A Direct Capture Method for Purification and Detection of Viral Nucleic Acid Enables Epidemiological Surveillance of SARS-CoV-2

SARS-CoV-2, COVID-19, wastewater, wastewater-based epidemiology, RT-qPCR, 12 nucleic acid purification 13 14 15 16 17 18 19 20 21 22 23 1 Contributed equally to this work 24 25 26 * Corresponding Author: 27 Subhanjan Mondal 28 Email: [email protected] 29 Address: Promega Corporation, 5430 E Cheryl Pkwy, Fitchburg, WI, 53711. 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint NOTE: This preprint reports new research that has not been certified by peer review and should not be used to guide clinical practice. 2

45 46 Studies have demonstrated that SARS-CoV-2 RNA can be detected in the feces of infected individuals. 47 This finding spurred investigation into using wastewater-based epidemiology (WBE) to monitor SARS-48 CoV-2 RNA and track the appearance and spread of COVID-19 in communities. SARS-CoV-2 is present 49 at low levels in wastewater, making sample concentration a prerequisite for sensitive detection and utility 50 in WBE. Whereas common methods for isolating viral genetic material are biased toward intact virus 51 isolation, it is likely that a relatively low percentage of the total SARS-CoV-2 RNA genome in 52 wastewater is contained within intact virions. Therefore, we hypothesized that a direct unbiased total 53 nucleic acid extraction method could overcome the cumbersome protocols, variability and low recovery 54 rates associated with the former methods. This led to development of a simple, rapid, and modular 55 alternative to existing purification methods. In an initial concentration step, chaotropic agents are added to 56 raw sewage allowing binding of nucleic acid from free nucleoprotein complexes, partially intact, and 57 intact virions to a silica matrix. The eluted nucleic acid is then purified using manual or semi-automated 58 methods. RT-qPCR enzyme mixes were formulated that demonstrate substantial inhibitor resistance. In 59 addition, multiplexed probe-based RT-qPCR assays detecting the N1, N2 (nucleocapsid) and E (envelope) 60 gene fragments of SARS-CoV-2 were developed. The RT-qPCR assays also contain primers and probes 61 to detect Pepper Mild Mottle Virus (PMMoV), a fecal indicator RNA virus present in wastewater, and an 62 exogenous control RNA to measure effects of RT-qPCR inhibitors. Using this workflow, we monitored 63 wastewater samples from three wastewater treatment plants (WWTP) in Dane County, Wisconsin. We 64 also successfully sequenced a subset of samples to ensure compatibility with a SARS-CoV-2 amplicon 65 panel and demonstrated the potential for SARS-CoV-2 variant detection. Data obtained here underscore 66 the potential for wastewater surveillance of SARS-CoV-2 and other infectious agents in communities. 67 68 69 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 3

70 71 In late December of 2019, Chinese health authorities examined a new respiratory virus that caused 72 unexplained cases of severe pneumonia1. Subsequent sequencing identified the virus as a member of the 73 Coronavirus family, a group of enveloped RNA viruses that commonly infect birds, mammals, and 74 humans2. The novel virus, designated Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), 75 is related to SARS-CoV and MERS-CoV, other respiratory viruses that can lead to fatal illness3. SARS-76 CoV-2 was determined to be the causative agent of the respiratory disease COVID-191,4. COVID-19 77 quickly evolved into a global pandemic, which at the time of this report has resulted in over 3.1 million 78 deaths worldwide (https://coronavirus.jhu.edu/map.html). 79 80 The rapid person-to-person spread of SARS-CoV-2 is in part due to the high infectiousness of viral 81 carriers5,6. Pre-symptomatic shedding is thought to drive a significant amount of viral spread, as the 82 highest risk of transmission occurs very early in the course of the disease5,6. Asymptomatic carriers of the 83 virus are also infectious and display similar viral loads in the respiratory system, despite exhibiting faster 84 viral clearance6-9. As traditional nasal swab testing approaches can be biased towards symptomatic viral 85 carriers10, viral surveillance methods that provide a widespread view of community infection are vital for 86 accurate monitoring and control of the ongoing pandemic. 87 88 Although infectious SARS-CoV-2 virions are rarely isolated from feces11, several reports early in the 89 COVID-19 pandemic demonstrated that SARS-CoV-2 can frequently (50%-70%) be detected in fecal 90 samples from both symptomatic and asymptomatic infected individuals9,12-15. Levels of SARS-CoV-2 91 RNA in feces are not correlated with the presence/absence of gastrointestinal illness or with overall 92 COVID-19 disease severity12, as is often the case with respiratory samples. Although SARS-CoV-2 viral 93 load is consistently higher in respiratory specimens, viral RNA in feces can be detected significantly 94 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 4 longer after initial symptom onset compared to nasal swab samples13,16. This aspect of SARS-CoV-2 fecal 95 shedding could be a useful feature for assaying the levels of overall infection in a community. 96 97 One method that has been proposed for use in monitoring population-level rates of SARS-CoV-2 98 infection is wastewater-based epidemiology (WBE). WBE is the quantitative detection of chemical and 99 biological signatures in wastewater to analyze the status of a human population in an area17. This 100 approach has been used extensively in the past to characterize community usage/prevalence of illicit 101 drugs such as cocaine, amphetamines, and opiates17. WBE detection methods have also been widely used 102 to monitor the trends of circulating human pathogens in wastewater streams around the world17,18. Of 103 particular relevance to SARS-CoV-2 are the efforts centering around environmental surveillance of the 104 enteric viruses Poliovirus, Hepatitis A, and Norovirus19-21 105 106 In the Spring of 2020, research groups around the world started reporting the detection of SARS-CoV-2 107 RNA in untreated wastewater22-25 using quantitative reverse transcription PCR (RT-qPCR). SARS-CoV-2 108 RNA levels in wastewater positively correlated with clinical COVID-19 case numbers22. Several studies 109 were able to demonstrate that an increasing trend of SARS-CoV-2 in wastewater preceded a rise in 110 clinical cases by up to one week25-27, establishing the potential utility of COVID-19 WBE as an “early 111 warning system” to mitigate viral spread and cost effective strategy for monitoring real-time changes in 112 viral prevalence. 113 114 As viral concentrations in wastewater tend to be low, sample concentration is often a pre-requisite for 115 sensitive detection and accurate quantitation. Concentration of viral matter can be performed using a 116 variety of methods such as charged membrane filtration, centrifugal ultrafiltration, 117 flocculation/precipitation using skimmed milk, and polyethylene glycol (PEG)/NaCl precipitation28. Most 118 of the viral concentration methods described in the literature were originally developed to concentrate 119 non-enveloped viral particles for use in downstream culture-based approaches, but they have been also 120 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 5 used for PCR detection. Many of these methods may work efficiently with several virus types but present 121 difficulties when applied to SARS-CoV-2. Notable issues have included inconsistent rates of viral 122 recovery after concentration, requirement for large sample sizes, and co-purification of PCR 123 inhibitors29,30. In addition, these viral concentration methods are labor intensive and time consuming, 124 requiring separate viral concentration and a nucleic acid extraction steps. 125 126 In this report, we describe the use of a novel column-based viral concentration and nucleic acid 127 purification system coupled to RT-qPCR detection for monitoring SARS-CoV-2 levels in wastewater. 128 The system allows for increased sample throughput due to the combination of viral purification and 129 nucleic acid purification steps. The effect of PCR inhibitors commonly found in wastewater is also 130 minimized. To demonstrate proof-of-concept, wastewater from three communities in Dane County, 131 Wisconsin was monitored for the presence of SARS-CoV-2 RNA over the course of three months. We 132 also successfully sequenced from a subset of samples to show compatibility with a commercially 133 available SARS-CoV-2 amplicon panel and demonstrated the potential for SARS-CoV-2 variant detection 134 from wastewater nucleic acid purified with the direct capture method. 135 136 137

138 139 Sample collection 140 Wastewater (untreated primary effluent) was collected from three wastewater treatment plants in Dane 141 County, Wisconsin: Oregon (WWTP-1), Madison (WWTP-2), Sun Prairie (WWTP-3). In all three 142 communities, the wastewater collection system is separate from the storm sewer system, minimizing the 143 dilution effect of precipitation events. 500-1000mL of a flow-paced (Oregon, Sun Prairie) or time-paced 144 (Madison) 24-hour composite sample were collected weekly using an autosampler (Madison: ISCO 145 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 6 FR3710; Sun Prairie: ISCO 5800; Oregon: ISCO 4700). Samples were kept at 4°C at all times during 146 transport and storage. All samples were processed within six hours of sampling. 147 148 MS2 (ATCC 15597-B1) and Escherichia coli ATCC 15597 were purchased from American Type Culture 149 Collection (ATCC) and propagated based on published methods31. The viral fraction in the culture 150 supernatant was collected by centrifuging at 3500 x g for 30 minutes to remove the bacterial fraction and 151 filtered through a 0.22 μm PES membrane filter. The purified MS2 virus (∼107 PFU mL–1) was stored at 152 4 °C. MS2 virus was spiked in wastewater samples to final concentrations of 4 × 104 PFU mL–1. Samples 153 were then briefly mixed and nucleic acid extracted as described below. 154 155 Isolation of Total Nucleic Acid from wastewater using the direct capture method 156 Total nucleic acid (TNA) was purified from collected wastewater using the Wizard® Enviro Wastewater 157 TNA kit (Promega Corp.) and/or the Maxwell® Enviro Wastewater TNA kit (Promega Corp.), both of 158 which use an initial concentration step composed of direct capture of nucleic acids on silica resin. Briefly, 159 0.5 mL of alkaline protease was added to 40 mL of untreated wastewater in triplicate, and samples were 160 incubated statically for 30 minutes at room temperature. The samples were then centrifuged at 3000 x g 161 for 10 minutes to remove suspended solids. The supernatant was transferred to a new vessel and 12 mL of 162 Binding Buffer 1 and 1 mL of Binding Buffer 2 were added, followed by gentle mixing. A 48 mL of 163 isopropanol was added to the mixture, gently mixed and passed through a PureYield™ Midi Binding 164 Column (Promega Corp.) using a VacMan® Vacuum Manifold (Promega Corp.). Nucleic acid captured on 165 the PureYield™ Midi Binding Column was washed with 5 mL of Column Wash 1 followed by 20 mL of 166 Column Wash 2. Nucleic acid was eluted with nuclease-free water (Fig. 2). 167 168 The eluted nucleic acid was further purified using a Mini spin column for the Wizard® Enviro Wastewater 169 TNA kit or with an automated nucleic acid purification system (Maxwell® RSC, Promega Corp) for the 170 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 7 Maxwell® Enviro Wastewater TNA kit (Fig. 2). For the manual Wizard® Enviro Wastewater TNA kit, 171 400 µL of Binding Buffer 1, 100 µL of Binding Buffer 2 and 1.5 mL of isopropanol are added to the 1 172 mL of nucleic acid extracted in the concentration step, and then passed through a spin column with a 173 silica resin. The column is washed with 350µL and 1 ml of Column Wash 1 and 2 respectively, and 174 nucleic acid is extracted in 80 µL of water. For the automated Maxwell® Enviro Wastewater TNA kit , 175 150 µL of Binding Buffer 1 and 50 µL of Binding Buffer 2 are added to 0.5 mL of nucleic acid extracted 176 in the concentration step. Then the total volume is added to well #1 of the Maxwell® Cartridge and 177 nucleic acid is eluted in 80 µL of nuclease-free water. 178 179 180 Isolation of Total Nucleic Acid from suspended solids using the direct capture method 181 The method to extract total nucleic acids from suspended solids is a variation on the direct capture 182 methodology described above. After the alkaline protease step, the pellet of suspended solids is 183 resuspended in 5 mL of nuclease-free water. To the resuspended pellet, 1.5 mL of Binding Buffer 1, 125 184 μL of Binding Buffer 2 and 6 mL of isopropanol were added and mixed. This step releases the nucleic 185 acid bound to the solids into the suspension. The mixture was centrifuged at 3,000 x g for 10 minutes. 186 The supernatant contains the nucleic acid from the solids. The supernatant is then added to the 187 PureYield™ Midi Binding Column for the respective sample and treated independently as described 188 above. 189 190 PEG/NaCl precipitation 191 A sample of 120 mL of wastewater was centrifuged at 3000 x g for 30 minutes to pellet any particulate 192 material. The supernatant was carefully decanted and then mixed with 12 g of PEG 8000 and 2.7 g of 193 NaCl that were dissolved by gentle mixing. Samples were then centrifuged at 11,400 x g for two hours to 194 pellet viral material. The supernatant was carefully removed via pipetting on the side of the tube opposite 195 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 8 of that which the pellet was formed (pellet was not visible). The pellets were resuspended by vortexing 196 with residual supernatant to a final volume of approximately 200 µL. Further nucleic acid purification 197 was performed using Maxwell® RSC PureFood GMO and Authentication Kit (Promega Corp.) using the 198 Maxwell® RSC Instrument (Promega Corp.) according to a modified version of manufacturer’s protocol. 199 Briefly, to 200μl of wastewater concentrate, 200 μL of CTAB and 40 μL of Proteinase K were added. 200 Samples were vortexed, then incubated at 56°C for 10 minutes. The entire sample volume and 300 μL of 201 Lysis Buffer were added to well #1 of the Maxwell® RSC cartridge. Nucleic acid was eluted with 50 μL 202 of Elution Buffer. Samples were processed on the Maxwell® RSC with the PureFood GMO and 203 Authentication method. 204 205 Oligonucleotides and quantification standards 206 Table S1 lists the primers and probes used in this study, including primer and probe sets for the detection 207 of SARS-CoV-2 (CDC-N1, CDC-N2, E_Sarbeco), PMMoV, and MS2 bacteriophage. The primers and 208 probe for amplifying the luciferase (luc) gene for the internal amplification control are also listed. 209 210 The SARS-CoV-2 quantification standard was created by inserting the envelope gene (NCBI: 43740570) 211 and the nucleocapsid gene (NCBI: 43740575) of SARS-CoV-2 into the pGEM-3z vector (Promega 212 Corp.) using the BamH I site. The plasmid was subsequently linearized using Xba I. The linearized 213 plasmid was in-vitro transcribed using T7 RiboMAX™ Large Scale RNA Production System (Promega 214 Corp.) to create Positive Control RNA. The linearized plasmid DNA containing the N and E genes, or the 215 in-vitro transcribed RNA were quantified using droplet digital PCR and used as a quantification control. 216 Linear quantification DNA and positive control RNA for PMMoV and MS2 were generated and 217 quantified using the same methods described above. 218 219 220 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 9 RT-qPCR and quantification of viral load in wastewater 221 TNA isolated from wastewater was used to perform RT-qPCR using the SARS-CoV-2 RT-qPCR 222 Detection Kit for Wastewater (Promega Corp.). RT-qPCR reactions targeted the nucleocapsid (N1, N2) 223 and envelope (E) regions of the SARS-CoV-2 genome. To allow for quantitation, log dilutions (2x101 – 224 2x105 Genome Units (GU)/µL) of DNA quantitation standard were amplified alongside experimental 225 samples and used to construct a standard curve for each SARS-CoV-2 target. 20 µL amplification 226 reactions were composed of 15 µL reaction mastermix and 5uL of nucleic acid. 5 µL of nuclease-free 227 water was used as a no-template-control (NTC). Reactions were performed on a Stratagene MX3005 228 Real-Time Thermocycler (Agilent) with the following cycling conditions: reverse transcription for 15 229 minutes at 45°C, initial denaturation for 2 minutes at 95°C, and 40 cycles of 3 seconds at 95°C and 30 230 seconds at 62°C. Standard curve reactions were run in triplicate and wastewater sample reactions were 231 run in duplicate. The standard curve approach described above was used in parallel to estimate the 232 concentration of MS2 and PMMoV in purified samples. 233 234 Concentration of viral load in wastewater was calculated using the equation: 235 236 237 238 239 240 Normalization of SARS CoV-2 amounts in wastewater was performed by using the following equation: 241 242 243 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 10 Statistical analysis 244 Concordance between trend analysis using the three SARS-CoV-2 targets (N1, N2 and E) was evaluated 245 using Kendall’s coefficient of concordance (W). Correlation between weekly normalized SARS-CoV-2 246 levels from wastewater and number of new clinical cases (7 day moving average) was determined using 247 Kendall’s tau correlation coefficient. 248 249 SARS-CoV-2 amplicon library preparation and sequencing 250 10 µL of each total nucleic acid sample was treated for 30 minutes at 37°C with 1.5 U of RQ1 RNase-251 Free DNase (Promega Corporation) in a 15 µL reaction, as specified in the manufacturer’s protocol, 252 except that the Stop Solution provided was not used. 10 µL of nuclease-free water was added to each 253 sample and cleaned up with the ReliaPrep™ RNA Clean-Up and Concentration System (Promega 254 Corporation), as specified in the manufacturer’s technical manual. RNA was eluted with 15 µL of 255 nuclease-free water. 256 257 cDNA was synthesized as follows. 10 µL of either total nucleic acid sample or DNase-treated sample was 258 used as a template for first strand synthesis with random hexamers and associated steps using 259 Invitrogen™ SuperScript™ IV First-Strand Synthesis System (Thermo Fisher Scientific). Manufacturer’s 260 instructions were followed with one exception: incubation time at 50°C was increased from 10 minutes to 261 30 minutes, as recommended by the Swift Biosciences SARS-CoV-2 Additional Genome Coverage 262 amplicon panel library preparation protocol. 263 264 10 µL of cDNA was used as input for the SARS-CoV-2 Additional Genome Coverage panel (Swift 265 Biosciences) using the low input version of the library preparation protocol. This 345-amplicon panel 266 covers 99.7% of the SARS-CoV-2 genome and has amplicons ranging from 116 to 255 bp (average 150 267 bp). Libraries were quantified by qPCR, pooled, and sequenced with 2 x 150 base-pair reads on an 268 Illumina MiniSeq Instrument with a MiniSeq Mid Output Kit (300-cycles). 269 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 11 270 Amplicon sequencing analysis 271 Compressed, demultiplexed reads were obtained from the Illumina MiniSeq instrument and assessed for 272 sequencing quality using FastQC 0.11.9-0 273 (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/). For the purpose of comparing 274 performance between libraries, random subsampling was performed to obtain the same number of input 275 reads for each library using seqtk v1.2. All available reads for each library were used in known variant 276 detection. Adapter trimming was done using Fastp v0.20.132 (sequences 277 AGATCGGAAGAGCACACGTCTGAACTCCAGTCA and 278 AGATCGGAAGAGCGTCGTGTAGGGAAAGAGTGT). Trimmed reads were mapped to the SARS-279 CoV-2 reference genome (NC_045512.2) using BWA v 0.7.17. To assess human genomic contribution, 280 trimmed reads were also mapped to a combined SARS-CoV-2 + GRCh38/hg3833 (UCSC Genome 281 Browser34 Dec 2013 assembly) reference genome. Amplicon primers were trimmed following alignment 282 using Swift’s Primerclip tool v 0.3.8 (https://github.com/swiftbiosciences/primerclip) with the Swift-283 provided master file for the 345 amplicon panel as input. QC metrics were generated using Picard (v 284 2.9.2; http://broadinstitute.github.io/picard) BedToIntervalList, CollectTargetedPcrMetrics, 285 CollectGcBiasMetrics. The GATK335 DepthOfCoverage tool was used for target base coverage 286 assessments (median, mean, percent of bases at or above 1X, 100X, 1000X, and 5000X). Snakemake 287 v5.17.0 was used for workflow management on a Microsoft Azure CycleCloud instance. 288 289 Detection of known signature variants 290 Genomic locations of known signature variants were obtained from Nextstrain36 (accessed 12 March 291 2021) and UCSC Table Browser37 (SARS-CoV-2 Jan. 2020/NC_045512.2 Assembly (wuhCor1)) and 292 formatted into a BED file. With aligned, primer-trimmed reads in BAM format, the bam-readcount tool 293 (https://github.com/genome/bam-readcount) was used to collect base composition information at the 294 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 12 signature variant locations. The bam-readcount output was filtered to look for the presence of signature 295 variants with at least 10% frequency at positions covered to at least 50X, so that no signature variant 296 would be called as detected with fewer than 5 sequencing reads as evidence. 297 298

299 300 RT-qPCR Reaction Formulation and Setup 301 Operating under the assumption that wastewater is likely to contain RT-qPCR inhibitors, we formulated 302 enzyme mixes containing MMLV-RT enzyme and Taq DNA polymerase (with hot-start chemistry) that 303 would be resistant to PCR inhibitors. Next, we designed separate multiplexed RT-qPCR assays to detect 304 the nucleocapsid gene fragment N1 and N2 (as described by the CDC: 305 https://www.cdc.gov/coronavirus/2019-ncov/lab/rt-pcr-panel-primer-probes.html) and the envelope gene 306 (as described by Corman et al.38). SARS-CoV-2 targets were detected using primers and a target-specific 307 hydrolysis probes (labeled with 5’ FAM/ZEN™/ 3’ Iowa Black™ FQ). Each assay was multiplexed with 308 primers and a hydrolysis probe (labeled with 5’ Quasar®670/ BHQ®-2) complementary for Pepper Mild 309 Mottle Virus (PMMoV). PMMoV is a single-stranded RNA plant virus that commonly infects pepper 310 products intended for human consumption39. PMMoV RNA is detectable in wastewater worldwide and is 311 considered an important indicator of human-derived fecal pollution40-42. Each assay also includes an 312 exogenous RNA template, primers and hydrolysis probe (labeled with 5’ CalFluor® Orange 560/ BHQ®-1) 313 comprising an internal amplification control (IAC). The IAC’s cycle threshold (Ct) provides information 314 on the presence of reverse transcriptase and DNA polymerase inhibitors in the extracted nucleic acid 315 sample. CXR, (Carboxy-X-Rhodamine, Promega Corp.) which has similar spectral properties as ROX 316 (Ex: 580nm, Em: 602nm) was used as a reference dye. 317 318 We sought to optimize an appropriate amplicon length from the IAC template to provide sufficient 319 sensitivity to detect RT-qPCR inhibition. An in vitro transcribed Luciferase RNA (Promega Corp .) was 320 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 13 used as a template and four different amplicon lengths (93 bp, 285 bp, 310 bp and 435bp) were tested. 321 Humic acid, a known reverse transcriptase and DNA polymerase inhibitor 43, was titrated from 0 µg/mL 322 to 125 µg/mL using a two-fold dilution series, and inhibition was assessed by the difference in Ct value 323 with or without humic acid (Ct). The same hydrolysis probe was used in all cases. As expected, we 324 found a correlation between the amplicon length and sensitivity to humic acid (Fig . 1A), with the 435 bp 325 and 315 bp amplicons exhibiting higher ΔCt values compared to the 285 bp and 93 bp amplicons. Since 326 the RT-qPCR amplicons for detection of SARS-CoV-2 and PMMoV are below 150 bp in length and the 327 ΔCt values for 285bp and 93bp amplicons were similar, the 285 bp amplicon length was used as IAC 328 unless otherwise specified. 329 330 For WBE, it is important to quantify the viral genome units per volume of wastewater to determine the 331 quantitative trend in viral load. To analyze the analytical sensitivity, efficiency, and linearity of the assay, 332 a log dilution series of the in-vitro transcribed SARS-CoV-2 RNA (N and E) was used to perform RT-333 qPCR analysis. PCR amplification efficiencies for all three targets were between 90-120%. The limit of 334 detection (LOD95) for the three multiplexed assays for detecting the SARS-CoV-2 targets was 5 copies 335 and a limit of quantification (LOQ) of 8 copies. The RT-qPCR assays are linear in the tested range of 20-336 200,0000 copies, as the observed R2 for all the three targets were ≥ 0.99 (Fig. 1B). 337 The assay was also tested for specificity with other coronaviruses and respiratory pathogens. The assay 338 was found to be specific for detection of SARS-CoV-2 and PMMoV (Table S2). 339 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 14 340 Figure 1. (A) Optimization of Internal Amplification Control (IAC) amplicon size for 341 indication of inhibitor tolerance. Amplification of luc in-vitro transcribed RNA 342 template of listed amplicon sizes was performed with different humic acid 343 concentrations present in the RT-qPCR reaction. (B) Performance characteristics of the 344 three multiplexed RT-qPCR assays for detection of SARS-CoV-2 in wastewater. Standard 345 curve reactions were performed in triplicate. 346 347 Description of the direct-nucleic acid capture method 348 Existing literature suggests that SARS-CoV-2 may not be infectious in wastewater samples44,45. It is 349 therefore unknown if the viral genetic signature present in the wastewater samples is entirely derived 350 from compromised virions or if some proportion is present as unpackaged SARS-CoV-2 nucleic acid. We 351 hypothesized that by utilizing a direct capture method to bind total nucleic acid (TNA) to a silica-based 352 affinity resin in place of a method that is selective for intact viral particles, we may be able to eliminate 353 the viral concentration step which is often a cause of technical variability and low recovery. Direct TNA 354 isolation would be unbiased toward intact, partially intact, or free viral RNA. We developed a simple, 355 rapid, highly efficient, and modular alternative to existing wastewater RNA purification methods. The 356 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 15 primary concentration method utilizes raw sewage, to which chaotropic agents are added to allow binding 357 on to a silica matrix (PureYield™ Midi Binding column) by applying vacuum. The captured nucleic acid 358 is then subjected to successive alcohol washes (to remove RT-qPCR inhibitors that may have co-purified 359 with the nucleic acid) and subsequently eluted in water (Fig. 2). The eluted nucleic acid can then be 360 further processed in a second step with either a spin column with Wizard® Enviro Wastewater TNA kit or 361 with an automated nucleic acid purification system for the Maxwell® RSC Enviro Wastewater TNA kit 362 (Fig. 2). 363 364 Figure 2. Schematic overview of the direct capture nucleic acid purification process. 365 366 The direct-nucleic acid capture method was first tested for its ability to eliminate RT-qPCR inhibitors. 367 Inhibition was analyzed by comparing the difference in Ct value (ΔCt) between IAC amplification in 368 reactions for wastewater sample and for no-template-control (NTC) reactions. The ΔCt values were 369 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 16 analyzed for samples processed using either the primary concentration step with the PureYield™ Midi 370 column only or processed through the complete workflow comprising the Wizard® 371 concentration/purification steps outlined above. ΔCt values > 1 indicated the presence of reverse 372 transcriptase and/or DNA polymerase inhibitors in the wastewater samples added to the RT-qPCR 373 reaction. ΔCt values of <1 were observed for both nucleic acid purified with the PureYield™ Midi 374 column only and samples purified with both ™ Midi and PureYield™ Mini column steps (Fig. 3A). In the 375 second step, total nucleic acid eluted from the first step is further purified in a smaller volume (80 µL), 376 concentrating the nucleic acid by 12.5-fold. This concentration is evident when MS2 (viral spike-in 377 control) or SARS-CoV-2 (N1) is analyzed, as we observed a 8.34- and 5.54- fold increase in the amount 378 of MS2 and SARS-CoV-2 RNA detected, relative to the sample that only underwent initial sample 379 concentration (Fig. 3B and C). This indicates that the second purification step successfully concentrates 380 the nucleic acid in the sample. 381 Different laboratories have different throughput needs. Because our workflow can be done manually or 382 using automation for the final sample concentration, it provides flexibility for different types of labs. This 383 flexibility allows scale-up of the nucleic acid concentration process using an automated nucleic acid 384 purification workflow after the initial concentration step using the PureYield™ Midi column. We 385 compared the extraction efficiency of the nucleic acid isolated by either the manual spin column (Wizard® 386 Enviro Wastewater TNA kit) or with an automated nucleic acid purification system for the Maxwell® 387 RSC Enviro Wastewater TNA kit. Purified concentrations of PMMoV, MS2 (viral spike), and SARS-388 CoV-2 RNA were similar with both purification methods (Fig. 3D). These data indicate that both manual 389 and automated nucleic acid purification procedures can be used to extract viral genetic material at similar 390 extraction efficiencies. 391 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 17 392 Figure 3. (A) The direct capture method can reduce presence of RT-qPCR inhibitors 393 present in extracted samples. “Full workflow” refers to the combination of 394 concentration and purification steps shown in Fig. 2. ΔCt is defined as the 395 difference in internal amplification control (IAC) amplification (Ct Value) between 396 the NTC wells compared to sample wells. Viral concentrations detected via RT-qPCR 397 using listed workflows for (B) MS2 phage (viral wastewater spike) and (C) SARS CoV-2. 398 (D) The extraction efficiencies of the purification step performed using the Wizard 399 spin column or Maxwell automated instrument are similar for the extraction of PMMoV, 400 MS2 and SARS-CoV-2. Results shown are means ± SD (n= 3) 401 402 Sample volume considerations 403 SARS-CoV-2 levels in wastewater are often very low, making sample concentration from a larger volume 404 a necessary part of any WBE monitoring workflow. To determine what starting sample volume allows for 405 convenient and sensitive viral detection, we processed different volumes of wastewater (80 mL, 40 mL, 406 20 mL, 10 mL, and 5 mL) using the column-based manual concentration/purification scheme (Wizard®) 407 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 18 outlined above and performed all final elutions in the same volume (80 µL) of nuclease-free water. The 408 adjusted volumes of Protease solution, binding buffers 1 and 2 and isopropanol used for each starting 409 volume are outlined in Fig. 4A. MS2 bacteriophage was used as a viral spike-in recovery control. 410 The shift in Ct values from the NTC reactions (ΔCt) for the IAC exhibited a sample volume-dependent 411 increase in Ct value indicating a higher concentration of inhibitors may be co-purifying when larger 412 volumes are used. However, the ΔCt was <1 Ct for all the volumes tested, indicating co-purification of 413 RT-PCR inhibitors was not high enough to significantly impact data interpretation (Fig 4B). We also 414 analyzed the amounts of PMMoV, MS2 and SARS-CoV-2 (N1). As expected, all targets showed a 415 volume-dependent enrichment of genetic material. (Fig 4C-E). Though we can accurately detect SARS-416 CoV-2 signal from 5 ml of sample during the current sampling period when SARS-CoV-2 clinical cases 417 are high, use of 40 ml sample volume will allow sufficient assay sensitivity when viral loads are lower. 418 419 420 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 19 Figure 4. Sample volume considerations. (A) Volumes of reagents required for different 421 sample volumes. (B) Increasing the sample volume from 5 ml to 80 ml does not lead to 422 significant inhibition of RT-qPCR performance as measured using the Internal 423 amplification control (IAC). DCt represents the difference in Ct values of samples 424 compared to the NTC reactions. (C-E) Increase in the amount of viral genetic material 425 extracted as sample volume is increased as measured by RT-qPCR detection of PMMoV (C), 426 MS2 (D) and SARS-CoV-2 (E). Extraction efficiency of MS2 (F) and SARS-CoV-2 (G) 427 genetic material from wastewater samples with and without protease treatment. Results 428 shown are means ± SD (n= 3). 429 430 Enhancement of viral nucleic acid recovery with protease treatment 431 Wastewater, in addition to containing fecal matter and water, is also composed of cellulosic material from 432 toilet and tissue paper that may act as a substrate on which nucleic acids and nucleoprotein complexes can 433 aggregate. This material may form a large part of the suspended solids present in wastewater. SARS-434 CoV-2 may not be present as intact virions in wastewater, but the observation that the genetic material is 435 readily detectable indicates that the viral RNA is likely present in ribonucleoprotein complexes, which 436 shield it from nucleases that are presumably present in wastewater. In addition, detergents and chaotropic 437 agents present in wastewater may also cause structural changes to proteins causing association with 438 suspended solids. Therefore, we reasoned that a proteolytic cleavage step may be able to release some of 439 the viral genetic material associated with solids. 440 We measured viral nucleic acid extraction efficiency using a procedure that included alkaline protease 441 treatment. 40 mL wastewater samples were either treated with alkaline protease or left untreated and 442 processed as described in the Methods section. MS2 phage was also spiked into the samples. Percent 443 recoveries of MS2 nucleic acid were found to be 20% for untreated samples and 40% for samples that 444 were treated with alkaline protease (Fig. 4F). Similarly, we also observed a two-fold increase in 445 extraction of SARS-CoV-2 viral genome units when samples were treated with protease (Fig. 4G). These 446 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 20

indicate that alkaline protease treatment increases yield of viral TNA extraction, most likely by 447 releasing a portion of the viral genomic material associated with suspended solids. 448 449 Amount of viral matter associated with solids 450 In the purification workflow described above, suspended solids are removed via a brief centrifugation 451 following the alkaline protease treatment step. Removing suspended solids prevents clogging of the 452 PureYield™ Midi column. We investigated how much viral matter is associated with the pelleted solids 453 using the procedure described in the methods section. The eluted nucleic acid was analyzed for the 454 quantity of PMMoV and SARS-CoV-2 RNA. We observed that about 11.79% of the total amount of 455 PMMoV and 19.27% of the total amount of SARS-CoV-2 genetic material associated with the solids 456 (Fig. 5A-B). If maximal nucleic acid yield is desired, the solids-associated supernatant fraction can be 457 combined with the wastewater supernatant fraction to purify nucleic acid from the total pooled sample. 458 459 Figure 5. Amount of viral matter associated with solids after protease treatment. 460 Wastewater samples were centrifuged after protease treatment and resulting solids and 461 suspension fractions were processed as described in the Methods section. 462 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 21 Concentrations of PMMoV (A) and SARS-CoV-2 (B) RNA present in the two fractions as 463 measured by RT-qPCR are shown. Results shown are means ± SD (n= 3). 464 465 Comparison of viral nucleic acid recovery: Direct Capture method vs PEG/NaCl precipitation 466 Using the optimized method described above, we compared the direct capture method with PEG/NaCl 467 precipitation method for the ability to purify SARS-CoV-2 genetic material from wastewater samples. 468 PEG/NaCl is a widely used method for precipitation and concentration of non-enveloped enteric virus 469 such as Poliovirus28. 40 mL of wastewater sample was processed for the direct capture method and 120 470 mL of wastewater was processed using PEG/NaCl precipitation. We observed a 20-fold increased 471 extraction efficiency for extracting SARS-CoV-2 RNA (Fig. 6A) when using the direct capture method 472 compared to the PEG/NaCl protocol. For MS2 (viral spike control) we observed an extraction efficiency 473 of 3.76 ± 1.88% for the PEG/NaCl method and 39.67 ± 10.66% for the direct capture method (Fig 6B). 474 We also determined the percentage recovery for two human coronaviruses (OC43 and 229E) for the direct 475 capture method and they were 63.13 ± 4.16%, 40.09 ± 10.89% respectively (data not shown). 476 The direct capture purification workflow utilizing 40 ml of wastewater sample and a 40ul elution volume 477 has a concentration factor of 1000. With a LOD of 5 copies for the detection of SARS-CoV-2 by RT-478 qPCR reaction, and 1000-fold concentration in the purification process, the assay sensitivity is around 1 479 viral genome copies/ml. This level of sensitivity will be sufficient for trend analysis using WBE. 480 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 22 481 Figure 6. Comparison of the direct capture method with PEG/NaCl precipitation. 482 Wastewater samples were processed using the two methods as described in the Methods 483 section. (A) Concentration of SARS-CoV-2 RNA extracted is represented as GU/L in the 484 wastewater sample, (B) Percentage recovery for extraction of MS2 phage (viral spike 485 control). Results shown are means ± SD (n= 3). 486 487 SARS-CoV-2 RNA in wastewater from Dane County, Wisconsin 488 The levels of SARS-CoV-2 RNA from three wastewater treatment plants in Dane County, Wisconsin 489 from mid-October, 2020 to early January, 2021 were determined using the Wizard® Enviro Wastewater 490 TNA kit described above. Samples were processed and analyzed weekly. The levels of RT-qPCR 491 inhibitors present in the TNA samples (as assessed by shift in the Ct value (Ct) of IAC in sample wells 492 compared to NTC wells) were not notable, as Ct values were less than 0.5 for all wastewater samples 493 (Fig. 7A). 494 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 23 We analyzed SARS-CoV-2 RNA levels using N1, N2, and E targets (Fig. 7B). All samples were positive 495 for all three SARS-CoV-2 targets within the period tested. We calculated the degree of concordance 496 between the three SARS-CoV-2 targets for the three WWTPs over the sampling period using Kendall's 497 coefficient of concordance, W. Kendall's W ranges from 0 (no agreement) to 1 (full agreement). We 498 observed statistically significant concordance between the three SARS-CoV-2 targets for the three 499 WWTPs (W=0.97 for WWTP-1, W=0.96 for WWTP-2 and W=0.78 for WWTP-3). 500 WWTP-1 served a small community of around 10,000 people. We normalized SARS-CoV-2 levels with 501 PMMoV, which has been proposed to account for differential dilution and degradation rates over time. 502 We compared SARS-CoV-2 RNA signals in wastewater to the level of COVID-19 cases (7-day moving 503 average) declared by the municipality and analyzed the correlation between the two, resulting in a 504 Kendall’s Tau coefficient of 0.33, with a p-value of 0.08 (Fig 7C). The peak of SARS-CoV-2 genetic 505 signal observed in the wastewater is concurrent with the peak of positive SARS-CoV-2 reported in mid-506 November of 2020. Even with this limited data set we see the potential for wastewater-based surveillance 507 in assessing community-wide spread of the disease. 508 509 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 24 Figure 7. WBE analyses of samples from three wastewater treatment plants in Dane 510 County, WI. (A) Assessment of PCR inhibitors in nucleic acid isolated from the three 511 WWTP. Results shown are means ± SD (n> 30). (B) Levels of SARS-CoV-2 analyzed over the 512 indicated duration with three SARS-CoV-2 targets (N1, N2, E) for the three WWTP. 513

shown are means ± SD (n= 3). (C) Comparison of the normalized SARS-CoV-2 RNA 514 levels with the 7-day moving average of the new reported cases for the community 515 serving WWTP-1. 516 517 Amplicon sequencing of SARS-CoV-2 genetic material in wastewater 518 To determine if our direct capture purification method yielded nucleic acid compatible with next-519 generation sequencing, we prepared sequencing libraries with a subset of wastewater samples using a 520 commercially available SARS-CoV-2 amplicon panel and associated library preparation kit46,47. 521 For this proof of concept, we focused on three samples: wastewater from WWTP-2 in November of 2020 522 and January 2021, as well as wastewater collected from WWTP-3 in December of 2020. Because of the 523 proximity of the two collection sites to one another, we reasoned that these samples could be compared 524 broadly for the purpose of identifying SARS-CoV-2 variants, while also demonstrating the robustness of 525 the method to different wastewater treatment regimens from different facilities. 526 We compared total nucleic acid or DNase-treated nucleic acid as input into the library preparation 527 workflow. Most libraries had greater than 1 million reads, but for comparison of depth metrics, the total 528 number of reads was normalized across libraries by randomly subsampling to 650,000 total paired reads. 529 We aligned the subsampled, filtered sequencing reads to the SARS-CoV-2 genome and measured the 530 percentage of filtered reads aligned and the depth of coverage across amplicons tiling the SARS-CoV-2 531 genome (Fig. 8A-B). Samples that were DNase-treated had higher percentage of filtered reads aligning to 532 the SARS-CoV-2 genome and greater depth of coverage compared to total nucleic acid libraries. Because 533 wastewater is a heterogeneous sample, it is not surprising that the amplification reaction resulted in off-534 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 25 target amplification. This data shows that DNase treatment of wastewater total nucleic acid improves 535 next-generation sequencing quality from the SARS-CoV-2 amplicon panel tested. 536 We aligned the reads a second time to a combined reference genome containing both SARS-CoV-2 and 537 human reference genomes to examine the proportion of aligned, filtered reads attributable to 538 contamination by human genetic material. Surprisingly, we found that the percentage of filtered reads 539 aligning to the human genome increased after DNase treatment (3.5%-9.4% for total nucleic acid and 540 6.3%-16.5% for DNase-treated samples). However, the proportion of non-human contamination reduced 541 by DNase treatment makes this the preferred protocol. 542 Finally, using all available reads, we analyzed the aligned sequencing reads from the DNase-treated 543 samples to look for the presence of previously reported SARS-CoV-2 variants of concern associated with 544 widespread viral strains. The variants that we detected (Supplemental file 1) were consistent with known 545 variants in Dane County, Wisconsin during the sampling period36. Virtually all the SARS-CoV-2 546 sequences contained mutations corresponding to Nextstrain clade 20A. Mutations found in Nextstrain 547 clades 20B, 20C, and 20G were also present, though mutations specific to strain 20B were present at 548 2.0% frequency in January, below the 10% frequency cutoff we had set for making a positive call. 549 Although merely suggestive due to small sample size, this does correspond to Nextstrain reported data 550 which also suggests a decrease in the prevalence of this strain over the sampling period. More 551 transmissible strains were not observed in this data, consistent with Nextstrain reports for such strains 552 initially appearing in this geographic area approximately 2 months after the sampling period. 553 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 26 554 Figure 8. Analysis of wastewater nucleic acid for SARS-CoV-2 sequencing. Nucleic acid 555 from the three wastewater collection sites and a pooled sample from the three 556 collection sites at the indicated timepoints was used to generate SARS-CoV-2 amplicon 557 libraries. Each dataset was randomly subsampled to 650,000 reads for comparison. (A) 558 Percent of filtered reads aligned to SARS-CoV-2 genome for DNase-treated samples (+ 559 DNase) or untreated total nucleic acid samples (- DNase). (B) Percent of target bases 560 on the SARS-CoV-2 genome covered at or above indicated depths. 561 562

563 In this study, we describe a convenient, high throughput, robust, and consistent method to directly 564 capture, concentrate, and detect total nucleic acids (TNA) from wastewater using silica based PureYield® 565 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 27 columns and optimized RT-qPCR. This method offers ease-of-use and minimizes the need for specialized 566 laboratory equipment. In addition, the method achieves consistent recovery rates and significant reduction 567 in RT-qPCR inhibitors. 568 During the course of this study, alternative direct capture methods for extraction of SARS-CoV-2 RNA 569 from wastewater have been published, highlighting the simplicity of the general workflow48. Direct 570 capture methods have also been applied in large-scale interlaboratory method assessment studies where 571 they have shown superior and consistent performance over other methods including PEG/NaCl 572 precipitation, centrifugal ultrafiltration, or charged membranes49. By using a protease treatment step, we 573 are able to release a significant portion of viral nucleic acid associated with the solids in the wastewater. 574 In addition, chaotropes and alcohols provide an effective nucleic acid binding environment for capture on 575 a silica matrix. The two-step, modular purification strategy described in this work simplifies the 576 workflow for users processing either small or large amounts of wastewater samples. The flexibility of the 577

and compatibility of the resulting nucleic acid with downstream analysis by RT-qPCR and 578 SARS-CoV-2 sequencing allow for straightforward adoption for WBE-based viral surveillance 579 approaches. 580 Throughout the early course of the COVID-19 pandemic, the ramifications of not having nationwide 581 surveillance systems in place were observed. WBE and clinical diagnostic testing each can provide 582 structured surveillance systems. WBE can complement clinical diagnostic testing by independently 583 confirming prevalence of disease communities and possibly providing an early warning for future viral 584 outbreaks. WBE also provides a low-cost tool to understand community spread in low resource areas. 585 Similar to diagnostic clinical testing, WBE has experimental limitations (uncertainties related to timing 586 and quantities of viral and viral nucleic acid shedding, RNA stability, effect of temperature, and sample 587 processing techniques) that need to be well understood before using acquired data to inform 588 epidemiological and public health efforts around the globe. 589 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 28 As the COVID-19 pandemic has progressed, SARS-CoV-2 genetic variants have arisen, often leading to 590 increased transmissibility, concern about immune evasion, and subsequent outbreaks50. With the 591 emergence of new strains, it will be informative to see how this rapidly evolving method is able to help 592 understand the spread of variants within communities and guide health authorities to take appropriate 593 measures51-53. 594 595 Data Statement 596 Sequencing data are available upon request of the corresponding author. 597

598 Amplicon library preparation guidance, basic analysis guidance, and MiniSeq sequencing were provided 599 by Swift Biosciences, Inc (https://swiftbiosci.com). We thank Paul Muschler for providing valuable 600 feedback. We acknowledge the staff of Madison Metropolitan Sewage District (MMSD), Village of 601 Oregon, WI and City of Sun Prairie, WI for their support and providing wastewater samples. We also 602 thank the various research groups, public health, and commercial labs with whom we have had many 603 productive discussions on this topic in these difficult times. 604 605 Authorship contribution statement 606 Subhanjan Mondal: Conceptualization, Data curation, Formal analysis, Supervision, Methodology, 607 Investigation, Roles/Writing - original draft, Writing - review & editing. Nathan Feirer: 608 Conceptualization, Data curation Formal analysis, Methodology, Investigation, Roles/Writing - original 609 draft, Writing - review & editing. Michael Brockman: Data curation, Formal analysis, Methodology, 610 Roles/Writing - original draft, Writing - review & editing. Melanie A. Preston: Investigation, Writing - 611 review & editing. Sarah J. Teter: Investigation, Writing - review & editing. Dongping Ma: 612 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 29 Investigation, Said A. Goueli: Resources, Writing - review & editing. Sameer Moorji: 613 Conceptualization, Resources. Brigitta Saul: Conceptualization, Resources. James J. Cali: Resources, 614 Supervision, Writing - review & editing. 615 616 Declaration of competing interest 617 All authors are employed by Promega Corporation. 618 619

620 621 1 Zhu, N. et al. A Novel Coronavirus from Patients with Pneumonia in China, 2019. N Engl J Med 622 382, 727-733, doi:10.1056/NEJMoa2001017 (2020). 623 2 Hartenian, E. et al. The molecular virology of coronaviruses. J Biol Chem 295, 12910-12934, 624 doi:10.1074/jbc.REV120.013930 (2020). 625 3 Cui, J., Li, F. & Shi, Z.-L. Origin and evolution of pathogenic coronaviruses. Nature Reviews 626 Microbiology 17, 181-192, doi:10.1038/s41579-018-0118-9 (2019). 627 4 Zhou, P. et al. A pneumonia outbreak associated with a new coronavirus of probable bat origin. 628 Nature 579, 270-273, doi:10.1038/s41586-020-2012-7 (2020). 629 5 He, X. et al. Temporal dynamics in viral shedding and transmissibility of COVID-19. Nat Med 630 26, 672-675, doi:10.1038/s41591-020-0869-5 (2020). 631 6 Cevik, M. et al. SARS-CoV-2, SARS-CoV, and MERS-CoV viral load dynamics, duration of 632 viral shedding, and infectiousness: a systematic review and meta-analysis. The Lancet Microbe, 633 doi:10.1016/S2666-5247(20)30172-5 (2020). 634 7 Furukawa, N., Brooks, J. & Sobel, J. Evidence Supporting Transmission of Severe Acute 635 Respiratory Syndrome Coronavirus 2 While Presymptomatic or Asymptomatic. Emerging 636 Infectious Disease journal 26, doi:10.3201/eid2607.201595 (2020). 637 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 30 8 Zhang, Z. et al. Early viral clearance and antibody kinetics of COVID-19 among asymptomatic 638 carriers. 2020.2004.2028.20083139, doi:10.1101/2020.04.28.20083139 %J medRxiv (2020). 639 9 Park, S. K. et al. Detection of SARS-CoV-2 in Fecal Samples From Patients With Asymptomatic 640 and Mild COVID-19 in Korea. Clin Gastroenterol Hepatol, doi:10.1016/j.cgh.2020.06.005 641 (2020). 642 10 Day, M. Covid-19: identifying and isolating asymptomatic people helped eliminate virus in 643 Italian village. 368, m1165, doi:10.1136/bmj.m1165 %J BMJ (2020). 644 11 Wang, W. et al. Detection of SARS-CoV-2 in Different Types of Clinical Specimens. JAMA 323, 645 1843-1844, doi:10.1001/jama.2020.3786 %J JAMA (2020). 646 12 Chen, Y. et al. The presence of SARS-CoV-2 RNA in the feces of COVID-19 patients. J Med 647 Virol 92, 833-840, doi:10.1002/jmv.25825 (2020). 648 13 Zheng, S. et al. Viral load dynamics and disease severity in patients infected with SARS-CoV-2 649 in Zhejiang province, China, January-March 2020: retrospective cohort study. BMJ 369, m1443, 650 doi:10.1136/bmj.m1443 (2020). 651 14 Wolfel, R. et al. Virological assessment of hospitalized patients with COVID-2019. Nature 581, 652 465-469, doi:10.1038/s41586-020-2196-x (2020). 653 15 Han, M. S. et al. Viral RNA Load in Mildly Symptomatic and Asymptomatic Children with 654 COVID-19, Seoul, South Korea. Emerging Infectious Disease journal 26, 2497, 655 doi:10.3201/eid2610.202449 (2020). 656 16 Sethuraman, N., Jeremiah, S. S. & Ryo, A. Interpreting Diagnostic Tests for SARS-CoV-2. JAMA 657 323, 2249-2251, doi:10.1001/jama.2020.8259 (2020). 658 17 Lorenzo, M. & Picó, Y. Wastewater-based epidemiology: current status and future prospects. 659 Current Opinion in Environmental Science & Health 9, 77-84, 660 doi:https://doi.org/10.1016/j.coesh.2019.05.007 (2019). 661 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 31 18 Sims, N. & Kasprzyk-Hordern, B. Future perspectives of wastewater-based epidemiology: 662 Monitoring infectious disease spread and resistance to the community level. Environ Int 139, 663 105689, doi:10.1016/j.envint.2020.105689 (2020). 664 19 Hellmer, M. et al. Detection of pathogenic viruses in sewage provided early warnings of hepatitis 665 A virus and norovirus outbreaks. Appl Environ Microbiol 80, 6771-6781, 666 doi:10.1128/AEM.01981-14 (2014). 667 20 Hovi, T. et al. Role of environmental poliovirus surveillance in global polio eradication and 668 beyond. Epidemiol Infect 140, 1-13, doi:10.1017/S095026881000316X (2012). 669 21 Asghar, H. et al. Environmental surveillance for polioviruses in the Global Polio Eradication 670 Initiative. J Infect Dis 210 Suppl 1, S294-303, doi:10.1093/infdis/jiu384 (2014). 671 22 Medema, G., Heijnen, L., Elsinga, G., Italiaander, R. & Brouwer, A. Presence of SARS-672 Coronavirus-2 RNA in Sewage and Correlation with Reported COVID-19 Prevalence in the Early 673 Stage of the Epidemic in The Netherlands. Environmental Science & Technology Letters, 674 doi:10.1021/acs.estlett.0c00357 (2020). 675 23 La Rosa, G. et al. First detection of SARS-CoV-2 in untreated wastewaters in Italy. Sci Total 676 Environ 736, 139652, doi:10.1016/j.scitotenv.2020.139652 (2020). 677 24 Ahmed, W. et al. First confirmed detection of SARS-CoV-2 in untreated wastewater in Australia: 678 A proof of concept for the wastewater surveillance of COVID-19 in the community. Sci Total 679 Environ 728, 138764, doi:10.1016/j.scitotenv.2020.138764 (2020). 680 25 Peccia, J. et al. Measurement of SARS-CoV-2 RNA in wastewater tracks community infection 681 dynamics. Nature Biotechnology 38, 1164-1167, doi:10.1038/s41587-020-0684-z (2020). 682 26 Randazzo, W. et al. SARS-CoV-2 RNA in wastewater anticipated COVID-19 occurrence in a 683 low prevalence area. Water Res 181, 115942, doi:10.1016/j.watres.2020.115942 (2020). 684 27 Wurtzer, S. et al. Evaluation of lockdown impact on SARS-CoV-2 dynamics through viral 685 genome quantification in Paris wastewaters. 2020.2004.2012.20062679, 686 doi:10.1101/2020.04.12.20062679 %J medRxiv (2020). 687 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 32 28 Bofill-Mas, S. & Rusiñol, M. Recent trends on methods for the concentration of viruses from 688 water samples. Current Opinion in Environmental Science & Health 16, 7-13, 689 doi:https://doi.org/10.1016/j.coesh.2020.01.006 (2020). 690 29 Ahmed, W. et al. Comparison of virus concentration methods for the RT-qPCR-based recovery of 691 murine hepatitis virus, a surrogate for SARS-CoV-2 from untreated wastewater. Sci Total 692 Environ 739, 139960, doi:10.1016/j.scitotenv.2020.139960 (2020). 693 30 Kitajima, M. et al. SARS-CoV-2 in wastewater: State of the knowledge and research needs. 694 Science of The Total Environment 739, 139076, 695 doi:https://doi.org/10.1016/j.scitotenv.2020.139076 (2020). 696 31 Pecson, B. M., Martin, L. V. & Kohn, T. Quantitative PCR for determining the infectivity of 697 bacteriophage MS2 upon inactivation by heat, UV-B radiation, and singlet oxygen: advantages 698 and limitations of an enzymatic treatment to reduce false-positive results. Appl Environ Microbiol 699 75, 5544-5554, doi:10.1128/AEM.00425-09 (2009). 700 32 Chen, S., Zhou, Y., Chen, Y. & Gu, J. fastp: an ultra-fast all-in-one FASTQ preprocessor. 701 Bioinformatics 34, i884-i890, doi:10.1093/bioinformatics/bty560 %J Bioinformatics (2018). 702 33 Lander, E. S. et al. Initial sequencing and analysis of the human genome. Nature 409, 860-921, 703 doi:10.1038/35057062 (2001). 704 34 Kent, W. J. et al. The Human Genome Browser at UCSC. 12, 996-1006, doi:10.1101/gr.229102 705 (2002). 706 35 McKenna, A. et al. The Genome Analysis Toolkit: A MapReduce framework for analyzing next-707 generation DNA sequencing data. 20, 1297-1303, doi:10.1101/gr.107524.110 (2010). 708 36 Hadfield, J. et al. Nextstrain: real-time tracking of pathogen evolution. Bioinformatics 34, 4121-709 4123, doi:10.1093/bioinformatics/bty407 (2018). 710 37 Karolchik, D. et al. The UCSC Table Browser data retrieval tool. Nucleic Acids Research 32, 711 D493-D496, doi:10.1093/nar/gkh103 %J Nucleic Acids Research (2004). 712 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 33 38 Corman, V. M. et al. Detection of 2019 novel coronavirus (2019-nCoV) by real-time RT-PCR. 713 Euro Surveill 25, doi:10.2807/1560-7917.ES.2020.25.3.2000045 (2020). 714 39 Colson, P. et al. Pepper Mild Mottle Virus, a Plant Virus Associated with Specific Immune 715 Responses, Fever, Abdominal Pains, and Pruritus in Humans. PLOS ONE 5, e10041, 716 doi:10.1371/journal.pone.0010041 (2010). 717 40 Kitajima, M., Sassi, H. P. & Torrey, J. R. Pepper mild mottle virus as a water quality indicator. 718 npj Clean Water 1, 19, doi:10.1038/s41545-018-0019-5 (2018). 719 41 Bivins, A. et al. Cross-assembly phage and pepper mild mottle virus as viral water quality 720 monitoring tools—potential, research gaps, and way forward. Current Opinion in Environmental 721 Science & Health 16, 54-61, doi:https://doi.org/10.1016/j.coesh.2020.02.001 (2020). 722 42 Rosario, K., Symonds, E. M., Sinigalliano, C., Stewart, J. & Breitbart, M. Pepper Mild 723 Mottle Virus as an Indicator of Fecal Pollution. 75, 7261-7267, doi:10.1128/AEM.00410-724 09 %J Applied and Environmental Microbiology (2009). 725 43 Schrader, C., Schielke, A., Ellerbroek, L. & Johne, R. PCR inhibitors – occurrence, properties 726 and removal. 113, 1014-1026, doi:https://doi.org/10.1111/j.1365-2672.2012.05384.x (2012). 727 44 Bivins, A. et al. Persistence of SARS-CoV-2 in Water and Wastewater. Environmental Science & 728 Technology Letters 7, 937-942, doi:10.1021/acs.estlett.0c00730 (2020). 729 45 Dada, A. C. & Gyawali, P. Quantitative microbial risk assessment (QMRA) of occupational 730 exposure to SARS-CoV-2 in wastewater treatment plants. Science of The Total Environment, 731 142989, doi:https://doi.org/10.1016/j.scitotenv.2020.142989 (2020). 732 46 Fuqua, J. L. et al. A rapid assessment of wastewater for genomic surveillance of SARS-CoV-2 733 variants at sewershed scale in Louisville, KY. 2021.2003.2018.21253604, 734 doi:10.1101/2021.03.18.21253604 %J medRxiv (2021). 735 47 Fontenele, R. S. et al. High-throughput sequencing of SARS-CoV-2 in wastewater provides 736 insights into circulating variants. 2021.2001.2022.21250320, 737 doi:10.1101/2021.01.22.21250320 %J medRxiv (2021). 738 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint 34 48 Whitney, O. N. et al. Sewage, Salt, Silica, and SARS-CoV-2 (4S): An Economical Kit-Free 739

for Direct Capture of SARS-CoV-2 RNA from Wastewater. Environ Sci Technol 55, 740 4880-4888, doi:10.1021/acs.est.0c08129 (2021). 741 49 Pecson, B. M. et al. Reproducibility and sensitivity of 36 methods to quantify the SARS-CoV-2 742 genetic signal in raw wastewater: findings from an interlaboratory methods evaluation in the U.S. 743 Environmental Science: Water Research & Technology 7, 504-520, doi:10.1039/D0EW00946F 744 (2021). 745 50 Mascola, J. R., Graham, B. S. & Fauci, A. S. SARS-CoV-2 Viral Variants—Tackling a Moving 746 Target. JAMA 325, 1261-1262, doi:10.1001/jama.2021.2088 %J JAMA (2021). 747 51 Crits-Christoph, A. et al. Genome Sequencing of Sewage Detects Regionally Prevalent SARS-748 CoV-2 Variants. 12, e02703-02720, doi:10.1128/mBio.02703-20 %J mBio (2021). 749 52 Izquierdo-Lara, R. et al. Monitoring SARS-CoV-2 Circulation and Diversity through Community 750 Wastewater Sequencing, the Netherlands and Belgium. Emerging Infectious Disease journal 27, 751 1405, doi:10.3201/eid2705.204410 (2021). 752 53 La Rosa, G. et al. Rapid screening for SARS-CoV-2 variants of concern in clinical and 753 environmental samples using nested RT-PCR assays targeting key mutations of the spike protein. 754 Water Research 197, 117104, doi:https://doi.org/10.1016/j.watres.2021.117104 (2021). 755 756 757 758 . CC-BY-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 10, 2021. ; https://doi.org/10.1101/2021.05.06.21256753doi: medRxiv preprint