The SIN lentiviral transfer vector plasmid pTY2-NP expressing the nucleprotein (NP, GenBank: AAM75159.1) from Influenza virus A/PR/8/1934 (H1N1) has been already described (13). For construction of SIN lentiviral transfer vector plasmid pGAE-HA expressing the hemagglutinin (HA, GenBank: ACP44189.1) from influenza virus A/California/7/2009 (H1N1), the codon optimized HA open reading frame (ORF) was chemically synthesized, inserted into pUC57 plasmid (TwinHelix, Milan, Italy), excised with AgeI/SalI restriction enzymes and cloned into pGAE-green fluorescent protein (GFP) lentiviral transfer vector (37) by replacing the GFP coding sequence. The IN defective packaging plasmid, producing the viral proteins necessary for vector particle production, and the phCMV-VSV.G plasmid, encoding the pseudotyping vesicular stomatitis virus envelope glycoprotein G (VSV.G) from Indiana serotype, necessary for IDLV entry into target cells, have been previously described (38, 39). Plasmids pCAGGS-TMPRSS2 and pCAGGS-HAT, expressing the transmembrane protease serine 2 (TMPRSS2, GenBank: U75329) and the human airway trypsin (HAT, GenBank: AB002134) proteins, respectively, were described previously (40). Plasmids pCMV3-H1N1-C-NA (Sino Biological Inc., North Wales, PA, USA), expressing the neuraminidase (NA, GenBank: ACP41107.1) derived from influenza virus A/California/4/2009 (H1N1), and pCMV-HACal09, expressing the codon optimized HA protein from the CMV promoter, were used during production of HA-pseudotyped IDLV-NP/HA.

Lenti-X human embryonic kidney 293T cell line was obtained from Clontech (Mountain View, CA, USA) and was used for IDLV production following established protocols (41). Cells were maintained in Dulbecco’s modified eagles medium (Gibco, Life Technologies Italia, Monza, Italy) supplemented with 10% fetal calf serum (Corning, Mediatech Inc., Manassas, VA, USA) and 100 units/ml penicillin/streptomycin/glutamine (PSG) (Gibco, Life Technologies Italia, Monza, Italy). For production of IDLVs, pseudotyped with HA and expressing either HA (IDLV-HA/HA) or NP (IDLV-NP/HA), 3.5 × 10⁶ 293 T Lenti-X cells were seeded on 10-cm Petri dishes (Corning Incorporated—Life Sciences, Oneonta, NY, USA) and incubated overnight. Cells were transiently transfected with lentiviral transfer vector expressing influenza HA or NP, along with IN defective packaging and VSV.G-envelope plasmids by Calcium Phosphate using the Profection Mammalian Transfection System (Promega Corporation, Madison, WI, USA) as previously described (13, 39, 41). Plasmids pCAGGS-TMPRSS2 or pCAGGS-HAT and plasmid pCMV3-H1N1-C-NA were included to express protease and NA during IDLV production. To pseudotype IDLV-NP/HA with HA protein, pCMV-HACal09 plasmid was added during IDLV-NP/HA production. After 48 and 72 h post-transfection, cell culture supernatants were collected, cleared from cellular debris by low-speed centrifugation and passed through a 0.45-µm pore size filter (Millipore Corporation, Billerica, MA, USA). To produce IDLV stocks for mouse immunization, vector containing supernatants were concentrated by ultracentrifugation (Beckman Coulter, Fullerton, CA, USA) on a 20% sucrose gradient (Sigma Chemical Co., St. Louis, MO, USA) at 23.000 rpm for 2.5 h at 4°C using a SW28 swinging bucket rotor (Beckman). Pelleted vector particles were resuspended in 1× phosphate-buffered saline (PBS, Gibco, Life Technologies Italia, Monza, Italy) and stored at −80°C until use. Each IDLV-HA/HA or IDLV-NP/HA stock was titred by the reverse transcriptase (RT) activity assay and the corresponding transducing units (TU) were calculated by comparing the RT activity to the one of IDLV-GFP virions with known infectious titers, thus allowing for the determination of their infectious titer units (42).

293 T Lenti-X cells were plated in six-well microplates for flow cytometry analysis or seeded in 24-well cluster plates onto 12-mm cover glasses previously treated with l-polylysine (SIGMA) for CLSM. Cells were then transfected by calcium phosphate with pCMVHACal09 using the Profection Mammalian Transfection System (Promega). Twenty-four hours following transfection, non-permeabilized cells were stained with antiserum from CB6F1 mice immunized with purified H1N1 subunit vaccine (0.1 µg of HA per dose) from influenza strain H1N1 A/California/7/2009 (provided by Novartis Vaccine & Diagnostics Srl, Siena, Italy) in the presence of MF59 adjuvant (provided by Novartis). After 45 min, cells were washed in PBS 1× and then incubated for 30 min with secondary antibody PE Goat antimouse IgG (Biolegend, San Diego, CA, USA), for flow cytometry analysis, or Alexa Fluor-488-F(ab′)2 fragments of goat antimouse (Molecular Probes, Life Technologies, Carlsbad, CA, USA), for CSLM analysis. For flow cytometry, cells were fixed with 1% paraformaldehyde and fluorescence was measured using the FACSCalibur (BD Biosciences, Milan, Italy), and data were analyzed using CellQuest (BD Biosciences). For CLSM, cells were fixed with methanol and the coverslips were mounted with Vectashield^® antifade mounting medium containing DAPI (Vector Labs, Burlingame, CA, USA) on the microscope slides. CLSM observations were performed on a Leica TCS SP2 AOBS apparatus (Leica Microsystems, Wetzlar, Germany), using excitation spectral laser lines at 405 and 488 nm, using the confocal software (Leica, Wetzlar, Germany) and Photoshop CS5 (Adobe Systems, San Jose, CA, USA). Signals from different fluorescent probes were taken in sequential scanning mode, several fields were analyzed for each labeling condition, and representative results are shown. Images represented a single central optical section taken in the center of each cell nucleus and a 3D reconstruction.

To evaluate HA presence on IDLV particles, pellets of IDLV concentrated preparations were resuspended in SDS loading buffer. Lysed virions were separated on 12% SDS polyacrylamide gel under reducing conditions and transferred to a nitrocellulose membrane (Sartorius Stedim Italy). Filters were saturated for 2 h with 5% nonfat dry milk in PBST (PBS with 0.1% Tween 20) and then incubated with HA antiserum for 1 h at room temperature followed by incubation for 1 h at room temperature with an antimouse horse radish peroxidase (HRP)-conjugated IgG (Sigma Aldrich, Milan, Italy). The immunocomplexes were visualized using chemiluminescence ECL detection system (Luminata Forte Western HRP Substrate, Millipore). H1N1 subunit vaccine from A/California/7/2009 virus was used as positive control.

CB6F1 female mice were purchased from Harlan (Harlan Laboratory, Srl, San Pietro al Natisone, Italy). Six- to eight-week-old CB6F1 mice, four mice per group, were injected once intramuscularly in the thigh with (i) 1 × 10⁷ RT units/mouse of IDLV-HA/HA, (ii) 1 × 10⁷ RT units/mouse of IDLV-NP/HA, (iii) H1N1 A/California/7/2009 (0.1 μg/mouse of HA protein, henceforth referred to as HAp) alone, and (iv) 0.1 μg/mouse of HAp in presence of MF59 adjuvant (HAp + MF59). Naïve, non-immunized mice were kept for parallel analysis. All immunized mice were boosted with HAp alone 24 weeks after the prime. Antibodies (Abs) were measured in serum at different time points, starting from 2 weeks after the prime and up to 18 weeks after the boost. The cellular immune responses were analyzed at 4, 12, and 24 weeks after the prime in blood samples and in splenocytes at sacrifice (18 weeks after the boost).

Serum samples were obtained from blood collected from the retro-orbital plexus of mice with glass Pasteur pipettes and stored at −20°C until assayed. Heparin-treated glass Pasteur pipettes were used to collect blood in order to perform IFNγ ELISPOT assay. Leukocytes, obtained after ammonium chloride potassium (ACK) treatment of whole blood, were counted, suspended in RPMI 1640 (Gibco) containing 10% fetal bovine serum (Lonza, Treviglio, Milan, Italy), 100 units/ml of PSG (Gibco), non-essential aminoacids (Gibco), sodium pyruvate 1 mM (Gibco), HEPES buffer solution 25 mM (Gibco), and 50 mM 2-mercaptoethanol (Sigma Chemicals). Splenocytes were prepared by mechanical disruption and passage through cell strainers (BD Biosciences) and resuspended in complete RPMI medium, as previously described (39).

Sera were tested for the presence of binding Abs by a standard ELISA. Ninety-six well plates (Greiner bio-one, Kremsmünster, Austria) were coated with H1N1 A/California/4/09 subunit vaccine (0.2 μg/well of HA) overnight at 4°C. After washing and blocking, serial dilutions of serum from individual mice were added to wells in duplicate and incubated for 2 h at room temperature. The plates were washed and biotin-conjugated goat antimouse IgG (Southern Biotech, Birmingham, AL, USA) was added to the wells for 2 h at room temperature. The plates were washed before the addition of HRP-conjugated streptavidin (AnaSpec, Fremont, CA, USA) for 30 min at room temperature, followed by the 3,3′5,5-tetramethylbenzidine substrate (SurModics BioFX, Edina, MN, USA). Endpoint titers were determined as the reciprocal of the highest dilution giving an absorbance value at least equal to twofold that of background (biological sample from naive mice). For each group of immunization, results were expressed as mean titer with confidence interval.

Hemagglutination inhibition (HAI) Abs to A/California/7/09 virus were measured, according to standard procedures (43). Briefly, all sera were treated with receptor-destroying enzyme (Sigma Aldrich) to remove non-specific inhibitors of hemagglutination. Serial twofold dilutions of treated sera were mixed with 4 hemagglutinin units of the A/California/7/09 virus and, after 1 h incubation at room temperature, with 0.5% Turkey red blood cells. The HAI titers were expressed as the reciprocal of the highest dilution of serum that inhibited virus-induced hemagglutination.

The sera were tested by MDCK cell-based microneutralization (MN) assay (44) to titrate influenza virus-specific neutralizing antibodies. Briefly, all samples were heat-treated at 56°C for 30 min. Twofold serial dilutions of samples, starting from 1:10 dilution, were performed across the 96-well plates and then the viral working solution of A/California/07/2009 (H1N1) live virus was added (200 TCID50/100 μl). After incubation of mixture sera-virus at 37°C for 1 h, a cell suspension of 2 × 10⁵ was added and the plates were incubated at 37°C for 5 days. Each plate was then checked under an optical microscope to assess the presence of local lesions and cytopathogenic effect. The neutralization titer (NT) for each serum was calculated according to the Spearman-Kärber formula.

Since NA plasmid was included during the IDLV preparation and NA protein may be present in subunit vaccine preparations, the heat-treated sera were also tested by enzyme-linked lectin assay (ELLA) (45) to measure the functional anti-NA Abs. Briefly, serial twofold dilutions of treated sera were mixed with antigen A/California/07/2009 (H1N1) treated with Triton X-100 as described elsewhere, in 96-well plates coated with fetuin. The plates were incubated at 37°C overnight (16–18 h). After washing the plates, the peanut agglutinin conjugate with peroxidase was added and the plates were incubated at room temperature for 2 h in the dark. Then the o-phenylenediamine dihydrochloride substrate was added and the plates were incubated at room temperature for 10 min in the dark. The reaction was stopped by adding 1 N H2S04. The optical density of all the test plates was read at 490 nm. The NA inhibition (NI) titer was expressed as the reciprocal of the last dilution that results in at least 50% inhibition of the maximum signal.

The temporal trend of antibody response (i.e., ELISA, HAI, MN, and ELLA) was analyzed using a system of piecewise linear equations in order to jointly evaluate the relationships of each outcome at each time point, allowing for correlated errors. The analysis was conducted in STATA13 (StataCorp LP, College Station, TX, USA) within a structural equation modeling frame work (46, 47). The log likelihood estimation procedure was used to fit the model to the data.

To produce IDLV expressing HA, 293 T LentiX cells were transfected with IN defective packaging plasmid, expressing the proteins required for IDLV assembly and release, the pseudotyping VSV.G-expressing plasmid, essential for IDLV entry into target cells, and HA-expressing lentiviral transfer vector plasmid, with the aim of pseudotyping the recombinant IDLV particles with the membrane tethered HA. Importantly, the presence of HA protein on the surface of the transfected cells was confirmed by flow cytometry and confocal microscopy (Figures 1A,B). However, the recovery of IDLV in the supernatant of the transfected cells was very low, indicating inhibition of release of IDLV pseudotyped with HA envelope (Figure 2A, left column). This was expected since, as with wild type influenza virus, NA protein is required for release of lentiviral vectors pseudotyped with influenza HA (17, 18, 48). To improve production of HA-pseudotyped IDLV, we cotransfected plasmids expressing NA protein, required for cleavage of surface sialic acid molecules on producer cells allowing release of the vector in the supernatant, and human serine transmembrane protease TMPRSS2 or HAT proteins, mediating proteolytic activation of influenza HA (18).

As shown in Figure 2A, recovery of IDLV-HA increased an average of eightfold in presence of TMPRSS2 but only twofold in the presence of HAT. Importantly, NA was fundamental for IDLV-HA release, increasing the amount of IDLV recovered in the supernatant an average of 24-fold over the IDLV produced in the absence of NA expressing plasmid. The inclusion of TMPRSS2 or HAT in addition to NA protein further improved, although not significantly, vector production (27-fold and 25-fold, respectively, compared to IDLV produced in the absence of NA and proteases). For producing concentrated IDLV preparations for immunization, TMPRSS2 was chosen over HAT protease expressing plasmid since IDLV production was generally higher and more consistent.

To produce IDLV expressing NP, vector was produced as described above and in the Section “Materials and Methods,” by including a NP-expressing lentiviral transfer vector plasmid and a plasmid expressing the HA protein for pseudotyping the IDLV produced in the 293 T cells. In Figure 2B, is shown a representative western blot (WB) analysis of lysates from concentrated stocks of IDLV showing the presence of HA protein in all purified IDLV preparations.

To mimic the immunization protocol of the seasonal influenza in humans, mice were given a single immunization and the immune responses were analyzed at various time points up to 24 weeks. As positive controls, groups of mice were immunized with H1N1 subunit vaccine alone (HAp) or in combination with MF59 (HAp + MF59). The schedule of immunization is described in Figure 3A.

Humoral response to H1N1 was assessed initially by ELISA. As shown in Figure 3B, 2 weeks after the prime all vaccinated animals developed anti-H1N1 IgG antibodies in serum. The titers increased and persisted in all groups starting from 8 weeks after immunization. The highest titers were recovered in animals vaccinated with MF59, while the lowest levels were detected in the HAp group. No significant difference was observed between HAp + MF59 group and IDLV groups. Importantly, in both groups of IDLV vaccinated animals sustained Abs were induced at significantly higher titers than HAp alone (p < 0.01).

To assess the recall response, mice were boosted with HAp alone 6 months after the prime. Two weeks after the immunization with HAp all groups showed a boost in terms of binding Ab titers (Figure 3B). In particular, IDLV vaccinated animals were able to increase the anti-H1N1 Abs when boosted with HAp, indicating that immunization with IDLV enables the animals to respond to the HA antigen delivered in a different way, as a subunit vaccine in this case. The responses were persistent up to 18 weeks after the boost in all vaccinated animals. After the boost the differences between IDLV immunized animals and HAp group remained statistically significant (p < 0.01). Again, no significant difference in Ab titers between both IDLV groups and the adjuvanted group was seen.

To assess the functional antibodies, HAI assay, and microneutralization (MN) assay were performed at 8 and 19 weeks post priming and 2 weeks and 18 weeks post boost (Figure 4). HAI titers were present in all groups at 8 weeks postimmunization and were still detectable at 19 weeks post priming (Figure 4A). The response significantly increased after the boost in all groups. The kinetics of HAI titers mirrored the kinetics of binding Abs, showing the highest titers in the adjuvant treated group and the weakest ones in the HAp immunized animals. Both groups of IDLV vaccinated animals showed similar levels of HAI titers (p > 0.05), significantly lower than HAp + MF59 (p < 0.05), but significantly higher compared to HAp vaccinated animals (p < 0.01). Interestingly, the MN assay showed absence of neutralizing activity in the HAp alone group of mice after the prime, while MN titers were always present in both groups of IDLV vaccinated animals (Figure 4B). Again the highest activity was present in serum samples from HAp + MF59 immunized group. The boost increased the MN titers in all groups, including the HAp immunized animals.

Since NA plasmid was included during the IDLV preparation and NA protein is present in subunit vaccine preparations (49, 50), we investigated the anti-NA response in all groups of immunized animals. Serum samples were thus assayed for NI activity. As shown in Figure 5, both IDLV groups showed a strong NI activity at all indicated time points after the prime that was significantly boosted after the immunization with HAp. A similar response was detected in the adjuvant vaccinated group, while the animals vaccinated with HAp alone did not generate detectable anti-NA response after the prime, showing low NI activity only after the boost.

In order to assess the NP specific CD8-restricted cellular response in mice immunized with IDLV-NP/HA, INFγ ELISPOT was performed using blood cells collected at 4, 12, and 24 weeks after the prime, stimulated with MHC Class I-restricted NP peptide (Figure 6A). As expected, a high number of IFNγ producing T cells was detected in animals vaccinated with IDLV-NP/HA overtime, confirming a strong and persistent CD8+ T cell response directed to the transgene delivered by IDLV. The NP-specific T cell response was also analyzed in splenocytes at 42 weeks after a single IDLV-NP/HA immunization further confirming the persistence of transgene-specific IFNγ producing CD8⁺ T cells (Figure 6A).

In order to evaluate the HA specific response, INFγ ELISPOT was also performed in splenocytes from all groups of mice at sacrifice 18 weeks after the boost (42 weeks after the prime). Ten pools of 15mer peptides spanning the entire HA protein were used and the cumulative mean response against the HA pools is shown in Figure 6B. HA-specific IFNγ producing cells were detected only in mice primed with IDLV-HA/HA, while no positive response was observed in mice immunized with HAp with or without the adjuvant. In particular, pool 1, pool 9, and pool 10 (mean of 165.4, 666.2, 158.1 SFC/10⁶ cells, respectively) were responsible for the HA-specific response.