Female BALB/c mice (5–7 weeks old) were purchased from Sankyo Laboratories (Tokyo, Japan) or SLC (Shizuoka, Japan) and maintained in the animal facility at Tokyo University of Science and Aichi Medical University. All animal experiments were performed according to the guidelines of the Tokyo University of Science and Aichi Medical University.

In our previous report (10), we generated plasmids encoding the genes for the heavy chain (IgG) and the light chain (kappa) of a neutralizing anti-HA antibody (19) (Data S1 in Supplementary Material). In the current study, all constructions are based on the pCADEST1 vector, which was constructed from pCA5, a CAG promoter-driven plasmid, and pDEST12.2 (Invitrogen) (20). Mouse IgA, Joining chain, and IgE were cloned from spleen cells. IgM and IgD cDNA were cloned from the mouse B-cell line, WEHI-231. All the sequence data were described in Data S1 in Supplementary Material. Each plasmid vector encoding the anti-HA antibody was constructed by overlap PCR from pCADEST1-anti-HA IgG1 using the primers indicated in Table S1 in Supplementary Material. The plasmids, pCADEST1-anti-HA IgG1, pCADEST1-anti-HA IgA, pCADEST1-anti-HA kappa chain, and pCADEST1-Joining chain were purified by CsCl-ethidium bromide gradient centrifugation (21). pCADEST1-anti-HA IgM, pCADEST1-anti-HA IgD, and pCADEST1-anti-HA IgE were purified using NucleoBond^® kits (Clontech, CA, USA) according to the manufacturer’s instructions.

The procedure for measuring the anti-HA antibody levels was described previously (10, 22). Briefly, a 96-well plate was coated with HA protein purified from influenza virus [A/Puerto Rico/8/34 (A/PR8); H1N1] using affinity columns constructed by coupling of CNBr-activated Sepharose 4B beads (GE Healthcare UK Ltd., Buckinghamshire, England) and anti-HA antibodies. The plate was then incubated with PBS containing 25% Block Ace^® (Snow Brand Milk Products, Tokyo, Japan) for blocking. After washing, the plate was incubated with serially diluted mouse serum samples. Sera containing HA-specific antibodies were detected using alkaline phosphatase-conjugated goat anti-mouse IgG1 (Southern Biotech Birmingham, AL, USA) or an anti-mouse kappa chain (Southern Biotech). p-Nitrophenyl phosphate (Wako, Tokyo, Japan) was used as the chromogenic substrate. Absorbance was measured using AUTO READER III (Sanko-junyaku, Tokyo, Japan). Anti-HA IgG purified from a hybridoma (19) was used as the standard (10, 22). HRP-conjugated goat anti-mouse IgG (Southern Biotech), anti-mouse IgA (Thermo Fisher Scientific), anti-mouse IgM (Southern Biotech), and anti-mouse IgE (Southern Biotech) were purchased to measure anti-HA IgG, anti-HA IgA, anti-HA IgM, and anti-HA IgE in serum and nasal wash, respectively. Anti-HA IgD was measured using APC-conjugated rat anti-mouse IgD (11-26c.2a Biolegend, San Diego, CA, USA), followed by incubation with HRP-conjugated rabbit anti-rat IgG (Thermo Fisher Scientific). Finally, binding was detected using an ELISA POD Substrate TMB Kit (Nacalai Tesque, Tokyo, Japan). Absorbance was measured using Spectramax M5 (Molecular Devices, CA, USA). Secretary IgA against HA was measured using rabbit polyclonal antibodies against mouse pIgR (Sino Biological, Boston, MA, USA), followed by incubation with HRP-conjugated goat anti-rabbit IgG (BD Pharmingen, San Diego, CA, USA). We could not determine the titer of anti-HA IgD, because appropriate standards were not available.

For competitive ELISA shown in Figures 4C,D and 5B, we used biotin-conjugated anti-HA IgG antibodies purified from a hybridoma (19), along with HRP-conjugated streptavidin (Figure S1 in Supplementary Material).

The supernatants and standards (IgG1, IgA, IgM; Southern Biotech) (IgE; BD Pharmingen) were separated under non-reducing conditions by SDS-PAGE (6%) and transferred to a PVDF membrane (Immobilon-P) (Millipore, Billerica, MA, USA). The membrane was blocked with Blocking One reagent (Nacalai Tesque) for 30 min, followed by incubation at room temperature with a mixture of HRP-conjugated goat anti-mouse IgG (Southern Biotech), IgA, IgM, and IgE. Anti-HA IgD was measured using APC-conjugated rat anti-mouse IgD, followed by incubation with HRP-conjugated rabbit anti-rat IgG. The specific bands were visualized using the enhanced chemiluminescence substrate (GE Healthcare) on the ImageQuant LAS4000 system (GE Healthcare).

We performed EP (10, 23–27) and HD (11, 12) using a previously described method. In EP, 30 µg of plasmid (1 µg/µl) was injected into the adductor and rectus femoris and cranial tibial muscle. Six pulses (100 V, 50 ms, polarity reversal every 3 pulses) were delivered into the injection site. All muscles were pre-treated by an injection of bovine hyaluronidase about 10 min before gene transfer to decrease the viscosity of the extracellular matrix and facilitate DNA diffusion (28). In HD (12), mice were injected in the tail vein with PBS containing plasmid (e.g., 5 μg/1.6 ml), where the DNA volume was 8–12% of the body weight. The injection was performed over less than 5 s using a 26-gauge needle. Mice in the control group were not treated (naïve group) or received gene transfer with pCADEST1-empty (vector control group). To eliminate serum from the bronchoalveolar lavage and nasal wash specimens, they were obtained directly from the respiratory tract with a 20 G indwelling needle or a bent oral sonde kindly gifted by Dr. Kiyoko Iwatsuski-Horimoto (Institute Medical of Science, University of Tokyo).

Virus challenge was carried out as described previously (10, 29). Thus, mouse adapted influenza virus (A/PR8) was grown in the allantoic cavities of 10- to 11-day-old fertile chicken eggs and stored at −80°C until used. The mice were anesthetized and intranasally challenged with 1,000 times the 50% tissue culture infective dose (TCID₅₀) of A/PR8 in 20 µl PBS, which causes lethal pneumonia (40 LD₅₀). Mice were subjected to gene transfer on the indicated day post-infection. Survival and weight changes were monitored for 20 days after virus challenge.

BALB/c mice that underwent gene transfer in the same manner one day after virus challenge, were prepared for estimation of virus titers in bronchoalveolar lavage specimens using a previously described method (10, 30). Three days post-infection, the bronchoalveolar lavage was obtained by washing isolated lungs with 2 ml of PBS containing 0.1% bovine serum albumin. Serial 10-fold dilutions of the specimens were inoculated onto Madin-Darby canine kidney (MDCK) cells. Three days post-infection, the MDCK cells were fixed and stained with naphthol blue black (Sigma-Aldrich, St. Louis, MO, USA). The plates were then washed and 0.1 M NaOH was added to each well. The cytopathic effect (CPE) resulting from IAV infection was evaluated by measuring the absorbance. The virus titer was calculated as TCID₅₀ by the Reed–Muench method.

BALB/c mice were sublethally infected with 2 µl of PBS containing an A/PR8 virus suspension with 1,000 plaque-forming units (PFU) into each nostril (total volume of 4 µl per mouse) (29, 31). Three days post-infection, the nasal wash was obtained in 1 ml of PBS containing 0.1% bovine serum albumin. The viral titer was determined by the MDCK-plaque assay as described previously (29, 32). Briefly, MDCK cells were cultured in 6-well tissue culture plates. Upon attaining confluence, the medium was removed and serial 10-fold dilutions of the specimens were added to the cells. After 60 min, the inoculum was removed and 2 ml of agar medium containing acetylated trypsin (Sigma) was overlaid. Two days post-infection, the cells were stained with crystal violet (Nacalai Tesque), followed by counting the number of plaques to determine the viral titer in terms of PFU.

Human embryonic kidney (HEK) 293T cells were maintained in Dulbecco’s modified Eagle’s medium supplemented with 10% heat-inactivated fetal calf serum and penicillin–streptomycin–glutamine (ThermoFisher Scientific, Waltham, MA, USA). HEK293T cells were co-transfected with the indicated vector using FUGENE HD Transfection Reagent (Promega, WI, USA). After 1 week, the supernatants were obtained and used for the neutralizing assay.

The neutralizing titer of the antibody-expressed supernatants was measured by a micro-neutralization assay as described previously (10). Briefly, 100 TCID₅₀ of A/PR8 viruses were mixed with an equal volume of serially diluted supernatants treated with or without a receptor destroying enzyme (RDE) (Denka Seiken, Tokyo, Japan), and then incubated for 30 min at 37°C. The mixtures were then inoculated onto MDCK cells and incubated for 3 days. The CPE for IAV infection was evaluated by measuring the absorbance at 630 nm. The neutralization titer was defined as the highest dilution that demonstrated no CPE.

All graphs were constructed using GraphPad Prism 7 (GraphPad Software). Data were analyzed using a non-parametric Kruskal–Wallis test considering p < 0.05 as statistically significant. Data are shown and compared as medians and ranges of 3 or 5 mice in different groups.

We previously demonstrated the potent prophylactic efficacy of EP against IAV infection (10). On the other hand, for a therapeutic effect after IAV infection, rapid and high-level production of neutralizing antibodies is indispensable. We first confirmed the expression level of anti-HA antibodies in the serum during 5 days (120 h) after EP (10) and HD. The hydrodynamics procedure involves rapid injection of a large volume of plasmid-DNA solution into the tail vein of mice (12). Several reports show that hydrodynamic injection induces peak expression at 6–8 h after gene transfer in the liver, which is a major tissue of gene expression (11, 12). As shown in Figure 1A, we could first detect anti-HA antibody production in the serum at 24 h after EP. The antibody concentration reached approximately 3,000 (= 10^3.53) ng/ml after 120 h. On the other hand, by HD, we could first detect serum antibodies after 4 h. After 24 h, the antibody concentration reached approximately 20,000 (= 10^4.35) ng/ml and this expression level was stable until after 120 h. Thus, HD can induce antibodies more rapidly and at a higher expression level than that by EP.

HD induced rapid and stable production of neutralizing antibodies in the serum. However, it was unknown whether HD could also induce anti-HA antibody production in the respiratory mucosa. Because IAV first attacks the host respiratory mucosa, production of neutralizing antibodies at this site is essential for a protective effect against influenza virus infection. We and other groups have previously reported that serum IgG antibodies were secreted in mucosal tissues and prevented virus infection (10, 33). To confirm these results, we conducted EP or HD in BALB/c mice. At 20 days (EP) (10) or 5 days (HD), which was the peak time point of anti-HA antibody production, serum, bronchoalveolar lavage without serum contamination, urine, and feces were obtained from the mice and anti-HA IgG levels were measured by quantitative ELISA. As shown in Figure 1B, both methods could induce anti-HA antibody expression in the respiratory mucosa. HD induced more than eightfold higher levels of serum antibodies (from 10^4.31 to 10^5.24) and approximately fivefold higher levels in the bronchoalveolar lavage (from 10^1.38 to 10^2.07) compared to those by EP. Interestingly, antibodies were detected even in the urine and fecal specimens from mice subjected to HD [approximately 20–30 (= 10^1.28–10^1.41) ng/ml] (Figure 1B). These results suggest that HD could induce high levels of neutralizing antibodies in several mucosal tissues.

HD significantly induces rapid and potent neutralizing antibodies in the serum and in several mucosal tissues. To quantitatively assess the therapeutic efficacy of HD after IAV infection, virus titers were measured in bronchoalveolar lavage specimens. We first confirmed anti-HA antibody production in the serum after HD. As shown in Figure 2A, the serum anti-HA antibody levels reached approximately 70,000 ng/ml (= 10^4.86 ng/ml). At this time, the viral titer in the bronchoalveolar lavage was reduced to approximately 1/400 (from 10^7.44 to 10^4.86) compared to that in the vector control group (Figure 2B). These data therefore indicate that a single HD administration induces neutralizing antibody production and that these antibodies control the proliferation of IAV.

To evaluate the therapeutic effect of HD after IAV infection, we infected BALB/c mice with a lethal dose of IAV. We then performed HD at one (Day 1 p.i.) or two (Day 2 p.i.) days post-infection. The changes in body weight (Figure 3A) and survival rates (Figure 3B) were monitored for 20 days. As shown in the left panels of Figures 3A,B (Day 1 p.i.), the body weight of all naive mice and empty vector transferred mice was decreased severely and all mice died within 7 days post-infection. In contrast, all mice subjected to HD survived even 20 days post-infection. These mice showed between 10 and 15% loss of body weight from 3 to 11 days post-infection, but by 12 days post-infection, all mice completely recovered from the weight loss. As shown in the right panels of Figure 3 (Day 2 p.i.), the body weight of all naive mice and empty vector transferred mice was decreased severely and all mice died within 15 days post-infection. In contrast, all mice subjected to HD survived even after 20 days post-infection. Some of these mice showed up to 25% loss of body weight from 1 to 19 days post-infection, but by 20 days post-infection, all mice had completely recovered from the weight loss. These results indicate that a single administration of HD can rescue mice from lethal IAV infection even at 2 days post-infection.

Our previous report and the current research indicate that anti-HA IgG induced by antibody gene delivery had a neutralizing effect against IAV infection (10, 34, 35). However, it is unknown whether other isotypes induced by antibody gene delivery can protect against IAV infection. To examine this, we first modified the plasmid encoding the anti-HA antibody. We genetically switched anti-HA IgG to anti-HA IgA, IgM, IgD, and IgE. In order to analyze the expression level of these isotypes in vitro, we transfected these plasmids into HEK293T cells and analyzed the antibody levels in the supernatants. As shown in Figure 4A, we observed specific bands for anti-HA IgG, anti-HA IgA, and anti-HA IgE. These bands were almost the same size as those of the respective standard samples. Although we could recognize a dimer of anti-HA IgA, polymeric IgM was not observed on this membrane (Figure 4A, lanes 4, 5, and 7), whereas the polymeric form was detected in standard IgM (Figure 4A, lane 6). On the other hand, in Figure 4B, we could recognize both monomeric and polymeric forms of IgD (Figure 4B, right panel, lane 3). We next quantified the antibody concentration in the supernatant by competitive ELISA. As shown in Figure 4C, production of anti-HA IgA, IgM, IgD, and IgE was found to be greater than that of IgG. The neutralizing titer of anti-HA IgA and IgM was almost equal to that of anti-HA IgG (Figure 4E). However, the neutralizing ability of anti-HA IgD and IgE was low and almost non-specific. Because influenza HA binds to sialic acid, RDE that is a sialidase from Vibrio cholerae, is usually added to avoid non-specific binding due to sialic acid present in the supernatant. Interestingly, the binding activity in anti-HA IgD and IgE was reduced with RDE (Figure 4D). Moreover, the neutralizing activity of anti-HA IgD and IgE was also completely reduced in the presence of RDE (Figure 4F). The reason for this inactivity was determined by western blot analysis. As shown in Figure 4B, anti-HA IgD (Figure 4B, right panel, lane 4) and IgE (Figure 4B, left panel, lane 6) were not detected upon treatment with sialidase. These data demonstrate that anti-HA IgG, IgA, and IgM possess neutralizing activity for IAV infection, whereas anti-HA IgD and IgE are easily inactivated by sialidase and consequently have no neutralizing activity.

As mentioned above, we succeeded in expressing all classes of anti-HA Igs in vitro. Next, we evaluated whether all classes of anti-HA Igs are equally expressed in vivo. We performed HD for each class of anti-HA Ig in mice and collected the serum 1 day after injection. Other than IgM and the already demonstrated IgG (Figure 2A), anti-HA IgA, IgD, and IgE antibodies could be detected in the serum (Figure 5A). Anti-HA IgM antibody could also be evaluated but with a high level of background (Data not shown). To compare the expression levels of all classes of anti-HA antibodies, we performed competitive ELISA using the serum samples as shown in Figure 5B. The expression levels of anti-HA IgA, IgM, IgD, and IgE in the serum were equal at around 10⁴ ng/ml compared with that of anti-HA IgG [approximately 5,700 (= 10^3.76) ng/ml] (Figure 5B). These data correlated with the results of the in vitro experiment (Figure 4C). We also evaluated the neutralizing antibody titers in the serum (Figure S2 in Supplementary Material). Experiments conducted using IgD and IgE were not effective in the in vitro neutralizing experiment (Figure 4E). Anti-HA IgG, IgA, and IgM in serum also indicated neutralizing potential.

In these experiments, we co-transfected the plasmid encoding the joining chain along with anti-HA IgA or anti-HA IgM. We then evaluated the importance of the joining chain for the expression of these antibodies. We transfected these plasmids with or without the joining chain expression vector into HEK293T cells and analyzed the antibody levels in the supernatants by western blotting. Each specific band was detected only under with the expression of joining chain (Figure S3A in Supplementary Material). This result suggests that the joining chain is important for anti-HA IgA and IgM expression. We also confirmed these results by performing HD with or without the joining chain in mice. As shown in Figure S3B in Supplementary Material, we could first detect serum antibodies both with and without the joining chain at 4 h after injection. After 24 h, the antibody concentration of IgG and IgA with the joining chain reached approximately 10⁴ ng/ml, but the level of IgA without joining chain could only be detected slightly, and was undetectable after 36 h (Figure S3B in Supplementary Material). These results suggest that the joining chain is important for the stable expression of anti-HA IgA.

Finally, we analyzed whether HD inhibits virus proliferation in the nasal cavity. We conducted HD in mice at 8 h post-infection and collected nasal wash specimens from these mice at 3 days post-infection. We then measured the virus titter in these specimens. As shown in Figure 6A, anti-HA IgG reduced the virus titer to 1/20 (from 10^5.27 to 10^4.06) compared with the vector control group. Interestingly, the virus titer was completely reduced with anti-HA IgA (Figure 6A). On the contrary, we could not recognize an inhibitory effect on the virus titer by the HD for anti-HA IgM, IgD, and IgE. To confirm protection in the nasal cavity, we analyzed the antibody concentration in nasal wash specimens. Anti-HA IgG (Figure 6B) and secreted HA-specific IgA (anti-HA sIgA) (Figure 6C) were detected in the nasal wash of several HD administered mice. We also confirmed that anti-HA IgA was detected in the serum only with the joining chain (Figure S4A in Supplementary Material) and induced almost complete reduction of virus titer in the nasal wash (Figure S4B in Supplementary Material). These data show that HD for anti-HA IgG and IgA could induce neutralizing antibodies in the nasal cavity and that these antibodies protected from IAV infection in the upper respiratory tract.