Effective and Safe Stimulation of Humoral and Cell-Mediated Immunity by Intradermal Immunization with a Cyclic Dinucleotide/Nanoparticle Combination Adjuvant Free

Intradermal (ID) vaccination delivers Ags into the dermis of the skin, which has an abundant population of dendritic cells (DCs) in contrast to skeletal muscle that is targeted by most conventional vaccines. This route of immunization is attractive because of the potential for needle-free delivery and ease of administration (1, 2). It may induce an enhanced immune response, allowing a reduced dose of Ag in vaccines without loss of efficacy. Dose sparing increases the number of available vaccine doses, which is especially important in response to pandemic infections, and can lower the cost of vaccines (3–5). ID immunization with dose-sparing influenza vaccines induced equivalent or improved vaccine efficacy compared with the full i.m. dose in humans (6–8). Although these influenza vaccines did not contain an adjuvant, local reactogenicity of adjuvants approved for use in conventional vaccines, such as aluminum-containing adjuvants and oil-in-water emulsions, and lack of compatibility with needle-free delivery devices has hampered wider application of ID vaccine delivery (2, 9). There continues to be a need for new adjuvants that can be added to the toolbox for the development of effective and safe vaccines and for different routes of administration (10, 11). Cyclic dinucleotides (CDNs) are hydrophilic, negatively charged molecules composed of two nucleotides linked via two phosphodiester bonds. They induce robust humoral and cell-mediated immune responses when used in experimental vaccines (12). Bacteria produce CDNs as second messengers, including cyclic-di-GMP (cdGMP), cyclic-di-AMP (cdAMP), and 3′,3′ cyclic guanosine-adenosine monophosphate (cGAMP). These molecules are widely distributed in bacterial species and play critical roles in their physiology (13). In mammalian and other eukaryotic cells, the cytoplasmic enzyme cGMP adenosine monophosphate synthase generates 2′,3′ cGAMP upon binding of dsDNA. Bacterial CDNs and 2′,3′ cGAMP function as danger signals in mammalian hosts and activate the immune system (13, 14). Bacterial and eukaryotic CDNs are recognized by the endoplasmic reticulum–bound protein stimulator of IFN genes (STING). This leads to the activation of the TANK-binding kinase 1 (TBK1), which phosphorylates and activates the transcription factor IFN regulatory factor (IRF) 3 and results in the expression of type I IFN. In addition, STING activation induces signaling via the NF-κB pathway. The affinity of human STING for 2′,3′ cGAMP is higher than for the bacterial CDNs (15). The bacterial second messenger cdAMP but not cdGMP nor eukaryotic 2′,3′ cGAMP can activate at least two other receptors that enhance NF-κB signaling. Binding of cdAMP to reductase controlling (RECON) NF-κB inhibits the activity of this oxidoreductase, which allows increased NF-κB signaling (16). Cyclic-di-AMP also has high affinity for the endoplasmic reticulum adaptor protein (ERAdP). Binding of ERAdP provides another mechanism for the induction of NF-κB signaling via activation of TGFβ-activated kinase 1 (TAK1) (17). Furthermore, it has been reported that cdGMP and cdAMP can directly activate the NLRP3 inflammasome, resulting in the secretion of the active form of IL-1β and IL-18 through cleavage by caspase-1 (18).

The application of CDNs as vaccine adjuvants is limited by rapid diffusion of these small molecules from the injection site, which lowers the effective concentration and can lead to unwanted systemic reactions. Furthermore, CDNs have to penetrate the cell membrane to bind and activate the intracellular receptors. Various synthetic nano- and microparticle formulations have been developed to enhance the delivery of CDNs to APCs and to limit systemic cytokine release following s.c., i.m. or intranasal vaccination (19–21). A recent study demonstrated that ID injection of 20 μg of 2′,3′ cGAMP, cdGMP, or 3′,3′ cGAMP in mice and 200 μg of 2′,3′ cGAMP in swine significantly enhanced the Th1 response without significant reactogenicity (22). In this study, we investigate the activity and safety of the combination of cdAMP with cationic α-D-glucan nanoparticles (Nano-11) following ID vaccination in mice and swine. Previous work has shown that Nano-11 alone effectively enhanced the immune response when combined with protein Ags following i.m. injection in mice and swine (23–25). We demonstrate that the combination of Nano-11 with cdAMP activates mouse, human, and porcine APCs in vitro; increases the uptake of Ag by skin migratory and lymph node–resident DCs; and enhances both humoral and cell-mediated immune responses following ID administration, including needle-free delivery.

Nano-11 was prepared as previously described (23). Briefly, phytoglycogen (PG) nanoparticles from sweet corn containing the sugary-1 mutant gene were conjugated with octenyl succinic anhydride (OS) and (3-chloro-2-hydroxypropyl)-trimethylammonium chloride (CHPTAC) in two successive chemical reactions to produce PG-OS-CHPTAC (Nano-11). Endotoxin-free OVA and cdAMP were purchased from InvivoGen (San Diego, CA). All reagents were resuspended in sterile 10 mM Tris–saline buffer (pH 7.4). Formulations were prepared by mixing OVA, Nano-11, and/or cdAMP for 1 h at room temperature. i.m. vaccines were formulated with 50 μg/ml OVA, 2 mg/ml Nano-11, 100 μg/ml cdAMP, and 1 mg AI³⁺/ml aluminum hydroxide adjuvant (AH) (Rehydragel HPA; Chemtrade, Berkeley Heights, NJ) or aluminum phosphate adjuvant (AP) (Adju-Phos; Brenntag, Essen, Germany). The ID vaccines were formulated with 250 μg/ml OVA, 4 mg/ml Nano-11, and 500 μg/ml cdAMP. For the dose-sparing experiments, vaccines were formulated with 10 or 50 μg/ml OVA. Ag uptake was determined by formulating the ID vaccines with Alexa Fluor 647 (AF647)–labeled OVA (Thermo Fisher Scientific, Waltham, MA).

Female BALB/c mice, 6 wk old, were purchased from Envigo (Indianapolis, IN). They were housed at three to four animals per box in ventilated racks with food and water ad libitum. Room temperature was maintained at 20 ± 2°C and relative humidity at 50 ± 15% with a 12/12 h light/dark cycle. Mice were acclimated for 1 wk before start of the experiments. All animal experiments were approved by the Purdue University Animal Care and Use Committee. For i.m. vaccination, mice were injected in both thigh muscles with 50 μl of vaccine formulations each. The injections were repeated after 16 or 21 d as indicated. For ID vaccination, mice were injected with 20 μl of the vaccine formulations in the ear pinna. Mice were anesthetized via isoflurane inhalation for the ID injections. Draining lymph nodes were collected 24 h after one injection with AF647–OVA. Blood and tissues (spleen, draining lymph nodes, and bone marrow) samples were collected 10 or 14 d after the first or second immunization.

ID vaccination in pigs was performed with 15 animals at 5 wk age at the Animal Sciences Research and Education Center of Purdue University. They were randomly assigned into three treatment groups (n = 5 per group, two males and three females). Pigs were manually restrained and injected ID in the neck with 100 μl of vaccine formulation using a PharmaJet Tropis jet injector (PharmaJet, Golden, CO). A second administration was performed after 3 wk, and serum samples were collected 14 d after the boost to determine OVA-specific Igs by ELISA.

Human THP-1 cells (TIB-202; American Type Culture Collection, Manassas, VA) were cultured in complete RPMI (RPMI 1640 with 25 mM HEPES, 2 mM l-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, 0.25 μg/ml amphotericin B, and 50 μM 2-ME) with 10% FBS. To inhibit NLRP3 activation, cells were seeded in a 24-well plate at a concentration of 1 × 10⁶ cells per well and pretreated with 1 or 10 μM MCC950 (Selleck Chemicals, Houston, TX) for 30 min, followed by LPS priming for 3 h. Unstimulated and primed cells were treated alone or together with Nano-11 (80 μg/ml) and cdAMP (5 μg/ml) for 48 h at 37°C with 5% CO₂. Supernatants were analyzed for IL-1β by ELISA (eBioscience, San Diego, CA). The NF-κB–SEAP and IRF-Luc reporter THP1-Dual cells (InvivoGen) were cultured in complete RPMI supplemented with 10% FBS, 10 μg/ml blasticidin, 100 μg/ml zeocin, and 100 μg/ml normocin. The cells were seeded in a 96-well plate at a concentration of 1 × 10⁵ cells per well and treated with Nano-11 (80 μg/ml) and cdAMP (5 μg/ml) for 48 h at 37°C with 5% CO₂. Supernatants were collected for quantification of NF-κB and IRF pathway activation, and cells were prepared for flow cytometry analysis.

Superficial parotid lymph nodes and spleens were collected from mice to generate single-cell suspensions, as previously described (24). Bone marrow–derived DCs (BMDCs) were generated from 6- to 7-wk-old BALB/c mice tibias and femurs, as previously described (23). Cultures for in vitro stimulation experiments contained >90% CD11c⁺ cells. Twenty-four well plates were used to culture BMDCs at a concentration of 1 × 10⁶ cells/ml and treated alone or together with Nano-11 (80 μg/ml) and cdAMP (5 μg/ml). BMDCs were incubated at 37°C with 5% CO₂ for 2 d before harvesting the supernatant and cells for flow cytometric analysis. The supernatant was screened for cytokines IL-1β and TNF-α, and cytokine quantification was achieved according to the ELISA kit protocols (eBioscience).

Porcine monocyte-derived DCs (Mo-DCs) were generated from peripheral blood of 8–10-wk-old cross-bred (Yorkshire × Landrace) pigs. Heparinized blood was diluted 1:1 with RPMI 1640 (Corning, Corning, NY), layered over Histopaque-1077, and centrifuged at 1400 rpm (400 × g) for 40 min at 25°C. Isolated mononuclear cells at a concentration of 25 × 10⁶ cells/ml were plated in 1 ml of prewarmed complete RPMI (RPMI 1640 with 25 mM HEPES, 2 mM l-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, 0.25 μg/ml amphotericin B, and 50 μM 2-ME) containing 5% FBS in a six-well plate overnight at 37°C with 5% CO₂. The media was replaced with 2 ml complete RPMI supplemented with 25 ng/ml swine GM-CSF (RP0940S; Kingfisher Biotech, Saint Paul, MN) and 10 ng/ml swine IL-4 (RP0300S; Kingfisher Biotech). Every 2 d, 1 ml of old media was replaced with 1 ml of fresh media. On day 8, immature porcine Mo-DCs were harvested by centrifugation at 1000 RPM for 5 min and resuspended at 1 × 10⁶ cells/ml. Cells were seeded in a 96-well plate and treated alone or together with Nano-11 (80 μg/ml) and cdAMP (5 μg/ml) for 48 h at 37°C with 5% CO₂ prior to analysis by flow cytometry. Supernatants were collected to determine the concentration of secreted TNF and IL-1β by ELISA (Sigma-Aldrich, St. Louis, MO).

Swine and mouse serum samples were collected 14 d after the primary immunization and 10–14 d after the booster injection. OVA-specific IgG titers were analyzed by ELISA in 96-well plates coated with 1 μg/ml OVA overnight at 4°C, as previously described (23). In short, 100 μl of serially diluted serum samples were added to the wells in duplicate for 1 h, followed by well washing and incubation with 100 μl of peroxidase-conjugated goat anti-mouse IgG (1030-05), IgG1 (1073-05) or IgG2a (1080-05; SouthernBiotech, Birmingham, AL) for 1 h. A final well wash was followed with 100 μl 3,3′,5,5′ tetramethylbenzidine substrate solution (Neogen, Lexington, KY) and allowed to react in the dark at room temperature for 5 min. The reaction was terminated with 50 μl of 2 M sulfuric acid, and absorbance at 450 nm (OD 450) was measured in a microplate reader (BioTek Instruments, Winooski, VT). Anti-OVA titers were determined in porcine serum samples by ELISA using mouse anti-swine IgG1 (clone K139 3C8) or mouse anti-swine IgG2 (clone K68 Ig2, both from Bio-Rad Laboratories, Hercules, CA) Abs. End-point titers were determined as the dilution at which OD 450 nm reached 0.2.

Isolated cells from lymph nodes and BMDCs were washed with Cell Staining Buffer (CSB; BioLegend, San Diego, CA) and treated with Zombie Violet (BioLegend) and anti-mouse CD16/32 (clone 93) for 30 min at 4°C. Cells were labeled with mAbs (all from BioLegend, unless specified otherwise) against CD3 (clone 145-2C11), CD4 (clone GK 1.5), CD8α (clone 53-6.7), CD11c (clone N418), CD80 (clone 16-10A1), CD86 (clone GL-1), CD45 (clone RA3-6B2), I-A/I-E (clone M5/114.15.2), CXCR5 (clone L138D7), PD-1 (clone 29F.1A12), GL-7 (clone GL7), and B220 (clone RA3-6B2) in cold CSB for 45 min at 4°C. Migratory and resident DCs were similarly isolated and prepared following ID immunization with AF647–OVA. Cells were labeled with mAbs against I-A/I-E (clone M5/114.15.2), CD11c (clone N418), CD11b (clone M1/70), CD8α (clone 53-6.7), CD205 (clone 205yekta; Thermo Fisher Scientific), CD103 (clone 2E7), and CD207/Langerin (clone eBioRMUL.2; Thermo Fisher Scientific) in cold CSB for 45 min at 4°C. Cells were washed and fixed with 1× PBS solution containing 4% paraformaldehyde. For intracellular cytokine staining, splenocytes were cultured in complete RPMI and 10% FBS with 25 μg/ml OVA for 72 h, followed by stimulation for 3 h at 37°C with PMA, ionomycin, and monensin in complete medium. Subsequently, the cells were labeled with Abs as mentioned above and permeabilized for intracellular staining with IFN-γ (clone XM G1.2) and IL-17A (clone TC11-18H10.1).

Porcine Mo-DCs were washed with CSB followed by blocking nonspecific binding sites with 1% normal rabbit serum for 30 min at 4°C. Cells were labeled with biotinylated human CD152 (CTLA-4)–mouse IgG (Fc) fusion protein (501-030; Ancell, Bayport, MN) followed by streptavidin–PE (405203; BioLegend) and anti-SLA-DR-FITC (clone 2E9/13; Bio-Rad Laboratories) for 45 min at 4°C. Human THP1-Dual cells were treated with human TruStain FcX (422301; BioLegend) for 30 min in 4°C, followed by labeling with anti-human CD11b (clone ICRF44), anti-human CD80 (clone 2D10), anti-human CD86 (clone BU63), and anti-human HLA-DR (clone L243).

Flow cytometry was performed with a FACSCanto II (BD Biosciences, San Jose, CA) or Attune NxT Flow Cytometer (Invitrogen, Carlsbad, CA). Data were analyzed with FlowJo software (FlowJo, Eugene, OR).

Bone marrow cells were collected from the tibias and femurs of mice to enumerate OVA-specific Ab-secreting cells (ASCs). Twenty-four hours prior to euthanasia, 96-well MultiScreen IP Filter Plates (MAIPS4510; Millipore Sigma) were prepared by treating the plates with 20 μl/well of 35% ethanol for 30 s, then washing the plate with 300 μl/well sterile PBS three times. The wells were coated with 100 μl/well OVA (10 μg/ml) in PBS overnight at 4°C. Following washing and blocking with complete RPMI with 10% FBS for 2 h, wells were incubated with serial dilutions of single-cell suspensions for 24 h at 37°C/5% CO₂. The plate was washed with PBS and PBS–0.05% Tween 20 (PBST), followed by biotin-labeled anti-mouse IgG (SouthernBiotech) in PBST with 1% FBS (1:5000) for 2 h. This was followed by washing, then the plate was incubated with avidin-HRP conjugate (Thermo Fisher Scientific) in PBST with 1% FBS (1:500) for 1 h. Enzyme activity was detected using 3-amino-9-ethylcarbazole (Sigma-Aldrich) in the dark for 10 min, and the plate was thoroughly rinsed with deionized water. The filter tray was allowed to completely dry in a dark place at room temperature before quantifying the red-colored spots in an ELISPOT reader (AID Diagnostika, Strassberg, Germany).

Cyclic dimeric AMP determination was performed on an Agilent 1290 UHPLC system coupled to an Agilent 6470 Triple Quadrupole Mass Spectrometer (Agilent Technologies, San Jose, CA). Filtrate solution was collected and diluted 100-fold in acetonitrile/water 1:1 (v/v). Five microliters of the diluted sample was delivered to an Agilent Poroshell 120 HILIC-Z PEEK-lined (2.7-μm particle size, 2.1 × 150 mm column; Agilent Technologies) heated at 40°C through a Multisampler (G7167B; Agilent Technologies). The high-speed binary pump flow rate was set at 0.3 ml/min. The mobile phase was composed of A equaled acetonitrile/2-propanol/200 mM ammonium acetate 90:5:5 (v/v) titrated with formic acid to pH 3; B equaled water/acetonitrile/200 mM ammonium acetate 90:5:5 (v/v) titrated with formic acid to pH 3. The ultraperformance liquid chromatography column was pre-equilibrated with 100% A for 1 min, then set in a linear gradient to 85% B in 8 min, held for 2.5 min, returned to 100% A in 0.5 min, and re-equilibrated for 4 min. The samples were analyzed in positive ion mode, and the jet stream electrospray ionization source conditions were set as follows: gas temperature 325°C, drying gas 8 l/min, nebulizer 40 ψ, fragmentation voltage 80 V, and capillary voltage 3500 V. The multiple reaction monitorings (parent-fragment) used for the acquisition of cdAMP were m/z 659.2 →524.2, m/z 659.2 →330.2, m/z 659.2 →136.2 with a collision energy of 10 V, dwell of 30, and a cell accelerator voltage of 4 W. Quantitation was based on a four point standard curve with concentrations ranging from 250 to 1000 μg/ml, fitted to a linear function with a calculated correlation coefficient (R²) of 0.94. Data processing was performed using MassHunter (B.10.00).

Statistically significant changes between experimental groups were determined by using a one-way ANOVA test with a Tukey multiple comparisons test using GraphPad Prism version 8.3.0 (GraphPad Software, San Diego, CA) for Windows. The p values < 0.05 were considered statistically significant (*p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001).

Nano-11 is composed of positively charged nanoparticles ∼70–80 nm in diameter that readily adsorb negatively charged proteins (23). The nanoparticles were mixed at 4°C with cdAMP, which carries a negative charge. The degree of adsorption of cdAMP to Nano-11 was determined by mass spectrometry of the supernatant after 1 and 24 h of continuous agitation. Over 99% of cdAMP was adsorbed to Nano-11 at both time points, suggesting stable association of the two compounds (Fig. 1A). Adsorption of cdAMP did not affect the adsorption of OVA (data not shown).

Incubation of mouse BMDCs with cdAMP or Nano-11 alone induced expression of CD80 and CD86, whereas the conventional adjuvants, AH and AP, had minimal effects consistent with earlier reports (23, 26, 27). The combination of Nano-11 with cdAMP greatly enhanced the expression of CD80 and modestly increased the expression of CD86 compared with cdAMP alone (Fig. 1B). In contrast, combining cdAMP with AP slightly increased the expression of CD80 and CD86, whereas AH decreased the expression of CD80 and CD86 compared with cdAMP alone (Fig. 1B).

The mAbs specific for swine CD80 and CD86 are not available. Instead, the analysis of the expression of these costimulatory molecules was determined using a human CD152–mouse IgG (Fc) fusion protein that binds both CD80 and CD86 (28). Nano-11 increased the expression of CD80/86 on porcine Mo-DCs, whereas cdAMP had no effect by itself (Fig. 1C). However, the cdAMP/Nano-11 adjuvant further enhanced the expression of CD80/86 compared with Nano-11.

Nano-11 induced significantly increased expression of CD80 and had a modest effect on CD86 expression in human THP-1 cells (Fig. 1D). Incubation with cdAMP alone had no effect, but the cdAMP/Nano-11 combination adjuvant greatly increased the expression of CD86, whereas it did not significantly alter CD80 expression compared with Nano-11 alone.

Activation of DCs and other APCs is associated with increased cell surface expression of MHC class II (MHC II) molecules. Nano-11 increased the expression of MHC II on mouse BMDCs and human THP-1 cells, whereas cdAMP had no effect. The cdAMP/Nano-11 combination adjuvant increased MHC II expression on the cells from all three species (Fig. 1E).

These results demonstrate that Nano-11 directly activates mouse, human, and porcine APCs and that this effect is greatly enhanced by combining Nano-11 with cdAMP.

Incubation of mouse BMDCs with cdAMP and Nano-11 alone induced modest secretion of TNF, which was markedly enhanced when cdAMP and Nano-11 were combined (Fig. 2A). The secretion of IL-1β was investigated in LPS-primed BMDCs, as LPS induces expression of pro–IL-1β and NLRP3. Both cdAMP and Nano-11 induced release of IL-1β, and this was further enhanced upon incubation with the cdAMP/Nano-11 combination adjuvant (Fig. 2A). Similar effects were observed in porcine Mo-DCs, although the overall concentration of TNF and IL-1β was lower (Fig. 2B).

Cyclic-di-AMP activates IRF3 and NF-κB signaling pathways by binding STING and other intracellular proteins. Using the NF-κB–SEAP and IRF-Lucia luciferase dual reporter THP-1 cells, we investigated the activation of these signaling pathways by cdAMP, Nano-11, and cdAMP/Nano-11. Incubation of THP1-Dual cells with cdAMP alone induced modest activation of IRF3 and no activation of NF-κB. Nano-11 activated NF-κB but not IRF3, whereas the cdAMP/Nano-11 combination adjuvant had a marked synergistic effect and induced a much stronger activation of both signaling pathways (Fig. 2C). We previously reported that secretion of IL-1β by Nano-11–treated mouse BMDCs is dependent on cathepsin B and caspase-1 signaling (23). To determine whether Nano-11 activates the NLRP3 inflammasome, the effect of MCC950, a selective inhibitor of the NLRP3 inflammasome (29), on IL-1β secretion by THP-1 cells was evaluated. Nano-11 and cdAMP/Nano-11 both induced the release of IL-1β (Fig. 2C). Pretreatment of THP-1 cells with MCC950 greatly reduced the secretion of IL-1β induced by Nano-11 and cdAMP/Nano-11, indicating that this process requires activation of the NLRP3 inflammasome.

We showed previously that Nano-11 enhances the immune response to i.m.-injected protein Ags similar to aluminum adjuvants (24). To compare the combinations of cdAMP with either the AP and AH or with Nano-11, the individual components or combination adjuvants were injected i.m. with OVA. All individual adjuvants enhanced the IgG response to OVA (Fig. 3A). The combination of cdAMP with Nano-11, AH, or AP significantly increased the Ab titer compared with the individual components (Fig. 3A). The majority of IgG belonged to the IgG1 subclass; however, the cdAMP/Nano-11 combination also induced a modest increase of the IgG2a titer. In addition, the combination adjuvants induced a larger number of OVA-specific ASCs (plasma cells) in the bone marrow (Fig. 3A). Thus, Nano-11 and aluminum adjuvants enhanced the immune response when used in combination with cdAMP for i.m. immunization.

Aluminum adjuvants are thought to be too reactogenic for ID vaccination because they induce granulomatous inflammation, which can persist for months, at the injection site (2, 30). As Nano-11 induces only a transient inflammatory reaction at the injection site after i.m. injection (23), we assessed the ability of Nano-11 alone and in combination with cdAMP to enhance the immune response following ID administration in the ear. Both Nano-11 and cdAMP alone significantly increased the anti-OVA IgG and IgG1 titers, whereas the cdAMP/Nano-11 combination induced the highest titers (Fig. 3B). Nano-11 and cdAMP alone did not increase the IgG2a titer, but the combination adjuvant induced a marked increase of IgG2a, indicating a synergistic interaction that results in a more balanced Th1/Th2 response (Fig. 3B). At 2 wk after a single injection, a significant increase of Ab titers was observed (Supplemental Fig. 1), but the levels were greatly enhanced after the second injection. The cdAMP/Nano-11 combination also induced a greater increase of the number of plasma cells in the bone marrow compared with either cdAMP or Nano-11 alone (Fig. 3B). No evidence of local or systemic adverse reactions were observed after either the primary or booster injections (Fig. 3C, 3D).

ID injection of the ear pinna with the combination adjuvant induced marked swelling of the draining (superficial parotid) lymph node (Fig. 4A). Images of fluorochrome-labeled section of the lymph nodes captured by laser scanning confocal microscopy show minimal formation of germinal centers (GCs) following injection with cdAMP, a moderate number following injection with Nano-11, and a robust induction upon injection with cdAMP/Nano-11 (Fig. 4B). This was confirmed by determining the number of GC B and follicular helper T (Tfh) cells in the draining lymph node by flow cytometry 10 d after the first and second injections (Fig. 4C, 4D). The largest increase of GC B and Tfh cells was observed 10 d after injections of cdAMP/Nano-11.

We showed previously that Nano-11 increases the uptake of Ag by migratory DCs in the draining lymph node following i.m. injection (24). At least five populations of DCs can be identified in lymph nodes draining the skin, including migratory conventional (c)DC1 cDC1, cDC2, and Langerhans cells and resident cDC1and cDC2. To investigate the effect of the cdAMP/Nano-11 combination adjuvant on the uptake and distribution of Ag, mice were injected ID with AF647–OVA alone or with the adjuvant. Mice that received cdAMP/Nano-11 had enlarged superficial parotid lymph nodes 24 h postinjection (Fig. 5A). The combination adjuvant greatly increased the uptake of OVA by lymph node–resident cDC1s (CD8⁺) and cDC2s (CD11b⁺) (Fig. 5B). In addition, mice injected with the combination adjuvant had a marked increase of OVA⁺ migratory DCs (Fig. 5C). Although the injection of AF647–OVA with cdAMP/Nano-11 significantly increased the total number of OVA⁺ DCs in each subset compared with AF647–OVA alone, it increased the percentage of OVA⁺ resident cDC2 and decreased the percentage of migratory cDC1 (Fig. 5D). Overall, these results demonstrate that the cdAMP/Nano-11 combination enhances the transport of Ag to the draining lymph node by migratory DCs and uptake of Ag by lymph node–resident DCs.

To investigate the potential dose-sparing effect of ID immunization with cdAMP/Nano-11, we immunized mice with a 5.0 μg/dose OVA injected i.m. or ID and compared this with reduced doses of OVA (0.2 and 1.0 μg/dose) injected ID. Mice immunized ID had significantly higher titers of OVA-specific IgG, IgG1, and IgG2a compared with i.m. vaccination, even at the 0.2 μg dose of OVA (Fig. 6). The number of OVA-specific ASCs in the bone marrow was greater after ID immunization with 1.0 and 5.0 μg/dose than the i.m. immunization with 5.0 μg/dose (Fig. 6). Furthermore, the number of GC B cells and Tfh cells in the draining lymph node after ID immunization was similar across the three doses (Supplemental Fig. 2). This suggests that ID immunization with the cdAMP/Nano-11 combination adjuvant is a promising dose-sparing strategy.

Intranasal and s.c. immunization with cdAMP as adjuvant can induce differentiation of CD4 T cells into Th1 and Th17 and induce activation of CD8 T cells (27, 31). The presence of CD4 effector cells in the spleen after ID injection was determined by restimulation of splenocytes with OVA and intracellular cytokine labeling. The cdAMP/Nano-11 combination adjuvant induced the largest number of Th1 and Th17 cells, whereas very few cells were obtained following injection of either cdAMP or Nano-11 (Fig. 7). In addition, the combination adjuvant stimulated the activation and differentiation of CD8 T cells as indicated by the increased number of IFN-γ–secreting CD8 T cells (Fig. 7). This suggests that the ID injection of OVA with cdAMP/Nano-11 induced cross-presentation of OVA-derived peptides.

The skin of pigs is a well-recognized model for dermatologic research and for cutaneous vaccine and drug delivery (32, 33). We used a needle-free injector for delivery of vaccine formulations into the dermis of domestic piglets. The adjuvanted formulations induced a significant increase of the anti-OVA IgG titer compared with the group that received OVA only (Fig. 8A). There was a modest increase of the Ab titer in pigs that received OVA with the combination adjuvant compared with Nano-11 only, but this difference did not reach statistical significance. Analysis of IgG1 and IgG2 subclasses did not reveal a bias toward one of these subclasses in pigs that received OVA with either Nano-11 or cdAMP/Nano-11. The injection induced a small visible swelling that disappeared quickly. No visible reactions were observed at 24 h or 14 d following injection with OVA only or OVA with Nano-11 (Fig. 8B). A small discoloration was observed 24 h after injection of OVA with cdAMP/OVA, which had completely disappeared at 14 d after injection.

More effective adjuvants are needed to enable the development of vaccines against diseases for which currently no effective vaccine exists, to increase the longevity and breadth of the immune response, to induce protection in individuals with a diminished immune system, and to allow the induction of protective immune response following alternative routes of administration (10, 11). The design of combination adjuvants is aimed at bringing together molecules in a single complex that targets different activation pathways of the immune system, resulting in a broader and more effective immune response and superior protection. The results presented in this study demonstrate that Nano-11, an adjuvant based on plant-based α-d-glucan nanoparticles, and the CDN cdAMP have marked synergistic effects that result in a significantly enhanced immune response. Nano-11 is composed of positively charged nanoparticles that readily adsorb cdAMP. The high degree of adsorption compares favorably with other nano- and microparticle formulations (19–21). The goal of adsorption or incorporation of CDNs in particles is to limit the diffusion from the injection site and to enhance the uptake by DCs. Nano-11 not only serves this function but also has inherent immunostimulatory properties, as indicated by the increased expression of costimulatory molecules and MHC II by APCs and the release of cytokines, including IL-1β.

We showed previously that Nano-11 increased the expression of CD80 and CD86 on BMDCs from BALB/c mice as well as TLR4-deficient C3H/HeJ mice (23). We confirm and extend these observations and show that Nano-11 also increases the expression of costimulatory molecules on porcine Mo-DCs and human THP-1 cells. Furthermore, Nano-11 increased the expression of MHC II molecules on BMDCs and THP-1 cells. These phenotypic changes are consistent with maturation of DCs as induced by LPS and other stimuli. The LPS-induced maturation of human DCs is at least partially dependent on NF-κB signaling (34, 35). Indeed, Nano-11 induced NF-κB activation based on experiments with dual reporter cells. The cdAMP/Nano-11 combination adjuvant induced a synergistic increase of costimulatory molecule expression on all cells. This was associated with a marked synergistic increase of NF-κB and IRF signaling induced by the combination adjuvant in THP-1 cells compared with either component alone.

Mouse BMDCs and porcine Mo-DCs secreted TNF following stimulation with cdAMP, consistent with the induction of NF-κB via activation of STING and ERAdP, which may be further enhanced by inhibition of RECON (16, 17, 36). Nano-11 alone induced a small amount of TNF but greatly increased the secretion of TNF when combined with cdAMP. The effect of cdAMP and Nano-11 on the secretion of IL-1β was determined in LPS-primed BMDCs. Priming by LPS or other TLR agonists induces the transcription and translation of biologically inactive pro–IL-1β. Proteolytic cleavage of pro–IL-1β by caspase-1 results in the release of the active form of IL-1β (37). Previous studies have shown that cdAMP can induce the release of IL-1β from LPS-stimulated bone marrow–derived macrophages and that this is dependent on the NLRP3 inflammasome (18). We have also shown that Nano-11 induces IL-1β from LPS-stimulated BMDCs by activation of caspase-1 (23). In this study, we show that the cdAMP/Nano-11 combination adjuvant is more effective in inducing IL-1β release than either component alone. The release of IL-1β induced by Nano-11 alone and cdAMP/Nano-11 was abrogated by treatment with MCC905, a specific inhibitor of NLRP3 (29), indicating that Nano-11 activates the NLRP3 inflammasome similar to other particulates (38). In aggregate, these in vitro experiments demonstrate a marked synergistic activity between cdAMP and Nano-11 in activating APCs across three different species. Our results suggest that the cdAMP/Nano11 combination adjuvant targets multiple signaling pathways, including IRF and NF-κB signaling, as well as inflammasome activation, which underlies its robust activation of the immune response.

Aluminum adjuvants are safe and effective adjuvants used in many licensed vaccines administered via i.m. injection (39). Their large adsorptive surface allows association with other immunostimulatory molecules, such as monophosphoryl lipid A and other TLR agonists, which results in enhanced activity (40–42). The two types of commercially available aluminum adjuvants, AH and AP, differ in structure and surface characteristics (43). Negatively charged CDNs would be expected to adsorb well to AH, which is positively charged at neutral pH, and adsorb poorly to AP, which is negatively charged (39). A few studies have investigated the effect of combining aluminum adjuvants with CDNs on the immune response with variable results (44–46), likely due to differences in the type of aluminum adjuvant and in the experimental design. In our experiments, both AH and AP enhanced the IgG1 titer but had little effect on the IgG2a response. The cdAMP/Nano-11 combination increased both IgG1 and IgG2a Ab titers, suggesting a stronger synergistic effect than with the aluminum adjuvants. This observation, as well as concerns about using aluminum adjuvants for ID immunization, led us to investigate the use of Nano-11 with cdAMP for this alternative route of vaccination.

ID immunization with Nano-11 alone significantly enhanced the Ab response to OVA in both mice and pigs without any evidence of local reactions. This is consistent with the moderate transient inflammation induced at the injection site after i.m. injection and underscores the safety of this adjuvant (23, 24). Previous work showed that ID immunization with cGAMP was sufficient to induce a robust Th1 response in mice (22). In contrast, we did not observe a significant IgG2a and Th1 response following ID injection of cdAMP alone, which may be due to the different localization (back skin versus ear) and different types of CDNs used in these studies. However, cdAMP induced an increase of IgG1 and greatly increased the IgG2a titer when formulated with Nano-11. This is suggestive of a more balanced immune response and was supported by the increased number of Th1 and Th17 cells in cdAMP/Nano-11–immunized mice. Intranasal immunization of mice with cdAMP with a protein Ag similarly induced a balanced IFN-γ and IL-17 response (27).

ID immunization with cdAMP/Nano-11 also greatly increased the number of plasma cells in the bone marrow. These are long lived plasma cells that contribute to the lasting protection induced by vaccination (47, 48). Their induction and differentiation depend on the formation of GCs and the function of differentiated Tfh cells (49, 50). Indeed, the cdAMP/Nano-11 adjuvant induced formation of robust GCs associated with a marked increase of the number of Tfh and GC B cells in the draining lymph node.

Several human studies have shown that ID immunization has the potential to induce a robust immune response with a reduced Ag dose compared with i.m. immunization (5–8). Because Ags are usually the most expensive component of vaccines and a limiting factor in vaccine production, such dose sparing could reduce the cost of vaccines and increase the number of available vaccine doses. ID immunization with 5- and 25-fold–reduced Ag doses with cdAMP/Nano-11 induced higher IgG1 and IgG2a titers compared with i.m. immunization, suggesting significant dose-sparing potential. This is further supported by the increased numbers of GC B cells and Tfh cells in the draining lymph node independent of Ag dose.

Restimulation of splenocytes of mice immunized with OVA with cdAMP/Nano-11 induced a significantly increased number of IFN-γ–secreting CD8 T cells in comparison with either cdAMP or Nano-11 alone. This indicates that the combination adjuvant increased cross-presentation of OVA peptides for presentation to CD8 T cells. The activation of CD8 T cells through cross-presentation is necessary for therapeutic subunit vaccines against chronic viral diseases and cancer (51). Previous studies have shown that CDNs, including cdAMP and cdGMP, increase the frequency of Ag-specific CD8 T cells (31, 52). The enhanced cross-presentation was dependent on the proteasome and transporter associated with Ag processing (TAP) molecules, demonstrating that it involves the cytosolic pathway (31). We show in this study that the stimulation of cross-presentation by cdAMP was markedly increased by its adsorption to Nano-11, suggesting that Nano-11 facilitates the entry of both cdAMP and adsorbed protein into the cytosol.

The Nano-11 particles have an average diameter of 70–80 nm (23), a size that allows entry into the lymphatics as well as efficient uptake by DCs (53). Direct entry into the lymphatics results in drainage to the lymph node and uptake by resident DCs. Indeed, analysis of DC subpopulations in the draining lymph node after injection of AF647-labeled OVA with cdAMP/Nano-11 revealed OVA⁺ migratory and resident DCs. The largest fraction of OVA⁺ DCs was the CD11b⁺ resident cDC2 subset. These cells are located in close association with cortical and medullary sinuses in the lymph node and are well positioned to take up particulate Ags (54). The two major subsets of OVA⁺ migratory cells were cDC2 and Langerhans cells, both of which were recently identified as playing an important role in the induction of Tfh cells after immunization (55–57). The relative role of migratory versus residential DCs in inducing Tfh cells and Ab responses remains to be determined, and cooperation between these DC populations may be required for optimal responses (58). Migratory CD103⁺ DCs and resident CD8⁺ cDC1 in lymph nodes are capable of cross-presentation of viral and soluble Ags (59, 60). Although these DC subsets comprised the smallest percentages of OVA⁺ DCs, they are likely to play a critical role in the activation of CD8 T cells observed after ID immunization with cdAMP/Nano-11.

Although the ID injections of mice were performed with small (30 gauge) hypodermic needles, one of the main advantages of ID vaccination is the possibility of needle-free delivery. The use of Nano-11 and cdAMP/Nano-11 as vaccine adjuvants in a needle-free device was evaluated in pigs, which have a thicker skin than mice, similar to human skin. The PharmaJet Tropis device delivers the vaccine by a high-velocity liquid jet. The inclusion of only 400 μg of Nano-11 per dose markedly increased the Ab response. Vaccines are exposed to higher shear forces during jet delivery compared with injection by hypodermic needle (61), which may affect the stability of Ags and adjuvants. The immunoenhancing effect of Nano-11 is consistent with the stability of the phytoglycogen nanoparticles that form the basis of this adjuvant. The cdAMP/Nano-11 combination adjuvant induced a modest, but NS increase over Nano-11 alone. The addition of cdAMP also did not affect the distribution of the IgG1 and IgG2 subclasses. IgG1 has been associated with a Th2 response, and IgG2 has been associated with a Th1 response in pigs (62, 63). Porcine cells are able to respond to cdAMP, as indicated by the synergistic effect of cdAMP/Nano-11 on CD80/86 expression by Mo-DCs. However, a recent study showed that porcine STING has a relatively low affinity for the bacterial CDNs cdAMP, cdGMP, and 3′,3′-cGAMP compared with 2′,3′-cGAMP (64). ID immunization with 2′,3′-cGAMP alone appeared to induce an enhanced Ab response in pigs, although the number of animals was too small to draw conclusions (22). As discussed above, cdAMP but not 3′,5′-cGAMP binds to other intracellular proteins besides STING in mouse cells, enhancing NF-κB signaling, but whether this is also the case in human and porcine cells remains to be determined.

In conclusion, a combination adjuvant composed of cdAMP and Nano-11 enhanced the immune response in a synergistic manner following ID delivery. The resulting immune response is balanced with increased production of IgG1 and IgG2a Abs as well as Th1-, Th17-, and IFN-γ–expressing CD8 T cells. These studies support the further evaluation of Nano-11 as a convenient and stable platform for the development of combination adjuvants and for different routes of vaccine delivery.

H.H. and Y.Y. are cofounders of a startup company aimed at further developing and commercializing the Nano-11 adjuvant technology. The other authors have no financial conflicts of interest.