This study was conducted in compliance with the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes (20). These studies were approved by the University of Adelaide Animal Ethics Committee under ethics approval number S-2018-013.

Influenza A/Puerto Rico/8/1934 [H1N1] (A/PR8) and A/California/07/2009 [H1N1] (A/California) were grown in the allantoic cavity of 10-day-old embryonated chicken eggs at 37°C for 48 hours. Eggs were then chilled at 4°C overnight, and infected allantoic fluid was harvested and clarified by centrifugation at 3272 × g for 10 minutes.

Vaccine concentration and purification was performed by haemadsorption as described previously (21). Briefly, infected allantoic fluid was incubated with chicken erythrocytes at 4°C for 1.5 hours to allow virus adsorption to red blood cells (RBCs). Samples were then centrifuged to pellet virus-RBC complexes, and allantoic fluid supernatant was removed. Pellets were resuspended in 0.85% saline and incubated at 37°C for 1.5 hours to allow virus release from RBCs. Samples were then centrifuged to pellet RBCs, and the virus-containing supernatant was collected, aliquoted and stored at -80°C until required. Titres of concentrated IAV stocks were estimated as 3 × 10⁹ TCID₅₀/mL and 4 × 10⁷ TCID₅₀/mL for A/PR8 and A/California, respectively, by TCID₅₀ assay.

Concentrated IAV stocks of A/PR8 at a TCID₅₀ of 3 × 10⁹ TCID₅₀/mL were inactivated by γ-irradiation at the following temperature conditions: frozen on dry-ice (DI, approximately -78.5°C), cold on ice water (ice, 4-8°C) or at room temperature (RT, 24-27°C), generating DI-γ-Flu, Ice-γ-Flu, and RT-γ-Flu respectively. Sterilising doses were calculated as described previously to be 35 kGy for DI-γ-Flu and 16 kGy for Ice- and RT-γ-Flu (12).

Irradiation was performed using a cobalt-60 batch-type gamma irradiator at the Australian Nuclear Science and Technology Organisation (ANSTO, NSW). Samples for irradiation were double-contained in cryovials within 10 ml falcon tubes and placed in a 45 litre cooler box, sited in a fixed, reproducible location within the irradiation room. Samples were placed in pre-determined positions in the cooler box at various distances from the radiation source so that multiple doses could be delivered simultaneously. The cooler box was then filled with water (RT), chilled water containing ice blocks, or powdered dry ice for the different temperature conditions. Radiation doses were measured using calibrated Fricke (22) and ceric-cerous dosimeters (23) and dose rates varied from 0.3-1.6 kGy/h.

Temperature was monitored with a calibrated digital temperature probe connected to a data logger (Novus LogBox-AA) for ice and RT samples for the duration of irradiation, and non-irradiated control samples were subject to the same temperature conditions, stored out of the irradiation room. After irradiation, all samples were stored at -80°C until required.

IAV was titrated by 50% tissue culture infectious dose (TCID₅₀) assay using Madin-Darby canine kidney (MDCK) cells. MDCK cells were maintained in Dulbecco’s Modified Eagle’s Medium (DMEM) with 10% foetal bovine serum (FBS) and 1% penicillin/streptomycin (P/S). MDCK cells were kept at 37°C with 5% CO₂ and were passaged with trypsin when they reached approximately 90% confluence. For TCID₅₀ assay, MDCK cells were seeded in 96-well round-bottomed plates at 5 × 10⁴ cells/well. After 24h incubation, confluent cell monolayers were infected with 10-fold serial dilutions of IAV in DMEM supplemented with 8% trypsin for virus activation. Plates were incubated at 37°C for 3 days, then amplified virus in culture supernatants was detected by the addition of 0.6% packed RBCs based on pellet or mesh formation, with a mesh being considered positive for IAV. 50% infectious doses (TCID₅₀/mL) were calculated using the Reed and Muench method (24).

For haemagglutination assays, serial dilutions of IAV were performed in 0.85% saline in a 96-well round-bottomed microtitre plate. 0.6% packed RBCs in 50μL were added to each well and plates were scored for mesh or pellet formation. The reciprocal of the highest virus dilution showing a mesh was used to determine the total haemagglutination units (HAU/mL).

Sterility testing was also performed after γ-irradiation of A/PR8 to ensure that the doses selected were sterile. MDCK cells were plated in 96-well flat-bottomed microtitre plates at 2 × 10⁴ cells/well. γ-Flu was activated with 10μg/mL TPCK-trypsin at 37°C for 30 minutes then diluted 1:10 in DMEM + 1% P/S + 0.5μg/mL TPCK-trypsin. Inoculum was added to MDCK cells at an MOI-equivalent of 600 and cells were then incubated at 37°C for 24 hours to allow virus replication (passage 1). Supernatant was then collected and used to infect fresh MDCK monolayers (passage 2). This was then repeated for passage 3. At the time of collecting supernatant, cells were washed with PBS then fixed and permeabilised with 1:1 acetone:methanol (v/v) at 4°C for 15 minutes. Cells were then stained with polyclonal mouse anti-A/PR8 serum (1:200 dilution in PBS) for 1 hour at 4°C followed by Alexa-fluor^® 488 goat anti-mouse IgG secondary antibody (Life Technologies, 1:500 dilution). DAPI was used to stain cell nuclei (1μg/mL in DAPI). Images were taken using the Nikon TiE inverted fluorescence microscope and analysed using NIS elements software (Tokyo, Japan).

Two-fold serial dilutions of live and irradiated IAV samples were performed in PBS in triplicate. Samples were then incubated with 0.125mM of 2’-(4-Methylumbelliferyl)-α-D-N-acetylneuraminic acid (4-MUNANA, Sigma M8639) at 37°C for 1 hour, facilitating cleavage of 4-MUNANA by active IAV neuraminidase (NA) into the fluorescent substrate 4-Methylumbelliferyl (4-MU). 4-MU (Sigma M1381) was also included at increasing concentrations to generate standard curves. After 1 hour the assay was stopped with ice-cold 0.5M Na₂CO₃ at pH 10.5 and read using a SpectraMax fluorescent plate reader with an excitation wavelength of 365nm and emission wavelength of 450nm.

Irradiated IAV at different temperatures was loaded onto 3mm formvar/carbon coated grids (approx. 3 μL/grid) and left to settle for 3 to 5 minutes. Grids were blotted to dry, washed, then stained with 2% uranyl acetate for 3 minutes. Grids were then washed with PBS and blotted to dry prior to visualisation using the FEI Tecnai G2 Spirit TEM (Adelaide Microscopy, University of Adelaide).

6-8 week old female BALB/c mice were vaccinated intranasally under ketamine anaesthetic (10% ketamine, 1% xylazil in sterile water, inject IP at 10μL/gram of body weight) with 32μL of either PBS (mock-vaccine control) or A/PR8-derived γ-Flu irradiated at different temperatures (9.6 × 10⁷ TCID₅₀-equivalent/mouse). Immune serum was collected 20 days post-immunisation by submandibular bleeding. Mice were then challenged intranasally with lethal IAV on day 21 (3 weeks post-immunisation), under ketamine anaesthetic as above. Lethal doses were determined by challenging mice with serially diluted IAV. The lowest virus concentration that gave 100% lethality in mice was selected (data not shown). Challenge doses used were 1.6 × 10² TCID₅₀/mouse for A/PR8 and 1.3 × 10⁵ TCID₅₀/mouse for the human isolate A/California. A higher dose was required to achieve lethality for A/California. Weight loss was measured daily for a period of 21 days post-challenge, with a 20% loss of starting body weight was used as a humane end point.

Enzyme-linked immunosorbent assay (ELISA) was used to measure IgG responses in serum samples from vaccinated and control mice. Plates were coated with A/PR8 in bicarbonate/carbonate coating buffer and incubated overnight at room temperature. Plates were then blocked with 2% skim milk for 2 hours. Serum samples were serially diluted then added to the plate for 2 hours at room temperature. Plates were washed and horseradish peroxidase-conjugated goat anti-mouse IgG antibody (1:10,000 dilution in blocking buffer, Thermo Scientific) was added to each well. After 2 hours at room temperature, plates were washed, and colour was developed using TMB peroxidase substrate in the dark for 30 minutes then the reaction was stopped with 2M H₂SO₄. Absorbance was measured at 450nm using a Bio-Tek Instruments plate reader. The reciprocal of the highest dilution to give absorbance readings higher than naïve mice + 3 standard deviations was considered the IgG titre.

To measure neutralising antibody responses, a focus-forming inhibition assay was used. Monolayers of MDCK cells were treated with 0.1 MOI of A/PR8 that has been pre-incubated with serial dilutions of immune serum. Virus was allowed 2 hours to adhere to cells then inoculum was removed, and cells were washed with PBS. Fresh media was added, and cells were incubated at 37°C for a further 22 hours. Staining procedure and visualisation were performed as described for sterility testing. For measuring A/California neutralisation, the primary antibody used was polyclonal murine anti-A/California serum (1:200 dilution). Secondary antibody was Alexa-fluor^® 488 goat anti-mouse IgG secondary antibody (Life Technologies, 1:500 dilution).

Cytotoxic T lymphocyte (CTL) assays were performed as described previously (15). Mice were vaccinated intravenously with 3 × 10⁸ TCID₅₀-equivalent of γ-Flu. 7 days later, spleens were harvested from naïve donor mice, minced, and pushed through a 70μm filter to generate a single-cell suspension. Cells were then split into equal populations, and one was pulsed with K^d-restricted influenza nucleoprotein (NP) peptide (NPP, sequence: TYQRTRALV) and stained with CFSE (NPP-Pulsed). The second population was stained with cell tracker red (CTR) only (Unpulsed). The cells were mixed at a 1:1 ratio and injected intravenously into vaccinated and non-vaccinated control mice at 10⁷ cells/mouse. 24 hours later, all mice were sacrificed, and spleens were harvested and processed into a single-cell suspension prior to fixing using 1% PFA. Labelled pulsed and non-pulsed cells were acquired using the LSRII flow cytometer (BD Biosciences), and data was analysed using FlowJo software (Treestar Incorporated).

Statistical analysis was performed using GraphPad Prism version 8 (GraphPad Software, La Jolla, CA, USA). Quantitative results were expressed as mean ± SEM. One-way ANOVA (with Tukey’s multiple comparisons test) was used for comparison of data from 3 or more groups. Survival data were analysed using Fisher’s exact test (two-tailed). P-values < 0.05 (95% confidence) were considered statistically significant.

The aim of this study was to compare immunogenicity of vaccines irradiated to the SAL across different temperatures. Sterilising doses required to reduce virus titre to an acceptable SAL of 10^-6 were calculated as described previously (12). For DI-irradiation, the DS_SAL was determined to be 35 kGy (DI-γ-Flu) and for ice- and RT-irradiation the sterilising dose was calculated to be 16 kGy (Ice-γ-Flu, RT-γ-Flu). Sterility testing based on multiple in vitro passages was performed to ensure complete inactivation of irradiated materials. Live and irradiated IAV samples were passaged three times in MDCK cells, with supernatants from each treated monolayer (or passage) used to treat the next MDCK monolayer. After 3 passages, monolayers were fixed and stained for IAV infection. No virus infectivity was detected in any of the MDCK monolayers treated with irradiated preparations for all 3 passages, whereas replication of live virus was amplified at each passage ( Figure 1 ). The irradiated materials were thus confirmed to be sterile and appropriate for subsequent in vitro and in vivo experiments.

The structural integrity of the IAV within each vaccine preparation was then assessed by HA and NA functionality assays. While hemagglutination assay show reduced HA activity for all γ-Flu preparations compared to live IAV ( Figure 2A ), no significant difference was detected between the three irradiated samples despite the highly varied temperature conditions used for irradiation. Furthermore, Figure 2B demonstrates that the functionality of NA proteins in each γ-Flu preparation was not affected by irradiation, with all three vaccine formulations showing comparable NA enzymatic activity to live IAV. Transmission electron microscopy was then used to examine whole virion structure. Representative images in Figure 2C show that virions within all three irradiated preparations were intact and retained spherical IAV structure. This shows that in addition to having minimal impact on surface proteins, exposing IAV to DS_SAL at relative temperature conditions does not cause substantial damage to viral envelopes.

Given that all three γ-Flu preparations appeared suitably intact in terms of virion structure and protein functionality, we next assessed their efficacy as vaccine candidates in animal models. Initially, mice were vaccinated intranasally with a single dose of each vaccine preparation (DI-γ-Flu, Ice-γ-Flu, or RT-γ-Flu), or with PBS as a mock-vaccine control. 20 days post-immunisation, sera was harvested from all animals and an ELISA was performed to determine IAV-specific IgG titres. As shown in Figure 3A , all three γ-Flu preparations induced strong IgG responses above PBS-mock control levels, and no significant difference was detected between IgG titres induced by the three γ-Flu preparations. Interestingly, whilst not significant, there was a trend towards lower IgG responses detected in serum samples from mice vaccinated with DI-γ-Flu.

Following this, a focus-forming inhibition assay was performed to determine the ability of γ-Flu-induced antibodies to inhibit receptor binding and IAV infection. Neutralising antibody responses are crucial for protection against homotypic IAV infection, thus it is important to assess antibody functionality in addition to overall titre. Serum samples from γ-Flu-vaccinated and control mice were used to treat live A/PR8, and virus:serum mixtures were used to infect monolayers of MDCK cells. After a 22h incubation period, cells were stained with DAPI to visualise cell nuclei, and with murine anti-APR8 and FITC-conjugated anti-murine antibodies to visualise IAV-infected cells. Fluorescence levels of each fluorophore were quantified, and FITC-fluorescence relative to DAPI-fluorescence was calculated to determine the average IAV infectivity per cell. Quantified fluorescence of serum-treated virus samples were then compared to untreated virus only controls. As shown in Figure 3B , no reduction in infectivity was detected for virus treated with PBS-mock control sera, indicating that the murine sera from naïve animals had no effect on IAV infectivity. Conversely, infectivity was significantly reduced when A/PR8 was treated with serum from DI-, Ice- and RT-γ-Flu vaccinated mice. Interestingly, immune sera from mice vaccinated with Ice- and RT-γ-Flu was significantly more effective at neutralising A/PR8 when compared to immune sera from mice vaccinated with DI-γ-Flu. Representative images of virus neutralisation were also taken at a 1:10 serum dilution, and similarly demonstrate a clear reduction in foci for all γ-Flu groups, with antibodies induced by Ice- and RT-γ-Flu vaccination being the most effective ( Figure 3C ). This trend is likely due to the higher titre of total IgG present in immune sera from Ice- and RT-γ-Flu vaccinated animals, compared to those immunised with DI-γ-Flu.

Given the observed differences in functionality of γ-Flu-induced antibodies, we challenged vaccinated animals with live IAV to assess if these variations would translate to detectable differences in protective efficacy. Initially, the ability of DI-, Ice-, and RT-γ-Flu to mediate homotypic protection was investigated. Three weeks post-vaccination, mice were challenged with a lethal dose of homotypic A/PR8. No clinical symptoms were observed and no weight loss was recorded for all vaccinated groups, in contrast to PBS-mock control mice that succumbed to A/PR8 challenge and showed progressive weight loss to reach the humane end point of 20% body weight loss by day 7 post-infection ( Figure 4A ). Importantly, all vaccinated mice, irrespective of vaccine irradiation temperature, show 100% survival based on using 20% bodyweight loss as the humane end point ( Figure 4B ). This indicates that the antibody responses shown in Figure 3 , though variable, were sufficient to induce robust homotypic protection.

Importantly, a key feature of γ-Flu is its ability to induce cross-protective CD8⁺ T-cell responses against vaccine and non-vaccine IAV strains. To assess the effect of the differential irradiation temperatures on the induction of CD8⁺ T-cell responses, an in vivo CTL assay was performed. Here, the killing of IAV NPP-pulsed splenocytes (target cells) was assessed in vaccinated and non-vaccinated animals. NP has been identified as a key CD8⁺ T cell IAV antigen (25). As shown in Figure 5 , splenocytes from naïve control mice show a 1:1 ratio of pulsed target cells to unpulsed cells, indicating no non-specific killing of targets cells in vivo. Conversely, we detected a substantial loss of NPP-pulsed cells relative to unpulsed cells in all three γ-Flu vaccinated groups. This demonstrates the ability of all γ-Flu preparations to induce a robust IAV-specific CTL responses as pulsed target cells were lysed within 24h of injection into immunised animals. Interestingly, animals vaccinated with Ice-γ-Flu and RT-γ-Flu showed significantly more effective CTL responses (97% and 93% killing of IAV-pulsed targets, respectively) compared to animals vaccinated with DI-γ-Flu (73% killing of IAV-pulsed targets).

Enhanced IAV-specific CTL responses should theoretically translate to enhanced cross-protection against newly emerging IAV strains. To assess this, mice were vaccinated intranasally with different γ-Flu preparations (based on A/PR8 [H1N1]), or PBS-mock control. Three weeks later, mice were intranasally challenged with a lethal dose of A/California, the pdmH1N1 strain. As shown in Figure 6 , all vaccinated and non-vaccinated mice experienced some weight loss following A/California infection, however mice vaccinated with Ice-γ-Flu showed less weight loss and faster recovery than the other vaccine groups. Furthermore, 100% survival was recorded for mice vaccinated with Ice-γ-Flu and RT-γ-Flu, whereas 86% survival occurred in mice vaccinated with DI-γ-Flu (1 out of 7 mice reached the humane end point of 20% weight loss). Overall, while γ-Flu vaccination was associated with significantly less weight loss and faster recovery time in all vaccinated groups, only Ice-γ-Flu and RT-γ-Flu was associated with significantly higher survival rates compared to the unvaccinated group. This outcome is consistent with the enhanced CTL responses ( Figure 5 ). To rule out the possibility that the protection demonstrated in Figure 6 was mediated by neutralising antibody responses, we tested the ability of serum generated by different γ-Flu preparations to neutralise A/California. Live A/California was treated with serial dilutions of serum generated by DI-, Ice- or RT-γ-Flu, or naïve serum as a control. Virus + serum was then added to confluent monolayers of MDCK cells and allowed to adhere for 2 hours before unbound virus was washed away. Cells were incubated for a further 2 hours at 37°C to allow virus growth then cells were fixed and stained with murine anti-A/California serum used as a primary antibody. As expected, we observed no cross-neutralisation generated by A/PR8 based γ-Flu preparations against A/California H1N1 ( Figure 7 ).