can be replicated with chemical MC activators such as compound 48/80 (c48/80), which does not require TLR ligands.[16–18]

c48/80 provides safe and effective adjuvant activity for i.n. delivery of anthrax protective antigen (PA), comparable to the gold-standard mucosal vaccine adjuvant cholera toxin (CT). A critical concern when targeting MCs is the potential for IgE-mediated allergic reactions, as IgE is the primary marker of MC-mediated anaphylaxis in humans. Importantly, intradermal (i.d.) injection and i.n. delivery of the MC activator c48/80 shows pronounced therapeutic efficacy without generating IgE antibodies. In contrast, CT generates an IgE response.[16] Despite the role of MCs in allergic reactions, the use of MC activators as vaccine adjuvants is safe, with no signs of allergic reactions or elevated antigen-specific IgE.[19] Moreover, commonly used MC activators, such as polymyxins, have been shown to be safe in several species, including humans, and are safe when delivered i.n. or as an aerosol.[20] Other MC agonists are derived from mastoparan, an insect-derived oligopeptide. Mastoparan served as the basis for the second-generation analog M7 (i.n.LKALAALAKALL-NH2).[21] Mastoparan activates MCs and other cells by stimulating Mrgprb2 and MRGPRX2 receptors, GTP-binding proteins, phospholipase A2, and phospholipase C.[22, 23] This activation has been shown to induce the release of pro-inflammatory cytokines TNF-α and IL-1β.

There are several FDA approved adjuvants to consider for i.n. use, one of which is CpG. CpG 1018 is FDA approved for intramuscular delivery in the subunit hepatitis vaccine Heplisav-B. Further, CpG 7909 is FDA approved for the anthrax vaccine Cyfendus. Toll-like receptor (TLR) ligands like CpG have been widely used as vaccine adjuvants. CpG 1826, a mouse TLR-9 ligand, has been shown to increase antigen-specific antibodies and provide a robust Th1 response in mice and rats.[24, 25] CpG shows promise as an intranasal adjuvant [24–26].

In addition to adjuvant selection, antigen selection can be improved to generate more robust vaccines against respiratory infections like influenza. Regardless of the type of seasonal influenza vaccine, its effectiveness fluctuates annually, primarily due to the match between circulating strains and the vaccine formulation, which is influenced by antigenic shift. New and more effective influenza vaccines are needed as exemplified during the 2023–2024 influenza season where the vaccine effectiveness was on average 42%.[26]

One strategy to improve influenza vaccine efficacy is the use of Computationally Optimized Broadly Reactive Antigens (COBRA). COBRA HA antigens elicit potent, broadly reactive antibody responses, protecting against both seasonal and novel pandemic strains, including those that have undergone genetic drift.[27] COBRA has been used to address the diversity of H5N1 highly pathogenic avian influenza,[28–32] seasonal H3N2 strains,[33] and seasonal and pandemic H1N1 subtypes.[27, 34] Additionally, H3N2 COBRA antigens were tested against contemporary and future co-circulating isolates, successfully neutralizing these viruses.[35, 36] Importantly, antigen imprinting—a limitation of FluMist—was evaluated through prior exposure to historical strains, and the protection afforded by COBRA was not significantly different with pre-exposure in a ferret model of influenza.[36]

One application of COBRA HA is for i.n. vaccination. We previously reported that the combination of M7 and CpG 1826 forms a nanoparticulate complex (NP) that, when combined with COBRA H3 HA, results in a robust and protective immune response when given i.n.[37] Mice vaccinated i.n. with COBRA HA and M7-CpG NPs showed significantly higher antigen-specific antibody responses (IgG, IgA) and increased cytokine production. This vaccination induced H3N2-specific hemagglutinin inhibition antibody titers across several H3 influenza strains and partially protected mice from H3N2 virus challenge.

Building on our success with the M7-CpG 1826 NPs we sought to incorporate more potent emerging adjuvants to replace both M7 and CpG 1826, in addition, since M7 showed in vitro cytotoxicity traits we aimed to create an adjuvant system with safer properties. To replace M7, we screened mastoparan analogs, evaluating them based on MC degranulation to identify analog MP12W. We also substituted CpG 1826 in the NPs with CpG 55.2, a new computationally optimized TLR-9 ligand. CpG 55.2 induces robust adjuvant activity when combined with Advax (a delta inulin polysaccharide adjuvant). Together CpG 55.2 and Advax are being assessed in preclinical and clinical vaccine studies (NCT04944368 and NCT05175625).[38–40] In this work, we evaluated the humoral, cellular, and protective response of our new NP formulations in combination with ovalbumin (OVA) or H3 COBRA (J4) in mice and mouse cell lines.

Materials and Methods

Materials

Unless otherwise stated, all chemicals, assays, and biologicals were purchased from Millipore Sigma (St. Louis, MO) and Thermo Fisher Scientific (Waltham, MA). H3 COBRA (J4) was obtained from a consensus sequence of different circulating H3N2 influenza strains using the methodologies previously described.[37, 41, 42]

M7 and analog generation and evaluation

To generate mastoparan analogs, a single amino acid substitution with all possible amino acids was introduced into the mastoparan sequence at amino acid position 6, 12 or 13. peptide sequence at positions. The peptides were synthesized by CPC Scientific (Sunnyvale, CA).

The mastoparan analogs were subsequently evaluated in vitro for their ability to induce degranulation of MC/9 cells. Mouse MCs MC/9 (ATCC) were cultured in DMEM medium supplemented with 2 mM L-glutamine, 0.05 mM 2-mercaptoethanol, 10% Rat T-STIM (BD, Franklin Lakes, NJ, USA), and 10% fetal bovine serum (VWR, Radnor, PA, USA) (5 × 10⁵ cells/well). MC/9 cells were treated with varying concentrations of each mastoparan analog.

MC degranulation levels were measured by β-hexosaminidase release with a method previously described.[37] Briefly, MC/9s were cultured in Tyrode’s buffer for the stimulation, 100% degranulation cells were treated with 0.1% Triton X-100 at 37°C, and supernatants were collected after 30 min. Supernatants were incubated with p-nitrophenyl-N-Acetyl-β-D-glucosaminidase (NAG) substrate solution for 1 h at 37°C and then carbonate buffer pH 10 was added to develop a colored substrate that was measured at 405 nm. The peptide concentration required to provide 10% and 50% mast cell degranulation were calculated.

NPs complexation and characterization

MP12W peptide (i.INLKALAALAKWIL-NH2) (10 nmol = 15 μg; Biomatik, Wilmington, DE, USA) from a stock of 1 mg/ml in PBS was combined with 10 μg of CpG ODN 1826 (Nitrogen/ Phosphate (N/P) ratio of 1/1) (Invivogen, San Diego, CA, USA) or CpG 55.2 (Vaxine, Adelaide, Australia) from a stock solution of 1 mg/ml of CpG in PBS. The mixture was sonicated in a water bath at 100% amplitude (Branson, Danbury, CT, USA) for 20 min and at room temperature (RT). Particle size of MP12W-CpG 55.2 complexes were determined using dynamic light scattering (DLS) (Brookhaven, Holtsville, NY, USA). Particle shape and size were also evaluated using scanning electron microscopy (SEM) (Hitachi S-4700, Japan). The endotoxin content of MP12W, CpG 1826, CpG 55.2, and MP12W-CpG combinations was analyzed via a Limulus amoebocyte lysate (LAL) endotoxin assay. All samples had undetectable endotoxin levels (<0.1 EU/ml at 1 mg/ml).

In vitro cytotoxicity and innate immune response

Mouse dendritic cells (DCs) DC2.4 (ATCC, Manassas, VA, USA) were cultured in RPMI 1640 medium (Corning, Corning, NY) supplemented with 10% fetal bovine serum (VWR) and 10 U/ml penicillin–streptomycin in non-TC treated flat-bottom 96-well plates (25K cells/well).

DC2.4 and MC/9 cells were stimulated with MP12W, CpG 1826, CpG 55.2, MP12W-CpG 1826 or MP12W-CpG 55.2 NPs at different dilutions. Culture supernatants were taken 24 h post-stimulation for a colorimetric lactate dehydrogenase (LDH) release cytotoxicity assay. The percentage of LDH release was normalized to 0% (untreated cells) and to 100% dead positive control (cells treated with 1× lysis buffer for 10 min). Cytokine ELISAs were also performed on supernatants according to the manufacturer’s protocols (Thermo) to assess TNF-α or IL-6 production.

Adjuvanticity evaluation of M7 analogs

All animal experiments were performed in accordance with the UNC or Duke Institutional Animal Care and Use Committee (IACUC). C57BL/6 mice were intranasally (i.n.) immunized with 2.5 μg of SARS-CoV-2 Spike protein alone or combined with 5 nmoles or 0.5 nmoles of M7, MP12W, MP12L, or MP12F in a total volume of 15 μl on days 0 and 14. Serum was collected on day 21 and evaluated for anti-Spike IgG titers using an ELISA, as explained in the next method’s section.

Safeness evaluation of MP12W-CpG vs M7-CpG NPs

CC027 (collaborative cross recombinant) mice were obtained from the UNC Animal Model Core Facilities and immunized with 10 μg of OVA combined with PBS, M7-CpG 1826 NPs or MP12W-CpG 1826 NPs (14.22 μg MC agonist + 10 μg CpG), or Cholera Toxin B subunit (CTB) on days 0, 21 and 35. Another group of mice was treated with CTB 3 times a week (3X) as a positive allergy control on days 0, 3, 5, 21, 23, 25, 35, 37 and 39. Rectal temperatures of mice were measured using a type T thermometer (Thermoworks, American Fork, UT) every 15 minutes and every hour after the prime or boost vaccinations.

Immunogenicity evaluation of MP12W-CpG 55.2 NPs

C57BL/6 mice were i.n. immunized on days 0, 21, and 35 with PBS, CpG (1826 or 55.2; 10 μg), M7 (10 nmol), MP12W (10 nmol), M7-CpG or MP12W-CpG NPs (10 nmol + 10 μg) combined with OVA or J4 COBRA. Submandibular blood and fecal samples were collected on days 14, 28, and 42 post-immunization. Depending on the experiment, mice were challenged with H3N2 virus or were sacrificed on day 42, when bronchoalveolar lavages (BALs), nasal washes, and spleens were collected.

Influenza challenge

Immunized mice were i.n. challenged on day 56 post-immunization with 2.5K PFU (50 μl) of the mouse-adapted virus A/Hong Kong/1/68 (HK/68; BEI resources). Mice were weighed daily for 14 days, according to previous reports. [43, 44] A 20% loss of body weight was used as survival endpoint.

Nasal samples collection

To perform BAL collection, a 22G catheter was inserted in each mouse’s trachea, and 1 ml of PBS supplemented with complete protease inhibitor (1 tablet/10 ml) (Roche) and 0.01% Triton X-100 (Sigma) was flushed 3X towards the lungs. To perform nasal washes, the catheter was re-inserted in the trachea and 1 ml of BAL buffer was flushed towards the nose. Wash and BAL fluids were collected in sterile 1.5-ml microcentrifuge tubes.

Antibody response (sera and mucosal titers)

Sera were obtained by centrifugating the blood from different days for 10 min at 3,000 × g, aliquoted, and stored at −80°C. BAL and nasal wash samples were aliquoted and stored at −80°C. Fecal samples were snap-frozen and stored at −80°C until analysis. On the day of analysis, fecal samples were defrosted, diluted to 10 mg / 100 μl in protein extraction buffer [10% goat serum (MP Biomedical, Santa Ana, CA) in PBS], and vortexed for 20 min at 1200 rpm or until fully dispersed. Fecal samples were centrifuged for 10 min at 13,000 × g, and supernatants were collected for ELISAs. For the evaluation of antibody titers, all frozen samples were defrosted and used directly for ELISAs

High-binding flat bottom 384-well plates (Greiner Bio-One) were coated overnight with COBRA J4 (100 ng/ml) or OVA (5 μg/ml) in PBS at 4°C, washed 3X with 0.05% Tween 20 in PBS (PBST) using a plate washer (Biotek ELX405). Plates were blocked for 2 h with 3% non-fat instant milk in PBS (blocking buffer) at RT and washed 3X with PBST. Samples (serum, BAL, nasal or fecal) were diluted 1:100 in blocking buffer and added to the plate and diluted 5X along consecutive rows. Plates were incubated for 1 h at RT and washed 3X (PBST). The different HRP goat anti-mouse secondary antibodies (IgM, IgA, IgE, IgG, IgG1, IgG2c) were diluted in blocking buffer to the highest dilution recommended by the manufacturer (Southern Biotech) and added to the plates. Plates were incubated for 1 h at RT, washed 5X with PBST, treated with one-component substrate tetramethylbenzidine (TMB) until developed. The enzymatic reaction was stopped with 25 μl of 2 N sulfuric acid and ODs were read at 450 and 570 nm in a plate reader (Biotek). 570nm ODs were subtracted from the 450 nm ODs, and endpoint titers were defined as the lowest serum dilution, with the absorbance three standard deviations above the average negative control.

Hemagglutinin inhibition (HAI) titers

Sera from the vaccinated mice were treated as previously described [45] and stored at 4°C until analysis. Sera were added to 96-well V-bottom microtiter plates and diluted 2X serially in PBS across the plate. 25 μl of a 8 HAU/50 μl H3 influenza virus stock was added and incubated for 20 min at RT. Horse red blood cells (HRBC) (Lampire Biologicals) in PBS were added and incubated at RT for 1 h. Hemagglutinin inhibition (HAI) titers were calculated as the dilution for the last non-agglutinated well.

Evaluation of the T cell response (Antigen Recall, ELISpot and Flow Cytometry)

Spleens were harvested from the vaccinated mice on day 42 and processed into single-cell suspensions. Splenocytes were seeded at 1,000,000 cells/well on a 96-well PVDF multiscreen plate to perform ELISpot (wells were previously coated with 100 μl of anti-IL-2 or anti-IFN-γ capture antibody) or a U-bottom cell culture 96-well plate to perform for flow cytometry. Cells were stimulated with 10 μg/mL of full OVA protein for 36 h at 37°C.

After antigen recall stimulation, the cell suspension on the ELISpot plates was discarded, and wells were washed. 100 μl of anti-IL-2 or anti-IFN-γ detection antibody diluted in 10% FBS in PBS (diluent buffer) were added to the plates according to the manufacturer’s protocol (BD). Plates were incubated for 2 h at room temperature and washed 3X with PBST. 100 μl of Streptavidin-HRP (BD) diluted in diluent buffer were added to the plates and incubated for 1 h at RT. Plates were washed 4X with PBST and then 1X with PBS. 100 μl of substrate (3-amino-9-ethylcarbazole (AEC)) were added until spots were observable. Plates were then washed 2X with milli-Q water and dried at RT for 48 h. Spots were counted using an ImmunoSpot plate reader (CTL).

After completing the antigen recall stimulation, the supernatants on the flow cytometry plates were discarded, and cells were stained for flow cytometry with the next anti-mouse antibodies: CD3-AF488 (BioLegend), CD44-APC (BioLegend), CD4-BV421 (BD), CD62L-BV785 (BD) and efluor 506 viability dye (Thermo). Cells were analyzed in an Attune NxT flow cytometer.

Statistics

All the studies evaluated across different concentrations were analyzed with a two-way ANOVA test with post-hoc Tukey’s multiple comparisons. The remaining studies were analyzed using a one-way ANOVA test with post-hoc Tukey’s multiple comparisons. Survival studies were analyzed using Kaplan–Meier survival statistics. All tests were performed using the software GraphPad Prism 10^® with 0.05 as the lowest p-value.

Results and Discussion

Increased Indicators of Safety and Efficacy with MP12W-CpG NPs compared to M7-CpG NPs

The new peptide analog sequences were selected based on the differences of the mastoparan sequence (INLKALAALAKKIL-NH2), M7 sequence (INLKALAALAKALL-NH2) and the sequence of its inactive analog, mastoparan-17 [46] (INLKAKAALAKKLL-NH2, modified amino acids in bold) where it was evident that positions 6, 12 and 13 might play a critical role in the peptide biological activity. New peptides were generated with amino acids substituted in the sixth, twelfth and thirteenth position of the mastoparan sequence. We tested the degranulation activity of the peptides in vitro with MCs, identifying that some amino acid substitutions on the twelfth position led to enhanced degranulation activity. When the twelfth-place amino acid was substituted with tryptophan (W), MP12W was formed, and resulted in greater degranulation activity in vitro (Fig. 1A). Substitutions on sixth and thirteenth position were also capable to degranulate MCs but with a lower activity than M7 (Table S1).

Figure 1. M7 derivatives, their MCD50 and adjuvant activity.

. (A) Table showing the name and sequence of the top 7 M7 analogs ordered by their capacity to in vitro degranulate 50% of mast cells (MCD50). (B) C57BL/6 mice (n=3) were i.n. immunized on day 0 and 14 with Spike protein (2.5 μg) alone or combined with a normal dose (5 nmol) or a low dose (0.5 nmol) of M7, MP12W, MP12L, or MP12F. Serum was collected on day 21 and tested for anti-Spike IgG titers. Data is shown as mean ± SD. Significant difference for a one-way ANOVA test compared versus “no adjuvant” group * p≤0.001.

The top three analogs (MP12W, MP12L and MP12F) were evaluated in vivo as an intranasal subunit vaccine against COVID. Mice were vaccinated i.n. on a prime-boost schedule, and day 21 serum was evaluated for anti-spike IgG. Although all analogs produced similar levels of IgG to M7 at a higher 5 nmol dose, MP12W induced significantly greater IgG titers than the other analogs when given at 0.5 nmol (Fig. 1B). Because MP12W outperformed M7 and the other analogs, we evaluated it further in combination with CpG.

Combining a T-helper type 2 (Th2) adjuvant, like an MC agonist, with a second Th1 skewing adjuvant can enhance protection against respiratory infections like influenza.[47] Zheng et al. showed that the MC activator, c48/80, induces protective immunity against influenza.[48] c48/80 promoted elevated type 1 CD4+ and CD8+ T cell reactivity coupled with an influenza-specific antibody response made up of both IgG2a and IgG1.[48] Additionally, induced memory Th1 and Th2 cells have been shown to provide long-term viral protection.[49] Our previous work showed that M7-CpG NPs invoke a balanced Th1 and Th2 humoral response significantly improved from the individual adjuvants alone with H3 COBRA (J4). Together with a Th1 skewing TLR ligand like CpG, an MC agonist can aid in promoting both Th1 and Th2 immune response.

After identifying MP12W as the top performing M7 analog, we combined it with CpG 1826 to form a nanocomplex particle (NP; Fig. 2A). When we cultured NPs with dendritic cells to evaluate cytotoxicity and pro-inflammatory cytokine production, we observed a significant decrease in LDH release (Fig. 2B; S1A) and increased levels of TNF-α and IL-6 when cells were stimulated with MP12W-CpG 1826 versus M7-CpG 1826 (Fig. 2C, D; S1B, C). Overall, the substitution of M7 with MP12W in the formulation induced higher proinflammatory cytokines and lower cytotoxicity.

Figure 2. Evaluation of cytotoxicity and induction of proinflammatory cytokines of MP12W-CpG 1826 NPs.

(A) SEM image showing nanocomplex particles (NPs) when MP12W and CpG 1826 are combined. DC 2.4 cells were stimulated with M7 or MP12W (64 μM), CpG 1826 (8 μM), and M7-CpG 1826 NPs or MP12W-CpG 1826 NPs (64 μM + 8 μM). Supernatants were tested for (B) LDH release, (C) TNF-α and (D) IL-6. Data is shown as mean ± range. Significant differences were determined using a one-way ANOVA test with Tukey’s multiple comparison ** p≤0.01, *** p≤0.005, **** p≤0.001.

To add to the reduced cytotoxicity observed in vitro, we evaluated M7-CpG NPs in a collaborative cross (CC) mouse that has increased hypersensitivity (mouse strain CC027).[50, 51] The use of MC activators as adjuvants often evokes safety concerns due to the possibility of an anaphylactic response (a type I hypersensitivity reaction). CC027 mice have uniquely shown enhanced MC-mediated responses in comparison to wild-type C57BL/6 mice.[52] The CC is a large panel of recombinant inbred mouse strains[53] derived from eight founder strains (A/J, C57BL/6J, 129S1/SvImJ, NOD/ShiLtJ, NZO/HiLtJ, CAST/EiJ, PWK/PhJ, and WSB/EiJ). The genome of each mouse line has equal contributions from the original founder strain, and the recombination’s that accumulate during the breeding process are independent between lines. CC027 has been found to be more sensitive to anaphylaxis because of a genetic variant in the Themis gene that results in an increase in immature T cells.[50, 51] To assess the potential for MC-CpG NPs to induce IgE and anaphylaxis, we vaccinated CC027 mice i.n. on a prime-boost-boost schedule. We measured mouse rectal temperatures after immunization because changes in body temperature are indicative of hypersensitivity.[54] Our NP formulation was compared to Cholera toxin B subunit (CTB) as a positive control, also given i.n.

We observed that the change in body temperature in mice immunized with MP12W-CpG 1826 NP prime-boost or prime-boost-boost was lower than in mice vaccinated with M7-CpG 1826 NPs. In addition, the temperature change for MP12W-CpG 1826 NP vaccinated mice was similar to the average temperature change observed in the OVA control group, a 1–2°C reduction (Fig. 3A-C). We corroborated the safer temperature profile with MP12W-CpG vaccination by comparing the Area Under the Curve (AUC) of the temperature curves, observing that M7-CpG 1826 showed the greatest AUC among all the groups (Fig. 3D). We then measured antigen-specific IgE titers and observed that neither M7 or MP12W-CpG 1826 induced this anaphylaxis associated antibody in the sera (Fig. 3E), whereas CTB vaccinated mice, given on a prime or prime-boost-boost schedule had significantly higher levels of serum IgE. Together, these data indicate that MP12W-CpG NPs are less prone to induce anaphylaxis than M7-CpG NPs or CTB.

Figure 3. Assessment of body temperature and specific IgE titers of MP12W-CpG1826 in allergy-prone mice.

CC027 mice (n=5) were intranasally immunized on days 0, 21, and 35 with OVA alone (non-adjuvanted) or OVA combined with M7-CpG 1826 NPs, MP12W-CpG 1826 NPs, CTB once a week (CTB 1X) or CTB thrice a week (CTB 3X). Rectal temperature was recorded at different time points (0–5 h) post-immunization during the (A) prime, (B) first boost, and (C) second boost. (D) Area under the curve of the rectal temperatures obtained on A-C. (E) Day 42 sera was tested for OVA-specific IgE titers. Data is shown as mean ± SD. (A-C) Significant differences were determined using a two-way ANOVA test with Tukey’s multiple comparisons vs. OVA alone groups at the same time point * p≤0.05, ** p≤0.01, *** p≤0.005. (D-E) Significant differences were determined using a one-way ANOVA test with Tukey’s multiple comparison * p≤0.05.

Cationic peptides, such as M7 and MP12W, have been reported as potential cytotoxic compounds. [55, 56] However, in MP12W, the presence of tryptophan (a bulky amino acid) adjacent to lysine (a positively charged amino acid) may create a masking effect that reduces its cationic properties, leading to lower cytotoxicity while maintaining or even enhancing its degranulation activity. Although increased degranulation activity could raise safety concerns for this adjuvant, MP12W has demonstrated a favorable in vivo safety profile in an hypersensitivity model. Notably, MP12W did not induce the drastic temperature fluctuations observed with M7 in CC027 mice and also prevented the generation of antigen-specific IgE titers, both key indications of MC driven immune responses [57, 58].

To evaluate the immune response of MP12W-CpG 1826 NPs in a mouse i.n. vaccination model, we immunized C57BL/6 mice with OVA and MP12W-CpG 1826 NPs and compared it to M7-CpG 1826 NPs. When we evaluated the antibody response, we observed no significant differences in serum IgG between both groups (Fig. S2A). However, the BAL IgG (Fig. 4A) and nasal IgA titers (Fig. 4B) were higher in mice immunized with MP12W-CpG 1826 NPs. Further, the antigen recall of splenocytes from vaccinated mice showed that the number of cells expressing IL-2 (Fig. S2B) and IFN-γ (Fig. 4C) was higher in mice immunized with MP12W-CpG 1826 NPs. In addition, increased numbers were observed, but for central memory CD4+ T cells in the spleen and draining lymph nodes (dLN) for the MP12W-CpG 1826 NP vaccinated mice compared to those with an M7 based NP (Fig. 4D, E). Replacing M7 with MP12W in the nanocomplex increased the generation of mucosal antibodies and an increased cellular response.

Figure 4. Analysis of mucosal antibody titers and T cell responses of M7/MP12W and CpG 1826 NPs.

C57BL/6 mice (n=5) were intranasally immunized on day 0, 21 and 35 with PBS, OVA alone (non-adjuvanted), or OVA combined with CpG 1826, M7, MP12W, M7-CpG 1826 NPs or MP12W-CpG 1826 NPs. (A) Anti-OVA IgG titers on BAL and (B) Anti-OVA IgA titers on nasal washes from the immunized mice at day 42. Spleen and LNs from the different groups of mice were collected and processed at day 42, splenocytes were stimulated with OVA for 36 h. (C) IFN-γ-producer cells were detected by ELISPOT and counted. Total counts of central memory CD4+ T cells on (D) spleen and (E) LN by flow cytometry. Data is shown as mean ± range. Significant differences were determined using a one-way ANOVA test with Tukey’s multiple comparisons * p≤0.05, ** p≤0.01.

When we immunized mice with MP12W-CpG 1826 NPs and H3 COBRA (J4), and as observed with OVA, there were similar serum IgG titers between vaccination groups with NPs containing MP12W or M7 (Fig. S3A). Previously, vaccination with M7-CpG 1826 NPs resulted in full protection from an H3N2 i.n. challenge.[37] For that reason, we increased the challenge dose with this study. At the increased dose, the challenge results showed a similar survival percentage between MP12W and M7 combined with CpG 1826 (Fig. S3B). The influenza virus used for our challenge (HK/68) resulted in a marked difference between the control and adjuvanted J4 groups. However, HK/68 is not included in the consensus sequence of COBRA J4. The J4 sequence was composed of HA sequences from virus strains dating from May 2013 to April 2016, whereas our challenge strain was H3N2 A/Hong Kong/1/1968 isolated in 1968. This strain was chosen because it is one of very few H3 strains that are lethal to mice. This displays how COBRA can be used to protect from strains outside the sequence range, and broad neutralization has been shown before for COBRA proteins that pre-date and post-date the consensus date range.[34–36, 59, 60] Similar protection levels observed in our study between vaccination with M7-CpG 1826 and MP12W-CpG1826 NPs suggest that the higher antibody isotypes and central memory T cell response observed in the MP12W group were not enough to see differences or they do not play an important role in this challenge model, nevertheless, they have been described as important protective mechanisms against influenza infection [61, 62].

Overall, the substitution of M7 with MP12W in the adjuvant combination resulted in higher local antibody responses. This is likely because MP12W is a more potent MC activator and MCs are primarily located in tissues with high exposure to foreign antigens (like the nose), therefore, their activation leads to an increase in the lymphatic permeability to activate T and B cells in the dLN [63–65]. This hypothesis also explains the higher number of central memory T cells in the dLN. Nevertheless, we also noticed a higher number of memory T cells in the spleen and higher numbers of IL-2 and IFN-γ-producing splenocytes, which tells us that the local lymphatic cell trafficking might expand to reach a systemic effect.

Replacement of CpG 1826 with CpG 55.2 Results in Increased Correlates of Protection

After choosing MP12W as a safer and more immunogenic option than M7, we then replaced CpG 1826 with the human optimized TLR-9 ligand CpG 55.2 and compared their adjuvant activity in combination with MP12W. Like CpG 1826, CpG 55.2 combined with MP12W formed complexed NPs, as observed via SEM micrographs, noting the size was around 200 nm (Fig. 5A) as verified with DLS (245.05 ± 2.0 nm). A particle size of around 200 nm diameter is desirable as it can target DCs and induce a robust CD4+ T cell response after an intranasal immunization [66, 67], in addition, particles in the nanoscale have better mucosal permeability and lymphatic drainage than particles bigger than 1 micron [68, 69].

Figure 5. Evaluation of sera IgM, HAI titers and protective response of MP12W-CpG 55.2 NPs.

(A) SEM image showing nanocomplex particles (NPs) when CpG 55.2 and MP12W are combined. C57BL/6 mice (n=5) were intranasally immunized on days 0, 21, and 35 with PBS (Saline) or OVA combined with CpG 1826, CpG 55.2, CpG 1826-M7 NPs, CpG 1826-MP12W NPs, CpG 55.2-M7 or CpG 55.2-MP12W NPs. (B) Sera was collected on days 14, 28, and 42 and run for anti-OVA IgM titers. (C) Splenocytes from the immunized mice were stimulated on day 42 with OVA for 36 h, and IL-2-producer cells were detected and counted using ELISPOT. C57BL/6 mice (n=5) were intranasally immunized on day 0, 21 and 35 with PBS (Saline), or COBRA J4 combined with CpG 1826, CpG 55.2, CpG 1826-MP12W NPs or CpG 55.2-MP12W NPs. (D) On day 42 post-immunization sera were collected and tested for anti-SW/13 HAI titers and (E) on day 56 mice were intranasally challenged with 2.5K pfu of A/Hong Kong/1/68 influenza virus and survival was analyzed for 14 days post-challenge. Data is shown as mean ± range (A-D) or survival percentage (E). Significant differences were determined using a one-way ANOVA test with Tukey’s multiple comparison ** p≤0.01, *** p≤0.005, **** p≤0.001.

We immunized C587BL/6 mice with MP12W-CpG 55.2 or MP12W-CpG 1826 NPs in combination with OVA as antigen in a prime-boost-boost schedule. Significant differences in total IgG serum or nasal IgA titers were not observed (Fig. S4), but significant differences in sera IgM titers were noted for the MP12W-CpG 55.2 NP vaccinated group (Fig. 5B). IgM is a crucial antibody to achieve protection in vaccination, it has been described as one of the main humoral protective mechanisms against bacterial and viral pathogens [70, 71] and plays an important role in the cross-reactivity against respiratory viruses like SARS-CoV-2 [72]. In addition to higher sera IgM, we noted a significant increase in HAI titers from serum isolated from the MP12W-CpG 55.2 NP vaccinated group compared to the formulation with CpG 1826 (Fig. 5D, S4D-F). Additionally, significant differences were observed in cytokine production with antigen recall for splenocytes. Cells from MP12W-CpG 55.2 NP vaccinated mice produced increased levels of IL-2 (Fig. 5C) and IFN-γ (Fig. S4C). Overall, we observed several significant correlates of protection, including increased sera IgM titers, HAI titers, and cytokine production, with the MP12W-CpG 55.2 NP formulation.

When we challenged the vaccinated mice with HK/68, we observed that mice vaccinated with MP12W-CpG 55.2 or MP12W-CpG 1826 NPs performed very similarly in terms of survival percentage (Fig. 5E). Comparing the survival rate between CpG 55.2 and CpG 1826 (alone or combined with MP12W), a 10% decrease in survival rate was noticed in the groups with CpG 55.2 versus CpG 1826, but this difference was not statistically significant. Based on the noted survival, it would appear that IgM and HAI antibody responses do not play a crucial role in this challenge model, as we noted with the M7/MP12W comparison. Additionally, there might be other immune responses induced by MP12W-CpG that we can evaluate in future studies, such as NK or innate lymphoid cells, which have been reported as important effector mechanisms in protection against influenza [73–75]. Additional rationale for the observed difference could be linked to the fact that CpG 1826 is optimized for mice TLR-9, and CpG 55.2 is a sequence optimized for the human TLR-9 [24, 38]. Even though the human cytokine profile induced by CpG 55.2 has not been published yet, another human-optimized CpG sequence (CpG 1018) has been able to induce the expression of pro-inflammatory cytokines in human cells and has robust adjuvant activity in mice and humans [76, 77]. Based on our challenge results, mice vaccinated with MP12W-CpG 55.2 or MP12W-CpG NPs showed similar survival rates, suggesting that IgM and HAI antibody responses may not be crucial in this model, and highlighting the need to explore other immune responses in future studies.

Conclusion

New mucosal adjuvant systems are essential for better protection against respiratory viruses like influenza and SARS-CoV-2. Intranasal vaccines offer significant advantages, including needle-free administration and stronger local immune responses. Current i.n. vaccines like FluMist have limitations, especially for immunocompromised individuals. More effective adjuvants are needed to enhance safer subunit vaccines. MC activators show promise as i.n. adjuvants but are not well evaluated in the literature. Additionally, Computationally Optimized Broadly Reactive Antigens (COBRA) can improve vaccine effectiveness against diverse influenza strains.

We have shown that we can develop a nanocomplex particle adjuvant formulation with an MC agonist and TLR9 ligand CpG. Our previous studies with M7-CpG NPs showed great promise, but MP12W was identified to be a more potent inducer of MCs degranulation. When we complexed MP-12W with CpG 1826, NPs were formed. MP12W-CpG 1826 NPs illustrated increased indicators of safety and efficacy compared to M7-CpG NPs, in vitro. Further, vaccination with MP12W-CpG NPs and OVA or H3 COBRA J4 induced higher local antibody responses and cytokine production. The MP12W-CpG 1826 NPs also displayed an in vivo reduced risk of anaphylaxis in an allergy-sensitive mouse strain. Replacing CpG 1826 with CpG 55.2 in the formulation further enhanced immune responses, including increased sera IgM titers, HAI titers, and cytokine production. These findings suggest that MP12W-CpG 55.2 NPs are a promising candidate for effective and safe mucosal vaccination.