A method for determining avian influenza virus hemagglutinin and neuraminidase subtype association

Methods for grouping specific avian influenza virus (AIV) hemagglutinin (HA) and neuraminidase (NA) subtype reverse-transcription polymerase chain reaction (RT-PCR) products into HA:NA subtypes when egg incubation is technically not feasible were evaluated. These approaches were adopted for use as post hoc methods after melt curve analysis. The methods are based on ratios obtained from amplicon copy count and amplicon molarity and were founded on the premise that infectious particles contain an equal copy count of single-stranded ribonucleic acid segments that encode HA or NA, and thus subtype-specific amplicons from a single AIV isolate should yield a theoretical HA:NA ratio of 1. Single and mixed HA:NA AIV subtype samples were evaluated to determine whether the calculated HA:NA ratios would approach the theoretical value. With these samples, preference was given to the molarity methods to better define and correct for the effects of multiple potential amplicons in the amplification mix. Further, the molarity method was used to evaluate pond sediment spiked with intact virus of known HA:NA subtype to determine whether the method is sufficiently robust to be used with complex samples, such as those acquired from waterfowl habitat. This was a proof-of-concept study intended to guide future methods development. The methods here are not meant to be applied in any other context.

From the analysis of fully characterized isolates of North American AIV, the HA:NA molarity-based ratios were found to be 1.63 ± 0.75 (mean ± standard deviation) when corrected for the difference in amplification strength and the production of multiple amplicons in some reactions using equations developed in this study. Copy count HA:NA ratios, obtained from HA and NA subtype (RT-qPCR), were 1.146 ± 0.124 (mean ± standard deviation) when corrected for amplification efficiency. Correct associations of HA:NA subtype sample composition were made with mixed samples containing 1 HA and 2 NA, and 2 HA and 2 NA. When spiked pond sediment was evaluated, the molar ratio obtained for the H4 and N6 identified in the sample was 1.28 with correction and 1.14 without correction.