Amplicon sequencing of pasteurized retail dairy enables genomic surveillance of H5N1 avian influenza virus in United States cattle

Abstract

Highly pathogenic avian influenza (HPAI) viruses with H5 hemagglutinin (HA) genes (clade 2.3.4.4b) are causing an ongoing panzootic in wild birds. Circulation of these viruses is associated with spillover infections in multiple species of mammals, including a large, unprecedented outbreak in American dairy cattle. Before widespread on-farm testing, there was an unmet need for genomic surveillance. Infected cattle can shed high amounts of HPAI H5N1 viruses in milk, allowing detection in pasteurized retail dairy samples. Over a 2-month sampling period in one Midwestern city, we obtained dairy products processed in 20 different states. Here we demonstrate that a tiled-amplicon sequencing approach produced over 90% genome coverage at greater than 20x depth from 5 of 13 viral RNA positive samples, with higher viral copies corresponding to better sequencing success. The sequences clustered phylogenetically within the rest of the cattle outbreak sequences reported. A combination of RT-qPCR testing and sequencing from retail dairy products can be a useful component of a One Health framework for responding to the avian influenza outbreak in cattle.

Fig 1. Genomic coverage of samples at 20X depth.

Coverage at each base is plotted per-segment, with each color line representing a different carton. The vertical axis is given in log base 10, with a dashed line at log₁₀(20) to indicate our depth cutoff. The sharp drop-offs in coverage depth that dip below the dashed line correspond to amplicons that did not amplify or sequence well. Gene segments are abbreviated as follows: polymerase-basic 2 (PB2), polymerase-basic 1 (PB1), polymerase acidic (PA), hemagglutinin (HA), nucleoprotein (NP), neuraminidase (NA), membrane protein (MP), and nonstructural (NS).

Fig 2. Percent coverage at 20x depth by segment for each sample.

The red colors correspond to higher coverage, while the blue colors correspond to lower coverage. All numbers are expressed as percentages.

Fig 3. Phylogenetic clustering in Nextclade.

The consensus sequences from each dairy product (represented by red diamonds and triangles) are superimposed on the background tree sourced from Nextclade (https://nextstrain.org/avian-flu/h5n1-cattle-outbreak/genome). Samples are colored by their admin division as reported in the sequence metadata, with missing metadata represented by gray color. The x-axis shows divergence, a measure of genetic distance between sequences (approximately number of nucleotide substitutions per nucleotide position).