Evaluation of sampling methods for genomic surveillance of SARS-CoV-2 variants in aircraft wastewater: Advancing global early-warning systems for future pandemics

Opeyemi U. Lawal^1,2*Valeria R. Parreira³Fozia Rizvi³Melinda Precious³Rebecca E.V. Anderson³Alyssa K. Overton⁴Jennifer J. Knapp⁴Brittany Maxwell⁵Steven Thomas⁵Marcos Zambrano⁵Chrystal Landgraff⁶Natalie C Knox⁶Manon D. Fleury⁶Trevor Carlos Charles⁴Lawrence Goodridge³

• ¹University of Windsor School of the Environment, Windsor, Canada
• ²University of Windsor Great Lakes Institute for Environmental Research, Windsor, Canada
• ³University of Guelph, Guelph, Canada
• ⁴University of Waterloo, Waterloo, Canada
• ⁵Greater Toronto Airports Authority, Toronto, Canada
• ⁶Public Health Agency of Canada, Ottawa, Canada

Background Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is an ongoing threat to global health. Wastewater-based surveillance (WBS) has proven to be an important tool for tracking the dissemination of SARS-CoV-2 variants of concern (VOCs) in the community. In Canada, metagenomic analysis of aircraft wastewater was adopted at an early stage of the pandemic to track importation of emerging variants into the country. However, the acute need to determine the presence of emerging SARS-CoV-2 sublineages meant that the sampling methods utilized were not adequately validated. Here, we compared two different sampling methods for genomic surveillance of SARS-CoV-2 VOCs in aircraft sewage samples. Methods Eighty-eight composite wastewater samples were collected over nine weeks using both autosampler and passive torpedo samplers at the same location. SARS-CoV-2 nucleic acid in the samples was quantified using RT-qPCR. RNA samples were extracted and sequenced with the MiniSeq system using the tiled-amplicon sequencing approach with ARTIC V4.1 primer sets. Raw reads were preprocessed and SARS-CoV-2 mutations, variants lineages, and other sequence metrics from the two sampling methods were compared. Results The two sampling methods yielded comparable viral load by RT-qPCR, but the autosampler produced higher genome coverage relative to the passive samplers. The Omicron lineages identified differed by sampling method. BQ.1* and BA.5.2*, which were the predominant lineages in wastewater and clinical samples at the time, were identified as dominant in the autosampler and passive sampler, respectively. Additionally, the autosampler captured higher diversity and relative abundance of VOCs, including emerging variants (XBB* and CH.1* lineages), as well as more clinically relevant mutations (S:K444T, T22942A, S:R346T) relative to passive sampler. Overall, the passive samplers produced concordant results with the autosampler for measuring SARS-CoV-2 load with RT-qPCR in aircraft wastewater. Conclusions Taken together, our results suggest underestimation of the diversity and abundance of SARS-CoV-2 VOCs and mutations in aircraft sewage using passive torpedo samplers. These data can be used to optimize genomic surveillance approaches for SARS-CoV-2 VOCs in aircraft wastewater samples.