Chicken embryo fibroblast (CEF) primary cells and Hep-2 (CCL-81; ATCC, Manassas, VA) cell lines were grown in Dulbecco's modified Eagle's medium supplemented with 8% fetal bovine serum and antibiotics. The previously established NDV vaccine (K148/08), KCTC 11570BP, was used to generate a vector vaccine platform (31). The modified vaccinia Ankara T7 recombinant virus (MVA-T7) for the supply of T7 reverse transcriptase was kindly provided by Dr. Bernard Moss. The MVA-T7 virus was propagated in CEF cells for further use. The H5 subtype HPAIV clade 2.3.4.4e strain H5N6 A/duck/Korea/ES2/16 (ES2), GISAID EpiFlu Isolate ID EPI_ISL_239262, was kindly provided by the Animal and Plant Quarantine Agency in Korea.

The viral RNA of the K148/08 NDV strain was extracted using the RNeasy kit (Qiagen, Valencia, CA, USA). The propagation and RNA extraction of the ES2 HPAIV was performed inside the biosafety level 3 (BSL3) facility at Konkuk University, as described in sections Virus Rescue and Propagation and Serological Analysis. cDNAs of the K148 NDV full genome and ES2 HPAIV HA gene were synthesized using the SuperScript IV first-strand synthesis system (Thermo Fisher Scientific, California, USA).

The cDNA of K148 NDV was divided into six fragments with ~300 nucleotides of overlapping regions between adjacent fragments (Figure 1). All six fragments were amplified with six sets of gene-specific primers and cloned into the RBC TA cloning vector (RBC Bioscience, Taipei, Taiwan; Supplementary Table 1). Each fragment was sequentially inserted into the pBluescript vector using the In-Fusion® PCR cloning kit (Clontech, Mountain View, CA). The pfuUltra^TM II fusion HS DNA polymerase (Stratagene, La Jolla, CA) was used to amplify the in-fusion fragments and linearizing vectors. The vector plasmid was linearized before each step of the fragment in-fusion. Nucleoprotein (NP), polymerase (P), and large polymerase (L) genes were cloned into an expression plasmid to construct the supporting plasmids. The primers used for vector linearization and in-fusion are listed in Supplementary Table 2. After construction of the full-length cDNA clone of the K148 virus (pK148), pK148 was transformed into HIT-DH5α competent cells (RBC Bioscience) at 37°C for 18 h and purified using the PureLink™ HiPure plasmid midiprep kit (Thermo Fisher Scientific).

pK148 was linearized to insert the HA gene of ES2 HPAIV virus and an additional cDNA fragment containing the untranslated region (UTR), including the gene end, and gene start (GS) signals, between polymerase gene (P) and matrix (M) gene of the full genome K148 cDNA clone. The additional UTR and cDNA of ES2 HA gene were amplified by PCR and cloned into RBC TA cloning Vector (RBC Bioscience, Taipei, Taiwan). The multibasic cleavage site of the HA gene (PLRERRRKR/GLF) was replaced by the cleavage site of LPAIV (PLRERR/GLF). To enhance the expression of ES2 HA gene, Kozak sequence (ACC) was inserted at the upstream of HA gene by using Muta-direct^TM site-directed Mutagenesis kit (iNtRON Biotechnology, Korea). The cDNA of UTR and the HA gene were ligated and inserted between P and M genes of pK148 by infusion PCR described above, forming pK148/ES2-HA clone. The pK148/ES2-HA clone was amplified and purified as described above.

The plasmid with full-genome cDNA of pK148 or pK148/ES2-HA and three supporting plasmids were co-transfected into Hep-2 cells for virus rescue. Briefly, Hep-2 cells seeded on six-well-plates were infected with the MVA-T7 virus to provide T7 transcriptase. A mixture of 8.8 μg of pK148 or pK148/ES2-HA, 1 μg of pSupport_NP, 0.1 μg of pSupport_P, and 0.1 μg pSupport_L plasmids were transfected into Hep-2 cells using Lipofectamine 3000 (Thermo Fisher Scientific). After 1 h of transfection, the transfected cells were washed three times with phosphate-buffered saline and incubated in Opti-MEM at 37 °C for 72 h. After incubation, the cells and supernatants were frozen and thawed three times to release rK148 and rK148/ES2-HA viruse particles. The harvested viruses were inoculated into 10-day-old SPF embryonated chicken eggs and incubated at 37°C. After 3 days of incubation, the inoculated eggs were chilled at 4°C for 1 h and allantoic fluids were used for the hemagglutination assay to detect the rescued virus. HA-positive samples were filtered through a 0.2 μm syringe filter (Ministart RC 15, Sartorius Stedim Biotech, Germany) and subsequently passaged in 10-day-old SPF embryonated chicken eggs. Allantoic fluids were harvested, aliquoted, and stored at −80°C for further experiments. To determine the 50% egg infectious dose (EID₅₀), rescued virus, or ES2 virus were 10-fold serially diluted and inoculated in 10-day of SPF embryonated chicken eggs. After 3 days of incubation, allantoic fluids were harvested and checked for hemagglutination reaction. The EID₅₀ endpoint was calculated according to Reed and Muench method.

To examine the expression of ES2 HA protein and the stability of the HA gene after consecutive passages, rK148 and rK148/ES2-HA passaged four times in 10-day-old SPF embryonated chicken eggs, were harvested and clarified using low centrifugation (2,000 × g, 10 min, 4°C). The supernatants of the clarified fluids were inactivated with 0.1% formalin. After inactivation, the supernatants were pelleted using centrifugation (30,000 × g, 1 h, 4°C). The pellets were resuspended in PBS, loaded on a 20–50% (w/v) discontinuous sucrose density gradient, and centrifuged (150,000 × g, 1.5 h, 4°C) for purification. Protein concentrations of the purified samples were measured using the Bradford protein assay (Pierce, USA). The purified samples with high concentration (2 μg/well) and low concentration (400 ng) were analyzed using Western blotting. The expression of the ES2 HA gene was detected using antiserum of chickens immunized with clade 2.3.4.4e H5 subtype AIV and HRP-conjugated goat anti-chicken IgY secondary antibody (Invitrogen, California, USA). The presence of NDV was identified with anti-NDV HN mouse IgG antibody, which was kindly provided by Bionote (Suwon, Republic of Korea), and HRP-conjugated goat anti-mouse IgG antibody (Komabio, Seoul, Republic of Korea).

The rK148/ES2-HA virus and K148 NDV were inoculated in 10-day-old embryonated SPF chicken eggs with 0.1 mL of serially diluted virus (10¹, 10², 10³, and 10⁴ EID50). Each serial dilution of two viruses were inoculated in three 10-day-old embryonated SPF chicken eggs. The allantoic fluid from inoculated eggs were harvested at 12, 24, 36, 48, 60, and 72 h post inoculation and viral titers were measured in HAU (Hemagglutination Unit).

Twenty-four 1-day-old commercially available Pekin ducks were kindly provided by the Moran Food & Breeding Company (Eum-seong, Republic of Korea). Eighteen 1-day-old SPF chickens were purchased from Namduk SPF (Icheon, Republic of Korea). The ducks and chickens were divided into seven groups (n = 6), as described in Table 1. The vaccination groups were vaccinated oculonasally or by the spray method with a 10⁷ EID50 dose of rK148 or rK148/ES2-HA. An automated spraying device (SK-MO-2000, Threeshine Inc, Korea) producing aerosol particles of 50–70 μm, was used for the spray vaccination. The mock vaccination group was vaccinated with PBS. Booster vaccination was performed 3 weeks after the priming vaccination. The clinical signs of the vaccinated chickens and ducks were assessed daily. Blood samples were collected from ducks and chickens for serological analyses. Six weeks after priming, all birds were challenged with 100 mean bird lethal dose (BLD₅₀) of ES2 HPAI virus. Each BLD₅₀ for chickens or ducks was established as described previously (9, 33). The clinical signs and mortality rates were recorded daily. Oropharyngeal and cloacal swab samples of the challenged birds were collected at 2, 4, 6, 8, and 12 days post-challenge. All animal experiments conducted in this study were observed and approved by the Institutional Animal Care and Use Committee of Konkuk University (permit number: KU21045). The vaccination experiment was performed in a BSL2 facility, and the challenge study was performed at the animal BSL3 facility at Konkuk University.

The antibody responses in chickens and ducks before and after vaccination with rK148/ES2-HA were determined using hemagglutination inhibition (HI) tests. As both the HA protein of AIV and HN protein of NDV expressed on rK148/ES2-HA can induce hemagglutination, inactivated K148 virus and ES2 HPAI virus were used as antigens. Briefly, sera collected before vaccination [0 weeks post-vaccination (wpv)] and at 1, 2, 3, 4, and 5 wpv were heat-inactivated (56°C for 30 min) to inactivate the complement system. Duck serum samples were treated overnight with receptor-destroying enzyme (Denka Seiken Co., Japan) at 37°C to eliminate non-specific HI factors. Inactivated serum was diluted two-fold in PBS and incubated with four hemagglutination units of K148/08 or ES2 antigen for 40 min. The incubated samples were then mixed with 1% chicken red blood cells in a 96-well V-bottom plate. The HI titer of each sample was determined following a standard protocol (2).

To quantify viral shedding after challenge, RNA was extracted from oral and cloacal swab samples collected on days 2, 4, 6, 8, and 12 post-challenge (pc). The RNA was extracted from 200 μL of the supernatant of swab samples using the Roche MagNa Pure 96 extraction system (Roche, Manheim, Germany) according to the manufacturer's instructions. The extracted RNA was quantified by analyzing the cyclic threshold (Ct) value using matrix gene-based real-time reverse transcription-polymerase chain reaction (rRT-PCR) (34). The Ct values were converted into an infectious unit equivalent to EID₅₀/mL using a standard curve.

Chickens and ducks were monitored and scored every 24 h until 14 dpi. Chickens and ducks were scored based on their health status as follows: 0 (normal); 1 (sick); 2 (severely sick), and 3 (dead). “Sick” birds showed one of the clinical signs whereas “severely sick” ducks showed more than one: respiratory involvement, ruffled feathers, lethargy, anorexia, prostration, swollen heads, and green diarrhea.

The data for viral shedding titers and HI antibody titers were analyzed using one-way and two-way ANOVA with the Tukey–Kramer post-hoc test, and the long-rank test were performed to analyze significant of survival curves between animal experiment groups in GraphPad Prism version 8.0 (GraphPad Software Inc., CA). Statistical significance was set at p ≤ 0.05.

The pBlueScript encoding full genome of K148/08 virus cDNA (pK148) was generated using in-fusion PCR (Figure 1). The ES2-HA gene was inserted between the P and M junctions and ligated into the pK148 plasmid to construct pK148/ES2-HA (Figure 1). The rK148/ES2-HA virus was rescued after transfection of pK148/ES2-HA into Hep-2 cells. The rescue of rK148 and rK148/ES2-HA viruses was verified by using the hemagglutination assay, and the HA gene was verified by using PCR after three consecutive passages in 9–11-day-old embryonated SPF chicken eggs (Figure 2). No mutations were observed in the HA gene of the passaged rK148/ES2-HA virus. The total length of the gene encoding the K148/ES2-HA gene is 17,190 bp, which follows “the rule of six” (35).

The expression of ES2-HA protein was evaluated using Western blot analysis after the concentration and purification of rK148/ES2-HA. The expression of ES2-HA protein in rK148/ES2-HA was demonstrated by detecting the ~70 kDa band representing the HA0 protein, the ~55 kDa band representing the HA1 protein, and the ~30 kDa band representing the HA2 protein. The HN proteins of rK148 and rK148/ES2-HA were detected in the ~70 kDa band (Figure 2B).

The titer of rK148/ES2-HA (10^9.5 EID₅₀/mL) was slightly lower than that of the K148/08 backbone virus (10^9.7 EID₅₀/mL) after 72 h of incubation 10-day-old embryonated SPF chicken eggs. To compare the viral growth of rK148/ES2-HA and K148/08 NDV, four serial dilutions (10¹-10⁴ EID50) were inoculated in 10-day-old embryonated SPF chicken eggs and the viral titers were measured in HAU every 12 h after the inoculation. The rK148/ES2-HA showed the highest mean HAU (2^10.67) at 72 h of incubation, while the K148/08 showed the highest mean HAU (2¹¹) at 72 h of incubation. No significant differences were observed between growth curves of rK148/ES2-HA and K148/08 NDV (Figure 3).

No antibodies against the ES2 virus were observed in the chickens and ducks before vaccination. The HI antibody titer against HPAIV of ducks immunized via the oculonasal route started to increase at 2 wpv with geometric mean titer (GMT) of 11.31. The HI titer of ducks immunized via the spray method increased 1 week after booster vaccination (GMT: 3.18), and ducks vaccinated via the oculonasal route showed a higher HI titer (GMT: 16) than the spray-vaccinated group (GMT: 3.18) at 4 wpv. The HI titer of the oculonasally vaccinated ducks decreased (GMT: 7.11), whereas the HI titer of the spray-vaccinated ducks increased (GMT: 7.11) at 5 wpv (Figure 4A). Since HI titer is regarded as positive if the inhibition at a serum dilution of 1/16 or more against 4 HAU of antigen, majority of ducks were HI negative against HPAIV after the vaccination, with only 2 and 3 ducks in the oculonasal and the spray group showed positive HI antibody titer against HPAIV at 5 wpv (Figure 4A). For chickens immunized with the rK148/ES2-HA vaccine, the HI titer to the homologous ES2 virus was elicited as early as at 1 wpv with a GMT of 7.1. After booster vaccination at 3 wpv, the HI titers increased with GMT of 17.96. The chickens vaccinated via the oculonasal route showed an increase in HI titer from 4 wpv (GMT:17.96) to 5 wpv (GMT: 22.63). Only at 5 wpv, all chickens in the rK148/ES2-HA group showed positive HI titer (≥4 log₂) against ES2 HPAIV (Figure 4B). Chickens and ducks vaccinated with rK148 or PBS did not show HI antibody titers against H5N6 HPAIV (Figures 4A,B).

Unlike HI antibody titer against HPAIV in ducks, low HI antibody titers against NDV were observed in all groups. The HI titer of the mock vaccinated group diminished after 3 weeks, while other vaccinated ducks showed slight increment in HI titer against NDV after the booster vaccination. The oculonasally vaccinated ducks showed the highest positive rate of 6/6 and GMT of 64 at 4 wpv, spray vaccinated ducks showed the highest positive rate of 5/6 and GMT of 35.93 at 5 wpv, and rK148 vaccinated ducks showed the highest positive rate of 6/6 and GMT of 23.97 at 3 wpv (Figure 4C). In the chicken experiment, preimmunized chickens and mock-vaccinated chickens did not show HI titers against NDV, whereas chickens immunized with rK148/ES2-HA and rK148 showed increases in HI titers after the priming and booster vaccinations (Figure 4D).

To examine the protective efficacy of the rK148/ES2-HA vaccine, we challenged ducks and chickens with 100 BLD₅₀ of ES2HPAI virus at 3 weeks after the booster vaccination. In the duck challenge study, the survival rates of the oculonasally vaccinated group (100%) was significantly higher than those of the mock-vaccinated group (33.3%) (p = 0.0195; Figure 5A). Interestingly, the rK148 vaccinated group and spray vaccinated group showed the same survival rate of 83%. In the chicken challenge study, all chickens in the rK148-vaccinated and mock-vaccinated groups succumbed on day 3 post-challenge, whereas the rK148/ES2-HA vaccinated group showed a 100% survival rate (Figure 5B). In the viral shedding study of challenged ducks, the mean viral titers in oropharyngeal swab during day 2–6 pc of two vaccinated groups were significantly lower than that of rK148 vaccinated group (10^1.49−2.81 EID₅₀/ml) and mock vaccinated group (10^2.11−2.88 EID₅₀/ml) (p < 0.01). Only one out of six ducks in the oculonasally vaccinated or spray-vaccinated group showed virus shedding in the oropharyngeal swab on day 2 pc (10^0.86 EID₅₀/ml and 10^0.56 EID₅₀/ml) with mean viral titers significantly lower than that of the rK148 (10^1.49 EID₅₀/ml) or mock vaccinated group (10^2.11 EID₅₀/ml) (p < 0.01 and p < 0.0001, respectively; Figure 6A). Cloacal viral sheddings of vaccinated ducks were significantly lower than rK148 vaccinated and mock vaccinated ducks at day 4 and 6 pc (p < 0.05). In addition, viral shedding by the rK148-vaccinated and mock-vaccinated groups was observed until day 6 pc, whereas most ducks vaccinated with rK148/ES2-HA did not show viral shedding after day 2 pc (Figure 6). Chickens vaccinated with rK148/ES2-HA via the oculonasal route showed lower viral shedding compared to the rK148-vaccinated and mock-vaccinated control groups (Figure 7).

Ducks vaccinated with rK148/ES2-HA by oculonasal and spray route did not show any clinical sign. Spray vaccinated ducks have higher clinical score due to one dead duck at day 5 pc. In K148/08 vaccinated group, one duck showed prostration at day 3 pc and died at day 4 pc. Three ducks in same group showed ruffled feather and lethargy starting from day 5 and 7 pc. In positive control group, 4 ducks started to showed lethargy, prostration and torticollis 1–2 days before deaths, and two survived ducks showed lethargy and ruffled feather. Chickens vaccinated with rK148/ES2-HA did not showed any clinical sign after the challenge. All chickens in mock vaccinated group and rK148 vaccinated group showed sudden death at day 2 and 3 pc (Supplementary Figure 1).