Extraction and Detection of Avian Influenza Virus From Wetland Sediment Using Enrichment-Based Targeted Resequencing

• Main Authors: Lauren C. Tindale, Waren Baticados, Jun Duan, Michelle Coombe, Agatha Jassem, Patrick Tang, Miguel Uyaguari-Diaz, Richard Moore, Chelsea Himsworth, William Hsiao, Natalie Prystajecky
• Format: Article
• Language: English
• Published: Frontiers Media S.A. 2020-05-01
• Series: Frontiers in Veterinary Science
• Subjects: avian influenza virus; next generation sequencing; nucleic acid extraction; RT-qPCR; surveillance; sediment
• Online Access: https://www.frontiersin.org/article/10.3389/fvets.2020.00301/full

Early virus detection and characterization is key to successful avian influenza virus (AIV) surveillance for the health of humans as well as domestic poultry. We explored a novel sampling approach and molecular strategy using sediment from wetlands and outdoor waterbodies on poultry farms as a population-level proxy of AIV activity in waterfowls. RNA was extracted using the MoBio RNA PowerSoil Total RNA isolation kit with additional chloroform extraction steps to reduce PCR inhibition. AIV matrix protein (MP) gene was detected in 42/345 (12.2%) samples by RT-qPCR; an additional 64 (18.6%) samples showed evidence of amplification below the threshold and were categorized as “suspect positive.” Enrichment-based targeted resequencing (TR) identified AIV sequences in 79/345 (22.9%) samples. TR probes were designed for MP, hemagglutinin (HA), and neuraminidase (NA), however PB2 and PA were also identified. Although RT-qPCR and TR only had fair-moderate agreement, RT-qPCR positivity was predictive of TR-positivity both when using only strictly positive RT-qPCR samples (OR = 11.29) and when coding suspect positives as positive (OR = 7.56). This indicates that RT-qPCR could be used as a screening tool to select samples for virus characterization by TR and that future studies should consider RT-qPCR suspect positives to be positive samples for subsequent resequencing when avoiding false negatives is the priority, for instance in a diagnostic test, and to consider suspect positives to be negative samples when cost efficiency over a large number of samples is the priority, for instance in a surveillance program. A total of 13 HA (H1-7, H9-13, H16) and 9 NA (N1-9) subtypes were identified, with a maximum of 8 HA and 8 NA subtypes detected in a single sample. The optimized RNA extraction and targeted resequencing methods provided increased virus detection and subtyping characterization that could be implemented in an AIV surveillance system.