Human pathogen nucleic acids in wastewater solids from 191 wastewater treatment plants in the United States

Abstract

We measured concentrations of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and its variants, influenza A and B viruses, respiratory syncytial virus, human metapneumovirus, enterovirus D68, human parainfluenza types 1, 2, 3, 4a, and 4b in aggregate, norovirus genotype II, rotavirus, Candida auris, hepatitis A virus, human adenovirus, mpox virus, H5 influenza A virus, and pepper mild mottle virus nucleic acids in wastewater solids prospectively at 191 wastewater treatment plants in 40 states across the United States plus Washington DC. Measurements were made two to seven times per week from 1 January 2022 to 30 June 2024, depending on wastewater treatment plant staff availability. Measurements were made using droplet digital (reverse-transcription–) polymerase chain reaction (ddRT-PCR) following best practices for making environmental molecular biology measurements. These data can be used to better understand disease occurrence in communities contributing to the wastewater.

Background & Summary

Wastewater-based epidemiology, or wastewater monitoring (hereafter WBE), is used across the globe to understand the spread of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)¹. Since coronavirus disease 2019 (COVID-19) testing availability and test seeking behavior has declined, wastewater remains one of the best methods for assessing virus circulation². Research on applications of WBE for surveillance for other diseases rapidly advanced since the start of the pandemic, and it is now known that a wide range of viral and other diseases can be tracked using wastewater^3–6. Wastewater represents a composite sample of the entire contribution community, and at times, may even contain contributions from other animals⁷. Biomarkers of infectious disease are shed in a variety of excretions (feces, urine, saliva, mucus, semen, sloughing skin cells)^8,9 that end up in wastewater via shower, sink, washing machine and toilet drains in households as well as businesses.

Wastewater is a complex environmental mixture containing both solid and liquid components. Solids are often defined as material larger than approximately 0.2 µm in diameter. They can be removed from wastewater through centrifugation or filtration, are composed of both organic and inorganic material and may be of fecal origin, or represent food, skin cells, bacterial cells, eukaryotic cells, or other aggregated colloids or debris. Given their size definition, prokaryotic and eukaryotic organisms will be part of the solid fraction⁵. In addition, extensive research indicates that numerous diverse virus biomarkers partition to solids in wastewater, likely driven by physico-chemical interactions between biomarkers and solids, as well as potentially steric interactions^10–12. Given the enrichment of biomarkers in wastewater solids, we implemented a national prospective sampling effort to provide real time data on concentrations of human pathogens in wastewater solids samples in early 2022. One hundred and ninety one (191) wastewater treatment plants (WWTPs) (Fig. 1) were enrolled in the study (exact number varied over time) and provided samples of wastewater solids two to seven times per week (median = three times per week). We processed the samples and made data available within 48 hours of sample receipt at the laboratory on a public facing dashboard. Initially, we measured two conserved gene targets found in the SARS-CoV-2 genome (N and S genes), influenza A virus (IAV), respiratory syncytial virus (RSV), pepper mild mottle virus (PMMoV), and characteristic mutations in SARS-CoV-2 BA.1 (Omicron) del143/145 and Delta del156/157. PMMoV is globally highly abundant in human stool and domestic wastewater¹³ and is used in the project as an internal recovery and fecal strength control^14,15. We later added additional SARS-CoV-2 genomic targets that detected mutations that were characteristic of various circulating variants of concern. We then added genomic targets of influenza B (IBV), mpox virus (MPOX), human metapneumovirus (HMPV), human parainfluenza 1, 2, 3, 4a and 4b (HPIV, in aggregate), norovirus GII (HuNoV), rotavirus (Rota), Candida auris (Cauris), hepatitis A virus (HAV), human adenovirus (HAdV), and influenza A H5 (H5) nucleic acids. A diagram illustrating the enrollment of WWTPs and duration of the different measurements in the program is provided in Figs. 2 and 3.

Fig. 1.

Map of wastewater treatment plants (WWTPs) enrolled in this study. States where there are enrolled sites are shaded in light purple. Each dot (191) represents a participating WastewaterSCAN site.

Fig. 2.

Timeline of when different targets were measured during the study.

Fig. 3.

Wastewater treatment plant enrollment count over the study period.

We previously collected data for a small subset of these sites located in the Greater Bay Area of California, USA (12 WWTPs) for a limited duration (15 Nov 2020 through 31 Dec 2022) available in a Data Descriptor in Boehm et al.¹⁶. The present Data Descriptor contains data for a different time duration (1 Jan 2022 through 30 June 2024), additional measurements, and a large number of additional sites (191 sites total). We are making these data available for others to use to better understand disease circulation in communities. Researchers have suggested that wastewater data can be used to inform public health decisions by increasing situational awareness, informing clinical decision making, and providing insights into disease epidemiology such as disease reproductive numbers¹⁷, disease incident cases¹⁸ and hospitalizations¹⁹. It is our hope that by providing full access to the data and methods to the community, others will be able to use the data to gain valuable insights into disease circulation.

Methods

Sampling locations

WWTPs were enrolled over the course of the study, so the number of WWTPs contributing samples changed over time (Fig. 3). At any one time, between 15 and 191 WWTPs were enrolled in the project (Table S1, Fig. 1). WWTPs served between 3,000 and 4,000,000 people (median = 86,000 people); in aggregate, the 191 WWTPs served 45,330,248 million people or 13.5% of the US population (estimated in 2023 year) (Fig. 4). The number of people served by and location of each WWTP, and the start and end date of WWTP participation in the study is provided in Table S1.

Fig. 4.

Distribution of populations served by the 191 wastewater treatment plants who participated in the study. The y-axis represents the binned population served and the x-axis represents the number of WWTPs in each bin. “k” after the number indicates times 10³ and “M” after the number indicates times 10⁶.

Sample collection

WWTP staff provided either “grab” samples from the primary clarifier or 24-hour composite samples from the headworks (“influent”) (Table S1 shows which WWTP collected which type of sample). We note that the “grab” samples mentioned above are in fact solids from the wastewater that are collected over 1-6 hours in the clarifiers. They are therefore essentially a composite of wastewater solids from the community over several hours. Once collected, the sample was stored at 4 °C and shipped to the laboratory on blue ice overnight and processed typically within 24 hours of collection. Our research suggests limited decay of a variety of nucleic acid targets in wastewater at 4 °C over weeks²⁰, so we do not expect the time between sample collection and analysis to affect target quantification. A total of 48,758 samples were collected and analyzed for this study.

Pre-analytical processing

The pathogen nucleic acid targets are strongly associated with the solid phase of wastewater based on previous empirical measurements across liquid and solid phases in wastewater by our team^4,8,11,21, and by other researchers^5,10,12, so in this study we made measurements in the solid phase of wastewater. Solids were obtained from the influent samples by either using an Imhof cone²², or allowing the influent to settle for 10–15 min, and using a serological pipette to aspirate the settled solids into a falcon tube. Subsequent methods for pre-analytical processing are described step-by-step on protocols.io²³. Samples were centrifuged at 24,000 × g for 30 min at 4 °C to dewater the solids. The supernatant was then aspirated using a vacuum and discarded. A 0.5- to 1-g aliquot of the dewatered, wet solids was dried at 110 °C for 19 to 24 h to determine its dry weight. Bovine coronavirus (BCoV) served as a positive recovery control. In the laboratory each day, attenuated bovine coronavirus vaccine (PBS Animal Health, Calf-Guard Cattle Vaccine) was spiked into DNA/RNA shield solution (Zymo Research) at a concentration of 1.5 μl/ml. Dewatered solids were resuspended to a concentration of 75 mg/ml in the BCoV-spiked DNA/RNA shield. This concentration of solids was chosen as we showed in previous work that it minimized inhibition while maintaining sensitivity of various RT-PCR assays^16,17. Between 5 and 10 stainless steel grinding balls (5/32-in., OPS Diagnostics) were added to each sample, and the sample was subsequently homogenized using a Geno/Grinder 2010 (Spex SamplePrep). Air bubbles introduced during the homogenization process were subsequently removed from samples by centrifuging briefly.

Nucleic acid extraction

The methods for nucleic acid extraction are described step-by-step on protocols.io²⁴. Nucleic acids were extracted from 6 to 10 replicate aliquots per sample (see Table S1). For each replicate, nucleic acids were extracted from 300 μl of homogenized sample using the Chemagic Viral DNA/RNA 300 kit H96 for the Perkin Elmer Chemagic 360 followed by PCR inhibitor removal with the Zymo OneStep-96 PCR Inhibitor Removal kit. The resultant extract had a total volume of 50 µl. Extraction-negative controls (water) and extraction-positive controls were extracted using the same protocol as the homogenized samples. The positive controls consisted of different targets (Tables 1 and 2) added into the BCoV-spiked DNA/RNA shield solution described above.

Table 1.

Positive controls used in the study for SARS-CoV-2 targets and variants, Gene blocks are dsDNA molecules with the target sequence purchased from Integrated DNA Technologies (IDT, Coralville, Iowa).

Target	Genomic target	Positive control
SARS-CoV-2 N gene	Conserved region of nucleoprotein (N) gene	SARS-CoV-2 gRNA (ATCC VR-1986D)
SARS-CoV-2 S gene	Conserved region of S gene	SARS-CoV-2 gRNA (ATCC VR-1986D)
SARS-CoV-2 Delta	Mutation in the S gene S:del156/157	Twist Synthetic Delta gRNA control 23
SARS-CoV-2 Omicron BA.1	Mutation in the S gene S:del143/145	Twist Synthetic Omicron gRNA control 48
SARS-CoV-2 BA.2 + BA.4 + BA.5	Mutation in the S gene S:LPPA24S	Twist synthetic Omicron BA.2 gRNA control 50
SARS-CoV-2 BA.4 + BA.5 + BQ*	Mutation in the S gene S:del69/70	SARS-CoV-2 B.1.117 gRNA (ATCC VR-3326HK)
SARS-CoV-2 BA.4	Mutation in ORF1a: del141/143	Gene block
SARS-CoV-2 XBB	Mutations in S gene S:Y445P, G446S, E484 A, F486S and F490S	Gene block
SARS-CoV-2 XBB*	Mutations in S gene S:Y445P, G446S, E484A, F486S/P and F490S	Gene block

The variant name(s) containing the specific mutation are provided and then the mutation location is provided. The abbreviation for the assay is provided in parentheses next to the target name when applicable. ATCC is American Type Culture Collection, Twist is a vendor located in South San Francisco.

Table 2.

Positive controls used in the study for non-SARS-CoV-2 targets, Gene blocks are dsDNA molecules with the target sequence purchased from Integrated DNA Technologies (IDT, Coralville, Iowa).

Virus	Target	Positive control
RSV	Nucleoprotein (N) gene	Intact RSV B virus (Zepto NATFVP-NNS)
Influenza A (IAV)	Membrane (M) gene	Twist Synthetic Influenza H3N2 RNA Control (Twist 103002)
Influenza B (IBV)	RNA directed RNA polymerase (L) gene	Twist Synthetic Influenza B RNA Control (Twist 103003)
Human parainfluenza (HPIV)	Hemagglutinin- neuraminidase (HN) gene (1,2,4b), Fusion protein (F) gene (3), Untranscribed region (4a)	Intact HPIV 1 (ATCC VR-94), HPIV 2 (ATCC VR-92), HPIV 3 (ATCC VR-93), HPIV 4A (ATCC VR-1378), HPIA 4B (ATCC VR-1377)
Human metapneumovirus (HMPV)	RNA directed RNA polymerase (L) gene	Gene block
Enterovirus D68 (EVD68)	Genome Polyprotein (VP1)	ATCC VR-1826DQ and ATCC VR-1823D Gene Blocks
H5 IAV (H5)	Hemagglutinin (HA) gene	Gene block
Norovirus GII (HuNoV)	ORF1-2 junction	Quantitative Synthetic RNA from Norovirus G2 (II) (ATCC VR-3235SD)
Adenovirus Group F (HAdV)	Hexon gene	ATCC VR-930DQ
Rotavirus (Rota)	Non-structural Region Protein 3 (NSP3)	ATCC VR-2018DQ
Candida auris (Cauris)	5.8S, all of ITS2, and a fragment of 28S rDNA	ATCC MYA-5001
Hepatitis A (HAV)	5’UTR	ATCC VR-3257SD
Mpox clade II (MPOX_WA)	Tumor necrosis factor (TNF) receptor gene	Gene block
Mpox clades I and II (MPOX_G)	TNF receptor gene	Gene block

The abbreviation for the assay is provided in parentheses next to the target name when applicable. ATCC is American Type Culture Collection, Twist is a vendor located in South San Francisco.

Digital-droplet RT-PCR

Nucleic acid extracts were used as the template in digital droplet RT-PCR assays. BCoV and PMMoV were quantified using a duplex assay in each sample; for samples that had 10 replicate nucleic acid extracts, all 10 were run for BCoV/PMMoV and for those that had 6, 2 nucleic acid extracts were randomly chosen for running the targets each in its own well. Over the course of the study, we added and removed different pathogen assays to and from the project (Fig. 5) as different variants of SARS-CoV-2 emerged and science became available to support the addition of various pathogen nucleic acid targets. The primers and probes used are provided in Tables 3–8, and have been described and tested previously in peer-reviewed publications with the exception of the XBB and XBB* assays. We changed the mpox virus assay on 15 December 2022 in response to a request from United States Center for Disease Control, switching from an assay that detected both mpox virus clades I and II to an assay that just detected clade II. We changed the assay that targeted the mutations characteristic of SARS-CoV-2 XBB on 1 February 2022 to target the mutations in SARS-CoV-2 XBB* as the virus evolved. All nucleic acid replicates were run in their own well (either 6 or 10 per sample). Undiluted nucleic acid extract was used as template in all assays for infectious disease targets, and a 1:100 dilution of the extract was used for the BCoV/PMMoV assay template. The one exception was for the WWTP Modesto for which we used a 1:10 dilution of the nucleic acid extract as template to alleviate suspected RT-PCR inhibition as this plant receives some cannery wastes.

Table 4.

Forward and reverse primers, and probe sequences for detection of influenza, RSV, HMPV, and enterovirus D68 virus nucleic acids in this study.

IAV39	Forward	CAAGACCAATCYTGTCACCTCTGAC
	Reverse	GCATTYTGGACAAAVCGTCTACG
	Probe	TGCAGTCCTCGCTCACTGGGCACG
IBV3	Forward	TCCTCAAYTCACTCTTCGAGCG
	Reverse	CGGTGCTCTTGACCAAATTGG
	Probe	CCAATTCGAGCAGCTGAAACTGCGGTG
RSV37	Forward	CTCCAGAATAYAGGCATGAYTCTCC
	Reverse	GCYCTYCTAATYACWGCTGTAAGAC
	Probe	TAACCAAATTAGCAGCAGGAGATAGATCAG
HMPV40	Forward	ACTTTATTGGAGAAGGAGCAGG
	Reverse	GGGTAATGRTGATCAAGRTCA
	Probe	AYTGGATGGCMAGAACAGCA
EVD6841	Forward	CACYGAACCAGARGAAGCCA CACTGAACCAGAGGAAGCTA
	Reverse	CCAAAGCTGCTCTACTGAGAAA CTAAAGCTGCCCTACTAAGRAA
	Probe	TCGCACAGTGATAAATCARCAYGG

Primers and probes were purchased from Integrated DNA Technologies (IDT, Coralville, Iowa). All probes contained fluorescent molecules and quenchers (5′ FM/ZEN/3′ IBFQ) where FM is the fluorescent molecule (see Fig. 2), ZEN, a proprietary internal quencher from IDT; IBFQ, Iowa Black FQ. The references for the primers and probes are provided in the first column after the name of the target. The name of the target is the same used in the first column of Table 2.

Table 5.

Forward and reverse primers, and probe sequences for detection of HPIV nucleic acids in this study.

HPIV 13	Forward	AAGTTCAGTACAAAGCGGGA
	Reverse	GCTCARTAGGGGTTCTCCTA
	Probe	AGCAAAGCAGAGATCTCACACA
HPIV 23	Forward	AATACAACAGGGCARTGGG
	Reverse	GATAAAATAGCGTGAGGACTGC
	Probe	TCCTGTATATGGTGGTCTCATAAATGG
HPIV 33	Forward	TGTGGTGACCAACAGATCAA
	Reverse	CCCTCCAAAGAATCGTCCTG
	Probe	TCCAATGAAAACACTGATCCCAGA
HPIV 4a3	Forward	TGAACGGTTGCATTCAGG
	Reverse	TTGACTGGTTGCACCTAATTCT
	Probe	CTGGCAATCTCAACATAGACCATG
HPIV 4b3	Forward	GGAGAACTTTGAAACCACCTCTAA
	Reverse	TCTCCTTTAACTACCCTATCTTTGC
	Probe	ACCCCCATAAGGCAAGAAGC

Primers and probes were purchased from Integrated DNA Technologies (IDT, Coralville, Iowa). All probes contained fluorescent molecules and quenchers (5′ FM/ZEN/3′ IBFQ) where FM is the fluorescent molecule (see Fig. 5), ZEN, a proprietary internal quencher from IDT; IBFQ, Iowa Black FQ. The references for the primers and probes are provided in the first column after the name of the target. The name of the target is the same used in the first column of Table 2.

Table 6.

Forward and reverse primers, and probe sequences for detection of enteric virus nucleic acids in this study.

HuNoV4	Forward	ATGTTCAGRTGGATGAGRTTCTCWGA
	Reverse	TCGACGCCATCTTCATTCACA
	Probe	AGCACGTGGGAGGGCGATCG
Rota42	Forward	CAGTGGTTGATGCTCAAGATGGA
	Reverse	TCATTGTAATCATATTGAATACCCA
	Probe	ACAACTGCAGCTTCAAAAGAAGWGT
HAdV42	Forward	CCTCCTGTGTTACGCCAGA
	Reverse	CAGGCTGAAGTASGTATCGG
	Probe	CTCGATGATGCCGCAATGGT

Primers and probes were purchased from Integrated DNA Technologies (IDT, Coralville, Iowa). All probes contained fluorescent molecules and quenchers (5′ FM/ZEN/3′ IBFQ) where FM is the fluorescent molecule (see Fig. 5), ZEN, a proprietary internal quencher from IDT; IBFQ, Iowa Black FQ. The references for the primers and probes are provided in the first column after the name of the target. The name of the target is the same used in the first column of Table 2.

Table 7.

Forward and reverse primers, and probe sequences for detection of mpox and HAV, and C. auris nucleic acids in this study.

MPOX_G43	Forward	GGAAAATGTAAAGACAACGAATACAG
	Reverse	GCTATCACATAATCTGGAAGCGTA
	Probe	AAGCCGTAATCTATGTTGTCTATCGTGTCC
MPOX_WA43	Forward	CACACCGTCTCTTCCACAGA
	Reverse	GATACAGGTTAATTTCCACATCG
	Probe	AACCCGTCGTAACCAGCAATACATTT
Cauris38	Forward	CGCACATTGCGCCTTGGGGTA
	Reverse	GTAGTCCTACCTGATTTGAGGCGAC
	Probe	CTTCTCACCAATCTTCGCGGT
HAV44	Forward	GGTAGGCTACGGGTGAAAC
	Reverse	AACAACTCACCAATATCCGC
	Probe	CTTAGGCTAATACTTCTATGAAGAGATGC

Primers and probes were purchased from Integrated DNA Technologies (IDT, Coralville, Iowa). All probes contained fluorescent molecules and quenchers (5′ FM/ZEN/3′ IBFQ) where FM is the fluorescent molecule (see Fig. 5), ZEN, a proprietary internal quencher from IDT; IBFQ, Iowa Black FQ. The references for the primers and probes are provided in the first column after the name of the target. The name of the target is the same used in the first column of Table 2.

Fig. 5.

Arrangement of multiplexed PCR plates for pathogen targets run in the study. There are two distinct plate arrangements (A and B). Arrangement A was used for the majority of sites and arrangement B was used for a small subset of, at most, 16 WWTPs (see Table S1 when and where 10 replicates were run) followed their own arrangement. Each box represents a schematic of how assays were multiplexed during different periods of the project. The date at the bottom left side of each box is the start date and the date near the right edge of the box is the approximate end date (based on the date the assay was stopped in the lab, the date associated with the last sample run could be different depending on the date it was processed in the lab). All probes contained fluorescent molecules (FM) and quenchers (5′ FM/ZEN/3′ IBFQ), as indicated. FAM, 6-fluorescein amidite; HEX, hexachloro-fluorescein; ATTO590; ROX, carboxyrhodamine; Cy5, Cyanine5; Cy5.5, Cyanine5; and ZEN, a proprietary internal quencher from IDT (Coralville, Iowa); IBFQ, Iowa Black FQ. If the box is light gray then the annealing temperature is 59 °C and if it is dark gray, it is 61 °C. Assay abbreviations are provided in Tables 1 and 2 except for the following: N and S are assays targeting those genes in SARS-CoV-2. HV69-70 is the assay targeting the deletion 69/70 in the S gene characteristic BA.4 + BA.5 + BQ*. del143/145 is the assay targeting the 143/145 deletion in the S gene characteristics of SARS-CoV-2 BA.1. del156/157 is the assay targeting the deletion 156/157 in the S gene characteristic of SARS-CoV-2 Delta. BA.4 is the assay targeting the deletion 141/143 in the ORF1a gene characteristic of SARS-CoV-2 BA.4. BA.2 is the assay targeting the set of deletions LPPA24S characteristic of BA.2 + BA.4 + BA.5. XBB is the assay targeting adjacent SNPs in the S gene characteristic of SARS-CoV-2 XBB. XBB* is the assay targeting adjacent SNPs in the S gene characteristic of SARS-CoV-2 XBB*. HPIV are the assays targeting human parainfluenza 1, 2, 3, 4a, and 4b. The dashed yellow line indicates when instrumentation switched from QX200 to QX600.

Table 3.

Forward and reverse primers, and probe sequences for detection of SARS-CoV-2 nucleic acids in this study.

Target	Primer/Probe	Sequence
SARS-CoV-2 N gene17	Forward	CATTACGTTTGGTGGACCCT
	Reverse	CCTTGCCATGTTGAGTGAGA
	Probe	CGCGATCAAAACAACGTCGG
SARS CoV-2 S gene17	Forward	CAGACTAAKTCTCVTCGGCG
	Reverse	TGCACCAAGTGACATAGTGT
	Probe	AGCTAGTCAATCCATCATTGCCT
SARS-CoV-2 BA.2 + BA.4. + BA.536	Forward	GCCACTAGTCTCTAGTCAGTGTG
	Reverse	TGTCAGGGTAATAAACACCACGT
	Probe	CAGAACTCAATCATACACTAATTCTTTCAC
SARS-CoV-2 BA.416	Forward	TAATAAAGGAGCTGGTGGCC
	Reverse	ATGAGTTCACGGGTAACACC
	Probe	CGGCGCCGATCTAGACTTAG
SARS-CoV-2 BA.4 + BA.5 + BQ*35	Forward	ACTCAGGACTTGTTCTTACCT
	Reverse	TGGTAGGACAGGGTTATCAAAC
	Probe	ATGCTATCTCTGGGACCAAT
SARS-CoV-2 XBB	Forward	GGCTGCGTTATAGCTTGGAA
	Reverse	GAGAGTAACAATTAGAACCTGCAAC
	Probe	CTTGATTCTAAGCCTAGTGGTAATTATAAT
SARS-CoV-2 XBB*	Forward	GGCTGCGTTATAGCTTGGAA
	Reverse	GAGAGTAACAATTAGRACCTGCAAC
	Probe	CTTGATTCTAAGCCTAGTGGTAATTATAAT

Primers and probes were purchased from Integrated DNA Technologies (IDT, Coralville, Iowa). All probes contained fluorescent molecules and quenchers (5′ FM/ZEN/3′ IBFQ) where FM is the fluorescent molecule (see Fig. 5), ZEN, a proprietary internal quencher from IDT; IBFQ, Iowa Black FQ. The references for the primers and probes are provided in the first column after the name of the target. The name of the target is the same used in the first column of Table 1.

Table 8.

Forward and reverse primers, and probe sequences for detection of BCoV and PMMoV nucleic acids in this study.

BCoV45	Forward	CTGGAAGTTGGTGGAGTT
	Reverse	ATTATCGGCCTAACATACATC
	Probe	CCTTCATATCTATACACATCAAGTTGTT
PMMoV45	Forward	GAGTGGTTTGACCTTAACGTTTGA
	Reverse	TTGTCGGTTGCAATGCAAGT
	Probe	CCTACCGAAGCAAATG

Primers and probes were purchased from Integrated DNA Technologies (IDT, Coralville, Iowa). All probes contained fluorescent molecules and quenchers (5′ FM/ZEN/3′ IBFQ) where FM is the fluorescent molecule see Fig. 5, ZEN, a proprietary internal quencher from IDT; IBFQ, Iowa Black FQ. The references for the primers and probes are provided in the first column after the name of the target.

Digital droplet RT-PCR was performed on 20 μl samples from a 22 μl reaction volume, prepared using 5.5 μl template, mixed with 5.5 μl of One-Step RT-ddPCR Advanced kit for Probes (catalog no. 1863021; Bio-Rad), 2.2 μl reverse transcriptase, 1.1 μl dithiothreitol (DTT), and primers and probes at a final concentration of 900 nM and 250 nM, respectively. Droplets were generated using the AutoDG Automated Droplet Generator (Bio-Rad). PCR was performed using Mastercycler Pro with the following protocol: reverse transcription at 50 °C for 60 min, enzyme activation at 95 °C for 5 min, 40 cycles with 1 cycle consisting of denaturation at 95 °C for 30 s and annealing and extension at either 59 °C or 61 °C (for human pathogen targets, Fig. 5) or 56 °C (for PMMoV/BCoV duplex assay) for 30 s, enzyme deactivation at 98 °C for 10 min, and then an indefinite hold at 4 °C. The ramp rate for temperature changes was set at 2 °C/s, and the final hold at 4 °C was performed for a minimum of 30 min to allow the droplets to stabilize. Droplets were analyzed using either the QX200 or QX600 Droplet Reader (Bio-Rad). All liquid transfers were performed using the Agilent Bravo (Agilent Technologies).

Assays for human pathogen targets were run in multiplex using the probe mixing approach. The manner in which the assays were multiplexed changed over time as we added and removed assays to detect different pathogen targets of public health importance. Figure 5 shows how assays were multiplexed over the duration of the study. The QX200 was used to read plates with three or fewer multiplexed assays and the QX600 was used for other multiplex arrangements.

Each sample was run in 6 to 10 replicate wells for human pathogen targets, and 2 or 10 wells for BCoV/PMMoV (Table S1), extraction-negative controls were run in 3 to 7 wells, and extraction-positive controls in 1 well. In addition, PCR-positive controls (Tables 1 and 2) were run in 1 well, and no-template controls (NTC) (negative PCR controls) were run in 3 to 7 wells. The positive controls that were used are provided in Table 1. Results from replicate wells were merged for analysis. Thresholding was done using QuantaSoft Analysis Pro software (Bio-Rad, version 1.0.596). In order for a sample to be recorded as positive, it had to have at least three positive droplets. Three positive droplets corresponds to a concentration between ∼500 and 1000 copies (cp)/g; the range in values is a result of the range in the equivalent mass of dry solids added to the wells and the number of wells merged.

Concentrations of nucleic acid targets were converted to concentrations per dry weight of solids in units of cp/gram (dry weight) using dimensional analysis. The errors are reported as 68% confidence intervals (standard deviations) and include errors associated with the Poisson distribution and the variability among the replicates. Error for the merged replicates is reported as output by the QuantaSoft Analysis Pro software. The recovery of BCoV was determined by normalizing the concentration of BCoV by the expected concentration given the value measured in the BCoV-spiked DNA/RNA shield. BCoV recovery was used solely as a process control and not used in the calculation of concentrations²⁵.

Influenza A Reruns

In early 2024, we discovered that the influenza A virus (IAV) assay run using a probe labeled with the fluorescent molecule Cy5.5 (Cyanine5.5) was not performing well using template obtained from actual wastewater samples. This is despite the assay performing well using positive controls as template. Therefore, we removed all IAV data measured using the Cy5.5 labeled probe between 26 September 2023 and 11 January 2024, and then re-tested approximately 3 samples per week for each WWTP for IAV. Samples that were run retrospectively in duplex with the assay for the N gene in SARS-CoV-2 (results not reported herein for SARS-CoV-2), and those results were added to the database. The IAV assay probe was labeled with the fluorescent molecule FAM (6-fluorescein amidite), nucleic acid extracts biobanked at −80 °C were thawed and used as template undiluted. We determined that the IAV measuring using this probe with template that underwent one freeze thaw provided the same concentration as samples freshly acquired with no freeze thaw (Figure S2). The QX200 droplet reader was used in conjunction with the software described above.

RT-PCR assay design

The design of all assays included in the study have been described previously in peer-reviewed publications with the exception of the assays targeting characteristic mutations in SARS-CoV-2 XBB and XBB* (Table 3) (XBB mutations are S:Y445P, G446S, E484A, F486S and F490S; and XBB* mutations are S:Y445P, G446S, E484A, F486S/P and F490S). To design these assays, sequences of circulating variants of interest were downloaded from the National Center for Biotechnology Information (NCBI) in the month prior to assay deployment and aligned to identify conserved, characteristic regions. Assays were developed in silico using Primer3Plus (https://www.primer3plus.com/). Parameters used in assay development (e.g., sequence length and GC content) are provided elsewhere³. Primers and probes were screened for specificity in silico, and in vitro against genomic or synthetic RNA from non-target variants that do not contain the mutations including “wild-type” gRNA from SARS-CoV-2 strain 2019- nCoV/USA-WA1/2020 (ATCC VR-1986D, American Type Culture Collection), synthetic gRNA from Twist Biosciences (South San Francisco, California, USA) for Delta (Twist control 23, Twist, South San Francisco, CA) and Omicron BA.2 (Twist control 50), and gene blocks purchased from IDT (Coralville, Iowa) for the S gene region where the mutations are located with sequences from variants BM.1.1.1 and BJ.1.

Data Records

All measurements made in this study are available at a permanent URL at the Stanford Digital Depository: 10.25740/hj801ns5929³⁰. The data are available in a CSV file called “wwscan_data_descriptor_SDR_2024_Update.csv” that includes the fields listed below as columns.

City: City in which the collection site represents

State: State in which the collection site resides

State_Abbr: State in which the collection site resides

Plant: Plant name as shown on the website

Site_Name: Full name of collection site

Population_Served: Human population contribution to the sewershed of the collection site

Month: Month (1 through 12) of the year of sample collection

Day: Day of the month (1 through 31) of sample collection

Year: Year of sample collection

Bcov_recovery: Fractional recovery between 0 and 1 of Bovine coronavirus. Sometimes the value will be greater than 1 owing to uncertainty in the denominator of the ratio.

C_auris_gc_g_dry_weight: Copies of Candida auris nucleic acids per gram of dried solid

EVD68_gc_g_dry_weight: Copies of rotavirus RNA per gram of dried solid

HAdV_F_gc_g_dry_weight: Copies of adenovirus group F DNA per gram of dried solid

HAV_gc_g_dry_weight: Copies of hepatitis A RNA per gram of dried solid

HMPV_4_gc_g_dry_weight: Copies of human Metapneumovirus RNA per gram of dried solid

HPIV_gc_g_dry_weight: Copies of Human Parainfluenza (HPIV) RNA per gram of dried solid; includes HPIV 1, 2, 3, 4a, and 4b

HV_69_70_Del_gc_g_dry_weight: Copies of SARS-CoV-2 RNA from the S:HV69-70 deletion (indicative of BA.4, BA.5, and BQ*) per gram of dried solid

Influenza_A_gc_g_dry_weight: Copies of influenza A RNA per gram of dried solid

Influenza_B_gc_g_dry_weight: Copies of Influenza B RNA per gram of dried solid

InfA_H5_gc_g_dry_weight: Copies of Influenza A H5 RNA per gram of dried solid

MPXV_G2R_G_gc_g_dry_weight: Copies of MPOX DNA (clade I and II) per gram of dried solid

MPXV_G2R_WA_gc_g_dry_weight: Copies of MPXV Clade II DNA per gram of dried solid

Noro_G2_gc_g_dry_weight: Copies of Norovirus GII RNA per gram of dried solid

PMMoV_gc_g_dry_weight: Copies of PMMoV RNA per gram of dried solid

Rota_gc_g_dry_weight: Copies of rotavirus RNA per gram of dried solid

RSV_gc_g_dry_weight: Copies of RSV RNA per gram of dried solid

SC2_BA_2_LPPA24S_gc_g_dry_weight: Copies of SARS-CoV-2 RNA from the S:LPPA24S deletion (indicative of BA.2, BA.4 and BA.5) per gram of dried solid

SC2_BA_4_ORF1a_Del141143_gc_g_dry_weight: Copies of SARS-CoV-2 RNA from the ORF1a:141-143 deletion (indicative of BA.4) per gram of dried solid

SC2_N_gc_g_dry_weight: Copies of SARS-CoV-2 RNA from the nucleocapsid gene per gram of dried solid

SC2_Omicron_Del143145_gc_g_dry_weight: Copies of SARS-CoV-2 RNA from the S:143-145 deletion (indicative of BA.1) per gram of dried solid

SC2_S_gc_g_dry_weight: Copies of SARS-CoV-2 RNA from the S spike gene per gram of dried solid

copies per gram of dry solid sample

XBB_bkpt_gc_g_dry_weight: Concentration of the 5 adjacent SNPs (found in XBB*, including XBB1.5) in gene copies per mass of wastewater solids as measured by dry weight.

SC2_Delta_156157: Copies of SARS-CoV-2 RNA from the S:156-157 deletion (indicative of Delta) per gram of dried solid

Additional variables:

All variables ending in _UCL represent the upper 68% confidence limit (standard deviation) of the variable.

All variables ending in _LCL represent the lower 68% confidence limit (standard deviation) of the variable.

If a cell is blank, it means the assay was not run for that day.

If a “0” appears, it means the assay was a non-detect. The detection limit varies by sample depending on the amount of solids by dry weight included, but is between 500-1000 cp/g dry weight.

Technical Validation

We followed the Environmental Microbiology Minimum Information (EMMI) guidelines²⁶ and Minimum Information for Publication of Quantitative Digital PCR Experiments (dMIQE2020 guidelines)²⁷ and Checklists are in the supplementary information (Figure S1 and Table S2).

As described in our previous Data Descriptor¹⁶, we used strict QA/QC procedures. These included negative and positive extraction and PCR controls, and recovery controls coupled to replicate analyses (n = 6 or 10 for human pathogens, n = 2 or 10 for PMMoV) for each sample. On each plate, we ran negative and positive extraction and RT-PCR controls.We required that for each plate, there had to be 2 or fewer droplets across negative controls, and positive controls had to have 3 or more droplets. Data included in this Data Descriptor passed theQA/QC metrics. If a plate did not meet these QA/QC criteria, then all the samples run on the plate were re-run.

Median bovine coronavirus (BCoV) recoveries (BCoV measured in sample / BCoV measured in DNA/RNA shield without sample) were, on average 1.1 (25th percentile = 0.81, 75th percentile = 1.4) across all 48,758 samples, and all were above 10%. At times, a recovery higher than 1 was recorded, likely a result of underestimation of the amount of BCoV spiked into the samples. We did not attempt to correct any of our measurements by BCoV recovery owing to the complexities associated with estimating endogenous viral recovery in environmental matrices²⁵. The recovery results indicate that there was no gross inhibition and we had fairly consistent recovery of nucleic-acids.

Details associated with the Minimal Information MIQE Experiments (dMIQE) for digital droplet PCR reporting are described here. All samples run in a randomly chosen week were chosen for this analysis. In total, this represented 465 samples; this represents ~1% of the samples processed in the study. As described in the methods section, each sample was run as template in multiple multiplexed PCR reactions. For the week chosen (early June 2024), three different multiplex PCR reactions were run for each sample; 1 for PMMoV and BCoV, 1 for rota, SARS-CoV-2, HuNoV, HAV, IBV, RSV, and IAV, and 1 for EVD68, Cauris, HMPV, MPOX, and H5. The average (standard deviation) number of partitions (droplets) for each of the three reactions (was 37692 (2470) for the reaction for PMMoV and BCoV (2 merged wells), 106648 (20958) for the reaction for rotavirus, SARS-CoV-2, HuNoV GII, HAV, IBV, RSV, and IAV (6 merged wells), and 100574 (21478) for the reaction for EVD68, C. auris, HMPV, MPOX, and H5 (6 merged wells). The volume of the partitions, as reported by the machine vendor is 0.00085 μL. The mean and standard deviation of copies per partition for each target is shown in Tables 9 and 10. Example fluorescence plots from the digital PCR machine can be found in Wolfe et al.²⁸ and Topol et al.²⁹, as well as in the data link (10.25740/hj801ns5929)³⁰.

Table 9.

For each human infectious disease target measured in this study during early June 2024, the mean and standard deviation (sd) of the total number of copies of target per partition.

Target	SC2	IAV	IBV	RSV	HMPV	EVD68	HuNoV	Rota	mpox	Cauris	HAV	H5
num	465	465	465	465	465	465	465	465	465	465	465	465
mean	7.96 × 10−4	2.73 × 10−5	1.33 × 10−5	3.32 × 10−6	7.68 × 10−5	3.62 × 10−5	9.01 × 10−2	3.13 × 10−3	5.81 × 10−8	9.06 × 10−6	1.91 × 10−5	2.94 × 10−6
sd	1.12 × 10−3	3.82 × 10−5	2.01 × 10−5	9.77 × 10−6	2.42 × 10−4	1.11 × 10−4	8.69 × 10−2	7.47 × 10−3	7.22 × 10−7	6.99 × 10−5	1.43 × 10−4	9.06 × 10−6

Num is the number of samples out of the random week of the study chosen for this analysis (n = 465 samples). Note that the mean and sd were calculated using from all samples, not just those with detectable target. Abbreviations for the targets are provided in the main text except SC2 is the N gene of SARS-CoV-2.

Table 10.

For PMMoV and BCoV, the mean and standard deviation (sd) of the total number of copies of target per partition.

Target	PMMoV	BCoV
num	465	465
mean	0.040	0.00131
sd	0.045	0.00041

Num is the number of samples out of the random week of the study chosen for this analysis (n = 465 samples).

Assays were conducted in only one lab, so reproducibility was not assessed. Sample standard deviations for measured targets were 15% (median, IQR: 7%-32%) of the measurement. Since samples were extracted 6 to 10 times and each extract was analyzed in 6 or 10 replicate wells (for human pathogens) and 2 or 10 wells (for PMMoV) which were merged, the variation from both nucleic acid extraction and reverse transcription-digital droplet PCR (RT-ddPCR) with a heterogeneous solid sample are incorporated into the reported variability.

The theoretical lowest measurable concentration was 3 positive droplets which translates into between 500 and 1000 copies/g dry weight. The range reflects that fact that the detection limit depends, in part, on the percent dry weight of the solids used in the extraction. This, in turn, depends on the how effectively the solid can be dewatered, which depends on the properties of the solid

The methods used to assess inhibition are provided in our previous Data Descriptor¹⁶ and are summarized briefly here. We tested inhibition quarterly in samples. We titrated the solids concentration used in the pre-analytical processing steps using 150 mg/ml, 75 mg/ml, 37.5 mg/ml, 15 mg/ml, and 7.5 mg/ml and compared the resultant measured concentrations (calculated considering the different solid concentrations used). One would expect concentrations measured using lower concentrations of solids to be higher than those measured using 75 mg/ml if the RT-PCR were inhibited at 75 mg/ml.

The data descriptor we previous published provided example inhibition titrations SARS-CoV-2 N, SARS-CoV-2 S, SARS-CoV-2 LPPA24S, SARS-CoV-2 del143/145, IAV, HuNoV, HMPV, and RSV, and showed lack of significant inhibition at 75 mg wet weight solids per ml DNA/RNA shield solution¹⁶. Here we show example inhibition titrations for 7 of the assays described herein (Fig. 6). For 7 of the assays (IBV, HMPV, RSV, HAV, HuNoV, IAV, and SARS-CoV-2 N gene), lower concentrations of solids resulted in non-detects, or the same or slightly lower concentrations than those measured with 75 mg/ml. For the remaining two assays, the concentrations obtained using solids concentrations lower than 75 mg/ml were slightly higher. Differences in concentrations were, at most, about 2X, in all cases. We interpret results to indicate that inhibition is not significantly affecting our analyses. We further conclude that the choice of using 75 mg/ml effectively balances alleviating inhibition while providing good sensitivity.

Fig. 6.

Example inhibition titrations for nine assays used for the study. The concentration of solids from wastewater solids samples placed in DNA/RNA Shield in the preanalytical step prior to homogenization and RNA extraction is shown on the x-axis. On the y-axis is the concentration of the target measured in the solids in units of copies per gram (cp/g) dry weight, corrected for the concentration of solids placed in the DNA/RNA shield. If a symbol is on the value of “0”, it means the result was “not detected”. This occurred in some cases when the mass of solids added to the DNA/RNA shield was small, thereby reducing the sensitivity of the method. Errors are shown as standard deviations across 6 replicates as described in the methods section. If error bars are not visible it is because they are smaller than the symbol.

Usage Notes

The HPIV data provided in the data descriptor is provided as total HPIV and is the sum HPIV 1, 2, 3, 4a, and 4b. Data for individual HPIV are not provided.

These measurements were made using wastewater solids and therefore, concentrations are reported in units of copies per gram dry weight. These measurements are different from those carried out using a liquid wastwater matrix which are commonly reported by other researchers³¹ and reported in units of copies per milliliter. It should be noted that it is not straightforward to combine measurements made using liquids and solids, and care should be taken when doing so.

For clinicians or epidemiologists unfamiliar with environmental data it is important to note that the variability in environmental data is distinct from that observed in clinical case or syndromic data. For example, nucleic acid concentrations in wastewater may vary owing to non-steady sources of nucleic acids (from human shedding), intermittent deliveries of septic wastes or industrial discharges to the system, among other factors. Variability may also be attributable to heterogeneity within the environmental sample matrix, or attributable to the analytical methods. Despite this variability, wastewater data have been shown to be highly associated with measures of disease occurrence in sewersheds^3,8,28,32,33. To address variability in this study, we processed 6 or more replicates and multiple controls to describe analytical error and ensure high quality measurements.

When we share these data with public health partners, we typically normalize nucleic acid concentrations by concentrations of PMMoV RNA to account for process variability and the fecal strength of the wastewater^14,15, and use 5-d trimmed average smoothing of the high-frequency data to reduce the influence of measurement outliers. In addition, we typically present presence/absence data for targets that are rarely detected including MPOX, H5 marker in IAV, HAV, and C. auris.

Some of the data described in this data descriptor has been analyzed in previous publications including Wolfe et al.²⁸, Wolfe et al.³⁴, Yu et al.³⁵, Boehm et al.³⁶, Wolfe et al.⁸, Zulli et al.³⁷, Zulli et al.³⁸, and Schoen et al.³⁹. The vast majority of the data in this descriptor has not been analyzed.

Author contributions

A.B. Boehm: Conceptualization, Methodology, Validation, Formal analysis, Investigation, Resources, Data Curation, Writing - Original Draft, Writing - Review & Editing, Visualization, Supervision, Project administration, Funding acquisition. M.K. Wolfe: Conceptualization, Methodology, Data Curation, Validation, Writing - Review & Editing, Supervision, Project administration. A.L. Bidwell: Data Curation, Writing - Original Draft, Writing - Review & Editing, Visualization. B. White: Methodology, Validation, Investigation, Resources, Data Curation, Writing - Review & Editing, Supervision. B. Shelden: Methodology, Validation, Investigation, Data Curation, Writing - Review & Editing. D. Duong: Methodology, Validation, Investigation, Data Curation, Writing - Review & Editing. A. Zulli: Validation, Investigation, Data Curation, Writing - Review & Editing. V. Chan-Herur: Methodology, Validation, Investigation, Data Curation, Writing - Review & Editing.

Code availability

The figures made in this Data Descriptor were generated using IGOR PRO v8 (Wavemetrics, Lake Oswego, Oregon, USA) RStudio (RCore Team, Boston, Massachusetts, USA) and Tableau Desktop (Salesforce, Seattle, Washington, USA).

Competing interests

Bradley White, Dorothea Duong, Bridgette Shelden, and Vikram Chan-Herur are employees of Verily Life Sciences. The other authors have no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41597-024-03969-8.