Parallel deployment of passive and composite samplers for surveillance and variant profiling of SARS-CoV-2 in sewage

Abstract

Wastewater-based epidemiology during the COVID-19 pandemic has proven useful for public health decision-making but is often hampered by sampling methodology constraints, particularly at the building- or neighborhood-level. Time-weighted composite samples are commonly used; however, autosamplers are expensive and can be affected by intermittent flows in sub-sewershed contexts. In this study, we compared time-weighted composite, grab, and passive sampling via Moore swabs, at four locations across a college campus to understand the utility of passive sampling. After optimizing the methods for sample handling and processing for viral RNA extraction, we quantified SARS-CoV-2 N1 and N2, as well as a fecal strength indicator, PMMoV, by ddRT-PCR and applied tiled amplicon sequencing of the SARS-CoV-2 genome. Passive samples compared favorably with composite samples in our study area: for samples collected concurrently, 42 % of the samples agreed between Moore swab and composite samples and 58 % of the samples were positive for SARS-CoV-2 using Moore swabs while composite samples were below the limit of detection. Variant profiles from Moore swabs showed a shift from variant BA.1 to BA.2, consistent with in-person saliva samples. These data have implications for the broader implementation of sewage surveillance without advanced sampling technologies and for the utilization of passive sampling approaches for other emerging pathogens.

Keywords: Moore swab; Passive sampling; Wastewater-based epidemiology.

Graphical abstract

Fig. 1

Passive and composite sample N1 concentrations normalized by PMMoV concentrations at the four sample locations over the study period. The secondary y-axes enumerates the cumulative number of positive cases confirmed by voluntary saliva testing during a seven-day sliding window preceding the sample collection (or start of sample collection) date (in gray). Data points plotted on the “BLD” dotted line represent data that were below the empirically defined 60 % confidence limit of detection.

Fig. 2

Estimated variant prevalence in sewage sampled from Dorm C and D during January to early February 2022 and the clade assignments of the corresponding clinical saliva samples. The underlying data is the variant lineage estimation by Freyja summarized by WHO designations (e.g., Omicron BA.1 or BA.2 clade) of sewage samples (columns) and the clade assignments of consensus sequences from clinical saliva samples (circles). The estimate of variant prevalence of Dorm D from January 20th was from a composite sample (marked with an asterisk) and the rest were from Moore swabs. The predicted presence of Delta/Omicron recombinant lineages such as XF, XD, and XS are grouped into a ‘Delta+Omicron’ group. The ‘Other’ class incorporates both the B.1.1.221 and B.1.1.225 lineages. The deconvolution estimation in grab samples did not capture any new variant clades.