Received date: September 23, 2015; Accepted date: November 18, 2015; Published date: November 21, 2015
Citation: Roos A, Roozendaal R, Riahi SD, Vreugdenhil J, Vaneman J, et al. (2015) H5N1 Protection by Seasonal Influenza Vaccine in Homologous and Heterologous Prime/Boost Vaccination. J Vaccines Vaccin 6:299. doi: 10.4172/2157-7560.1000299
Copyright: © 2015 Roos A, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Visit for more related articles at Journal of Vaccines & Vaccination
A number of different approaches aimed at broadening the cross-protective ability of seasonal influenza vaccines are being explored today. Priming a seasonal vaccine with three administrations of DNA encoding H1 HA corresponding to the HA in the seasonal vaccine has been shown to confer protection against heterologous H1N1 influenza. Here we evaluated the heterosubtypic protection induced by a seasonal influenza vaccine when primed with H1 HA DNA and, in parallel, when given as a homologous prime/boost regimen.
Balb/c mice were immunized three times with vaccine homologous H1 HA DNA prior to a boost with seasonal influenza vaccine (season 2009/2010; Northern Hemisphere), or immunized three times with the seasonal influenza vaccine. To assess cross-protection, mice were subsequently challenged with either heterologous H1N1 or heterosubtypic H5N1 influenza virus.
The level of heterologous H1N1 protection elicited by the seasonal influenza vaccine was enhanced by priming with H1 HA DNA. In contrast, priming with H1 HA DNA did not enhance the level of heterosubtypic H5N1 protection. The heterologous prime boost regimen showed to be less efficient than multiple immunizations with seasonal vaccine in conferring protection against H5N1. Neither the DNA-priming vaccination regimen, nor the homologous prime/boost regimen induced detectable H5N1 cross-reactive anti-HA or anti-NA antibodies. Homologous prime /boost vaccination did induce higher levels of anti-NP antibodies.
Here we demonstrate that priming a seasonal influenza vaccine with vaccine homologous H1 HA encoding DNA enhances the level of heterologous H1N1 but not heterosubtypic protection induced by the vaccine alone. Homologous prime/boost vaccination resulted in higher levels of heterosubtypic protection. Of the immunogenicity parameters tested for both heterologous and homologous prime/boost regimens only anti-NP responses follow the same pattern as heterosubtypic protection.
Annual influenza epidemics are responsible for up to 5 million cases of severe illness and 250 000 to 500 000 deaths as reported by the world health organization (WHO) . Seasonal influenza vaccines are the most effective way to reduce the impact of influenza epidemics. To keep pace with the genetically drifted circulating strains the composition of seasonal influenza vaccines must be updated almost annually . The development of an influenza vaccine which confers protection against a wide range of influenza viruses is therefore of great importance.
The efforts to generate broadly protective influenza vaccines have largely focused on designing vaccine compositions and regimens able to induce immune responses, either humoral, cellular or a combination thereof, against proteins that are conserved across divergent strains of influenza [3,4]. Novel influenza vaccines that aim to elicit cross protective T-cell responses are often based on highly conservedviral proteins such as nucleoprotein (NP) [5-7] and matrix protein 1 (M1) , while vaccines that aim to induce broad humoral immunity have largely focused on surface exposed viral proteins such as hemagglutinin (HA) [9-13], neuraminidase (NA) [14,15] or matrix protein 2 (M2) [16-18]. Though HA and NA are highly variable proteins, they contain conserved epitopes that can elicit broadly protective antibodies [19-24], and as such are very interesting targets in the quest for universal influenza vaccines. B cells producing HAspecific broadly neutralizing antibodies (bnAbs) have been isolated from human volunteers after vaccination with seasonal vaccines [21,25,26], indicating that HA in seasonal vaccines can form the basis for developing a broadly protective immune response. While HA is the key immunogen in all current seasonal vaccines, and is used for vaccine dosing and evaluation of the seasonal vaccine potency, the majority of current vaccines also contain other viral proteins, such as NA, M2 and NP [27,28], albeit amounts thereof are neither quantified nor standardized. The immune response to these proteins may potentially contribute to cross-protective ability of seasonal influenza vaccines.
Different approaches aimed at broadening cross-protective ability of seasonal influenza vaccines are being explored [29-31]. We reported previously  that multiple immunizations with seasonal influenza vaccine of season 2011-2012 can enhance the heterosubtypic (H5N1) protection in mice. In other approaches, seasonal influenza vaccine has been administered together with adjuvants [31,33,34] or primed with DNA vectors . Wei et al.  demonstrated that heterologous H1N1 protection can be induced in both mice and ferrets by priming a seasonal vaccine with three administrations of DNA encoding H1 HA corresponding to the H1 present in the seasonal influenza vaccine (H1- DNA). The safety and efficacy of this vaccination regimen, expanded to include DNAs encoding H3 and B HA, is being further explored in healthy human volunteers .
In the current study, we expanded the evaluation of a heterologous prime/boost vaccination regimen (H1-DNA, followed by seasonal influenza vaccine) to induce heterosubtypic protection against H5N1. In parallel, we assessed the breadth of protection and immunogenicity induced by a seasonal influenza vaccine of season 2009/2010 when given in a homologous (e.g., 3x seasonal vaccine)  or heterologous prime/boost vaccination regimen.
Statement of ethics
All mouse experiments were performed in accordance with Dutch legislation on animal experiments and approved by DEC Animal Sciences Group, Wageningen UR when performed at CVI Lelystad and an independent Animal Ethics Committee (TNO, Zeist, The Netherlands) (permit number 3387) when performed at TNO Triskelion. In all experiments six-to-eight-week-old female Balb/c (H2d) mice (Charles River, Sulzfeld, Germany) were used. Mice were kept under specific pathogen-free conditions.
Groups of mice (n=10 or 8 for challenge studies, n=8 for immunogenicity study) received either: 1x or 3× intramuscular (i.m.) immunizations with Inflexal® V (Crucell, Bern, Switzerland), a trivalent seasonal vaccine (SV) (composition for the 2009-2010 season: H1N1 A/Brisbane/59/07, H3N2 A/Brisbane/10/07 and B/Brisbane/60/08) (3 μg HA per strain per immunization); 3x i.m. immunizations with DNA encoding RSV_F_A2 protein (15 μg per immunization) followed by an i.m. immunization with SV; 3× i.m. immunizations with DNA encoding HA of H1 A/Brisbane/59/07 (15 μg per immunization); 3× i.m., immunizations with DNA encoding HA of H1 A/Brisbane/59/07 (15 μg per immunization) followed by one i.m., immunization with SV; or 4× with PBS (Gibco®, Life TechnologiesTM, Paisley, UK). Immunizations were scheduled at 3 week intervals. All final immunizations with vaccine were performed at the same time. Mice immunized only once with SV, received three immunizations with PBS prior to vaccine. Mice immunized either 3× with SV or 3x with DNA encoding H1 HA A/Brisbane/59/07 received one immunization with PBS prior to vaccine.
In immunogenicity experiments blood and spleens were harvested four weeks after final immunization. Blood was collected via heart puncture under isoflurane anesthesia (IsoFlo®, Abbott Park, IL, USA) followed by cervical dislocation and collection of the spleen. Serum was collected after centrifugation for 4 minutes at 1699× g followed by 1 minute at 20817× g. The serum was isolated and stored at -20≡C.
For challenge experiments mice were infected with influenza virus 4 weeks after the final immunization. On the day of challenge a prechallenge blood sample (to assess pre-challenge antibody titers) was obtained via submandibular bleeding. Mice were challenged intranasally (i.n.) with 25 × LD50 of influenza virus (total 50 μl, 25 μl per nostril) under anesthesia with ketamine/xylazine (100 mg/kg ketamine (Nimatek® 100 mg/ml, Eurovet, Cuijk, the Netherlands); 20 mg/kg xylazine (Sedamun® 20 mg/ml, Eurovet, Cuijk, Netherlands)). Virus stocks of H1N1 A/Puerto Rico/8/34 (ATCC-VR-95, American Type Culture Collection, Manassas, VA, USA), H1N1 A/WSN/33 (Central Veterinary Institute, Wageningen University, the Netherlands) and wild-type H5N1 A/Hong Kong/156/97 (Central Veterinary Institute, Wageningen University, the Netherlands) were grown on embryonated chicken eggs. Groups of mice receiving 4× PBS i.m., were used as negative control and groups receiving broadly protective monoclonal antibody (CR6261, 15 mg/kg in PBS intravenously (i.v.) 24 hours prior to challenge) were used as positive control for the challenge. After challenge mice were monitored daily for weight-loss and clinical score for up to 21 days or until a humane endpoint based on clinical score or found dead. The challenge experiments were performed at two different locations at which different clinical score systems were used. The H1N1 A/WSN/33 and H5N1 A/Hong Kong/ 156/97 challenge studies were performed at CVI Lelystadt at which a 4-point clinical scoring system was used: 0=no clinical signs, 1=rough coat, 2=rough coat, less reactive, passive during handling, 3=rough coat, rolled up, labored breathing, passive during handling, 4=rough coat, rolled up, labored breathing, unresponsive. CS4 was defined as moribund based on unresponsiveness and used as a humane endpoint. The H1N1 A/Puerto Rico/8/34 challenge study was performed at TNO Triskelion at which a 5-point clinical scoring system was used: 0 = no clinical signs, 1=rough coat, 2=rough coat, labored respiration 3=rough coat, labored respiration, hunched posture and/or blepharospasm, 4=rough coat, labored respiration, hunched posture, blepharospasm, lethargic and/or thin/dehydrated, 5=lethargic behavior (CS4) is observed during four consecutive observations leading to euthanasia. CS5 was used as humane endpoint.
Influenza challenge studies
Differences between immunization regimens relative to negative control group receiving 4× PBS i.m., were statistically evaluated using survival proportion, survival time, change in bodyweight and clinical scores. Survival proportion and survival time after challenge were analyzed using Fisher’s exact test and log-rank test, respectively. Repeated measurements in the challenge phase (i.e. bodyweight and clinical scores) were summarized as a single outcome per animal using an Area Under the Curve (AUC) approach where missing values for animals that died early were imputed with a last-observation-carried-forward method. Body weight data are expressed as the change relative to the day 0 measurement. The AUC was then defined as the summation of the area above and below the baseline. An ANOVA on AUC’s was done with group as explanatory factor. Clinical scores were summarized as AUC per mouse and groups were compared using a generalized linear model with a cumulative logit distribution to compare area under the curves for ordinal variable. Statistical analysis was planned upfront and adjustments for multiple comparisons were done using a Bonferroni correction for (i) the H1N1 A/Puerto Rico/ 8/34 challenge: H1-DNA/SV=4 comparisons, SV=2 comparisons (clinical score data for this group were not adjusted), (ii) the H1N1 A/WSN/33 challenge: for all groups 4 comparisons, (iii) figure 1C H5N1 A/Hong Kong/156/97 challenge: H1-DNA/SV=4 comparisons, SV=2 comparisons (clinical score data for this group were not adjusted), (iv) figure 2 H5N1 A/Hong Kong/156/97 challenge: for all groups 2 comparisons. The studies were considered valid only when there was a statistically significant difference in survival proportion (Fisher’s exact-test, 2-sided) between negative and positive challenge model control groups (data not shown for positive controls).
Figure 1: Priming seasonal influenza vaccine with H1-DNA enhances protection against heterologous H1N1 but not against heterosubtypic H5N1 challenge. Mice ( n=10 ) were immunized either 1x with Inflexal (SV), 3x with H1 HA DNA (H1-DNA), 3x with RSV DNA followed by a boost of Inflexal (RSV-DNA/SV), 3x with H1-DNA followed by a boost of Inflexal (H1-DNA/SV) or 4x PBS (PBS). Shown are Kaplan-Meier survival curves (left) and mean bodyweight change (right) graphs following challenge. Mice were challenged with A) H1N1 A/Puerto Rico/8/34 B) H1N1 A/WSN/33 C) H5N1 A/Hong Kong /156/97, Error bars indicate 95% confidence interval (bodyweight).
Figure 2: H1-DNA/SV09 confers only partial protection while three times SV09 protects against heterosubtypic H5N1 challenge. Mice were immunized either 3x with Inflexal 2009 (3xSV09), 3x with H1 HA DNA followed by a boost of Inflexal 2009 (H1-DNA/SV09) or 4x PBS (PBS) followed by challenge with H5N1 A/Hong Kong/ 156/97 (n=8 for PBS and 3x SV09; n=7 for H1-DNA/SV09). Kaplan-Meier survival curves (left) and means bodyweight change (right). Error bars indicate 95% confidence interval (bodyweight).
Statistical analyses were performed using SAS version 9.2 (SAS Institute Inc. Cary, NC, USA) and SPSS version 20 (IBM, USA). Statistical tests were conducted two-sided at an overall significance level of α=0.05. Only p values less than 0.05 are reported in the Result section. A summary of all statistical tests and respective p values are presented in supplementary Table S1.
|Description Vaccine group||Day|
|H1-DNA||H1- DNAb||H1- DNAb||H1- DNAb||Challenge|
|H1-DNA/SV09||H1- DNAb||H1- DNAb||H1- DNAb||SV09c||Challenge|
|aRSV-DNA: pcDNA2004 (Neo-) containing RSV-F-AS cDNA insert. Dose: 15 µg i.m./immunization
bH1-DNA: pcDNA2004 (Neo-) containing H1 A/Brisbane/59/07 cDNA insert. Dose: 15 µg i.m./immunization
CSV09: Seasonal Influenza vaccine ,Inflexal® V, of season 2009-2010. Dose 3 µg HA/strain/ immunization
dFor immunogenicity experiments mice were sacrifised at day 91 and serum and spleens were collected. For influenza
challenge experiments mice were challenged at day 91
eChallenge: 25xLD50 of H1N1A/Puerto Rico/8/34, H1N1 A/WSN/33 or H5N1 A/Hong Kong/156/97
Table 1: A schematic overview of the immunization regimens for the various vaccine groups is presented. The constructs used in these experiments were based on the pcDNA2004 (Neo-) plasmid backbone containing either RSV-F-A2 or H1 A/Brisbane/59/07 cDNA inserts.
Virus neutralization assay
Madin-darby canine kidney (MDCK) cells were seeded in a 96-well plate at 15,000 cells/well in growth medium (Dulbecco's Modified Eagle Medium (DMEM) containing 200 mM L-glutamine, 3 μg/ml trypsin and 1% (w/v) penicillin/streptomycin stock solution, all Gibco, Invitrogen Ltd, Life Technologies, Paisley, UK) and allowed to attach for a minimum of 3 hours. Duplicate serial dilutions of heatinactivated (30 minutes at 56°C) serum samples (0.01-20%) were prepared in DMEM with or without trypsin/EDTA (0.6% of a 0.05% stock solution) and mixed with 120 TCID50 of H1N1 A/Brisbane/ 59/07 or 200 TCID50 of the H5N1 A/Hong Kong/156/97 (reassortant rgPR8-H5N1) virus per sample, respectively, for 1 hour at 37°C, 10% CO2. Mixes were subsequently added to the MDCK cells and incubated for 18 hours at 37°C, 10% CO2. Cells were fixed with 80% acetone, labeled with mouse anti-NP (H16-L10-4R5, produced inhouse), followed by goat anti-mouse HRP-coupled antibody (KPL, Gaithersberg, MD, USA) for one hour each. TMB substrate (Roche, Basel, Switzerland) was added, and absorbance was read in a BioTek® reader (PerkinElmer, Groningen, the Netherlands) after 5-15 minutes. Monoclonal antibody CR6261 (human IgG2a, produced in-house) and naïve mouse serum were used as positive and negative controls, respectively. Samples without detectable neutralization at the lowest dilution are indicated as the lowest dilution (i.e. background level). The IC50 values were calculated after 4-parameter logistic curve fit.
For ELISPOT analysis of T-cell responses ten amino acids overlapping 15-mer peptides covering the whole HA protein sequence of H1 A/Brisbane/59/07, a total of 104 peptides, were used (Pepscan, Lelystad, The Netherlands). For analysis of the total T-cell response against the full length H1 A/Brisbane/59/07 a pool of all 111 peptides was made (total pool). For analysis of T-cell response against a known 9-mer epitope IYSTVASSL , highly conserved among number of strains, including H1 A/Brisbane/59/07 and H5 A/Hong Kong/156/97, two 15-mer peptides containing this epitope were used. The concentration per peptide was 0.4 mg/ml, diluted in DMSO.
Ninety-six-well multiscreen plates (Millipore, Bedford, MA), coated overnight with rat anti-mouse IFNγ (Pharmingen, San Diego, CA) (1 μg per well in PBS pH 7.4), were washed with Dulbecco's PBS (Life Technologies, Gaithersberg, MD) containing 0.05% Tween-20 (D-PBS/ Tween) and blocked with D-PBS containing 5% FBS for 2 hours at 37°C. Splenocytes were prepared in R10 medium (RPMI 1640 (Gibco/ Invitrogen, Breda, The Netherlands) containing 10% heat inactivated FBS (HyClone, Logan UT), 1% Pen/Strep (Gibco/Invitrogen, Breda, The Netherlands), 1% MEM non-essential amino acids (Gibco/ Invitrogen, Breda, The Netherlands) and 13 μM 2-mercaptoethanol (Fluka Chemie, Buchs, Switzerland) and plated in duplicates at 5 μ 105 cells/well and 2 × 105 cells/well in a 100-μl reaction volume containing 2 μg/ml 15-mer peptides (total pool or two 15-mer peptides). Following 18 hour incubation at 37°C, the plates were washed with DPBS/ Tween and incubated for 1.5 hour with a biotinylated rat antimouse IFNγ (Pharmingen, San Diego, CA). Plates were washed and incubated for 1.5 hour with streptavidin–alkaline phosphatase (Southern Biotechnology Associates, Birmingham, AL). Upon final washing, specific staining was developed with nitro blue tetrazolium-5- bromo-4-chloro-3-indolyl-phosphate chromogen (Pierce, Rockford, IL), stopped by washing with tap water, air dried, and analyzed using an AELVIS ELISPOT reader (AELVIS GmbH). The cell concentration for which the average of spot forming units (SFU)/well was within the linear range, 50-225 spots/well, was selected. The average SFU/well count was adjusted to SFU’s per 106 cells in accordance with cell dilution.
Recombinant protein (0.05μg per well in PBS pH7.4) (i) HA of H1N1 A/Brisbane/59/07 (Protein Sciences Inc., CT, USA), H1N1 A/ Puerto Rico/8/34 (Protein Sciences Inc., CT, USA) or of H5N1 A/Hong Kong/156/97 (manufactured in-house on HEK293F cells), (ii) NP of H1N1 A/Puerto Rico/8/34 (Sino Biologics Inc. Beijing, China), or (iii) NA of H1N1 A/Brisbane/59/07 (manufactured in-house on HEK293F cells) or of H5N1 A/Hong Kong/156/97 (manufactured in-house on HEK293F cells) were coated onto Maxisorp 96-well plates (Nunc™, Thermo Scientific) O/N at 4°C. Plates were washed with PBS (Gibco®, Life Technologies™, Paisley, UK) containing 0.05% Tween-20 (Calbiochem®, Merck Millipore, Darmstadt, Germany) (PBS-T) and subsequently blocked with for rHA and rNP ELISA; PBS containing 2% dried skimmed milk (Difco™, BD, Breda, the Netherlands) and for rNA ELISA; PBS containing 2% BSA (Sigma-Aldrich, USA) for 1 hour at RT. Following a wash with PBS-T serum was added to the plate. The serum was in duplicate serially diluted (2-fold, 0.002-2%) and incubated for 1 hour at RT. Following a wash with PBS-T a 1:2000 dilution of Goat-anti-Mouse IgG-HRP (KPL, Gaithersburg, MD, USA) was added to the plate and incubated for 1 hour at RT. After washing with PBS-T OPD substrate (Thermo Scientific, Bremen, Germany) was added to the plate. The colorimetric reaction was stopped after 10 minutes by adding 1M H2SO4. The optical density (OD) was measured at 492 nm and standard curves were created using a four parameter logistic curve. The OD of each sample dilution was then quantified against the standard curve and the final concentration per sample (in Elisa Units, EU/ml) calculated by a weighted average, using the squared slope of the standard curve at the location of each quantification as weight. Negative samples were set at the limit of detection (LOD), defined as the lowest sample dilution multiplied by the lowest standard concentration with an OD response above the lower asymptote of the standard curve and background. All ELISA titers presented in the figures have been log10 transformed.
Priming seasonal influenza vaccine with H1-DNA enhances protection against heterologous H1N1 but not against heterosubtypic H5N1 challenge: We first tested whether a virosomal seasonal influenza vaccine of season 2009/2010 (SV09) primed three times with DNA encoding vaccine homologous H1 HA (H1-DNA) can provide heterologous H1N1 protection in mice. Albeit not significant, a single immunization with SV09 alone elicited partial survival (40%) against heterologous H1N1 A/Puerto Rico/8/34 with significantly prolonged survival time (p=0.003 compared to PBS) and with reduced clinical scores (p=0.001 compared to PBS). When primed with H1-DNA, SV09 elicited a significant increase in survival proportion compared to nonvaccinated mice (60% survival, p=0.043 compared to PBS) with prolonged survival time (p=0.044 compared to PBS), reduced bodyweight loss (p=0.050 compared to PBS) and reduced clinical scores (p<0.001 compared to PBS) (Figure 1A and S1.A).
To confirm the effect of DNA priming in another heterologous challenge model, we assessed protection by H1-DNA/SV09 upon challenge with H1N1 A/WSN/33. Two additional groups were included in this experiment to further delineate the contribution of H1-DNA to protection: one group of mice was immunized with H1-DNA only and another group three times with DNA encoding an irrelevant antigen (RSV_F_A2 protein) (RSV-DNA) followed by a boost of SV09 (RSVDNA/ SV09). H1-DNA/SV09 induced 70% survival after H1N1 A/WSN/33 challenge (not statistically significant), with prolonged survival time (p=0.038 compared to PBS) and reduced clinical scores (p=0.001 compared to PBS) (Figure 1B and S1.B), confirming protective ability of the heterologous prime/boost vaccination regimen. Partial survival seen in groups immunized with H1-DNA or RSVDNA/ SV, 60% and 50% respectively, was neither significant in survival proportion or reduction of disease symptoms. Thus, the improved level of cross-protection achieved with the heterologous prime/boost regimen requires specific antigen-expressing DNA as a prime and is not due to possible non-specific effect of DNA administration. The level of heterologous H1N1 cross-protection induced by the H1-DNA/ SV09 vaccination regimen was comparable to the cross-protection induced by three vaccinations with SV09 (3xSV09) (Figure S2), a regimen previously shown to be broadly protective .
To further evaluate whether the heterologous prime/boost vaccination regimen could elicit also heterosubtypic protection we challenged mice immunized with SV09 alone or H1-DNA/SV09 with H5N1 A/Hong Kong/156/97. Partial protection against H5N1 was elicited already by SV09 alone with a significant increase in survival proportion, survival time and reduced clinical scores compared to PBS-vaccinated mice (p=0.022, p=0.015 and p=0.005, respectively). Priming SV09 three times with H1-DNA did not improve the heterosubtypic cross protection induced by SV09. H1-DNA/SV09 induced partial survival, 44%, albeit not statistically significant, with reduced clinical scores (p=0.007 relative to PBS) (Figure 1C and S1.C).
H1-DNA/SV09 confers partial protection while three times SV09 protects against heterosubtypic H5N1 challenge: A homologous prime/ boost vaccination regimen in which a SV of season 2011-2012 was given 3 times was previously shown to elicit 80% survival against H5N1 challenge . Here, we assessed the ability of the SV of season 2009 to induce protection against H5N1 when given in either a homologous or heterologous prime/boost immunization regimen.
Consistent with our previous experiment, the H1-DNA/SV09 vaccination regimen partially protected mice. The survival proportion of 43% was not significantly increased compared to PBS but clinical scores were reduced (p=0.021 compared to PBS). A homologous prime/boost regimen consisting of 3×SV09, however, induced significant survival (80%, p=0.003 compared to PBS) with prolonged survival time (p<0.001 compared to PBS), reduced bodyweight loss and clinical score (p=0.001 and p<0.001 compared to PBS, respectively) (Figure 2 and S3).
Neither neutralizing antibodies nor cross-reactive T cells reflect heterosubtypic protection induced by H1-DNA/SV09 and 3xSV09: To elucidate possible mechanism of heterosubtypic protection, we characterized the neutralizing antibody and HA T-cell responses induced with H1-DNA/SV09 and 3×SV09 regimens against heterosubtypic (challenge) influenza strain (H5N1) and the SV09 H1N1 strain (A/Brisbane/59/07). Neither vaccination regimen elicited detectable cross-neutralizing antibodies, while both vaccination regimens induced high titers of vaccine homologous H1N1 neutralizing antibodies, with H1-DNA/SV09 regimen eliciting approximately 5.3-fold higher mean titer compared to 3×SV09 (Figure 3A). With respect to HA specific T-cell response, only the H1-DNA/ SV09 regimen elicited a T-cell response against both a highly conserved T-cell epitope in the stalk of the HA (IYSTVASSL, conserved in both H1 and H5) and the total pool of peptides from HA of A/Brisbane/59/07 (Figure 3B). Thus, neither cross-neutralizing antibodies nor HA-specific T-cell responses can explain the high heterosubtypic protection seen by 3×SV09 versus suboptimal protection obtained with H1-DNA/SV09 vaccination regimen.
Figure 3: Neither neutralizing antibodies nor cross-reactive T cells reflect heterosubtypic protection induced by H1-DNA/SV09 and 3xSV09. Characterization of humoral and cellular immune response induced by 3x Inflexal 2009 (3xSV09), 3x H1 HA DNA followed by a boost of Inflexal 2009 (H1-DNA/SV09) or 4x PBS (PBS), 4 weeks after final immunization (n=8). A) Neutralizing Ab titer against H5N1 A/Hong Kong/156/97 and H1N1 A/Brisbane/ 59/07. B) INF-γ secreting T cells measured in ELISPOT against a pool of 215-mer peptides carrying conserved epitope; IYSTVASSL and a pool of 15-mer peptides of HA of H1N1 A/Brisbane/59/07. Group medians are shown.
We further characterized the humoral immune responses induced with the two vaccination regimens by measuring the total IgG response against HA, NA and NP in the pre-challenge serum. Pre-challenge antibody titers were used to assess the relationship between humoral immunogenicity and challenge outcome. Neither H1-DNA/SV09 nor 3xSV09 elicited significant rH5 A/Hong Kong/156/97 binding antibodies (except in 2 out of 8 mice from 3xSV09 group) (Figure 4A). Both regimens elicited high titer antibodies against vaccine rH1 (A/ Brisbane/59/07).
Figure 4: High titers of antibodies against NP, but not against H5 HA or NA, are elicited with both vaccination regimens and reflect observed difference in heterosubtypic protection. Characterization of humoral immune response induced by 3x Inflexal 2009 (3xSV09), 3x H1 HA DNA followed by a boost of Inflexal 2009 (H1- DNA/SV09) or 4x PBS (PBS) in pre-challenge serum of mice challenged with H5N1 A/Hong Kong/156/97 (Immunogenicity of animals in challenge is reported; n=8 for PBS and 3x SV09; n=7 for H1-DNA/SV09). Binding Ab titer measured in ELISA against A) rHA of H5N1 A/Hong Kong/156/97 and rHA of H1N1 A/Brisbane/ 59/07 B) rNA of H5N1 A/Hong Kong/156/97 and rNA of H1N1 A/ Brisbane/59/07 C) rNP of A/Puerto Rico/8/34 (due to lacking serum n=5 for PBS; n=5 for 3x SV09; n=6 for H1-DNA/SV09). Group medians are shown.
The mean rH1 titer elicited with H1-DNA/SV09 was approximately 5.3-fold higher than the corresponding titer elicited with 3xSV09. Neither H1-DNA/SV09 nor 3xSV09 elicited significant rN1 A/Hong Kong/156/97 binding antibody titers (Figure 4B). A moderate titer of antibodies against vaccine rN1 (A/Brisbane/59/07) was detected after 3xSV09, but not after H1-DNA/SV09.
Since HA and NA specific assays did not appear to correlate with the observed protection against H5N1, we assessed the level of NP antibodies. As NP is highly conserved between influenza a group 1 virus (e.g., an amino acid homology above 90%) we did not differentiate between vaccine homologous and challenge strain NP. Pre-challenge serum of mice immunized with H1-DNA/SV09 contained moderate level of anti-NP antibodies, and comparable to the level observed after single immunization with SV09 (data not shown) while the mean titer of anti-NP antibodies elicited with 3xSV09 was approximately 11.4-fold higher (Figure 4C).
In this study we confirm the previous finding  that the breadth of heterologous protection elicited with seasonal influenza vaccine can be enhanced by priming the vaccine with DNA expressing vaccine corresponding H1 HA. In addition, we demonstrate that this regimen is less efficient than multiple immunizations with seasonal vaccine in conferring protection against heterosubtypic H5N1. Among tested immune parameters only the anti-NP antibody titer followed the pattern of the protection between the two vaccination regimens.
Influenza specific antibodies can protect against influenza infection via a number of different immunological mechanisms. Neutralizing antibodies can prevent viral attachment to target cells [37,38], inhibit release from endosomes [22,37,39] or block egress [37,40]. Binding antibodies can mediate protection via indirect mechanisms such as antibody-dependent cellular cytotoxicity (ADCC) or antibodydependent complement mediated cytotoxicity (ADCMC) [18,41]. Influenza specific T-cells play a role in clearing infected cells and thereby limit the virus spread and host morbidity [4,42-44]. While vaccine- or infection-elicited protection against closely related viruses predominantly relies on hemagglutination-inhibiting antibodies that block the viral attachment to sialic acid receptors at cell surface, the protection against distant viruses is less well understood. It is likely that for different vaccines different mechanisms contribute to heterosubtypic protection, depending on the vaccine composition, formulation and schedule.
DNA plasmids as vectors for antigen delivery have been used in several fields to enhance the level and the breadth of the immune response [45-47]. The mechanism underlying the increased breadth of the humoral immune response induced when priming with DNA is suggested to be due to the increased number and diversity of induced CD4 T cells which can increase the expansion of antigen specific B cells [45,48,49]. In the influenza field, this strategy has been implemented primarily for pandemic H5N1 vaccines [50-52] but also in the development of cross-protective seasonal influenza vaccines [30,35]. In our hands, priming virosomal seasonal influenza vaccine with DNA expressing vaccine-corresponding H1 HA enhanced the heterologous protection in two different influenza challenge models, confirming the results from Wei et al.,  for a split influenza vaccine. However, when evaluating the regimens’ ability to confer heterosubtypic protection we found that priming the virosomal vaccine with H1-DNA did not improve the vaccines ability to confer H5N1 protection. Though eliciting a higher cross reactive HA specific T cell response than the homologous prime/boost regimen neither H1- DNA/SV09 nor 3xSV09 elicited detectable cross reactive HA Ab responses. Nevertheless, the homologous prime boost regimen, 3×SV09, was able to confer a higher level of survival after challenge with the heterosubtypic H5N1 strain.
We reported previously that vaccinating mice with SV11 according to the same homologous prime/boost regimen induced a comparable level of survival against H5N1. Unlike 3xSV09, the protection induced by3xSV11 correlated with the vaccine mediated H5 HA binding antibodies induced. The difference seen in protective capacity and immunogenicity between SV09 and SV11 is likely due to difference in vaccine composition. These results further emphasize the difficulty in predicting the cross protective capacity of a seasonal influenza vaccine based solely on immunogenicity.
In the current study, T-cell immunity was elicited only with H1- DNA priming, so the superior protection against H5N1 by homologous vaccination could not be explained by HA-directed cellular immunity. Though we cannot rule out the level of cellular immunity against NA, our historical data indicate that these seasonal influenza vaccines do not induce NA-directed T-cell immunity. In combination with a lack of HA and NA cross-reactive antibodies, these findings suggest that responses to seasonal vaccine components other than HA or NA may contribute to the heterosubtypic H5N1 protection. Highly conserved influenza core proteins, such as nucleoprotein (NP), have frequently been investigated for their ability to induce cross-protection [5,7,53,54]. While primary mechanism of protection was considered to be via CD8 effector cells, a role for anti- NP antibodies has recently been suggested [55,56]. Vaccine candidates containing NP have recently been shown to induce cross-protection in mice, and passive transfer experiments of anti-NP IgG have confirmed the ability of NP antibodies to confer protection [55-57]. Here we demonstrate that anti-NP antibodies are elicited by both vaccination regimens and it is possible that they contribute to the heterosubtypic (H5N1) protection conferred with vaccination modules comprised of SV09. The anti-NP titer induced with the heterologous prime/boost vaccine (H1-DNA/SV), the regimen that conferred weaker protection against H5N1, was lower and comparable to the titer level elicited with a single vaccination with SV (data not shown). This was as expected considering that DNA used for priming encodes only HA and not NP.
As we did not measure NP-specific T cells, we cannot formally prove that they do not play a role in the heterosubtypic protection elicited with 3xSV09. However, considering that 3xSV09 was not very efficient in eliciting HA-specific T cells, and our historical data indicating that this type of vaccines does not induce strong NP-specific T cells (data not shown), we consider it unlikely that NP T cells play a significant role for 3xSV09 vaccine-mediated heterosubtypic protection. It is also unlikely that NP-specific T cells played a role in protection elicited with H1-DNA/SV09 considering that the DNA encoded only HA and not NP.
The exact mechanism by which NP-specific antibodies mediate protection is not completely understood. In recent studies it has been established that NP is presented on the surface of virus-infected cells during budding of new viruses [58,59] and the mechanism by which the anti-NP antibodies mediate protection has been shown to be FcγR dependent . Furthermore, Jegaskanda et al.,  have shown, using NK-cell activation assay, that a trivalent inactivated influenza vaccine induce NP-mediated ADCC responses. Thus, it is possible that antibodies directed against influenza NP contribute to the heterosubtypic H5N1 protection observed in our study through these mechanisms.
In conclusion, we demonstrate that priming a seasonal influenza vaccine with HA encoding DNA and thereby improve its cross reactive HA T-cell response does not improve its ability to cross protect against H5N1 influenza virus. Despite the lack of detectable cross reactive HA and NA Ab titers a homologous prime /boost regimen of the vaccine improved its ability to confer cross protection against H5N1. While HA and NA immunogenicity appears to play a minor role, NP-binding antibodies are the only immunogenicity parameter tested, which follows the same pattern as the heterosubtypic protection between the two studied vaccination regimens. We suggest that adding NP encoding DNA in the combination vaccine schedule may be an interesting approach to further broaden the protective activity of this heterologous prime/boost vaccine regimen in future studies.