The Rapid Assessment of Aggregated Wastewater Samples for Genomic Surveillance of SARS-CoV-2 on a City-Wide Scale

Abstract

Throughout the course of the ongoing SARS-CoV-2 pandemic there has been a need for approaches that enable rapid monitoring of public health using an unbiased and minimally invasive means. A major way this has been accomplished is through the regular assessment of wastewater samples by qRT-PCR to detect the prevalence of viral nucleic acid with respect to time and location. Further expansion of SARS-CoV-2 wastewater monitoring efforts to include the detection of variants of interest/concern through next-generation sequencing has enhanced the understanding of the SARS-CoV-2 outbreak. In this report, we detail the results of a collaborative effort between public health and metropolitan wastewater management authorities and the University of Louisville to monitor the SARS-CoV-2 pandemic through the monitoring of aggregate wastewater samples over a period of 28 weeks. Through the use of next-generation sequencing approaches the polymorphism signatures of Variants of Concern/Interest were evaluated to determine the likelihood of their prevalence within the community on the basis of their relative dominance within sequence datasets. Our data indicate that wastewater monitoring of water quality treatment centers and smaller neighborhood-scale catchment areas is a viable means by which the prevalence and genetic variation of SARS-CoV-2 within a metropolitan community of approximately one million individuals may be monitored, as our efforts detected the introduction and emergence of variants of concern in the city of Louisville. Importantly, these efforts confirm that regional emergence and spread of variants of interest/concern may be detected as readily in aggregate wastewater samples as compared to the individual wastewater sheds. Furthermore, the information gained from these efforts enabled targeted public health efforts including increased outreach to at-risk communities and the deployment of mobile or community-focused vaccination campaigns.

Keywords: COVID-19; SARS-CoV-2; mutants; sequencing; variants; variants of concern; wastewater.

Figure 1

The Quantitative Detection of SARS-CoV-2 in the Aggregated Wastewater of Water Quality Treatement Plants, Louisville, Kentucky. On the left Y-axis (Blue) are the averaged ΔCt values obtained from the 5 water quality treatment centers (WQTCs) that service the Louisville metropolitan area. Metadata regarding the selection of the WQTCs and the populations they serve may be found in a companion publication Yeager et al., 2021 [29]. The ΔCt values represent the levels of SARS-CoV-2 N1 amplicon normalized to the PPMOV control amplicon. Note that the left Y-axis is plotted in an inverse manner. Individual data points represent the means of three technical replicates per WQTC per timepoint, with the error bars representing the standard deviation of the means. Plotted on the right Y-axis (Red) are the numbers of confirmed SARS-CoV-2 cases reported to the city of Louisville public health authorities (as per publically available health data) on a per week basis. The X-axis represents the particular week to which the data belongs, with week 1 pertaining to the first week of October 2020, and week 28 being the last week of April 2021. The week pertaining to that of the 3rd and 4th weeks of December are omitted from the above data set, as the underlying public health and wastewater data sets were incomplete for that period of time due to the holiday season. The relevant public health data may be found in the supplemental data files accompanying this manuscript.

Figure 2

The Quantitative Detection of SARS-CoV-2 in Wastewater Tributaries of the Morris Forman and Derek R. Guthrie WQTCs. On the left Y-axes (Blue) are the average ΔCt values obtained from wastewater tributaries associated with the Shawnee Park B wastewatershed (A), and the aggregate wastewater treatment center to which it feeds- Morris Forman WQTC (B). Panels (C,D) are similar, but the areas represented are the Ashby Lane and Mill Creek tributary and the Derek R. Guthrie WQTC. The ΔCt values represent the levels of SARS-CoV-2 N1 amplicon normalized to the PPMOV control amplicon. Note that the left Y-axis is plotted in an inverse manner. Individual data points represent the means of three technical replicates per WQTC per timepoint, with the error bars representing the standard deviation of the means. Plotted on the right Y-axis (Red) are the numbers of confirmed SARS-CoV-2 cases reported to the city of Louisville public health authorities on a per week basis. The X-axis represents the particular week to which the data belongs, with week 1 pertaining to the first week of October 2020, and week 28 being the last week of April 2021. The week pertaining to that of the 3rd and 4th weeks of December are omitted from the above data set, as the underlying data sets were incomplete for that period of time.

Figure 3

Quantitative Analysis of Per-Base Genome Coverage and Percent Genome Covered for the Aggregate Wastewater Sites. (A) Shown above is a circle plot depicting the depth of coverage per nucleotide (circle size) and the color of the datapoint corresponds to the Ct value of the associated sample, as detected by qRT-PCR. (B) Similar to that presented in A, but the circle size represents the depth of coverage, and the color of the circle represents the percentage of the genome covered by at least ten reads (circle color). The individual aggregate wastewater sites are listed on the left vertical axis, and the sampling week listed on the bottom horizontal axis.

Figure 4

The Identification of Polymorphism Markers Informs About the Presence of Variants of Concern/Interest in the Community. Panels (A,C) detail the specific polymoprhisms associated with each of the currently identified VOCs and VOIs. Specific polymorphisms are found on the upper horizontal axis, and the SARS-CoV-2 gene to which they belong may be found on the lower horizonatal axis. Specific VOC/Is are listed on the left axis. Sequence analyses with respect to time for the aggregated wastewater samples obtained from each of the 5 water quality treatment centers (CC, Cedar Creek; DRG, Derek R. Guthrie; FF, Floyd’s Fork; HC, Hite Creek; and MF, Morris Forman) for each of the identified VOCs (panel B) and VOIs (panel D). Data shown represents the proportion of polymorphism markers which meet or exceed the 5% prevalence threshold that were identified for each listed variant.

Figure 4

The Identification of Polymorphism Markers Informs About the Presence of Variants of Concern/Interest in the Community. Panels (A,C) detail the specific polymoprhisms associated with each of the currently identified VOCs and VOIs. Specific polymorphisms are found on the upper horizontal axis, and the SARS-CoV-2 gene to which they belong may be found on the lower horizonatal axis. Specific VOC/Is are listed on the left axis. Sequence analyses with respect to time for the aggregated wastewater samples obtained from each of the 5 water quality treatment centers (CC, Cedar Creek; DRG, Derek R. Guthrie; FF, Floyd’s Fork; HC, Hite Creek; and MF, Morris Forman) for each of the identified VOCs (panel B) and VOIs (panel D). Data shown represents the proportion of polymorphism markers which meet or exceed the 5% prevalence threshold that were identified for each listed variant.

Figure 5

Realtime Identification of the Emergence of the P.1 Variant in a Northwestern Louisville Community. Sequence analyses with respect to time for the wastewater tributary from the Shawnee Park neighborhood in Louisville, KY. Data shown represents the proportion of polymorphism markers which meet or exceed the 5% prevalence threshold that were identified for each listed variant. Week 21 corresponds to the week period/date of the initial detection of the P.1 variant in a clinical sample.

Figure 6

S Protein Polymorphisms Detected in Aggregate Wastewater Over an 8 Week Period. Above are heatmaps for each of the individual wastewater treatment facilities that service the Louisville metropolitan area. Time in weeks is listed on the horizontal axis. Individual polymorphisms that met or exceeded the threshold of 5% prevalence are listed on the left axis. Prevalence, as determined by the number of reads associated with a given polymorphism over total is indicated above in the figure. Silent mutations are not shown.