Mass vaccination with reassortment-impaired live H9N2 avian influenza vaccine

Abstract

Avian influenza poses a severe threat to poultry production and global food security, prompting the development of vaccination programs in numerous countries. Modified live virus (MLV) vaccines, with their potential for mass application, offer a distinct advantage over existing options. However, concerns surrounding reversion, recombination, and unintended transmission have hindered the progress of MLV development for avian influenza in poultry. To address these concerns, we engineered reassortment-impaired, non-transmissible, safe, immunogenic, and protective MLVs through the rearrangement of internal gene segments and additional modifications to the surface gene segments HA and NA. The unique peptide marker aspartic acid-arginine-proline-alanine-valine-isoleucine-alanine-asparragine (DRPAVIAN) was incorporated into HA, while NA was modified to encode the chicken interleukin-18 (ckIL18) gene (MLV-H9N2-IL). In vitro, the MLV-H9N2 and MLV-H9N2-IL candidates demonstrated stability and virus titers comparable to the wild-type H9N2 strain. In chickens, the MLV-H9N2 and MLV-H9N2-IL candidates did not transmit via direct contact. Co-infection studies with wild-type virus confirmed that the altered HA and NA segments exhibited fitness disadvantages and did not reassort. Vaccinated chickens showed no clinical signs upon vaccination, all seroconverted, and the inclusion of ckIL18 in the MLV-H9N2-IL vaccine enhanced neutralizing antibody production. A significant decrease in viral loads post-challenge underscored the protective effect of the MLVs. The MLV-H9N2-IL vaccine, administered via drinking water, proved immunogenic in chickens in a dose-dependent manner, generating protective levels of neutralizing antibodies upon aggressive homologous virus challenge. In summary, this study lays the groundwork for safe MLVs against avian influenza suitable for mass vaccination efforts.

Fig. 1. The MLV-H9N2 and MLV-H9N2-IL vaccine candidates exhibit stability.

a RT-PCR fragment profiles of MLVs before serial passages in eggs (E1) or after serial passages in eggs (E5) targeting the PB1-M2 (top), HA-DRPAVIAN (middle), and NA-IL-18 (bottom). Wild-type segments were included as positive controls to demonstrate the differences in size between modified and non-modified segments. b Chicken IL-18 expression from the MLV-H9N2-IL virus by western blot. Protein lysates of MDCK cells inoculated with MLV-H9N2 (black) or MLV-H9N2-IL (blue) at an MOI of 10 or 1 as indicated. Positive control corresponds to protein lysates of HEK293T cells transfected with the control expression plasmid pCAGGS expressing N2WF10-Furin-2A-IL-18-HIS. Negative control corresponds to non-transfected HEK293T cell lysates. The arrows indicate the predicted molecular weight of chicken IL-18 (~23 KDa) and the host cellular protein glyceraldehyde-3-phosphate dehydrogenase (GAPDH; ~36 KDa) used as a gel loading control. Molecular weight marker (MWM) with numbers on the right in KDa of marker protein bands. Western blot images derived from the same set of protein lysates. Uncropped figures are shown in Supplementary Fig. 3.

Fig. 2. Reassortment data upon co-infections in vitro and in vivo.

MDCK cells were infected with H9N2 WT (WT) virus and either (a) MLV-H9N2 or (b) MLV-H9N2-IL viruses. MDCK cells were coinfected with a 10:1 MLV/WT virus ratio. Supernatants collected at 72 hpi underwent two rounds of limiting dilution followed by NGS sequencing. FASTQ files were used to calculate the proportion of NGS reads that matched MLV or WT for each specific segment. Reassortment data upon co-infection of 2-week-old chickens with H9N2 WT virus and either (c) MLV-H9N2 or (d) MLV-H9N2-IL viruses at a 10MLV:1WT ratio. OP swabs collected at 5 dpi were sequenced by NGS and analyzed as the in vitro data. Data is graphed as % proportion of reads versus gene segment. WT segments (red), MLV-H9N2 segments (black), MLV-H9N2-IL segments (blue) in the progeny viruses are shown. Gray shading rectangles represent unmodified viral gene segments in WT and MLV strains. White shading rectangles represent modified viral gene segments in the MLV strains but unmodified viral gene segments in WT.

Fig. 3. MLVs transmissibility in 2-weeks-old chickens.

Animals (n = 4/group) were directly inoculated with either the MLV-H9N2 (black) or MLV-H9N2-IL (blue) strains at 1 × 10⁶ EID₅₀/chicken. Naïve direct contacts were added 24 hpi. a OP swabs and (b) CL swabs were collected (n = 4/group/day) every day from 1 to 6 days after inoculation or after contact to analyze transmission between directly inoculated and direct contact animals. Viral loads by RT-qPCR shown as the mean ± SD Log₁₀ TCID₅₀ equivalent/mL, LOD 1.199 Log₁₀ TCID₅₀ equivalent/mL. c Blood samples were collected (n = 4/group) at 14 dpi and 14 dpc and sera prepared for HI assays to analyze seroconversion using 2-fold serum dilutions.

Fig. 4. MLVs are safe, generate a strong immune response, and protect chickens upon homologous challenge.

a After virus inoculation, OP and CL swabs were collected (n = 6/group) at 3- and 5-dpp to establish levels of virus shedding shown as the mean ± SD Log₁₀TCID₅₀/mL, LOD 0.699 Log₁₀ TCID₅₀/mL. Virus shedding for MLV-H9N2-IL (blue), MLV-H9N2 (black), H9N2 WT virus (red) and mock inoculated controls (white) are shown. Note that the MLV-H9N2-IL (blue) replicates significantly less than the H9N2 WT virus (red). b–d Blood samples were collected (n = 8/group/timepoint) at 12 dpp and 12 dpb and sera prepared to establish levels of seroconversion from the MLV-H9N2-IL (blue), MLV-H9N2 (black), WIVadj-H9N2 (gray), and mock-vaccinated controls (mock). b HI titers expressed as 2-fold serum dilution versus timepoint of serum collection. c VNluc titers plotted as arbitrary Log₁₀ relative light units (Log₁₀ RLU (AU)) versus the Log₂ sera dilution. Nluc activity was measured at 48 hpi. The Log₂ inhibitory sera dilution 50 (Log₂ ISD50) for each serum is shown. d Levels of NP antibodies as measured by a commercial ELISA and expressed as Signal-to-Noise (S/N) ratio versus timepoint of serum collection. S/N ≥ 0.6 is considered positive (e) Sinuses, trachea, lungs, pancreas, and cloaca samples (n = 4/group) were collected at 3 dpc and subjected to virus titration and shown as the mean ± SD Log₁₀ TCID₅₀/mL. f OP swabs and (g) CL swabs were collected (n = 6/group/day) every other day from 1 to 7 dpc. Virus shedding measured by RT-qPCR and expressed as the mean ± SD Log₁₀ TCID₅₀ equivalent/mL versus day post-challenge with LOD 1.199 Log₁₀ TCID₅₀ equivalent/mL. Data analysis and graphs were prepared using Prism v10 (GraphPad). Ordinary two-way ANOVA was performed to calculate P values followed by Tukey’s multiple comparison tests. Only statistically significant differences are shown.

Fig. 5. MLV-H9N2-IL is stable and immunogenic when given via drinking water.

a Chickens were vaccinated with MLV-H9N2-IL via drinking water for 2 h. Low dose: 10⁴ EID₅₀/mL (light blue). High dose: 10⁶ EID₅₀/mL (dark blue). Mock vaccinated controls included (white). OP and CL swabs were collected (n = 6/group) at 3- and 5- dpp. b MLV-H9N2-IL vaccine stability in drinking water analyzed by incubating the two vaccine doses at 25 °C, 30 °C, 40 °C, and 50 °C. Timepoints were collected at 0 h, 1 h, 2 h, and 24 h. Virus in samples were then titrated and shown as the mean ± SD Log₁₀ TCID₅₀/mL, LOD 0.699 Log₁₀ TCID₅₀/mL. Sera prepared from blood samples collected at 12 dpp (n = 6/group), 26 dpp (n = 4/group), and 12 dpb (n = 4/group) from the low dose (light blue) and high dose (dark blue) MLV-H9N2-IL vaccine groups were used to establish seroconversion. c HI titers (d), VNluc titers, and (e) NP antibody titers were established as detailed in Fig. 4.

Fig. 6. Prime-Boost with 10⁶ EID₅₀/ml of MLV-H9N2-IL induces sterilizing immunity.

a Sinuses, trachea, and cloaca samples (n = 4/group) were collected at 3 dpc and viruses titers established as the mean ± SD Log₁₀ TCID₅₀/gr of tissue homogenate, LOD 0.699 Log₁₀ TCID₅₀/gr. b OP swabs and (c) and CL swabs were collected (n = 4/group/day) every other day from 1 to 5 dpc. Virus titers shown as the mean ± SD Log₁₀ TCID₅₀/mL, LOD 0.699 Log₁₀ TCID₅₀/mL. Data analysis and graphs performed with Prism v10 (GraphPad) using ordinary two-way ANOVA, P values calculated by Tukey’s multiple comparison tests. Only statistically significant differences are shown.