Materials

Reagents for diacid and polymer synthesis are listed with supplier as follows: 1,6-dibromohexane, 1-methyl-2-pyrrolidinone, 4-p-fluorobenzonitrile, hydroxybenzoic acid, N,N-dimethylacetamide, N,N-(diethylamino)ethyl methacrylate (DEAEM), polyvinylpyrrolidone (MW = 40,000 Da), pyridine-2-carboxaldehyde, sebacic acid, and triethylene glycol were purchased from Sigma Aldrich (St. Louis, MO). Acetic acid, acetic anhydride, acetone, acetonitrile, alpha-bromoisobutyryl bromide, basic alumina powder, chloroform, copper acetate monohydrate, dimethyl formamide, ethyl ether, hexane, magnesium sulfate, methylene chloride, n-propylamine, pentane, petroleum ether, Pluronic® F127, potassium carbonate, sea sand, sodium borohydride, sodium hydroxide, sulfuric acid, triethylamine, and toluene were purchased from Fisher Scientific (Waltham, MA). Recombinant influenza A/HK/1/68 nucleoprotein (NP) was acquired from Sino Biological (Beijing, China). CpG ODN 1668 and cyclic dinucleotide (CDN) di-guanosine monophosphate were purchased from InvivoGen (San Diego, CA).

Polymer Synthesis

CPTEG and CPH diacids were synthesized and used to prepare 20:80 CPTEG:CPH as previously described [36]. Briefly, the 20:80 CPTEG:CPH copolymer was synthesized by melt polycondensation by reacting appropriate molar amounts of CPTEG and CPH at 140 °C for 6 h under a vacuum of 0.1 torr [36]. The resulting copolymer was consistent with previous work in terms of molecular weight (8963 Da), purity, and molar composition (24:76), as measured via [1]H nuclear magnetic resonance spectroscopy (NMR) (Varian MR-500, 500 MHz, Varian, Palo Alto, CA) in deuterated chloroform [37]. Copolymer glass transition temperature (Tg) was determined by differential scanning calorimetry (Q2000, TA Instruments, New Castle, DE) with a heating rate of 10 °C/min and a Tg of 29 °C for the 20:80 CPTEG:CPH copolymer was obtained.

Pentablock copolymers (PBC) based on Pluronic® F127 and DEAEM (PDEAEM5-PEO100-PPO65-PEO100-PDEAEM5) were synthesized via one-pot macroinitiation and atom transfer radical polymerization as previously reported [13, 38]. Briefly, Pluronic® F127 was converted to a macroinitiator by reaction with bromoisobutyryl bromide in toluene overnight. The resulting macroinitiator was used directly without purification. Addition of diethylamino ethyl methacrylate groups utilized atom transfer radical polymerization and was carried out under inert atmosphere at 70 °C for 20 h. The final polymer was obtained by passing the crude reaction mixture through an alumina column with a 50:50 methylene chloride:toluene mobile phase and precipitation of the final product in pentane. The molecular weight, purity, and composition of PDEAEM-Pluronic® F127-PDEAEM PBC were determined using 1H NMR in deuterated chloroform and ACQUITY Advanced Polymer Chromatography System (APC) (Waters Corporation, Milford, MA) with a polystyrene standard. The average molecular weight was ca. 15,500 Da, as measured by 1H NMR and APC. The number average molecular weight was measured by APC and found to be ca. 12,200 Da, with a polydispersity index of 1.27. All measurements were comparable with previous work [37]. In vitro cellular cytotoxicity of the PBC micelles was determined via MTT assay and was found to be comparable with previous work [38].

Recombinant Protein Production and Purification

Recombinant HA trimer (H3N8 rH33), based on influenza virus A/equine/1/KY/91 strain, was expressed as previously described [8]. Briefly, H3N8 rH33 was produced using a commercial baculovirus system (Mirus Bio, Madison, WI) and HiFive cells (Invitrogen, Carlsbad, CA). H3N8 rH33 was isolated from the supernatant via a 0.45 μm filter and by collecting retentate from a 100 kDa filter stirred cell system (Millipore Sigma). The protein was then characterized by native and denaturing/reducing polyacrylamide gel electrophoresis using Oriole staining (Bio-Rad, Hercules, CA) with specificity confirmation by western blotting using serum from H3N8 IAV vaccinated horses, as previously reported [8].

Polyanhydride Microparticle Synthesis and Characterization

20:80 CPTEG:CPH microparticles were prepared as previously described using a Büchi B-290 Mini Spray Dryer (Flawil, Switzerland) equipped with a two-fluid nozzle [23]. A suspension was prepared with 10 mg/mL 20:80 CPTEG:CPH in strictly anhydrous methylene chloride to which a mix of either 1 wt.% H3N8 rH33, 1 wt.% A/HK/1/68 NP, and 2 wt.% CpG 1668 or 1 wt.% H3N8 rH33 and 1 wt.% A/HK/1/68 NP was added and sonicated for 30 s. The resulting suspension was directly spray dried at a rate of 10 mL/min, aspirator 70, and nitrogen supplied at 30 psi. The resulting particles were collected and stored at − 20 °C in vacuum sealed containers over desiccant until use. Size and morphology were determined using scanning electron microscopy (FEI Quanta 250, FEI, Hillsboro, OR). Particle zeta potential was measured using Zetasizer Nano ZS (Malvern, Southborough, MA).

Antigen release kinetics were determined as previously described [8]. Briefly summarizing, three different mixtures (n = 3 per mixture) were prepared: (1) ~ 10 mg of polyanhydride particles in 300 μL of phosphate buffered saline (PBS, pH 7.2); (2) 300 µL of a 100 mg/mL pentablock copolymer solution in PBS with 100 µg H3N8 A/equine/KY/91 HA and 100 µg H3N2 A/HK/68 NP; (3) ~ 10 mg of polyanhydride particles were suspended in 300 μL of 100 mg/mL pentablock copolymer solution in PBS with 100 µg H3N8 A/equine/KY/91 HA and 100 µg H3N2 A/HK/68 NP. All nine samples were incubated at 37 °C and were periodically centrifuged to pellet the particles, whereupon the supernatant PBS was removed and replaced with fresh PBS. The protein content of the supernatant PBS was determined using a micro-bicinchoninic acid assay (mBCA) (Pierce, Rockford, IL). Actual loading percent was determined similarly, with ~ 10 mg of microparticles incubated at 37 °C in 40 mM NaOH. Samples were collected every 24 h until the entire particle mass was degraded; total protein content was determined via mBCA and by summation of all release samples. Encapsulation efficiency was determined by comparing the actual loading with the theoretical loading, as described previously [8].

Animal Vaccinations

Female C57BL/6 mice were purchased from Jackson Laboratories (Bar Harbor, ME). Animal procedures were conducted with the approval of the Iowa State University Institutional Animal Care and Use Committee. Two immunization studies were carried out to evaluate safety and efficacy, with vaccine groups summarized in Table 1 and study timelines shown in Fig. 1.

Table 1 Treatment groupsFig. 1

Experimental timelines for safety (A) and efficacy (B) studies. Sera, liver, kidney, and lungs (from intranasally vaccinated animals) were collected at each time point during the safety evaluation

In the safety studies, mice were euthanized at a predetermined time point following immunization. Whole blood, kidney, and liver samples were collected at all time points from all groups. Lung tissue samples and bronchoalveolar lavage fluid (see below) were also collected from the mice receiving the intranasal microparticle vaccine or intranasal saline.

For the efficacy studies, each group contained ten mice, four of which were euthanized at 4 days post-infection to assess viral load in the lungs. The remaining six mice were monitored for weight loss, lung function (penH), and survival.

Plethysmography

Whole body plethysmography (WBP) (Buxco Electronics, St. Paul, MN) was used to measure enhanced pause (PenH) and minute-box volume (MVb). Baseline measurements were taken prior to intranasal vaccination and daily for 4 days afterward.

Analysis of Serum Biomarkers

Blood was collected via cardiac puncture and centrifuged at 5000 rpm for 5 min at 4 °C to separate sera. Sera samples were stored overnight at 4 °C prior to analysis with Abaxis Comprehensive Profile (Zoetis, Union City, CA) at the Iowa State Clinical Pathology Laboratory. Results were compared with literature values provided by the manufacturer.

Analysis of Bronchoalveolar Lavage Biomarkers

Bronchoalveolar lavage was collected at the time of necropsy. Mice were euthanized, and a sterile catheter was inserted into the exposed trachea of each mouse. One mL of sterile PBS was instilled into the lungs and then extracted while massaging the thorax of the mouse; this was repeated three times, with samples from each mouse combined. Samples were then centrifuged at 5000 rpm for 5 min at 4 °C and then stored at − 80 °C until analysis with a custom MILLIPLEX multiplex kit (MilliporeSigma, Darmstadt, Germany) targeting IP-10 (CXCL10), KC (CXCL1), MIG (CXCL9), IFN-γ, IL-10, TNF-α, MIP-2 (CXCL2), RANTES (CCL5), MCP-1 (CCL2), IL-6, IL-12, MIP-1a (CCL3), and IL-1β.

Liver, Kidney, and Lung Histopathology

Whole organ tissue samples were collected and fixed in 10% neutral buffered formalin and transferred to 70% ethanol after 24 h. Liver and kidneys were collected from all groups, and lungs were also collected from the intranasal particle and intranasal saline groups. Sections from each organ were embedded in paraffin, sectioned at 5 µm thickness, and routinely stained in hematoxylin and eosin (H&E). H&E sections were evaluated, and photomicrographs were taken using an Olympus BX43 trinocular microscope equipped with a DP27 colored camera and CellSens Standard Software (Olympus Corporation, Japan). Tissues were blindly evaluated by a board-certified pathologist (TA Harm) using a histopathologic scoring system similar to those previously described [5]. The scoring system for the lung samples consisted of six independent scoring parameters. Parameters included inflammation, distribution of the inflammation, necrosis, BALT hyperplasia, edema, and hemorrhage. The scoring system for the liver and kidney samples consisted of six independent scoring parameters that included inflammation, distribution of inflammation, inflammatory distribution, necrosis, edema, hemorrhage, and extramedullary hematopoiesis (EMH). Each parameter was scored on a scale ranging from 0 to 5, and the sum of the scores was added per animal for a total possible score of 30. For inflammation, necrosis, edema, and hemorrhage, a score of 0 indicated no tissue affected, a score of 1 indicated 1–15% of the tissue was affected, a score of 2 indicated 16–30% of the tissue was affected, a score of 3 indicated that 31–45% of the tissue was affected, a score of 4 indicated that 46–60% of the tissue was affected, and a score of 5 indicated that greater than 60% of the tissue was affected. For inflammation distribution, a score of 0 indicated no lesion, a score of 1 indicated a focal lesion, a score of 2 indicated multifocal lesions, a score of 3 indicated coalescing lesions, a score of 4 indicated a locally extensive lesion, and a score of 5 indicated a diffuse lesion. For BALT hyperplasia, a score of 0 indicated no hyperplasia, a score of 1 indicated 2–3 hyperplastic BALTs, a score of 2 indicated 4–5 hyperplastic BALTs, a score of 3 indicated 6–10 hyperplastic BALTs, a score of 4 indicated 11–15 hyperplastic BALTs, and a score of 5 indicated greater than 15 hyperplastic BALTs. For EMH, a score of 0 indicated no EMH, a score of 1 indicated less than 5 EMH aggregates, a score of 2 indicated 6–10 aggregates, a score of 3 indicated 11–15 aggregates, a score of 4 indicated 15–20 aggregates, and a score of 5 indicated locally extensive infiltrates.

Anti-H3N8 HA ELISA

Anti-H3N8 HA antibody titers were measured using an enzyme-linked immunosorbent assay (ELISA) with horseradish peroxidase linked secondary antibodies. Blood was collected from the saphenous vein at 14 and 35 days (Fig. 2B), centrifuged at 5000 rcf for 15 min; sera were separated and stored overnight at 4 °C. High-binding, flat-bottom 96-well plates were coated with 100 μL 0.5 μg mL−1 H3N8 HA in pH 9.6 carbonate/bicarbonate buffer at 4 °C overnight. Wells were blocked with 1% (w/v) gelatin in PBS with 0.05% Tween-20 (PBS-T) for 2 h at room temperature. Plates were washed three times in PBS-T before adding 100 μL of PBS-T to each well. Sera samples were added to the first well in the microtiter plate at a dilution of 1:100; subsequently, two-fold serial dilutions were made across the plate and incubated at 4 °C overnight. Plates were again washed three times in PBS-T before adding 100 μL of PBS-T with secondary horseradish peroxidase-conjugated, goat anti-mouse IgG (H + L) at a dilution of 1:20,000. Plates were incubated for 2 h at room temperature and then washed four times in PBS-T followed by the addition of 75 μL ultra-TMB-ELISA substrate solution (Thermofisher, Waltham, MA). The reaction was allowed to proceed for 20 min and halted by the addition of 75 μL of 2 N sulfuric acid. The optical density (450 nm) was recorded using a plate reader (Spectramax, San Jose, CA). For analysis, the background is defined as the average optical density of the wells treated with sera from saline administered mice. Titer is defined as the reciprocal of the last dilution with an optical density greater than two times the background.

Fig. 2

Schematic diagram of the intranasal formulation and the subcutaneous formulation. Particle scale is not to scale, with polyanhydride particles shown 20 × larger and actual size being 200 × larger

Viral Challenge

For influenza challenge, mice were anesthetized with isoflurane to effect and challenged intranasally with an infectious dose of 15 LD50 of A/HK/1/68 (H3N2) in 25 μL sterile PBS (pH 7.6). The infectious viral dose for this viral culture in young, female C57BL/6 mice was determined previously in a challenge titration study (data not shown) finding 1 LD50 = 140 tissue culture infectious units (TCIU).

Statistical Analyses

Statistical comparisons were performed using Graphpad (Prism 10, Graphpad Software, La Jolla, CA). Statistical comparisons of PenH and MVb measurements were performed using multiple nonparametric Mann–Whitney tests. Comparisons of serum biomarkers and bronchoalveolar lavage markers were performed using nonparametric Kruskal–Wallis and Mann–Whitney tests. Numerical data are reported as averages with the standard error of the mean. Comparisons of weight loss profiles after challenge utilized a one-way ANOVA with repeated measures, and survival curve comparisons were performed using Mantel-Cox log-rank tests comparing against naïve controls.