*3.5. Systemic Cell-Mediated Immunity*

T cell-mediated immune responses to influenza A/H1N1/SI, NP-peptides, and A/H3N2/Wuhan virus showed significantly higher IFNγ spot reactivity in animals that were immunized with WIV with N3 and pFliC(-gly) adjuvant. Already, after one immunization (Figure 7A, day 21) IFNγ secreting spleen cells responding to A/H1N1/SI were significantly increased in NMRI and C57BL/6 mice that were immunized with WIV and N3 or WIV with N3 and pFliC(-gly) adjuvants.

After one booster immunization, IFNγ responses to A/H1N1/SI in both strains of mice were significantly increased in the WIV immunized groups if adjuvant was used (Figure 7B, day 90) as compared to non-adjuvanted immunizations.

Peptides from the conserved nucleoprotein (NP) from influenza A were selected to characterize the responses to a defined CTL-epitope. A H2d NP (aa 147-156) binding peptide was chosen to restimulate NMRI mice, and NP (aa 55-64) was chosen to restimulate H2b C57BL/6 mice. Mice in the groups given WIV with N3 responded by developing significantly higher numbers of IFNγ secreting cells than animals receiving WIV alone (Figure 7C). Adjuvant effects were further enhanced in WIV with N3 groups by the addition of pFliC(-gly).

A significant reduction in IFNγ ELIspot reactivity was only observed animals in groups where pFliC(-gly) was used as adjuvant when CD8+T cell depletions were performed (Figure 7D). These results reveal the proportion of responses that are derived from CD8+ cells, but also indicate that all of the influenza responding mice develop IFNγ secreting CD4+ T cells against influenza. In all experiments, mice that were immunized with WIV alone developed no or very few IFNγ ELIspot secreting cells

*Vaccines* **2019**, *7*, 64

(7–55 spots/spleen million cells), which was not significantly higher than influenza naïve control mice (Figure 7A–D).

**Figure 7.** Splenocyte anti-Interferon-γ (IFNγ) producing cell frequency (ELIspot) in response to Influenza A antigen recall. (**A**) Influenza A/H1N1/SI specific IFNγ producing spleen cells in NMRI mice (left side) and C57BL/6 mice (right side), three weeks after primary immunization. (**B**) Influenza A/H1N1/SI specific IFNγ producing spleen cells in NMRI mice (left side) and C57BL/6 mice (right side), four weeks after booster immunization. (**C**) Influenza A/NP-peptide specific IFNγ producing spleen cells in NMRI mice (left side) (against peptide; NP147-156/aa TYQRTRALV) and C57BL/6 mice (right side) (against peptide, NP 366-374 aa ASNENMDAM), four weeks after booster immunization. (**D**) Influenza A/H1N1/SI specific IFNγ producing CD8-depleted spleen cells in C57BL/6 mice, four weeks after booster immunization. Bar-height shows geometric mean (GMT) spots/million cells, and error-bars show 95% confidence intervals for each study group. Significant differences are indicated by nonparametric Mann-Whitney U analysis p-values.

Heterosubtypic cell-mediated IFN-γ secreting immunity was tested while using A/H3N2/Wuhan influenza as a recall antigen (Figure 8).

**Figure 8.** Splenocyte IFNγ producing cell frequency (ELIspot) in response to influenza A/H3N2/Wuhan stimulation. Influenza A/H3N2/Wuhan specific IFNγ spleen cells in NMRI mice (left side) and C57BL/6 mice (right side), 4 weeks after booster immunization. Bar-height shows geometric mean (GMT) spots/million cells, and error-bars show 95% confidence intervals for each study group. Significant differences are indicated by nonparametric Mann–Whitney U analysis *p*-values.

In general, the total numbers of IFNγ secreting cells were two-fold lower than those that were elicited by the homologous influenza strain (H1N1). Among the NMRI mice, all groups of immunized with WIV and N3 adjuvant developed significantly higher numbers of IFNγ secreting cells than the mice receiving WIV alone. A similar trend was observed among the C57BL/6 mice. The addition of the pFliC(-gly) adjuvant lead to even greater numbers of IFNγ secreting cells responding to H3N2.

Together, these results demonstrate that the addition of N3 adjuvant significantly elicits splenic T cell responses to homotypic whole influenza A after just one vaccination and after a single boost enhances the responses even further. The addition of pFliC(-gly) to a WIV and N3 immunization was, in nearly all cases, able to greatly enhance the immune responses when compared to WIV and N3 alone. Analysis of T cell responses after WIV and N3 boosting also revealed an ability to respond to conserved Class I T cell epitopes, as well as heterotypic influenza A strains. The addition of pFliC(-gly) to WIV and N3 immunizations were also able to greatly enhance these responses, but the analysis of specific T cell populations additionally revealed that a significant portion of immune reactivity came from both CD8 as well as CD4-expressing cells.

Summarizing the cell-mediated immune responses that were obtained at day 90, prior to influenza virus challenge, suggest that the used adjuvants were all capable of supporting both influenza-antigen stimulated IFN-γ secreting and IL-5 secreting immunity in vitro. The highest levels of IFN-γ secreting responses were detectable in the animals (of both strains of mice) given WIV and FliC-DNA/N3 adjuvant, of which around 50% seem to be from CD8+ T cells against the tested NP-epitope in the spleen of C57BL/6 animals (Figure 7D) and with cross-reactivity towards influenza A/H3N2 virus antigen (Figure 8). Thus, the most pronounced and broad influenza-specific immune responses were obtained through the combination of WIV A/H1N1/SI with FliC (-gly) and N3 adjuvant administered twice nasally.

The potential correlates of protective immunity on long-term periods were tested after in vivo challenge with a heterologous influenza A strain (California H1N1/2009pdm strain) and systemically analyzed in spleen cells, and in mucosal samples from lung wash samples.

Nasal challenge with influenza A virus resulted in significantly better survival in mice immunized with Influenza A vaccine and adjuvants N3 and N3 + FliC(-gly) DNA. This was seen in both strains of mice (Figure 9A–D).

**Figure 9.** Kaplan-Meier graphs from day 180 post challenge show significantly better survival (100% to 87.5%) in NMRI mice (**A**) and in C57Bl/6 mice (**B**) if the WIV influenza vaccine was combined with N3 and FliC-DNA. as adjuvant and 80–75% survival when WIV with N3 adjuvant was used. (**C**,**D**). Kaplan-Meier graphs from day 270 post-challenge show significantly better survival (100% to 50%) in NMRI mice (**C**) and in C57Bl/6 mice (**D**) if the WIV influenza vaccine was combined with N3 and FliC-DNA. as adjuvant and 20–37.5% survival when WIV with N3 adjuvant was used.

Interestingly, the group survival data in both in- and outbred animals fit well with the pre-immunization immune responses that were measured, where especially elevated IFN-gamma levels after influenza-peptide stimulation in vitro and subclass IgG pattern with higher influenza-antigen ELISA binding IgG2 levels, seem to be associated with increased survival. In this study, the number of survivors post challenge was prolonged with at least three months in comparison with the inbred C57BL/6 mice. N3 with FliC-DNA plasmids provided the most elevated levels of both Th1 and Th2 type immune signaling, since almost all the studied immune parameters in the study (humoral responses: influenza virus specific HAI (Table 3). lung-IgG and IgA ELISA binding antibodies, subclass IgG pattern in serum and lung wash and IFNgamma ELIspot and cytokine release pattern in vitro) show that the adjuvant combination. The data may indicate that all of these immune parameters may need to be activated in elderly animals, since, in groups immunized with single adjuvant, obtaining good Th2-type humoral immune responses, at a higher age were not as efficiently protected when challenged with pathogenic influenza virus nasally.



There is a variable test sample timepoint difference between the animals that are presented in Figures 9–11. Animals from groups that more rapidly became ill after nasal influenza challenge where spleen cells were collected at day 9 to 15, at the day when they had to be sacrificed due to pathogenicity. From the groups where animals better resisted the influenza virus challenge (among NMRI mice groups 2 and 4 and C57BL/6 mice groups 7 and 9), the spleens were collected at day 29–30.

The IFN-gamma ELIspot analyses in splenocytes that were collected from mice challenged with heterologous A/H1N1/pdm09 virus without a FluA vaccination with potent Th1-type enhancing adjuvant illustrate that, if mice are allowed to reach old age, their cell-mediated immunity responds to slowly to protect from disease. Even though they have previously responded with a substantial influenza A neutralizing (HAI) response and at least a detectable influenza-antigen binding serum and mucosal IgG and IgA ELISA response, with time due, to aging on poor stimulation the levels may drop to low levels. This seems to be the case both for in- and outbred mouse strains.

Avidity Index (AI) against the recombinant HA of A/H1N1/Ca09pdm was significantly higher in the serum samples that were collected from mice immunized with N3 + FliC-DNA adjuvant (Groups 4 and 9, *p* < 0.01). Prior to challenge, the median AI in group 4 was 0.97 (0.86–1.11) and in group 9 median AI 0.86 (0.78–0.91) in comparison with the other influenza vaccinated groups with median AI 0.34 (0.09–0.46).

An attempt to perform mucosal influenza A/H1N1neutralization assays was performed against the challenge virus. However, the amount of IgA was quite low in each individual washing solution, so to perform the assay pooling and the concentration of samples was needed. Thus, we obtained one single pool from each study group for a single assay effort. The obtained results showed a HAI titer of 40, and only in the lung wash pool from the group of NMRI mice that were immunized with the N3 and FliC-DNA adjuvant.

It seems clear that the analyzed influenza-antigen stimulated cell-mediated immunity, both before challenge (Figures 7 and 8) and after influenza challenge (Figures 10–12), the later in old animals, the vaccine regimen containing adjuvants that enhance both humoral, virus-neutralizing, and binding antibodies in serum and respiratory mucosa, together with interferon-gamma secreted cell-mediated immunity seem to result in long-lasting protective immunity in both strains of mice, but perhaps more in out-bred than in inbred animals.

**Figure 10.** Cell-mediated immunity against influenza antigens after A/H1N1/California.pdm09 challenge was shown as IFN-gamma responses in stimulated spleen cells in vitro in three study groups of vaccinated NMRI mice (Figure 10A–C). (**A**) illustrates the GMT (95% C.I) IFN-gamma ELIspot responses in WIV vaccinated mice (no adjuvant) at day the day of sacrifice day 9–18, (**B**) the WIV with N3 as adjuvant, at the day of sacrifice, at days 12–30, and (**C**) the WIV with N3 and FliC-DNA as adjuvant, at day of sacrifice, at day 29–30. The frequencies of spots/million were evaluated against WIV/Influenza A/H1 and A/H3, as well as against two CTL-peptides from the NP-protein. Significant differences are indicated by nonparametric Mann–Whitney U analysis *p*-values.

**Figure 11.** Cell-mediated immunity against influenza antigens after challenge was shown as IFN-gamma responses in stimulated spleen cells in vitro in three study groups of vaccinated C57BL6 mice (**A**–**C**). (**A**) illustrates the GMT (95% C.I) IFN-gamma ELIspot responses in WIV vaccinated mice (no adjuvant) at day of sacrifice at days 6, 9, and 12, (**B**) the WIV with N3 as adjuvant, at day of sacrifice at days 9, 18, and 30 and (**C**) the WIV with N3 and FliC-DNA as adjuvant, at day of sacrifice at days 12, 21, and 30. Frequencies of spots/million were evaluated against WIV/Influenza A/H1 and A/H3, as well as against one CTL-peptide from the NP-protein. Significant differences are indicated by nonparametric Mann–Whitney U analysis *p*-values.

**Figure 12.** Mucosal lung IgA in lung wash fluids from NMRI and C57BL6 mice, post challenge. IgA anti-recombinant HA/H1/CA09 antigen (**A**,**B**) and anti-recombinant HA/H3/CA (**C**,**D**) ELISA titer analysis. Significant differences are indicated by nonparametric Mann–Whitney U analysis *p*-values.

#### **4. Discussion**

A single-dose vaccination would be highly useful, especially in emergency situations with rapid spread of influenza, as most inactivated influenza vaccines require two immunizations to provide full and protective immunity [30,31]. Previous mouse studies with formalin-inactivated WIV have shown that the rapid development of B cell responses in serum and upper airway mucosa is sufficient to protect from mortality. We assessed the use of two potential mucosal adjuvants in combination with WIV intranasal vaccination to improve upon these results and observed the induction of mucosal B cell responses as well as heterosubtypic systemic cellular responses that were detectable after only one dose. Immune reactivity is further enhanced after a nasal booster immunization. Thus, to investigate the development of long-term immunity, vaccinated animals were studied over nine months post-booster immunization before a heterologous influenza A/H1N1 challenge experiment was performed, in mice reaching 21–23 months of age.

Nasal administration of inactivated influenza A antigen with N3 and DNA-encoded flagellin from *Salmonella typhimurium* or from flagellin isolated from human flora *Escherichia coli* significantly increased the influenza A specific immune responses. The observed immunogenicity of this low HA dose, influenza vaccine (25 μg total protein/mouse containing 1.5 μg HA) in each mouse strain induced similar levels of serum antibody, which suggested that the basic antibody responses were comparable. With addition of either N3 or N3 and pFliC(-gly) adjuvants responses were significantly enhanced in both NMRI and C57BL/6 mice after a single immunization. However, the ability of N3 and pFliC(-gly) to promote enhanced serum responses over that of N3 alone was not observed.

After the booster immunization, significantly higher serum IgG or IgA titer toward influenza antigens was only seen in the mice receiving N3 adjuvant alone or N3 and pFliC(-gly). An analysis of serum IgG isotypes revealed that the inclusion of pFliC(-gly) in the adjuvant composition enhanced IgG2a responses when compared to the use of N3 alone, demonstrating its ability to promote a Th1-type or a mixed Th1/Th2-type immune response. Apart from being an indicator of the degree of cell-mediated cytotoxic immune responses, the IgG2 antibodies have also been described to be more beneficial antiviral isotype antibodies. They may enhance antigen uptake and cross-presentation due to their higher capacity to bind to Fc-receptors on antigen-presenting cells, as well as demonstrating better C'-activating properties via the C1q-pathway [32]. The capacity of non-coding plasmid DNA formulated with vaccine antigens to improve or strengthen Th1- or mixed Th1/Th2-type antibody and cell mediated immune responses has been documented [33,34]. However, when compared to our previous study that demonstrated a specific adjuvant effect by the FliC open-reading frame [25], our results suggest that FliC is exerting similar immune enhancing responses observed here.

It is not a straightforward task to understand how such a vaccine formulation functions due to the complex nature of our antigen adjuvant formulation(s). However, we favor a model where the nasal delivery of a cationic adjuvant N3 affects the mucus at mucosal surface layers to better penetrate and deliver the influenza vaccine, together with pFliC(-gly). This would allow for larger amounts of WIV antigen to be taken up by antigen-presenting cells, leading to antigen processing as well as the potential for viral ssRNA to stimulate TLR7 [35]. This enhancement may account for the adjuvant effect of N3 alone. When combined with pFliC(-gly), the N3 may allow for pFliC(-gly) "transfection" of cells in the mucosa triggering multiple innate immune receptors by way of activating dsDNA sensors, as well as sequence-specific CpG DNA sensors [35]. Additionally, as a consequence of FliC polypeptide production, NLRC4-triggered inflammasome activation and inflammatory cell death could occur in the permissive cells [36]. The secretion of flagellin from living or release from dying cells may also activate TLR5 [36,37]. Irregardless of the mechanism, we clearly observe that adjuvant the inclusion of pFliC(-gly) stimulates a mixed Th1/Th2-immune response pattern in both inbred and outbred mice, indicating its benefit in broadening immune responses to WIV.

The use of adjuvants in influenza vaccinations may be especially valuable in elderly individuals that otherwise fail to respond or lack strong influenza virus-specific cell-mediated immunity after traditional adjuvant free vaccination. However, one need to bear in mind is that influenza vaccines with adjuvants have also previously resulted in significant side effects, such as the Berna vaccine incidents where bacterial toxoid subunits, with the ganglioside GM1-binding properties were used as adjuvants for nasal immunization with the risk of resulting in Bells palsy [38,39]. The most recent serious side-effects with the intramuscularly administered influenza A/H1N1/pdm09 vaccine being given with the ASO3-adjuvant, resulting in hundreds of children responding with narcolepsy as a side-effect [40]. Obviously, every new adjuvant combined with influenza-vaccine will need to be analyzed for these undesirable side effects. One should bear in mind that this study has the main goal of providing long-lasting immunity that persists in protective immune responses at higher ages, instead of in children. The influenza strain that was used in the primary immunization belongs to the seasonal influenza strains, but the data obtained show that these influenza strains can provide protection against pandemic influenza A strains, such as A/H1N1/pdm09. Thus, the suggestion in this work is to immunize adults, to develop a robust enough immunity with the capacity to remain protective at higher ages, as exemplified in this report.

The humoral mucosal immunity that was observed in the respiratory tract of immunized mice was clearly enhanced by the presence of cationic N3 adjuvant in the inactivated influenza vaccine mixture. However, the highest lung wash IgA and IgG titers against influenza was seen in the mice of both mouse strains after immunization with N3 and pFliC(-gly). This combination of adjuvant N3 with pFliC(-gly) significantly enhanced the systemic and the mucosal immune responses (41). However, slightly different immune reaction patterns were observed between the two strains of mice when the samples were analyzed by HAI. The putatively protective and neutralizing serum HAI titer among the outbred NMRI mice initially (Day 21) shows equivalent titers for the group vaccinated with WIV alone and groups vaccinated with WIV and adjuvant (Figure 4A). After the booster immunization, the HAI-titers increased to significantly higher titers in N3, and N3 and pFliC(-gly) combination immunized groups (day 90). C57BL/6 mice more robustly responded to WIV with adjuvant after one immunization, which was likely because of strain-specific differences. Although these results indicate that immunization in outbred populations may require two doses, importantly they indicate that these adjuvants have the potential to function effectively in outbred veterinary animal populations as well as humans to promote WIV mucosal immune responses. The inclusion of these adjuvants in an intra-nasal non-living vaccine could provide enhanced protection of the upper mucosa. The most common route of influenza A transmission [41].

In vitro analysis of IL-5 secretion by splenocytes that were stimulated with WIV H1N1 from WIV and N3 immunized mice showed that cells from NMRI mice secreted significantly higher amounts of IL-5 than C57BL/6 mice. This interleukin is normally considered to be a Th2-type, which suggests that the NMRI mouse strain may be more Th2-skewed than C57BL/6. Direct comparison of our trends of IL-5 secretion to the serum IgG and IgA titers and IgG isotypes also established that higher IL-5 levels correlate with Th2-type immunity. Importantly, our observed IL-5 levels peaked at different time points in the two mouse strains. The NMRI spleen cell secretion of IL-5 peaked at 44–48 h while cells from C57BL/6 mice peaked at 70–72 h after stimulation. These differences could be due to natural differences in the immune kinetics in each strain, differences in the total responding cell populations, or fewer spleen cells being capable of IL-5 secretion in the C57BL/6 mice. The IL-5 response following mucosal vaccination illustrates its correlation to immunological aging that could not be seen in younger mice in previous reports looking at the effects of aging and cytokine responses [42] (McDonald JU et al. 2017). Despite these strain-specific differences in their ability to produce IL-5 in response to immunization with WIV and N3, this trend could be altered by the addition of pFliC(-gly) adjuvant, demonstrating its ability to skew the immune responses towards a broader and possibly more robust anti-influenza A immunity.

Preclinical influenza vaccine studies use the influenza virus HAI as a gold standard test to identify the immunization regimes that reveal most favourable vaccine candidates. Although this is not a true influenza A neutralization assay, it correlates well with clinical antibody mediated protection from infection and mortality in man and animals. Ideally, the best measure would be to perform a

challenge study with the live influenza virus on influenza vaccinated ferret or man to investigate if protection can be obtained. In our study, we did not have access to a mouse-attenuated pathogenic influenza A/H1N1/SI strain, which is why homologous challenge studies were not performed. Instead, a mouse pathogenic heterologous influenza A/H1N1/pdm09 was performed at days 180 and 270 post final immunization. In lieu of this, HAI analyses with the Salomon Island strain was performed with serum from the various influenza vaccinated groups of mice. Sera with the highest HAI titers against the homologous H1N1 influenza strain were also subjected to an HAI assay against A/H3N2/Wuhan, but they were negative. Putatively protective (≥40) average serum HAI titers to the influenza vaccine strain was already seen after one immunization in the groups that received WIV with N3 or N3 and pFliC(-gly), suggesting a more rapid B cell response toward the HA antigen. These HAI titers were further elevated after the booster immunization and after boost also present in the WIV vaccinated mice given N3 together with pFliC(-gly). Interestingly, the serum and lung wash antibody ELISA titers are shown to increase, even when the HAI titers do not change or increase in a similar manner. Thus, it is important to remember that the binding antibody titers, as measured with ELISA, do not provide any information regarding how their functional activities against influenza A virus or influenza-infected cells. Instead, they may function as surrogate biomarkers of immunogenicity or (at least in mice) suggest Th1/Th2-type immune patterns against the studied antigens.

The choice of vaccination route is important to induce an effective immune response at a specific site. The inactivated vaccines that are currently used are usually intramuscularly injected, but it would be beneficial to administer the vaccines without needles. It would entail easier and faster administration, less expense, and discomfort with needles would be avoided. However, the alternative administration methods have limitations and they do not always lead to an adequate antibody response [31]. The injectable influenza vaccines are protective, because of their ability to induce influenza neutralizing serum antibodies as well as ADCC active antibodies, mainly IgG [33]. They can prevent the pathogen from spreading and be protective at the mucosal surfaces of the lower respiratory airways. However, resistance to influenza infection is connected to systemic and mucosal immunity. Serum IgA antibodies are produced to HA and NA in the upper respiratory tract, and IgG antibodies are also protective in the lower respiratory tract. In comparison to an injectable vaccine, the amount of local IgA antibodies in nasal washings is higher after nasal vaccination, but the amount of serum antibody titers is lower [36,37,41].

Our observations indicate that intra-nasal immunization with a formulation of WIV and N3 lipid alone is able to greatly enhance HAI-specific IgG and IgA in the upper respiratory tract, demonstrating that this approach can elicit effective antibodies at the location where influenza A infects. This effect can be enhanced with the addition of pFliC(-gly). Importantly, pFliC(-gly) strongly promotes a skewing of IgG isotypes towards IgG2a/c, as well as an enhancement in homotypic and heterotypic cellular immune responses. A subclass IgG1/IgG2 pattern mirroring what was seen in serum was seen also in the lung wash IgG content. However, it is unclear how well this indicates the local cell-mediated immunity in the lung tissues. Cellular immunity to influenza A has been long speculated to provide a role in limiting the pathology of influenza A infection, and recent evidence indicates that this is indeed the case [16]. Although it remains to be seen whether the promising anti-viral immune responses that we observe in mice can also be induced in humans, the approach that is delineated here indicates that it is possible to develop a non-living influenza A intranasal vaccine that is capable of eliciting both mucosal humoral antibody, as well as strong systemic cellular responses.

We kept groups of mice during a six and nine months follow up time to address the issue of long-term immunity as well as the protective capacity from disease and illness, and then nasally challenged them with a heterologous influenza A/H1N1 virus. At nine months, especially among the out bred NMRI mice, a complete protection (100%) was obtained, while the inbred C57Bl6 mice obtained 75% protective immunity, long-term, if both cationic N3 and FliC-DNA was used as adjuvant. The immune parameters that were seen in the best protected animals were cell-mediated IFN-gamma responses against a broader repertoire of influenza antigens, than in the other groups. Furthermore,

*Vaccines* **2019**, *7*, 64

mucosal IgA and serum IgG with a broad influenza A hemagglutinin antigen binding capacity was only seen in these groups and significantly more among the NMRI mice than in C57Bl6 mice, post-challenge. The fact that we used inactivated full virion antigen may explain this phenomena, as, in other studies where the FluMist vaccine was tested together with ritatolimoid (a TLR3-agonist), a broadened secretory IgA response was obtained [43]. Otherwise, no inactivated whole virion influenza vaccines are available on the vaccine markets. Thus, it is possible that the combined Th2/Th1 balanced immunity that was developed in the WIV N3+FliC groups provided a lasting protective immune mixture containing not only a cell-mediated IFN-gamma immunity against conserved NP-epitopes and mucosal broadly HA-binding IgA, but also serum IgG with influenza-specific ADCC-activity, often requiring antibodies with high binding affinity. ADCC have been described as an important immune parameter in protecting mice from severe influenza infection under experimental conditions [44]. In conclusion, the results suggest that a seasonal influenza A/H1N1/Salomon Island/2006, whole inactivated virion combined with cationic oil-in-water adjuvant with DNA-expressed Flagellin C *S. Typhimurium* given twice nasally provide protective immunity in 22 months old in- and out-bred mice against heterologous influenza A/H1N1/California/2009 challenge over nine months. However, with the recent experiences with side-effects when new adjuvants were combined with parenteral influenza vaccination, a thorough safety monitoring will be essential before aiming towards large-scale vaccination campaigns.

**Author Contributions:** For research articles with several authors, a short paragraph specifying their individual contributions must be provided. The following statements should be used conceptualization, J.H. and S.E.A.; methodology, J.H., S.E.A., S.N., A.B., C.D.; software, J.H., S.E.A., S.N., A.B.; validation, J.H., S.E.A. and A.B.; formal analysis, J.H., S.E.A., S.N., C.D.; investigation, S.E.A., J.H.; resources, S.E.A., J.H., C.D.; writing—original draft preparation, S.E.A., J.H.; writing—review and editing, S.E.A., J.H.; visualization, J.H., C.D.; supervision, J.H., S.E.A. CD; project administration, J.H., S.E.A.; funding acquisition, J.H.

**Funding:** This research was funded by Vinnova, grant number D09-01 and Hälsofonden 09-4.

**Acknowledgments:** Ulf Schröder and Jorma Hinkula share the IP rights for the N3 adjuvant. We acknowledge Eurocine Vaccines for the supply of N3 adjuvant. We thank Mona Hedenskog at Swedish Institute for Infectious Disease Control, Solna Sweden, and Erik Hedman, Kristina Kinnunen and Sofie Berger at Linköping University for assistance with HAI, ELISA and ELIspot analyses. This work was financially supported by Hälsofonden and VINNOVA.

**Conflicts of Interest:** Eurocine Vaccines AB have IP-rights and economical interests in the N3 adjuvant. J.H. declares financial support and IP share. S.E.A. declares interests of patents pending for DNA-encoded flagellin adjuvant technology. The remaining authors declare no conflict of interest.
