Spike Binding and Replication of Kidney-Cell-Adapted Infectious Bronchitis Virus in Chickens

Viral subpopulations in Arkansas Delmarva Poultry Industry (ArkDPI)-derived IBV vaccines are rapidly selected in inoculated chickens, resulting in tracheal damage and virus persistence. These subpopulations with distinct spike (S) amino acid sequences were previously eliminated from an ArkDPI-derived vaccine strain during serial passage in chicken embryonic kidney (CEK) cells, and a new viral population with an S protein harboring three amino acid differences relative to the original vaccine major population became predominant. Recombinant S protein representing the CEK-adapted vaccine strain showed severely reduced or no detectable binding to CEK cells. The two amino acid substitutions in the S1 subunit and the single amino acid substitution in the S2 subunit independently reduced binding. Structural models predicted that each of the amino acid substitutions involved an amino acid located at the interface between monomers in the trimer, suggesting that the conformation and/or stability of alternative conformations of the trimer might be affected. Recombinant S protein representing the CEK-adapted virus also exhibited reduced or nondetectable binding to most relevant chicken tissues in vitro. We compared replication after ocular inoculation in 1-day-old chickens between our CEK-adapted ArkDPI virus and a commercial ArkDPI-derived vaccine. Viral replication in individual chickens was monitored by relative levels of viral RNA in tears 3, 5, and 8 days postvaccination (DPV) and in choanal and tracheal swabs 5 and 8 DPV, determined by reverse transcription quantitative polymerase chain reaction. At 3 DPV, tears of chickens inoculated with 10⁵ 50% embryo infectious doses (EID₅₀) of CEK-adapted vaccine and 10⁴ EID₅₀ commercial ArkDPI vaccine had similar viral loads, while those of chickens given 10⁴ EID₅₀ CEK-adapted vaccine were two orders of magnitude lower. Viral loads in tears of both chicken groups given the CEK-adapted vaccine remained at levels similar to 3 DPV through days 5 and 8 DPV, while those in chickens given the commercial vaccine increased three orders of magnitude during the same period. Statistically significant differences among groups were also observed in viral loads in choanal and tracheal swabs at both 5 and 8 DPV, though less pronounced. The sharp increase in viral load in tears of chickens given the commercial vaccine was likely due to efficient replication of vaccine subpopulations selected in chickens, which are lacking in the CEK-adapted vaccine. Selection of vaccine subpopulations in chickens given the commercial vaccine, but not in chickens given the CEK-adapted vaccine, was verified by S1 gene sequencing of the IBV genomes in samples from individual chickens in all vaccinated groups. Protection following ocular challenge at 19 DPV with virulent Ark IBV, as assessed by reductions in viral loads in tears, correlated with vaccine viral loads at 5 and 8 DPV, with the greatest correlation (r = 0.52) with vaccine viral load in choanal swabs at 8 DPV.