Madin-Darby canine kidney (MDCK) cells were maintained in minimum essential medium (MEM) (Gibco) supplemented with 5% newborn calf serum (Sigma) at 37°C in 5% CO₂. MDCK cells were used for plaque assays to determine virus titers.

Mouse-adapted A/California/04/2009 virus (H1N1; MA-CA04), generated in our laboratory as previously described (Sakabe et al., 2011), was used to challenge mice. A/California/07/2009 virus (H1N1; CA07), which was isolated early in the 2009 pandemic, was used as an antigen for the ELISA after purification and inactivation to determine the virus-specific antibody titers of sera obtained from the immunized mice.

Quadrivalent split influenza HA vaccines were obtained from DENKA SEIKEN Co., Ltd. (Japan). The quadrivalent split influenza HA vaccine (for the 2016–2017 season), which contains the HA proteins (equivalent to 30 μg of HA protein for each virus included in a vaccine vial) of CA07 (H1N1), A/Hong Kong/4801/2014 (H3N2), B/Phuket/3073/2013 (Yamagata lineage), and B/Texas/2/2013 (Victoria lineage), was used. Aluminum hydroxide gel Alhydrogel^® adjuvant 2% (alum), purchased from InvivoGen, was used as a positive control [antigen: alum = 1:1 (v/v)] (approximately equal to 500 μg of alum/dose). The 7DW8-5 was purchased from Funakoshi Co., Ltd. (endotoxin level, <100 EU/mg), suspended in sterile water at a concentration of 1 mg/ml, heated at 80°C for 10 min, and then sonicated in a water bath for 15 min at room temperature. Stocks were stored at −20°C until use. Before being mixed with the split influenza HA vaccine, the 7DW8-5 suspension was sonicated again for 5 min after thawing. The 7DW8-5 and the HA split vaccine were diluted with the endotoxin-free D-PBS (Millipore, Cat. TMS-012-A, Endotoxin <0.005 EU/mL).

Five-week-old female BALB/c mice were purchased from Japan SLC Inc. After one week of adaptation, the mice (10 mice per group) were immunized with a suboptimal dose of influenza HA vaccine [0.001 μg/dose (2016–2017 season) calculated on the basis of the amount of HA from CA07] with or without 7DW8-5 into the gastrocnemius muscle. Two weeks later, the mice were boost immunized (second immunization) intramuscularly. On day 14 after the boost-immunization, blood was collected via the facial vein by using a goldenrod animal lancet (5 mm), and sera were obtained to measure virus-specific antibody titers. Three weeks after the boost-immunization, the immunized mice were challenged intranasally, under anesthesia, with 10 MLD₅₀ (Fifty Percent Mouse Lethal Dose) of MA-CA04 virus. Body weight and survival of four mice from each group were monitored daily for 14 days after virus challenge. Mice that lost more than 25% of their original body weight were euthanized.

The remaining mice (6 mice per group) were used to assess virus replication. To determine virus titers in mice, organ samples were harvested on days 3 and 6 post-challenge and were homogenized and titrated on MDCK cells by using a plaque assay.

Virus-specific antibody titers in the sera were determined by using a modified ELISA as previously described (Uraki et al., 2014). Briefly, 96-well ELISA plates (IWAKI) were coated with 6 μg/ml of inactivated, purified CA07 virus solution overnight at 4°C (50 μl/well). The plates were then blocked with 200 μl of 20% Blocking One (Nacalai) in water at room temperature for 1 h. After blocking, the plates were washed once with PBS containing 0.05% Tween-20 (PBS-T), and then 2-fold serially diluted serum samples were added to the plates, followed by a 1-h incubation at room temperature. Bound total IgG was detected by using peroxidase-labeled goat anti-mouse IgG (gamma) antibody, F (ab′) 2 fragment (Kirkegaard & Perry Laboratory Inc.) and horseradish peroxidase-conjugated anti-mouse IgG1, or IgG2a antibodies (Southern Biotech). After the plates were washed four times with PBS-T, 100 μl of 2, 2′-azino-bis (3-ethylbenzothiazoline-6-sulfonic acid) diammonium salt substrate solution was added to each well to initiate the color reaction, and the OD was measured at a wavelength of 405 nm. The antibody titer was defined as the reciprocal of the highest serum dilution that produced an OD₄₀₅ > 0.1 after correcting for the negative serum control.

Animal tissues were fixed in 4% paraformaldehyde phosphate (PFA) buffer solution for pathologic examination. They were then processed for paraffin embedding and cut into 3 μm-thick serial sections. The sections were stained using a standard hematoxylin and eosin procedure, and serial sections were processed for immunohistological staining with a rabbit polyclonal antibody for type A influenza virus nucleoprotein (prepared in the Department of Pathology, National Institute of Infectious Diseases, Tokyo, Japan) or a rat monoclonal antibody for mouse F4/80 antigen (CL: A3-1, Bio Rad). Specific antigen-antibody reactions were visualized with 3, 3′-Diaminobenzidine (DAB) staining by using a Dako Envision system (Dako Cytomation).

We used R³ and lme4 (Bates et al., 2015) to perform a linear mixed effects analysis of the body weight data, which were normalized to the initial weight of each individual animal. As fixed effects, we used the different treatment groups (i.e., vaccine alone, vaccine plus 7DW8-5 and vaccine plus alum), and the time of the measurement (with an interaction term between those fixed effects). As random effects, we used the intercepts for the individual animals. We used the lsmeans (Lenth, 2016) package to compare the groups at different time points for each model separately, and the p-values were adjusted using Holm’s method. For the comparisons of antibody titers and virus titers, we used a one-way ANOVA, followed by Tukey’s Post Hoc tests. The data were log-transformed before the comparisons were made to stabilize the variance. For the virus titer data, each timepoint was analyzed separately. For the antibody titers, values <10 were arbitrarily set to 5, because it was not possible to determine whether antibodies were completely absent or were present below the detection limit. For the analysis of the survival data, we used the Log-rank test, comparing the vaccine plus 7DW8-5 or alum to the vaccine alone group. We used OASIS 2 (Han et al., 2016) software for this analysis. P values of <0.05 were considered statistically significant.

All experiments with mice were performed in the biosafety level 2 containment laboratory in the Institute of Medical Science, the University of Tokyo (Tokyo, Japan) in accordance with the Regulations for Animal Care of the University of Tokyo and the Guidelines for Proper Conduct of Animal Experiments by the Science Council of Japan, and were approved by the Animal Experiment Committee of the Institute of Medical Science, the University of Tokyo (approval no. PA 14-38).

To evaluate the adjuvant effect of the glycolipid 7DW8-5 on a commercial HA split vaccine, mice were immunized with PBS, 7DW8-5 (1 μg/dose or 10 μg/dose), HA vaccine alone (0.001 μg of HA for each virus/dose), or HA vaccine plus 7DW8-5 (1 μg/dose or 10 μg/dose) via intramuscular administration in a 100 μl volume twice with a 2-week interval between the vaccinations. Commercially available alum adjuvant was used as a positive control, as described in the Materials and Methods, because alum is the most frequently used adjuvant worldwide and has been used in many clinical studies (Tetsutani and Ishii, 2012). Two weeks after the boost immunization, sera samples were obtained from the immunized mice and examined for the presence of virus-specific antibody in an ELISA. No antibody against CA07 virus was detected in the groups of mice that receive PBS only or 7DW8-5 only. Most mice immunized with the HA vaccine alone produced no or very low levels of virus-specific antibodies except for one mouse whose virus-specific IgG titer was 1280 (Figure 1). The mean antibody titer in the sera of the group of mice immunized with the HA vaccine plus 1 μg of 7DW8-5 (i.e., 1440.0) was significantly higher than that in the vaccine alone group (i.e., 240.5) (Figure 1). The vaccine plus 10 μg of 7DW8-5 also induced much higher levels of virus-specific antibody compared with the vaccine alone group, and the mean antibody titer was comparable to that induced by HA vaccine plus alum (Figure 1). These results demonstrate that 7DW8-5 significantly enhances the immunogenicity of the HA split vaccine.

We also measured the influenza virus-specific IgG1 and IgG2a titers in the sera by using an ELISA because IgG1 and IgG2a are stimulated during Th2-type and Th1-type immune responses, respectively (Stevens et al., 1988; Mosmann and Coffman, 1989; Hauge et al., 2007). The IgG1 titers in the HA vaccine plus 1 μg of 7DW8-5, the HA vaccine plus 10 μg of 7DW8-5, and the HA vaccine plus alum groups were significantly higher than that in the HA vaccine alone group (Figure 1B). Seven of the ten mice immunized with the HA vaccine plus 10 μg of 7DW8-5 expressed a high level of IgG2a antibodies, whereas no IgG2a antibody was detected in the other groups (Figure 1C). These results suggest that the HA vaccine plus 10 μg of 7DW8-5 induced both Th1-type and Th2-type immune responses, whereas the HA vaccine plus 1 μg of 7DW8-5 and the HA vaccine plus alum induced only the Th2-type immune response.

To examine whether 7DW8-5 enhances the protective efficacy of the HA split vaccine, we challenged the immunized mice with 10 MLD₅₀ of MA-CA04 virus 3 weeks after the boost-immunization and monitored their body weight changes and survival for 14 days. All mice given PBS or 7DW8-5 alone experienced body weight loss upon virus challenge and died by 6 days post-challenge, whereas 2 of 4 mice were protected from lethal challenge in the vaccine plus alum group (Figure 2 and Supplementary Table 1). More importantly, 3 of 4 mice immunized with the vaccine plus 1 μg of 7DW8-5 survived and all four mice that received the vaccine plus 10 μg of 7DW8-5 were protected from the lethal infection (Figure 2 and Supplementary Table 1). These results demonstrate that the adjuvanticity of the glycolipid adjuvant 7DW8-5 was sufficient to protect mice from lethal challenge with MA-CA04 virus.

To examine the effect of 7DW8-5 on virus replication in the immunized mice after challenge, the mice immunized with HA vaccine plus 7DW8-5 were challenged with 10 MLD₅₀ of MA-CA04 virus 3 weeks after the second immunization, and organ samples (i.e., nasal turbinates and lungs) were collected from the sacrificed mice on days 3 and 6 post-infection for virus titration. On day 3 post-challenge, over 10⁶ PFU of virus was detected in both the nasal turbinates and lungs of mice in all of the groups (Table 1). In contrast, on day 6 post-challenge, the mean virus titers in the lungs for the vaccine plus 7DW8-5 and the vaccine plus alum groups were lower than that of the vaccine alone group, although the difference was not statistically significant (Table 1).

Pathological analysis of mice immunized with PBS, 7DW8-5 (10 μg/dose), HA vaccine alone (0.001 μg/dose), or HA vaccine plus 7DW8-5 (10 μg/dose) after challenge with 10 MLD₅₀ MA-CA04 virus revealed that infection with MA-CA04 resulted in detectable viral antigens in bronchial epithelial cells and in infiltration of inflammatory cells into the lungs of all mice tested on day 6 post-challenge (Figure 3A). No apparent difference in inflammation, viral antigen distribution pattern, or the number of viral antigen-positive cells was observed across all groups (Figure 3A). Moderate inflammation, including infiltration of neutrophils, monocytes/macrophages, or lymphocytes, was observed in more than half of the sections from each mouse across all groups; however, we detected more F4/80-positive cells in the lungs of immunized mice after virus challenge in the HA vaccine alone and the 7DW8-5 plus HA vaccine groups compared with the PBS and 7DW8-5 groups (Figure 3B). These results suggest that immunization of mice with the HA vaccine induces infiltration of macrophages into the lungs after virus infection, although no difference in the number of F4/80-positive cells was observed between the HA vaccine alone and the 7DW8-5 plus HA vaccine groups.

Taken together, our results demonstrate that although 7DW8-5 did not facilitate a reduction in virus replication by the HA vaccine, it did improve vaccine efficacy as evaluated by lethality.