H5-based DNA constructs derived from selected highly pathogenic H5N1 avian influenza virus induce high levels of humoral antibodies in Muscovy ducks against low pathogenic viruses

Abstract

Background: H5 low pathogenic avian influenza virus (LPAIV) infection in domestic ducks is a major problem in duck producing countries. Their silent circulation is an ongoing source of potential highly pathogenic or zoonotic emerging strains. To prevent such events, vaccination of domestic ducks might be attempted but remains challenging. Currently licensed vector vaccines derived from H5N1 HPAIV possess clade 0, clade 2.2 or clade 2.3.4 HA sequences: selection of the best HA candidate inducing the largest cross protection is a key issue. For this purpose, DNA immunization of specific pathogen free Muscovy ducks was performed using different synthetic codon optimized (opt) or native HA genes from H5N2 LPAIV and several H5N1 HPAIV clade 2.1, 2.2.1 and 2.3.4. Humoral cross-immunity was assessed 3 weeks after boost by hemagglutination inhibition (HI) and virus neutralization (VN) against three French H5 LPAIV antigens.

Findings: Vaccination with LP H5N2 HA induced the highest VN antibody titre against the homologous antigen; however, the corresponding HI titre was lower and comparable to HI titres obtained after immunization with opt HA derived from clades 2.3.4 or 2.1. Compared to the other HPAIV-derived constructs, vaccination with clade 2.3.4 opt HA consistently induced the highest antibody titres in HI and VN, when tested against all three H5 LPAIV antigens and H5N2 LPAIV, respectively: differences in titres against this last strain were statistically significant.

Conclusion: The present study provides a standardized method to assess cross-immunity based on HA immunogenicity alone, and suggests that clade 2.3.4-derived recombinant vaccines might be the optimal candidates for further challenge testing to vaccinate domestic Muscovy ducks against H5 LPAIV.

Figure 1

Compared serological responses following immunization with different H5 DNA constructs. HI test (left) and VN test (right) were performed against LP H5N2 A/duck/France/05057b/2005 [20,21]: geometric mean titres (log₂) are displayed with corresponding standard deviations (error bar). A dashed line indicates the positivity threshold (≥4 log₂) in HI test, and the detection threshold (≥3 log₂) in VN test. The number of positive sera in HI test and of detected sera in VN test, over the total number of testable sera, is given above the corresponding bar for each group. For a given histogram, bars with different capital letters are statistically different, using Student-Newman-Keuls test following GLM procedure (SAS version 9.1, SAS Institute Inc., Cary, NC, USA) at risk α = 0.05. For the overall analysis of variance, p-values are mentioned in brackets above each histogram. For statistical analysis, sera with undetected antibody levels were given an arbitrary integer value immediately below the detection threshold of the respective test: respectively equal to 1 log₂ and 2 log₂ for HI and VN tests. *: HA in the DNA construct and in the antigen strain are homologous.

Figure 2

Compared serological responses in HI tests, following immunization with different H5 optimized DNA constructs. HI tests were performed against LP H5N1 A/duck/France/05066b/2005 and LP H5N3 A/Muscovy duck/France/070090b/2007 [20,21]: see Figure 1 for legend. NS indicates overall non statistically significant differences between groups.