Redesign and Validation of a Real-Time RT-PCR to Improve Surveillance for Avian Influenza Viruses of the H9 Subtype

Abstract

Avian influenza viruses of the H9 subtype cause significant losses to poultry production in endemic regions of Asia, Africa and the Middle East and pose a risk to human health. The availability of reliable and updated diagnostic tools for H9 surveillance is thus paramount to ensure the prompt identification of this subtype. The genetic variability of H9 represents a challenge for molecular-based diagnostic methods and was the cause for suboptimal detection and false negatives during routine diagnostic monitoring. Starting from a dataset of sequences related to viruses of different origins and clades (Y439, Y280, G1), a bioinformatics workflow was optimized to extract relevant sequence data preparatory for oligonucleotides design. Analytical and diagnostic performances were assessed according to the OIE standards. To facilitate assay deployment, amplification conditions were optimized with different nucleic extraction systems and amplification kits. Performance of the new real-time RT-PCR was also evaluated in comparison to existing H9-detection methods, highlighting a significant improvement of sensitivity and inclusivity, in particular for G1 viruses. Data obtained suggest that the new assay has the potential to be employed under different settings and geographic areas for a sensitive detection of H9 viruses.

Keywords: H9Nx; avian influenza; molecular diagnosis; real-time RT-PCR; validation.

Figure 1

Performance comparison of different amplification kits. Dots represent individual Ct values. Mean Ct and Spearman’s rank correlation coefficient (r) are reported. ns: not significant.

Figure 2

Diagnostic sensitivity of the pan-H9 rRT-PCR compared to M-gene screening rRT-PCR (Heine et al., 2015) [60] and three existing H9 detection methods [29,48,52]. Dots represent individual Ct values. Negative samples were arbitrarily assigned a value of 45. Mean Ct and Spearman’s rank correlation coefficient (r) are reported. **** p < 0.0001; *** p < 0.0002; ns: not significant.