Chicken ANP32A-independent replication of highly pathogenic avian influenza viruses potentially leads to mammalian adaptation-related amino acid substitutions in viral PB2 and PA proteins

Abstract

Acidic nuclear phosphoprotein 32 family member A (ANP32A) is an important host factor that supports the efficient replication of avian influenza viruses (AIVs). To develop an antiviral strategy against Gs/Gd-lineage H5 highly pathogenic avian influenza (HPAI) viruses in chickens, we established chicken ANP32-knockout (chANP32A-KO) DF-1 cells and evaluated their antiviral efficacy through in vitro validation. The replication of all HPAI viruses tested in chANP32A-KO cells was significantly lower compared to that of wild-type DF-1 cells. However, when HPAI strains A/mountain hawk-eagle/Kumamoto/1/2007 (H5N1; MHE) and A/chicken/Aichi/2/2011 (H5N1; H5Aichi) were passed in chANP32A-KO cells, mutant viruses were generated, which exhibited comparable replication levels in both chANP32A-KO and wild-type DF-1 cells. Sequence analysis revealed that mammalian-adaptive amino acid mutations PB2_D256G and PA_T97I were present in the MHE mutant virus, and the PB2_E627K mutation was identified in the H5Aichi mutant virus. These mutations have also been reported to enhance the polymerase activity of AIVs in mammalian cells; however, the minigenome assay in the present study showed that the polymerase activity of mutant viruses in chANP32A-KO cells was not restored to levels comparable to those in wild-type DF-1 cells. These findings suggest that ANP32A-independent viral replication may induce amino acid substitutions associated with mammalian adaptation in AIVs. They also imply that the high efficiency of viral replication mediated by these amino acid mutations may not result from enhanced polymerase activity but rather involve other undefined mechanisms.IMPORTANCEDuring the host-switching of avian influenza viruses (AIVs) to mammalian hosts, introducing adaptive mutations into viral proteins is essential to ensure optimal functionality through virus-host protein interactions in mammalian cells. However, the mechanisms leading to adaptive mutations in viral proteins remain unclear. Among several host proteins that promote viral growth, acidic nuclear phosphoprotein 32 family member A (ANP32A) is known to be an important factor for efficient viral replication. Here, we generated mutant highly pathogenic avian influenza viruses capable of ANP32A-independent replication in a chicken-derived cell line. We demonstrated that several amino acid mutations found in the mutant viruses correspond to those associated with the mammalian adaptation of AIVs. These results suggest that ANP32A-independent viral replication is one of the mechanisms for introducing amino acid mutations that are reportedly involved in the mammalian adaptation of AIVs.

Keywords: avian viruses; veterinary pathogens; viral replication.

Fig 1

Generation of chANP32A-KO DF-1 cells. (A) Nucleotide sequence analysis of a gene-editing target site. Comparison between wild-type DF-1 and chANP32A-KO cells showed that a cytosine base located four bases upstream of the 5′-side of the PAM sequence (yellow box) was deleted in exon 2 of chANP32A in the genomic DNA of chANP32A-KO cells, resulting in the insertion of a termination codon (*) immediately after amino acid position 29 of chANP32A. (B) Detection of the chANP32A protein in cell lysates from wild-type DF-1 and chANP32A-KO cells by Western blotting using anti-ANP32A-specific polyclonal antibodies. A band with a molecular weight of approximately 32 KDa, corresponding to the size of chANP32A, was observed in the lysate sample from wild-type DF-1 but not in chANP32A-KO cells.

Fig 2

Resistance to HPAI virus infections in chANP32A-KO cells. Cells were seeded in 48-well plates, inoculated with HPAI viruses, and observed for cytopathic effects at 3 days post-inoculation. Viral titers were calculated using the Reed and Muench method and expressed as the mean of three independent replicates. Each bar is marked with the corresponding standard deviation. Statistical analysis of the differences in viral titers between wild-type DF-1 and chANP32A-KO cells was performed using an unpaired t-test. Asterisks indicate statistically significant differences (P < 0.05). NDV, Newcastle disease virus; ARV, avian reovirus.

Fig 3

Virus titers of mutant HPAI virus in wild-type DF-1 and ANP32A-KO cells. Comparison of the viral titers of the mutant MHE (A) and H5Aichi strains (B) in wild-type DF-1 and chANP32A-KO cells. These mutant viruses were generated via serial passage (P1 to P5) in chANP32A-KO cells. After each virus was propagated once in embryonated chicken eggs, the virus titer was determined by a TCID₅₀ assay. Sequence analysis revealed amino acid mutations at positions 256 of the PB2 and 97 of the PA proteins in the MHE-mutant strains (A) and positions 627 of the PB2 and 339 and 561 of the PA proteins in the H5Aichi-mutant strains (B). Statistical analysis of the differences in viral titers between wild-type DF-1 and chANP32A-KO cells was performed using an unpaired t-test. Asterisks denote statistically significant differences (P < 0.05).

Fig 4

Effect of amino acid mutations on viral polymerase activity in HPAI virus mutants generated via serial passage in chANP32A-KO cells. (A) Comparison of the polymerase activities of the MHE, H5Aichi, and Km1-7 strains in wild-type DF-1 and chANP32A-KO cells. PB2, PB1, PA, and NP protein-expressing pCAGGS plasmids, the SEAP gene-expressing pPolIGG plasmid, and the secreted Metridia luciferase protein-expressing plasmid (pMetLuc2-vector) as a transfection control were co-transfected into cells seeded in 12-well plates. After 72 h, the cell supernatants were harvested, and SEAP and Metridia luciferase bioluminescence was measured. Differences in transfection efficiency between wells were corrected using the luciferase expression levels. The polymerase activity of each sample was determined from triplicate test results. Each bar is marked with the corresponding standard deviation. Statistical analysis of the differences in polymerase activity between wild-type DF-1 and chANP32A-KO cells was performed using an unpaired t-test. Asterisks indicate statistically significant differences (P < 0.05). (B) The effect of amino acid mutations found in the PB2 and PA proteins of MHE mutants on viral polymerase activity. The statistical significance of the differences in the polymerase activities between the parental MEH, MHE-P5, and their combinations in DF-1 or chANP32A-KO cells was determined via one-way ANOVA with the Tukey-Kramer test. Asterisks indicate statistically significant differences (P < 0.05). (C) The effect of amino acid mutations found in the PB2 and PA proteins of H5Aichi mutants on viral polymerase activity. (D) The effect of amino acid mutations found in the PB2 and PA proteins of MHE and H5Aichi strains on viral polymerase activity in HEK-293 cells.