This study was conducted as per the guidelines and approval of the Institutional Animal Ethics Committee (Approval No: IAEC/HEP-14/NIV-54/2012 dated 03/08/2012). H5N1 virus challenge experiments were done at the BSL-3+ biosafety containment facility following the approval of the Institutional Biosafety Committee (Approval No: NIVIBSC/27.07.2012/4). The mortality aspects of the protocols were reviewed and approved by the Institutional Animal Ethics Committee. To minimize animal suffering and distress, isoflurane and ketamine–xylazine anesthetics were used during experimental procedures.

H5N1-Navapur-A/chicken/India/33487/2006 (clade 2.2) and JC-2-1 NIVAN 1117307 (clade 2.3.2.1) influenza viruses isolated at National Institute of Virology were used for homologous and heterologous challenge, respectively. The clade 2.2 virus was used for the generation of recombinant proteins as described earlier (22).

The HA gene was cloned in pFastBac1 and expressed in Spodoptera frugiperdii insect cells, and recombinant HA protein was purified using lentil lectin affinity chromatography (GE healthcare, USA). NP gene cloned in pET15b bacterial expression system was used to transform BL21 codon plus (RIL) cells. Recombinant NP protein was purified using Ni⁺⁺ chelated resin (Invitrogen, USA). Synthetic M2e peptide, SLLTEVETPTRNEWECRCSDSSD, was obtained from INBIOS S.r.l, Italy.

Mw (5 × 10⁹ cells/ml, heat-killed M. indicus pranii) was purchased from Cadila Pharmaceuticals India in the form of “Immuvac.” Mw formulations were prepared by mixing adjuvant and immunogens (1:1 v/v ratio) to make a total of 100-µl dose per mouse containing 50 µl of Mw and 50 µl of immunogens.

Six to eight week old female inbred BALB/c mice were immunized intramuscularly with 100 µl (50 µl per quadriceps muscle) of single or double doses of respective vaccine formulations given at 3-week intervals (Table 1; Figure 1). Mice were bled before immunization, 3 weeks post immunization, and 10 days post-second dose under isoflurane anesthesia. Control mice received PBS.

For immunophenotyping experiment, the immunized mice were sacrificed at 24 h/3 weeks post-dose-1 and 24 h/10 days post-dose-2. Harvested spleens were used for immunophenotyping by surface staining, cytokine estimation by cytometric bead array (CBA), and gene expression profiling by Taqman low density array (TLDA). Spleens from unimmunized mice were used for control.

For the detection and titration of IgG-anti-HA, NP, and M2e antibodies, recombinant protein-based ELISA was used as described earlier (22, 27). IgG-subtype analysis was done as described previously (28).

The HI titers were measured on 10 days post-dose-2 as per the guidelines described in WHO manual (29) using receptor destroying enzyme (Denka Seiken UK Ltd., UK), 0.5% turkey erythrocytes, inactivated H5N1-Navapur-A/chicken/India/33487/2006 virus (homologous HI titers), and JC-2-1 NIVAN 1117307 virus (heterologous HI titers).

Splenocytes were prepared using standard methods. One million cells were used for antibody labeling. Anti-mouse fluorochrome-labeled antibodies used for different cell types were T cells: anti-CD3-FITC (Clone 17A2), anti-CD4-APC/PE (Clone RM4-5), anti-CD8-APC (Clone 53-6.7); B cells: anti-CD19-APC-Cy7 (Clone ID3), anti-CD45-FITC (Clone RA3-6B2); macrophages: anti-F4/80-FITC (Clone BM8), anti-CD105-APC (Clone MJ7/18); granulocytes: anti-Ly6G-FITC (Clone 1A8), anti-JAML-PE (Clone 4E10); natural killer cells: anti-CD11b-FITC (Clone M1/70), anti-CD49b-APC (HMα2), anti-NKG2D-PE (CX5); and dendritic cells: anti-CD11c-APC (Clone HL3), anti-CD33D1-PE (Clone 33D1). Costimulatory and activation markers used were anti-CD69-PE/FITC (Clone H1.2F3), anti-MHCII-eFluor (Clone M5/114.15.2), anti-CD80-PE (Clone 16-10A1), anti-CD86-FITC (Clone GL1), anti-CD134-PE (Clone OX-86), and anti-CD137L-PE (Clone TKS-1). These anti-mouse fluorochrome-labeled antibodies were purchased from BD Pharmingen, USA/e-Bioscience, USA/Biolegend CA. A minimum of 10,000 events were acquired for every sample after gating splenocyte clusters. Acquisition of samples was carried out on BD FACS ARIA-II flow cytometer, and data were analyzed using BD FACS DIVA software (BD Biosciences, USA).

Splenocytes (one million) from mice groups at 10 days post-dose-2 were cultured and stimulated with 1 µg HA + 1 μg NP + 5 μg M2e at 37°C for 4/24/72 h. The culture supernatants were tested for Th1 cytokines (IFN-γ, IL-2, and TNF) and Th2 cytokines (IL-4, IL-6, IL-10, and IL-5) by using CBA kit (BD Biosciences, USA) as described previously (30). As positive control, splenocytes stimulated with 2.5 µg/well concanavalin A (Sigma, USA) and as negative control, unstimulated cells were used. Cells collected after 4 and 24 h of stimulation were used for surface staining and gene expression analysis by TLDA.

Frozen spleen samples were processed for total RNA extraction and gene expression analysis as described previously (31). cDNA prepared from 500 ng RNA was mixed with 2×PCR master mix (Life technologies, USA) and loaded on the Taqman mouse immune array panel card and run on 7900 Real-Time PCR System (Applied Biosystems, USA). cDNAs from the spleens of naive mice and the mock-immunized (PBS) control mice were used as calibrators. 18s gene was used as the endogenous control. Relative quantification (RQ) values obtained were used to calculate fold upregulation or downregulation. RQ values between 0.5 and 2 were considered normal.

Control and immunized mice with single dose or two doses of the respective formulations (n = 10/group) were challenged intranasally with 50 µl of 100 LD50 homologous (H5N1-Navapur-A/chicken/India/33487/2006) or heterologous (JC-2-1 NIVAN 1117307) influenza virus under ketamine–xylazine anesthesia. Single-dose immunized mice were challenged post 3 weeks of immunization, while those receiving two doses were challenged on 10 days post-dose-2. Since the aim of this study was to assess the efficacy of different vaccine formulations against the challenge of H5N1 virus, euthanasia was not used till experimental endpoint [post infection day 14 (PID-14)] to find out if reversal of sickness occurs resulting into survival. Mice were monitored daily for 14 days for weight loss and mortality.

Three mice per group were sacrificed 72 h post challenge (PID-3) and harvested lungs stored in RNA later at −80°C until tested. RNA was extracted from the lung homogenates using QIAmp viral RNA mini kit (Qiagen, Valencia, CA, USA) according to the manufacturer’s instructions. Quantitation of viral RNA copy number was done by VLA Taqman wet assay real-time Influenza A/H5/H7 detection kit (Applied Biosystems, USA).

Statistical analyses were performed using the SPSS 20 software (SPSS Inc., IL, USA). To compare antibody titers and proportions of immune cell types, Student’s t-test and ANOVA with Tukey’s post hoc test were used, while Pearson correlation coefficients were determined by bivariate and partial correlation analyses. PCC values in the range 0.6–0.8 and >0.8 were considered to depict moderate and high correlation, respectively, at p < 0.05. Kaplan–Meier survival analysis was performed in GraphPad Prism software. The differences at p value <0.05 were considered significant for all the analyses.

Table 1 and Figure 1 depict immunogen-specific IgG antibody titers (ELISA) and isotype analysis conducted prior to virus challenge. After first dose, all the immunized mice showed 100% seroconversion against HA, NP, and M2e.

Highest anti-HA-IgG titers were recorded with Mw-HA formulation (20,318 ± 4,267, Figure 1A), while the addition of other antigens, i.e., NP or NP + M2e led to a significant decrease (p < 0.05). No adjuvant effect was observed for the combination formulations. With a single dose at the same (p < 0.01) or higher (50 µg, p < 0.05) Mw-HNM concentration, the titers were lower than two doses of Mw-HNM and were concentration dependent (p < 0.05).

Response to NP was distinctly different (Figure 1B). The addition of Mw to NP lowered anti-NP antibody titers (p < 0.05), no change was recorded with Mw-HN formulations while the titers increased with Mw-HNM formulation (p < 0.05). Among the only-protein formulations, NP elicited significantly higher titers than HNM (p < 0.001). Anti-NP titers did not increase with the number of doses for Mw-HNM or HNM. In fact, higher immunogen concentration (50 µg) led to a significant decrease in the antibody titers (p < 0.001). Addition of Mw to HNM led to a significant reduction in anti-M2e IgG titers (p < 0.05) (Figure 1C). A significant lowering of anti-M2e titers was associated with increasing the number of doses or antigen concentration (p < 0.05).

The Th2 response exhibited by non-adjuvanted HA, HN, and HNM formulations was switched to a balanced Th1/Th2 response in Mw-HA, Mw-HN, and the single dose-Mw-HNM/HNM formulations (Figure 1D). As compared to a Th2 response with Mw-HNM, a single dose with fivefold higher antigen concentration led to a distinct Th1 response. Only NP, Mw-NP, Mw-HN, and Mw-HNM induced Th2 response, while a balanced Th1/Th2 profile was seen with HN, HNM, and Mw-adjuvanted single dose with higher antigen formulations (Figure 1E). Since anti-M2e-IgG titers were low in all the groups, isotype analysis was not done.

Addition of Mw to HA led to significantly higher HI titers (64 ± 13) (Figure 2). Similar to ELISA, addition of NM led to significant decrease in the Mw-HNM group (4.3 ± 6, p < 0.05). For HN and HNM formulations, no adjuvant effect was seen. When tested against the clade 2.3.2.1 virus, no HI activity was exhibited by these mice groups.

The mice groups immunized with two doses of different formulations were initially challenged with 100 LD50 homologous viruses. All the three adjuvanted formulations (Mw-HA, Mw-HN, and Mw-HNM) provided complete protection (Figure 3B). The weight loss was minimum in these groups (Figure 3A). Thus, the presence of HA was associated with complete protection. The protection conferred by the respective protein alone formulations was 83, 83, and 67%, respectively (Figure 3D). At 72 h post infection (PID-3), the control mice exhibited 3 × 10⁸ RNA copies/ml lung suspension. Correlation of the reduction in lung viral load and 100% protection was seen with Mw-HA (14-fold, p < 0.001) and Mw-HN (29-fold, p < 0.001) groups, while protection in Mw-HNM group was associated with no reduction in the viral load (p > 0.05, Figure 4A).

Partial protection and considerable weight loss was observed with formulations containing only NP (Figures 3A–D). Interestingly, Mw-NP formulation leading to 21-fold reduction in the lung viral load provided only 17% protection (Figures 4A and 3B). A significant decline in viral load was recorded with HNM (29-fold, p < 0.05), while no reduction (p > 0.05) was noted with HA, NP, and HN formulations (Figure 4A). Clearly, protection did not correlate with the reduction in lung viral RNA load. With a single-dose HNM immunization, addition of the adjuvant led to higher protection (from 17 to 33%, Figure 3F) with appreciable weight loss (Figure 3E) and comparable lung viral load (Figure 4B).

We selected Mw-HA and Mw-HNM formulations providing 100% protection against the homologous challenge for assessing efficacy against the inter-clade challenge. Irrespective of the adjuvant, HA led to 50% protection (Figure 3H), while 100% protection was achieved by the addition of the conserved antigens; only HNM conferred 33% survival. Importantly, a single dose with higher antigen concentration could provide 80% protection. Weight loss was inversely proportional to the survival rate for all the groups (Table S1 in Supplementary Material, Figure 3G). No reduction in viral load was observed for any of the groups studied (p > 0.05) (Figure 4C).

We further attempted to understand the correlation of observed protection and different parameters studied. Bivariate correlation analysis documented highest positive correlation with anti-HA-IgG2a (PCC = 0.918, p = 0.001) followed by anti-HA-IgG (PCC = 0.892, p = 0.003), anti-HA-IgG1 (PCC = 0.824, p = 0.012), and HI titers (PCC = 0.797, p = 0.018). Partial correlation analysis identified anti-HA IgG2a (PCC = 0.790, p = 0.034), i.e., Th1 response as the only marker correlating with protection. Importantly, HI titers showed high correlation with anti-HA-IgG (PCC = 0.867, p = 0.005), anti-HA IgG1 (PCC = 0.827, p = 0.011), and anti-HA IgG2a (PCC = 0.840, p = 0.009) titers noted in ELISA.

Based on the above results, Mw-HNM was identified as the most efficacious formulation and investigated further.

As a first step, we determined the proportion of different immune cell types in the spleens of the immunized mice. For this, a few cell types playing crucial role in the generation of innate (macrophages, granulocytes, DCs, and NK cells) and adaptive (T helper, T cytotoxic, and B cells) immune response were studied. Figure S1 in Supplementary Material provides gating strategy.

Culture supernatants of antigen-stimulated splenocytes were used for cytokine estimation. No cytokines were detected at 4 and 24 h post-stimulation in Mw-HNM and HNM groups, while higher levels of IFN-γ (689.5 and 165.8 pg/ml), TNF (260.3 and 210.1 pg/ml), and IL-6 (235.3 and 100.5 pg/ml) were detected at 72 h. A marginal (1.8 pg/ml) and no increase in IL-2 levels was noted in Mw-HNM and only-protein groups, respectively. Mw-induced induction of the signature Th1 cytokine, IFN-γ, was evident. None of the Th2 cytokines tested (IL-4, IL-5) were detected in both Mw-HNM and HNM groups.

Gene expression analysis was done using splenocytes from the immunized mice (1) prior to virus challenge (unstimulated splenocytes) and (2) in vitro cultured and antigen-stimulated splenocytes (4 and 24 h). As evident from the heatmap (Figure 7), stimulated splenocytes from the HNM group formed a distinct cluster, while the unstimulated splenocytes from Mw-HNM and HNM-alone groups and stimulated splenocytes from the Mw-HNM group formed two branches of the second cluster. Table S2 in Supplementary Material presents modulation of mRNA levels of the immune response genes at different time points. Overall, Th1/Th2 and proinflammatory cytokine genes were upregulated in Mw-HNM and downregulated in HNM-alone groups. CD19 was upregulated in the unstimulated Mw-HNM and HNM groups, while the other immune cell surface marker genes were expressed normally. At 4 h post-stimulation, CD19, CD3, CD4, CD8, CD40, and CD80 genes were downregulated in the HNM group. Of these, CD19, CD40, and CD80 genes remained downregulated at 24 h post-stimulation. The mRNA levels of CD19, CD4, CD8, and CD80 were upregulated at both time points in the Mw-HNM group. Of note, FACS (protein) and TLDA (gene) correlated well (Figures 5–7).