Recombinant parainfluenza virus 5 expressing clade 2.3.4.4b H5 hemagglutinin protein confers broad protection against H5Ny influenza viruses

ABSTRACT

The global circulation of clade 2.3.4.4b H5Ny highly pathogenic avian influenza viruses (HPAIVs) in poultry and wild birds, increasing mammal infections, continues to pose a public health threat and may even form a pandemic. An efficacious vaccine against H5Ny HPAIVs is crucial for emergency use and pandemic preparedness. In this study, we developed a parainfluenza virus 5 (PIV5)-based vaccine candidate expressing hemagglutinin (HA) protein of clade 2.3.4.4b H5 HPAIV, termed rPIV5-H5, and evaluated its safety and efficacy in mice and ferrets. Our results demonstrated that intranasal immunization with a single dose of rPIV5-H5 could stimulate H5-specific antibody responses, moreover, a prime-boost regimen using rPIV5-H5 stimulated robust humoral, cellular, and mucosal immune responses in mice. Challenge study showed that rPIV5-H5 prime-boost regimen provided sterile immunity against lethal clade 2.3.4.4b H5N1 virus infection in mice and ferrets. Notably, rPIV5-H5 prime-boost regimen provided protection in mice against challenge with lethal doses of heterologous clades 2.2, 2.3.2, and 2.3.4 H5N1, and clade 2.3.4.4h H5N6 viruses. These results revealed that rPIV5-H5 can elicit protective immunity against a diverse clade of highly pathogenic H5Ny virus infection in mammals, highlighting the potential of rPIV5-H5 as a pan-H5 influenza vaccine candidate for emergency use.

IMPORTANCE

Clade 2.3.4.4b H5Ny highly pathogenic avian influenza viruses (HPAIVs) have been widely circulating in wild birds and domestic poultry all over the world, leading to infections in mammals, including humans. Here, we developed a recombinant PIV5-vectored vaccine candidate expressing the HA protein of clade 2.3.4.4b H5 virus. Intranasal immunization with rPIV5-H5 in mice induced airway mucosal IgA responses, high levels of antibodies, and robust T-cell responses. Importantly, rPIV5-H5 conferred complete protection in mice and ferrets against clade 2.3.4.4b H5N1 virus challenge, the protective immunity was extended against heterologous H5Ny viruses. Taken together, our data demonstrate that rPIV5-H5 is a promising vaccine candidate against diverse H5Ny influenza viruses in mammals.

INTRODUCTION

H5Ny highly pathogenic avian influenza viruses (HPAIVs) bearing the clade 2.3.4.4b HA gene caused unprecedented damage to both domestic and wild birds worldwide (1). Since 2020, the increasing number of H5Ny influenza virus detections among multiple mammalian species, such as ferrets, cats, and sea lions (2), which are biologically closer to humans, raises concern that the virus might adapt to infect humans more easily. Notably, between 2022 and 2023, nine cases of human infection with clade 2.3.4.4b H5N1 viruses have been identified in Europe, Asia, and the Americas. Five of these cases resulted in severe/critical lower respiratory tract disease with one fatality (https://www.cdc.gov). Previous studies have revealed that a genetically modified H5N1 virus could acquire mammalian adaptive mutations during serial passage in ferrets and ultimately became airborne transmissible among ferrets (3–5). In natural conditions, the occurrence of H5N1 outbreaks in mink farms in Spain (6) and mass sea lion deaths in Peru (7) underline that H5 HPAIVs may transmit among mammals, and some mammals may act as mixing vessels for influenza viruses (8, 9), leading to the emergence of novel viruses that could be more harmful to animals and humans. All these indicate that the threat of clade 2.3.4.4b H5Ny viruses pose in human public health is increasing. Action must be taken to prepare for outbreaks or even the pandemic of H5 subtype viruses.

Vaccination is a key measure to prevent emerging infectious diseases and plays a significant role in combating pandemics. Although inactivated influenza vaccines, the most widely used vaccines, demonstrated safety and efficacy, they fall short in preventing infection and transmission due to the blood-olfactory barrier that keeps serum antibodies from assessing the olfactory mucosa (10). Consequently, the pervasive use of inactivated vaccines may inadvertently contribute to the generation of immune-evasive viral mutants. Therefore, there is an urgent demand for vaccines, particularly for respiratory diseases, that not only ensure protection but also prevent viral transmission. Ideally, a pan-H5 vaccine that can elicit mucosal immunity to provide cross-immunoprotection for different clade viruses is optimal. Previous studies have shown that PIV5 is a superior viral vector for developing vaccines against respiratory pathogens including influenza virus (11–17), SASR-CoV-2 (18), MERS-CoV (19), respiratory syncytial virus (20–22), and Mycobacterium tuberculosis (23). PIV5-based vaccines have several advantages, can be administered intranasally (i.n.) or intramuscularly, and prior exposure to PIV5 does not compromise the immune response of a PIV5-vectored vaccine (24). In this study, we generated a PIV5-based vaccine candidate expressing HA protein of the current prevailing clade 2.3.4.4b H5 HPAIV and evaluated its safety and efficacy in mammals.

RESULTS

Generation of recombinant PIV5 expressing HA protein of clade 2.3.4.4b H5 HPAIV

Clade 2.3.4.4b H5N8 virus A/Astrakhan/3212/2020, responsible for human infections in Russia was selected as the vaccine candidate (https://www.who.int). For biosafety, the polybasic cleavage site PLREKRRKR in HA, as a marker of pathogenicity potential, was mutated to PLRETR. The mutated HA gene was inserted into a full-length cDNA clone of the PIV5 genome to replace the small hydrophobic (SH) gene. The rPIV5 expressing H5 HA gene was successfully rescued and was designated as rPIV5-H5 (Fig. 1A). The expression of HA protein in rPIV5-H5-infected cells was verified by immunofluorescence (IF) and western blotting (WB) (Fig. 1B and C). To examine whether HA could be incorporated into rPIV5 virions, we purified PIV5-WT and rPIV5-H5 virions (Fig. 1D) and visualized by immunoelectron microscopy (IEM). Influenza HA protein was detected in the surface of rPIV5-H5 virions and PIV5 phosphoprotein (P) protein was detected in both PIV5-WT and rPIV5-H5 virions (Fig. 1E). Taken together, rPIV5-H5 was successfully rescued and HA protein was efficiently expressed in recombinant virions.

Fig 1.

Generation and characterization of rPIV5-H5. (A) Schematic of rPIV5-H5. The HA gene with polybasic amino acid residues-PLREKRRKR-at the HA cleavage site replaced with-PLRETR-was inserted into the PIV5 genome to replace the SH gene to generate rPIV5-H5. (B) Immunostaining of MDBK cells infected with 0.1 MOI rPIV5-H5 and PIV5-WT or mock infected for 48 h using DAPI and antibodies specific for influenza HA (green) and PIV5-P (red). Scale bar = 100 µm. (C) Immunoblotting of MDBK cells infected with rPIV5-H5 and PIV5-WT or mock-infected and incubated with antibodies against influenza HA and PIV5-P. (D) Negative staining of purified PIV5-WT and rPIV5-H5 virions. rPIV5-H5 and PIV5-WT were purified by density gradient ultracentrifugation, and examined using an electron microscope. (E) Incorporation of HA into purified rPIV5-H5 virions. Purified rPIV5-H5 and PIV5-WT were stained with anti-HA monoclonal antibody or PIV5 P-specific antibody, followed by IEM. Red arrows indicate the influenza HA protein or PIV5 P protein in virions. Scale bar = 100 nm. (F) Detection of the HA gene during serial passage. To evaluate the genetic stability of rPIV5-H5, MDBK cells were infected with rPIV5-H5 at MOI = 0.1 for 10 serial passages. Viruses were collected for detection of HA gene in the viral genome. (G) WB analysis of HA protein expression during serial passage. Cell lysates were collected at the completion of each passage and subjected to WB analysis. The passage numbers are indicated above each lane. (H) Multistep growth curves of PIV5-WT and rPIV5-H5. MDBK cells were infected with PIV5-WT or rPIV5-H5 at MOI = 0.1 and 0.01. Supernatants were collected at 24 h intervals for 5 days and titrated by TCID₅₀ on MDBK cells. Data are presented as mean ± SD. Statistical significance was based on two-way ANOVA (ns, not significant, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001).

Genetic stability of rPIV5-H5

We evaluated the genetic stability of the inserted HA gene in rPIV5-H5 via serial passage in Madin-Darby bovine kidney (MDBK) cells for a total of 10 passages. The HA gene in rPIV5-H5 remained considerably stable through 10 serial passages, with no detectable mutations in HA gene at P10 rPIV5-H5 (Fig.1F). Additionally, HA expression was confirmed by WB through 10 passages (Fig.1G). These results demonstrate the genetic stability of HA gene insertion and the stable expression of HA protein.

In vitro and in vivo characterization of rPIV5-H5

To test whether the insertion of H5 HA gene affected PIV5 replication, rPIV5-H5 replication was determined in vitro and in vivo. To evaluate viral replication efficiency, multistep growth curve in MDBK cells was determined. As shown in Fig. 1H, for different infection doses (multiplicities of infection, MOIs = 0.1 and 0.01), the replication of rPIV5-H5 was delayed compared to that of PIV5-WT, indicating that insertion of the HA gene delayed virus replication in vitro.

To evaluate the replication and safety of rPIV5-H5 in vivo, groups of BALB/c mice were inoculated i.n. with 10⁶ 50% tissue culture infectious dose (TCID₅₀) of rPIV5-H5, PIV5-WT, or phosphate-buffered saline (PBS). The weight loss and mortality of mice in each group were monitored daily for 2 weeks. Mice inoculated with rPIV5-H5, PIV5-WT, or PBS did not cause any detectable clinical signs or weight loss during the 14-day period (Fig. 2A) and all mice survived (Fig. 2B). To monitor virus production, three mice per group were euthanized at 2, 4, 6, and 8 days post-immunization, turbinate and lung samples were collected for virus titration and histopathology assay. As shown in Fig. 2C and D, PIV5-WT and rPIV5-H5 showed similar replication levels in the turbinates and lungs at 2, 4, and 6 days post-immunization (P > 0.05). No histopathological findings were detected in turbinate and lung tissues of all groups of mice (Fig. 2E). Immunohistochemical (IHC) staining revealed that abundant PIV5 antigen-positive nasal epithelial cells and alveolar epithelial cells from PIV5-WT and rPIV5-H5 virus-infected mice (Fig. 2F). In summary, the insertion of HA gene did not alter the replication of rPIV5-H5 in vivo, and rPIV5-H5 showed safety profile in mice.

Fig 2.

Replication and safety of PIV5-WT and rPIV5-H5 in mice. BALB/c mice were i.n. inoculated with PBS or 10⁶ TCID₅₀ of rPIV5-H5 or PIV5-WT in a 50-µL volume. (A) Body weight changes and (B) survival rates of mice inoculated with PBS, rPIV5-H5, or PIV5-WT. The body weights of the mice were plotted as percentages of the weight relative to pre-inoculation weights, and survival rates were monitored and recorded daily for 14 days. (C and D) Replication of rPIV5-H5 and PIV5-WT in mice. BALB/c mice were i.n. infected with 10⁶ TCID₅₀ of PIV5-WT or rPIV5-H5. Three mice per group were sacrificed on 2, 4, 6, and 8 days post-immunization. Viral titers in the turbinates and lungs were determined by TCID₅₀ assay. The detection limit (1.75 log₁₀ (TCID₅₀ per mL)) is shown by the dashed line. (E) H&E staining of turbinate and lung sections of mice inoculated with either sterile PBS, rPIV5-H5, or PIV5-WT. (F) Detection of PIV5 P protein in turbinates and lungs (indicated by arrows) by IHC. Data are presented as mean ± SD. Statistical significance was determined by two-way ANOVA (ns, not significant).

rPIV5-H5 induces anti-H5 humoral, mucosal antibodies, and cellular immune responses in mice

To evaluate the immunogenicity of rPIV5-H5, BALB/c mice were i.n. immunized with 10⁶ TCID₅₀ rPIV5-H5 using either a prime or prime-boost immunization regimen with a 3-week interval, mice immunized with 10⁶ TCID₅₀ PIV5-WT were served as controls (Fig. 3A). Sera samples were collected at specific time points to measure clade 2.3.4.4b H5N1 HA-specific or PIV5-specific antibodies. PIV5-specific antibody titers showed no significant difference between rPIV5-H5 and PIV5-WT groups (P > 0.05) (Fig. 3B and C). H5N1-specific IgG responses were detected as early as 2 weeks after rPIV5-H5 prime immunization. Significantly, rPIV5-H5 prime-boost groups elicited almost 100-fold higher titers of H5N1-specific IgG compared to rPIV5-H5 prime groups after 3 weeks post-boost immunization (Fig. 3B and C). Consistently, rPIV5-H5 can elicit H5N1-specific neutralizing antibody up to endpoint titer of 160 after 3 weeks post boost immunization, which was significantly higher than that induced by rPIV5-H5 prime immunization (NAb titers: 40) (P < 0.0001) (Fig. 3D). ELISpot assays of splenocytes were used to evaluate the antigen-specific cellular immune response in mice after 3 weeks post-prime or prime-boost immunization. Compared to PIV5-WT-immunized mice, rPIV5-H5 prime and prime-boost immunization mice induced H5N1-specific T-cell response. IFN-γ-producing cells in prime-boost group mice were significantly higher than that of prime group mice (P < 0.05) (Fig. 3E). Mucosal antibody responses were evaluated by IgA production in bronchoalveolar lavage fluid (BALF) after 3 weeks post-prime and prime-boost immunization. Low levels of anti-H5 IgA titers were detected in BALF of prime group, after prime-boost immunization, IgA titers were approximately three times higher than those in the prime group (P < 0.0001) (Fig. 3F). These results indicate that immunization with rPIV5-H5 elicited a robust anti-H5 humoral, cellular, and local mucosal immune responses in mice, boost immunization could mount stronger immune responses as compared with prime immunization.

Fig 3.

Immunogenicity of rPIV5-H5 in BALB/c mice (A) Experimental overview of mice immunization and challenge studies. Six-week-old BALB/c mice were i.n. immunized with PBS or 10⁶ TCID₅₀ of rPIV5-H5 or PIV5-WT using a prime or prime-boost regimen at a 3-week interval. Weekly serum collections were conducted post-immunization. BALF, and the spleen were collected 3 weeks after both the prime and boost immunization. At day 21 post-prime and prime-boost immunization, the mice were challenged with 10⁶ TCID₅₀ H5N1 virus (A/wild goose/Jiangxi/1228/2022). PIV5-specific (left) and influenza HA-specific (right) IgG titers in BALB/c mice immunized with rPIV5-H5 in prime (B) or prime-boost (C) regimen were determined using ELISA. (D) Neutralization titers in BALB/c mice immunized with rPIV5-H5. (E) Cellular immune responses in mice immunized with rPIV5-H5. Splenocytes were stimulated with inactivated H5N1 virus or PIV5 virus. Results are presented as the mean number of IFN-γ-secreting cells per 10⁶ splenocytes. (F) Analysis of anti-HA IgA titers in the BALF of mice immunized with rPIV5-H5 using ELISA. Dashed lines indicate the lower limit of detection. Data are presented as mean ± SD. Statistical significance was determined by two-way ANOVA (ns, not significant, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001).

rPIV5-H5 immunization protects against 2.3.4.4b H5N1 influenza virus challenge in mice

To examine the immune efficacy of rPIV5-H5 immunization, rPIV5-H5 immunized mice were i.n. challenged with 10 50% mouse lethal doses (MLD₅₀) of clade 2.3.4.4b H5N1 virus (A/wild goose/Jiangxi/1228/2022) at 21 days post-prime or post-boost immunization, mice immunized with PIV5-WT or PBS were used as controls (Fig. 3A). In response to the lethal H5N1 virus challenge, mice in PBS and PIV5-WT immunized groups exhibited scruffy coat, sluggishness, anorexia, and hunched posture, with a mortality of 100% (Fig. 4A and B). In contrast, no clinical signs and mortality were observed in mice immunized with rPIV5-H5, although transient body weight loss was observed in rPIV5-H5 prime group (Fig. 4A and B).

Fig 4.

Protective efficacy of rPIV5-H5 immunization against clade 2.3.4.4b H5N1 virus challenge in mice. BALB/c mice i.n. immunization with 10⁶ TCID₅₀ of rPIV5-H5 or PIV5-WT using a prime or prime-boost strategy at a 3-week interval. Three weeks post-immunization, the mice were challenged with 10 MLD₅₀ of H5N1 virus (A/wild goose/Jiangxi/1228/2022). Body weight loss (A) and survival rate (B) were monitored and recorded daily for 14 days. Dashed line indicates 75% of initial body weight. (C) Viral loads in tissues were determined at 5 dpi by TCID₅₀ assay. Dashed lines indicate the lower limit of detection. (D) Representative pathological findings in the lungs of mice challenged with H5N1 virus, shown via H&E and IHC staining highlighting influenza NP antigen (brown, pointed by black arrows). (E) Histopathological scores of H5N1-challenged mouse lung tissues from indicated immunization treatments. Lung tissue sections were scored based on pathological changes: 0, no visible lesions; 1, lesion-affected area: <10%; 2, lesion-affected area: <30%, ≥10%; 3, lesion-affected area: <50%, ≥30%; and 4, lesion-affected area: ≥50%. Data are presented as mean ± SD. Statistical significance was determined by two-way ANOVA (ns, not significant, ****P < 0.0001).

To assess virus replication, five mice per group were euthanized at 5 dpi and turbinate, trachea, lung, and spleen were collected for virus titration and histopathology examination. As shown in Fig. 4C, mice in PBS and PIV5-WT immunized groups produced high virus titers from turbinate, trachea, lung, and spleen. However, mice in rPIV5-H5 prime group, low level of virus was detected in the turbinate, trachea, and lung, mice in rPIV5-H5 prime-boost group had no detectable virus in all tissues. Histopathological examination showed that the lungs of PBS and PIV5-WT immunized groups exhibited severe peribronchiolitis and bronchopneumonia, only mild bronchitis was observed in the lungs of the rPIV5-H5 prime group mice. Microscopic lesions of the lung from rPIV5-H5 prime-boost immunization group appeared normal. IHC staining revealed that no influenza NP-positive cells were detected in the lung of prime-boost group mice (Fig. 4D and E). Collectively, these results indicate that single-dose rPIV5-H5 immunization could protect mice against lethal H5N1 virus challenge, and rPIV5-H5 prime-boost immunization could provide sterile immunization against H5N1 virus infection.

rPIV5-H5 immunization protects against clade 2.3.4.4b H5N1 influenza virus challenge in ferrets

Ferrets (Mustela putorius furo) are ideal animal models for studying influenza virus infection as they exhibit clinical symptoms similar to humans (25). Ferrets (n = 3) were i.n. immunized with 10⁷ TCID₅₀ of rPIV5-H5, a booster immunization was performed 21 days later. Ferrets immunized with PIV5-WT or PBS were served as controls (Fig. 5A). No clinical signs and body weight loss were observed in all ferrets after immunization. After 3 weeks post-boost immunization, hemagglutination inhibition (HAI) (HAI titers: 160) and neutralization antibody (Nab titers: 40) against clade 2.3.4.4b H5N1 virus were detected in rPIV5-H5 immunized ferrets (Fig. 5B and C). Next, the protective efficacy of rPIV5-H5 immunization was assessed by i.n. challenged with 10⁶ TCID₅₀ H5N1 virus (A/wild goose/Jiangxi/1228/2022) at 21 days post-boost immunization. Ferrets immunized with PIV5-WT or PBS exhibited severe clinical signs, including sneezing, wheezing and coughing, while rPIV5-H5 immunized ferrets showed no clinical signs. Nasal washes were collected at 2, 4, and 6 dpi and ferrets were euthanized at 6 dpi and tissues including turbinate, trachea, and lung were collected for viral load detection and pathological analysis. High virus titers were detected in nasal washes, turbinate, trachea, and lung tissues of ferrets in the PIV5-WT and PBS groups, while no virus was detected rPIV5-H5 immunized ferrets (Fig. 5D and E). Histopathology, ferrets in the PIV5-WT and PBS groups exhibited severe bronchopneumonitis and interstitial pneumonia. The lungs of ferrets in the rPIV5-H5 immunization group appeared normal and no influenza NP-positive cells were detected in the lung tissue (Fig. 5F and G). In summary, these results showed that rPIV5-H5 immunization could effectively prevent the infection of clade 2.3.4.4b H5N1 virus in ferrets.

Fig 5.

Protective efficacy of rPIV5-H5 immunization against clade 2.3.4.4b H5N1 virus challenge in ferrets. (A) Schematic diagram of ferret immunization and challenge studies. Ferrets were i.n. inoculated with PBS, 10⁷ TCID₅₀ of rPIV5-H5 or rPIV5-WT in a prime-boost regimen at a 3-week interval, serum samples were collected for antibody analysis. (B) Anti-HA and anti-PIV5 antibodies in ferret serum were evaluated via HAI assay. (C) Neutralizing antibodies in serum against H5N1 and PIV5 were determined using MN assay. At day 42 post-prime immunization, the ferrets were i.n. challenged with 10⁶ TCID₅₀ of H5N1 virus (A/wild goose/Jiangxi/1228/2022). (D) Viral loads in nasal wash at 2, 4, and 6 dpi were determined using TCID₅₀ assay on MDCK cells. (E) All ferrets were euthanized at 6 dpi, and tissues including the turbinate, trachea, and all lung lobes were harvested for virus titration on MDCK cells. (F) Representative images of H&E and IHC stained lung sections of ferrets. Black arrows indicate influenza viral NP antigen. (G) Histopathological scores of the H5N1-challenged ferret lungs from indicated immunization treatments. Lung tissue sections were scored based on pathological changes: 0, no visible lesions; 1, lesion-affected area: <10%; 2, lesion-affected area: <30%, ≥10%; 3, lesion-affected area: <50%, ≥30%; and 4, lesion-affected area: ≥50%. Dashed lines indicate the lower limit of detection. Data are presented as mean ± SD. Statistical significance was determined by two-way ANOVA (ns, not significant, **P < 0.01, ***P < 0.001, ****P < 0.0001).

rPIV5-H5 immunization provides cross-protection in mice against lethal challenges with heterologous viruses

Due to the rapid antigenic drift of influenza viruses, it is imperative for vaccine candidates to confer broad protection against heterologous viruses. Since the first cases of H5N1 infection in humans were reported in 1997 (26), multiple clades of H5Ny viruses have been reported to cause human infections (27–29). As a result of genetic drift, there exists considerable antigenic diversity in the HA protein among H5Ny viruses. First, we determined the breadth of rPIV5-H5-induced immunity in ferrets by testing the antibody response against an extended panel of heterologous viruses (clades 2.2, 2.3.2, 2.3.4 H5N1, and clade 2.3.4.4h H5N6). Serum antibodies in ferret at 3 weeks post boost immunization were determined by HAI and micro-neutralization (MN) assays. As shown in Fig. 6A and B, antibodies from rPIV5-H5 immunized ferrets showed a significant increase in HAI and MN titer against clades 2.2, 2.3.2, 2.3.4 H5N1 and clade 2.3.4.4h H5N6 viruses (P < 0.01). Then, we determined cross-reactive antibodies in mice against heterologous H5Ny viruses. Cross-protective antibodies could be detected in rPIV5-H5 immunized mice serum, suggesting that rPIV5-H5 induced cross-reactive antibodies against H5 viruses from different clades (Fig. 6C).

Fig 6.

Cross-protection of rPIV5-H5 against heterologous H5 influenza viruses. Ferrets were i.n. inoculation with PBS or 10⁷ TCID₅₀ of rPIV5-H5 or rPIV5-WT in a prime-boost regimen at a 3-week interval, and serum samples were collected at 3 weeks post-boost immunization. HA-specific antibodies in ferret serum against clade 2.2 H5N1, clade 2.3.2 H5N1, clade 2.3.4 H5N1, and clade 2.3.4.4h H5N6 viruses were determined using HAI assay (A) and MN assay (B). BALB/c mice were i.n. inoculation with 10⁶ TCID₅₀ of rPIV5-H5 or PBS in a prime-boost regimen at a 3-week interval. Sera were collected at 3 weeks post-boost immunization for detection of HA-specific IgG antibodies (C). Groups of mice were i.n. challenged with five MLD₅₀ of H5Ny viruses from different clades. Body weight loss (D) and survival rates (E) were monitored and recorded daily for 14 days. Dashed line indicates 75% of initial body weight. (F) Turbinates and lungs of mice were collected for virus titration on MDCK cells. Dashed lines indicate the lower limit of detection. Representative images of H&E (G) and IHC (H) stained lung sections of mice. Black arrows indicate influenza viral NP antigen. Data are presented as mean ± SD. Statistical significance was determined by two-way ANOVA (ns, not significant, *P < 0.05, **P < 0.01, ****P < 0.0001).

Subsequently, rPIV5-H5 immunized mice were i.n. challenged with 5 MLD₅₀ heterologous H5Ny viruses at 21 days post-boost immunization, mice immunized with PBS were used as controls. In response to the lethal H5Ny virus challenge, mice in PBS-immunized groups with a mortality of 100%. In contrast, only a slight decrease in body weight was observed in rPIV5-H5 immunized mice (Fig. 6D and E). In all virus challenge groups, high viral loads were detected in the turbinate and lung tissues of PBS-immunized mice. While no virus was detected in the rPIV5-H5 group, except low level of viral loads was detected in lung tissues challenged with clade 2.3.4 H5N1 virus (Fig. 6F). Histopathology, the lungs of rPIV5-H5 group mice showed mild bronchitis, while those of control mice exhibited severe peribronchiolitis and bronchopneumonia (Fig. 6G). IHC results showed that more abundant influenza NP-positive cells were detected in lung tissues of the PBS group mice (Fig. 6H). Collectively, immunization with rPIV5-H5 protected mice from death, reduced clinical symptoms and virus shedding, and attenuated lung injury, highlighting the potential of rPIV5-H5 as a pan-H5 influenza vaccine candidate.

DISCUSSION

Since 2020, clade 2.3.4.4b H5N8 HPAIVs has led to an unprecedented outbreak in wild birds and poultry in Africa, Asia, and Europe (30). Subsequently, clade 2.3.4.4b H5N1 virus spread to North, Central, and South America (31). There is a recent paradigm change in the ecology and epidemiology of clade 2.3.4.4b H5Ny HPAIVs which has heightened global concern as the disease caused unusual epidemics among wild and domestic birds, and alarming rise in mammalian infection. Till now, clade 2.3.4.4b H5N1 virus has been detected in 26 species of mammals (2). Efficient transmission of the H5N1 virus among minks has already occurred, which signified that the virus was highly susceptible to accumulate adaptive mutations in mammals and even acquire human-to-human transmissibility (6). Therefore, it is urgent to develop vaccines for mammals or even humans to control H5Ny HPAIVs.

Many countries control highly pathogenic influenza by culling infected and suspected birds, whereas some countries, including China, have adopted a “cull plus vaccination” strategy. Inactivated vaccine has been used in poultry in China since 2004, and the prevalence of H5Ny HPAIVs in poultry decreased dramatically since then (32). However, H5Ny viruses have still been isolated from domestic poultry that received H5Ny inactivated vaccine immunization (33). The failure of vaccination based on inactivated vaccines can be due in part to the antigenic drift of H5Ny, besides, inactivated vaccines that only induce humoral immune responses. PIV5 vector vaccines can stimulate humoral, cellular, and mucosal immune responses, which provide greater cross-protection and longer-lasting immunity (34). Furthermore, PIV5 vector vaccines can be administered using intranasal (35). In this study, rPIV5 expressing H5 HA gene of clade 2.3.4.4b H5N8 virus rPIV5-H5 was successfully rescued and termed as rPIV5-H5. Safety and genome stability are critical considerations for viral-vectored vaccine, PIV5-based vaccine candidates exhibit distinct advantages. PIV5 virus can infect nearly all mammalian cells by binding to the ubiquitous sialic acid residues on cell surface proteins, which means it can infect a wide range of hosts (36). PIV5 does not produce the typical cytopathic effect that results in host cell damage or death (35). There is currently no evidence to demonstrate that PIV5 infection in humans or animals caused typical symptoms and diseases, suggesting the excellent safety profile of PIV5 as a viral vector. Phan et al. found that insertions of RSV F and G genes were stably maintained in the PIV5 genome and there was no mutation that abolished the expression of foreign genes (37). Our study confirmed that HA gene inserted into PIV5 vector was stable and HA protein expression through 10 passages in MDBK cells. In mice and ferrets, rPIV5-H5 could efficiently replicate in the respiratory tract, and did not cause disease. Taken together, rPIV5-H5 is a safety and stability vectored vaccine candidate that can be used in mammals.

Pre-existing immunity can restrict the replication of vector vaccines thus weaken immune response, such as adenovirus-vectored vaccines (38, 39). PIV5-vectored vaccines were found to be unaffected by pre-existing immunity to PIV5 in canines (24), suggesting that PIV5-vectored vaccines may have the potential to overcome preexisting immunity. In our studies, we found enhanced immune response after the second immunization, indicating that the immune response generated against the PIV5 vector during the primary immunization was not critically interfering with the booster dose, rPIV5-H5 could be immunized repeatedly.

Inactivated influenza virus vaccine effectiveness is low compared to natural infection, and the induced immune response is narrow and short-lived, as it focuses on induction of systemic IgG responses but does not effectively induce T-cell immunity and mucosal IgA responses (40). Cytotoxic T-lymphocytes recognize and eliminate virus-infected cells and thereby contribute to viral clearance and recovery (41). Since circulating neutralizing antibodies may not reach the site of virus infection in the upper and lower respiratory tract, it is crucial for antibodies to be present in the respiratory tract to effectively prevent influenza infection events (10). Our data showed that 3 weeks after boost immunization, IFN-γ-producing T cells were significantly elevated in the spleen of rPIV5-H5 immunized mice and high levels of IgA antibody titers were detected in BALF. Although rPIV5-H5 immunization elicited modest humoral immune response in ferrets, it provided sterile immunity against clade 2.3.4.4b H5N1 virus challenge, hence, T-cell immunity and mucosal immune responses also played an important role in protection.

The continuous circulation of H5Ny viruses has led to the emergence of distinct viral lineages with significant antigenic variations in the surface HA protein (42). The biggest challenge for the vaccination strategy is ensuring that the vaccine matches the circulating virus. Results from immunization of ferrets and mice show that rPIV5-H5 can induce antibodies against clades 2.2, 2.3.2, 2.3.4, and 2.3.4.4h H5Ny viruses, albeit at lower level than clade 2.3.4.4b H5N1 virus. Importantly, rPIV5-H5 immunized mice could protect against lethal heterologous H5Ny viruses challenge, indicating that rPIV5-H5 is subjected to minor antigenic variations.

In conclusion, we developed an rPIV5-H5 expressing clade 2.3.4.4b HA protein, the vaccine candidate was safe and genome stable, and could elicit a robust humoral immune response as well as cellular and mucosal immunity in mammals. We conclude that rPIV5-H5 is an excellent vaccine candidate for combating H5Ny viruses and holds promise for application in both human and diverse animal populations. Our data advocate for the continued development of this highly promising vaccine candidate, positioning it as a critical component of future pandemic preparedness efforts against H5Ny viruses.

MATERIALS AND METHODS

Cells and viruses

BSR (a clone of BHK)-T7/5 (BHK cells stably expressing T7 RNA polymerases), MDBK, and Madin-Darby canine kidney (MDCK) cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM) (Thermo Fisher Scientific Gibco, Beijing, China) containing 10% fetal bovine serum (FBS) and 100 U/mL penicillin and 100 mg/mL streptomycin.

Influenza virus clade 2.3.4.4b H5N1 (A/wild goose/Jiangxi/1228/2022) (GISAID accession no. EPI_ISL_18692934), clade 2.3.4.4h H5N6 (A/duck/Northern China/22/2017) (GISAID accession no. EPI_ISL_697782), clade 2.3.4 H5N1 (A/Anhui/1/2005) (GISAID accession no. EPI_ISL_96870), clade 2.2 H5N1 (A/bar-headed goose/Qinghai/1/2005) (GISAID accession no. EPI_ISL_136423), and clade 2.3.2 H5N1 (A/chicken/HuaBei/0513/2007) (GISAID accession no. EPI_ISL_116930) were previously isolated in our laboratory. All the viruses were propagated in 9-day-old embryonated SPF chicken eggs and titrated in MDCK cells. All experiments with H5 subtype AIVs were performed under BSL-3 containment. PIV5 strain CHNSD-011 was previously isolated and preserved in our laboratory.

Plasmid construction

Whole-genome sequences of PIV5 were obtained through PCR amplification and Sanger sequencing, ensuring coverage of the entire genome with overlapping regions. The entire cDNA was cloned into a transcription vector flanked by the T7 promoter, followed by the hepatitis D virus ribozyme sequence and T7 terminator using overlapping PCR and In-Fusion cloning (Takara Bio Inc., Shiga, Japan), and the full-length cDNA vector was designated as pPIV5. PIV5 nucleoprotein (NP), P, and large (L) genes were amplified from pPIV5 and cloned into the eukaryotic expression vector pcDNA 3.1(+), and designated as pcDNA-NP, pcDNA-P, and pcDNA-L, respectively. The HA gene from H5N8 virus A/Astrakhan/3212/2020 (GISAID accession no. EPI_ISL_1038924) was selected and the polybasic cleavage site (PLREKRRKR) between HA1 and HA2 was mutagenized to amino acid PLRETR using overlapping PCR for biosafety reasons and then inserted to replace the SH open reading frame in the PIV5 genome to construct pPIV5-H5.

Virus rescue

BSR-T7/5 cells were grown overnight to 80% confluence in a six-well plate and co-transfected with 5 µg of pPIV5-H5, 1 µg of pcDNA-NP, 0.4 µg of pcDNA-P, and 1.5 µg of pcDNA-L using 16 µL of JetPRIME (Polyplus-transfection, Strasbourg, France). After 6 h, the culture medium was replaced with DMEM containing 2% FBS, and the cells were cultured at 37°C for another 72 h. Next, supernatants were collected and inoculated onto MDBK cells for blind passage for two to three times. The insertion of the HA gene into rPIV5-H5 was detected by RT-PCR using specific primers (forward, 5′-CAACTCTTGGAACAAGATAAGACAGT-3′; reverse, 5′-CATCTTCTGCAACCATTGTAGTGT-3′) and confirmed by Sanger sequencing. Rescued virus was designated as rPIV5-H5, purified, propagated, and stored at −80°C for further study.

Growth kinetics of viruses

To determine the replication kinetics of the recombinant rPIV5-H5, a multistep growth curve was conducted in MDBK cells. Cells were infected with either PIV5-WT or rPIV5-H5 at an MOI of 0.1 and 0.01. After 1 h of viral absorption, infectious medium was discarded, and the cells were washed with PBS for three times to remove unbound virus. DMEM containing 2% FBS was added, and the cells were incubated at 37°C in a 5% CO₂ atmosphere for 120 h. Supernatants were collected at 24-h intervals for 5 days, and viral titers were determined using TCID₅₀ assay on MDBK cells. Reed and Muench’s method was used to calculate TCID₅₀ (43).

IF assay

MDBK cells in six-well plates infected with PIV5-WT or rPIV5-H5 at MOI = 0.1. At 48 hpi, the cells were washed with PBS and fixed with 4% paraformaldehyde for 20 min, followed by permeabilization with Triton X-100 for another 20 min. After blocking with QuickBlock blocking buffer (Beyotime, Beijing, China) for 10 min, the cells were incubated overnight with primary mouse anti-V5 tag antibody (CUSABIO, CSB-MA000161M0m, Wuhan, China), followed by rabbit anti-HA antibody (Sino Biological, 11048-RM07, Beijing, China). Subsequently, the cells were incubated with secondary antibodies: TRITC-conjugated goat anti-mouse IgG antibody and FITC-conjugated goat anti-rabbit IgG antibody. Fluorescent labeling was visualized under a fluorescence microscope (Olympus, Tokyo, Japan).

WB

MDBK cells were infected with PIV5-WT or rPIV5-H5 for 48 h, then cells were lysed with RIPA lysis buffer. Proteins were separated by SDS-PAGE and then transferred onto a PVDF membrane. After blocking with 5% skim milk in PBS containing 0.05% Tween (PBST), the membranes were incubated with mouse anti-V5 tag antibody (CUSABIO, CSB-MA000161M0m, Wuhan, China) or rabbit anti-HA antibody (Sino Biological, 11048-RM07, Beijing, China) at 4°C overnight and Horseradish Peroxidase (HRP)-conjugated goat anti-mouse IgG (Beyotime, A0192, Beijing, China) or goat anti-rabbit IgG (A0208; Beyotime) at room temperature for 1 h. Bands on the membrane were detected using a Western 451 Lightning chemiluminescence kit (Millipore, WBLUC0500, USA) and visualized using a Tanon 4600SF Chemiluminescent Imaging system (Tanon, Shanghai, China).

IEM

To verify HA incorporation into rPIV5-H5 virions, IEM was conducted as described previously (44). MDBK cells were infected with rPIV5-H5 or PIV5-WT at 0.1 MOI, and supernatants were harvested at 72 hpi. After centrifugation at 10,000 × g for 30 min to remove cell debris, supernatants were collected and subjected to ultracentrifugation at 120,000 × g for 2 h. The precipitates were resuspended in PBS, pooled, layered onto a 30–60% sucrose gradient, and centrifuged at 120,000 × g in an SW40 rotor (Beckman Coulter, Brea, CA, USA) for 70 min at 4°C. The precipitates were then coated onto glow-discharged carbon-coated copper grids and negatively stained with 2% phosphotungstic acid (pH 7.2) for 120 s. The grids were incubated with 1% bovine serum albumin (BSA) and 15% glycine in PBS blocking solution for 1 h at room temperature for immunogold labeling. After blocking, the grids were incubated with rabbit anti-HA antibody (11048-RM07; Sino Biological) or mouse anti-V5 antibody (Invitrogen, Carlsbad, CA, USA) for 1 h. The grids were washed and then incubated with goat anti-rabbit IgG H&L-conjugated gold particles or goat anti-mouse IgG H&L-conjugated gold particles (10 nm gold; Abcam, Beijing, China). Finally, the grids were washed with blocking solution, fixed with 2.5% glutaraldehyde for 5 min, stained with 2% phosphotungstic acid, and examined using an HT-7800 electron microscope (Hitachi, Tokyo, Japan) at 80 kV.

Genetic stability of rPIV5-H5 in vitro

MDBK cells were used to evaluate the genetic stability of the HA gene in the rPIV5-H5 virus for 10 serial passages, MDBK cells in a six-well plate were infected with rPIV5-H5 harvested and each virus passage was titrated and then used to infect a new plate of cells at MOI 0.1. Cells at each passage were subjected to immunoblotting of HA protein. Viral RNAs were also extracted from the supernatants using a RaPure Total RNA Mini Kit (Magen, Beijing, China), reverse transcribed into cDNA and followed by RT-PCR using PIV5-specific primers, PCR products were subjected to Sanger sequencing to examine potential HA mutations during serial passage.

Virus titration of PIV5 viruses

PIV5 titers were determined by TCID₅₀ assay on MDBK cells. Briefly, MDBK cells were seeded in 96-well plates, PIV5 viruses were 10-fold serial diluted, 100 µL from 10⁻¹ to 10⁻⁸ of each sample was inoculated onto MDBK cells, and six replicates per dilution were used. After 48 h, cells were fixed and permeabilized, followed by incubation with 30 µL mouse anti-V5 tag antibody (CUSABIO, CSB-MA000161M0m) at 4°C overnight and incubation with 30 µL FITC-conjugated goat anti-mouse IgG as secondary antibody at 37°C for 1 h.

Infection and replication of rPIV5-H5 in mice

Six-week-old BALB/c mice were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. (Vital River, Beijing, China). Groups BALB/c mice were anesthetized with Zoletil and inoculated i.n. with 10⁶ TCID₅₀ of rPIV5-H5, PIV5-WT, or PBS in a volume of 50 µL. At 2, 4, 6, and 8 dpi, three mice per group were euthanized, and nasal turbinates and lungs were harvested to evaluate virus titers. The weight loss and mortality of mice (n = 5) in each group were monitored daily for 2 weeks. For histopathology, mice from each group were euthanized at 5 dpi (n = 2), and nasal turbinates and lungs were analyzed by H&E and IHC.

Immunogenicity and immune efficacy of rPIV5-H5 in mice

Six-week-old BALB/c mice were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. (Vital River, Beijing, China). Following anesthesia with Zoletil, mice were i.n. inoculated with PBS, 10⁶ TCID₅₀ of PIV5-WT, or 10⁶ TCID₅₀ of rPIV5-H5 in a 50-µL volume, respectively. At 3 weeks post-prime immunization, booster immunization was administered to PIV5-WT and rPIV5-H5 groups at same dose. Blood was collected via the retro-orbital route weekly, and sera were prepared for serological analysis. At both 3 weeks post-prime and post-boost immunization, five mice per group were euthanized, BALF and splenocytes were collected for ELISA assay and ELISpot assay, respectively.

At 3 weeks post-last vaccination, mice were i.n. challenged with 10 MLD₅₀ clade 2.3.4.4b H5N1 virus (A/wild goose/Jiangxi/1228/2022) or challenged i.n. with 5 MLD₅₀ clade 2.3.4 H5N1 (A/Anhui/1/2005), clade 2.2 H5N1 (A/bar-headed goose/Qinghai/1/2005), clade 2.3.2 H5N1 (A/chicken/HuaBei/0513/2007), or clade 2.3.4.4h H5N6 (A/duck/Northern China/22/2017) virus. The body weight changes and mortality of mice were recorded daily for 14 days. At 5 dpi, mice were euthanized and tissues (turbinate, trachea, lung, and spleen for homologous challenge) were harvested and homogenized in sterile DMEM containing 1% streptomycin/penicillin. Supernatants were collected and titrated using the TCID₅₀ assay on MDCK cells. A portion of the lungs was collected and subjected for histopathological examination.

Ferret study

Nine 6-month-old Angora ferrets (Angora Ltd, Jiangsu, Chian) were divided into three groups and i.n. inoculated with PBS, 10⁷ TCID₅₀ of PIV5-WT or rPIV5-H5 in a 500-µL volume in a prime-boost regime with a 21-day interval. Clinical signs and body weight loss were observed. At 21 days post-boost immunization, serum samples were collected for HAI assay and neutralization analysis. Subsequently, all ferrets were challenged i.n. with 10⁶ TCID₅₀ of clade 2.3.4.4b H5N1 virus (A/wild goose/Jiangxi/1228/2022) in a 500-µL volume. Nasal wash samples were collected at 2, 4, and 6 dpi for viral detection. At 6 dpi, all ferrets were euthanized, and the lungs, trachea, and nasal turbinate were collected for viral titration and histopathological examination.

ELISA

The binding of serum and BALF antibodies to H5Ny and PIV5 viruses was analyzed using ELISA. Clade 2.3.4.4b H5N1 virus (A/wild goose/Jiangxi/1228/2022) was used as homologous antigen, and clade 2.3.4.4h H5N6 (A/duck/Northern China/22/2017), clade 2.3.4 H5N1 (A/Anhui/1/2005), clade 2.2 H5N1 (A/bar-headed goose/Qinghai/1/2005), and clade 2.3.2 H5N1 (A/chicken/HuaBei/0513/2007) viruses were used as heterologous antigens. In brief, ELISA plates (42592; Corning, Beijing, China) were coated with indicated inactivated virus using pH 9.6 carbonate-bicarbonate buffer (C1055; Solarbio, Beijing, China) at 4°C overnight. The next day, the plates were washed with PBS and blocked using 100 µL of PBS containing 5% skimmed milk. Serum samples were serially diluted twofold, and 100 µL of the diluted solution was loaded. The plates were then incubated at 37°C for 2 h and followed by incubation with HRP-conjugated goat anti-mouse IgG (1:5,000, Beyotime) at 37°C for 1.5 h. After aspiration and washing with PBST, the plates were developed using a TMB two-component substrate solution (for ELISA) (PR1210; Solarbio) following the manufacturer’s instructions, and the OD was read at 450 nm using a Biotek microplate reader (Biotek, Winooski, USA). The endpoint titer was defined as the highest reciprocal dilution of serum to give an absorbance greater than 2.1-fold of the background values.

HAI and MN assays

Antibodies in serum were determined using HAI and MN assay. Clade 2.3.4.4b H5N1 (A/wild goose/Jiangxi/1228/2022), clade 2.3.4.4h H5N6 (A/duck/Northern China/22/2017), clade 2.3.4 H5N1 (A/Anhui/1/2005), clade 2.2 H5N1 (A/bar-headed goose/Qinghai/1/2005), and clade 2.3.2 H5N1 (A/chicken/HuaBei/0513/2007) were used as antigens. The MN and HAI assays were performed as described previously (45, 46).

Histological and IHC assays

Tissues collected from mice or ferrets were fixed overnight with 4% paraformaldehyde, paraffin-embedded, and cut into 3-μm-thick sections. For histological analysis, the sections were stained with H&E using standard protocols (47). For IHC analysis, viral antigens were detected using polyclonal rabbit anti-NP antibody (ab91648; Abcam, Beijing, China) or mouse anti-V5 tag antibody (CUSABIO, CSB-MA000161M0m), followed by visualization using 3,3ʹ-diaminobenzidine tetrahydrochloride staining. The scoring criteria of H&E assay were conducted as previously described (48).

ELISpot assay

To detect T-cell responses in splenocytes against H5N1 HA and PIV5, ELISpot assays were conducted using a mouse IFN-γ-precoated ELISpot kit (2210005; Dakewe, Beijing, China) following the manufacturer’s instructions. In brief, mice i.n. immunized with PIV5-WT or rPIV5-H5 were sacrificed at 3 and 6 weeks post-prime, and splenocytes were prepared by pushing the spleen through a 70-µm cell strainer (352350; Corning, Beijing, China). The cells were cultured with inactivated H5N1 virus (16 HAU per well) or inactivated PIV5 virus (16 HAU per well), Phorbol 12-Myristate 13-Acetate (PMA)/ionomycin, or without stimulation in mouse IFN-γ pre-coated plates at 37°C in 5% CO₂ for 48 h. Splenocytes were removed, and the plates were washed and incubated with biotinylated anti-mouse IFN-γ at 37°C for 1 h. After washing, the plates were incubated with streptavidin-HRP at 37°C for 1 h. AEC solution was then added to the plates, followed by incubation at room temperature for 20 min. Finally, the number of spots was determined using an automatic ELISpot reader. The antibody titer was defined as the highest dilution that gave an absorbance greater than 2.1-fold of the background absorbance of negative control.

Statistical analysis

Graphing and statistical analyses were performed using GraphPad Prism 8 (GraphPad Software, San Diego, CA, USA; www.graphpad.com). Each data point represents the mean ± SD. All data were analyzed by two-way ANOVA, the asterisks represent significant differences between groups (*P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001). ns, no significant difference between groups.

ETHICS APPROVAL

All animal experiments were approved by the Beijing Association for Science and Technology [approval SYXK (Beijing) 2007-0023] and conducted in accordance with the Beijing Laboratory Animal Welfare and Ethics guidelines issued by the Beijing Administration Committee of Laboratory Animals and the China Agricultural University Institutional Animal Care and Use Committee guidelines (SKLAB-B-2010-003).

DATA AVAILABILITY

All study data are included in the manuscript.