Temporal Viral Viability Data from Avian Influenza A Viruses Maintained in North American Wetlands Under Experimental and Environmental Conditions

Sample collection from wild birds: Swab samples for viral screening were collected by applying a polyester-tipped applicator to the cloaca and oropharynx of hunter harvested or live captured ducks. All swabs were placed in cryovials with chilled filtered surface water from wetlands near where duck samples were collected. Two aliquots were made from each sample. One aliquot (A) was shipped to the University of Georgia for IAV screening. The other aliquot from each sample (B) was placed in mesh bags weighted with rocks and sealed in steel drums in approximately one meter of water.

Sample IAV screening: All swab samples were initially screened at the University of Georgia using real-time reverse transcriptase polymerase chain reaction (rRT-PCR) targeting gene products for the influenza A virus matrix gene (Spackman et al. 2002). All samples were also inoculated into specific pathogen-free embryonated chicken eggs for virus isolation (Fouchier el al. 2000). Amnio-allantoic fluids were collected after 120 hours and tested via hemagglutination assay. RNAs extracted from hemagglutination assay positive amnio-allantoic fluids were tested for the IAV matrix gene by rRT-PCR and resultant isolates were genomically sequenced at the Alaska Science Center using next-generation sequencing technology (Illumina platform, Illumina Inc., San Diego, CA).

Wetland water temperature: During September–November 2018, a perforated steel drum containing a Hobo® TidbiT® temperature logger (Onset Computer Corp; Bourne, Massachusetts) was anchored at a depth of approximately 0.5–1.0 m, as measured from substrate to the top of the water column, within each of the six field sites. The temperature loggers were programed to record temperature in Celsius every hour on the hour for the duration of the study. Once retrieved (March–April 2019) the data was uploaded via to a mobile device and downloaded as a comma-separated values file.

Wetland water collection: To determine pH and specific conductance of water at each field site, 250 ml water samples were collected from each wetland in air-tight acid-washed plastic bottles. These samples were sent immediately on ice packs to the U.S. Geological Survey California Water Science Center (CAWSC) for testing. In addition to untreated samples, surface water was collected from each field site and pre-filtered through 100 micron and 1 micron pour-through filters (Duda Energy LLC; Decatur, Alabama) and final filtered through a 0.22 µm vacuum filter (Corning®; Corning, New York). Duplicate 250 ml samples of filtered water originating from each site were collected in air-tight acid-washed plastic bottles for pH and conductivity determinations with one sample sent immediately to the lab and the duplicate sample submerged within the steel perforated drum until the termination of the experiment. Upon retrieval of the filtered water sample submerged in the drum, a fresh unfiltered sample was collected at each wetland. At this time both fresh unfiltered samples and the filtered samples that were submerged were immediately shipped on ice packs to the CAWSC for testing.

Wetland water measurements: Measurement of pH, and specific conductance was taken at the CAWSC. An Orion model 250A pH meter with automatic temperature compensation (accuracy +/- 0.02) was used to measure pH of water samples. A Fisherbrand™ Traceable™ Conductivity Meter model 15-077-977, which automatically corrects temperature to 25 ºC (accuracy: ±1% of full scale +1 digit), was used to measure specific conductance (SC) of water samples. The devices were calibrated for pH and SC before use on the same day measurements were recorded.

Lab experiment: To compare results for our field-based experiment to a laboratory-controlled approach, remaining inoculated water from all initial (A) samples from Agassiz NWR, Minnesota that tested positive for IAVs via rRT-PCR or virus isolation were maintained in their original tube in a tray of cold water at 4º C. Approximately 1 ml of water from each sample was removed and tested by virus isolation in embryonated chicken eggs monthly through April 2019. Samples which were isolation negative in two successive months were considered to no longer contain viable IAV and not subjected to further testing.

Challenge: To evaluate the infectivity of IAVs in filtered surface water after a time period during which samples were held within wetlands, we experimentally inoculated 10 groups of three mallards (Anas platyrhynchos) with virus isolation positive field samples (i.e., replicate samples of filtered surface water inoculated with CL/OP swabs) or filtered water from the Farmes Pool field site at Agassiz NWR, Minnesota inoculated with A/mallard/Minnesota/Sg-00169/2007 (H3N8) to serve as a positive inoculation control. Procedures were performed at the U.S. Department of Agriculture Southeast Poultry Research Laboratory. Mallard ducks were provided at one day of age by a commercial hatchery and held for two weeks at the USNPRC. Birds from each group were inoculated by the intranasal (choanal cleft), intracloacal, oral, and ocular (both eyes) routes with 100, 100, 200, and 50 µl of inoculum, respectively. Ducks were observed daily for clinical signs of disease. Paired CL and OP swabs were collected at 2, 4, 7, 9, and 15 days post inoculation (dpi). Swabs were placed in 1.0 ml of brain heart infusion broth with penicillin (2000 units/ml; Sigma Aldrich, St. Louis, Missouri), gentamicin (200 µg/ml; Sigma Aldrich) and amphotericin B (5 µg/ml; Sigma Aldrich), and stored at -80˚C. At 15 dpi, ducks were bled to evaluate seroconversion, and euthanized. All CL and OP swabs were screened for IAVs using rRT-PCR and samples with rRT-PCR cycle threshold (Ct) values ≤ 38 were subjected to virus isolation in ebryonated chicken eggs. Serum collected prior to inoculation (two weeks of age) and at 15 dpi were tested for antibodies to type A influenza with the Avian Influenza MultiS-Screen blocking enzyme-linked immunosorbent assay (bELISA; IDEXX Laboratories, Westbrook, ME) in accordance with the manufacturer's instructions. A signal-to-noise ratio (i.e., ratio of optical density of the test sample/ optical density of negative control) less than 0.7 was interpreted as a positive assay result.

Total RNA was also extracted from at least one swab and at least one isolate for every unconfirmed virus-isolation positive mallard identified as part of our challenge study using TRIzol (Invitrogen, Carlsbad, CA) and RNA Clean and Concentrator-5 (Zymo Research, Irvine, CA). Swab and isolate samples from the challenge study were selected for sequencing by identifying samples with the lowest rRT-PCR Ct values. Complementary DNA was amplified from virus RNA extracts, visualized, and purified using methods reported in Reeves et al. (2018) at the USGS Alaska Science Center. Using a Nextera XT DNA library preparation kit (Illumina, Inc., San Diego, CA), indexed libraries were pooled and sequenced on an Illumina MiSeq using either 500 or 600-cycle reagent kits with paired-end reads. Sequence reads were assembled using a customized workflow on Geneious R11 (Biomatters Ltd. Auckland, New Zealand) using reference data for IAVs obtained from GenBank (Clark et al. 2016). Duplicate independent extractions of RNA were performed on all isolates identified as mixed infections to confirm sequencing results.