key: cord-0919762-u22p05u5
authors: Rubio-Acero, Raquel; Beyerl, Jessica; Muenchhoff, Maximilian; Roth, Marc Sancho; Castelletti, Noemi; Paunovic, Ivana; Radon, Katja; Springer, Bernd; Nagel, Christian; Boehm, Bernhard; Böhmer, Merle M.; Graf, Alexander; Blum, Helmut; Krebs, Stefan; Keppler, Oliver T.; Osterman, Andreas; Khan, Zohaib Nisar; Hölscher, Michael; Wieser, Andreas; Emad, Alamoudi; Jared, Anderson; Abhishek, Bakuli; Maxilmilian, Baumann; Marc, Becker; Franziska, Bednarzki; Olimbek, Bemirayev; Jessica, Beyerl; Patrick, Bitzer; Rebecca, Böhnlein; Isabel, Brand; Jan, Bruger; Friedrich, Caroli; Noemi, Castelletti; Josephine, Coleman; Lorenzo, Contento; Alina, Czwienzek; Flora, Deák; Diefenbach Maximilian, N.; Jana, Diekmannshemke; Gerhard, Dobler; Jürgen, Durner; Ute, Eberle; Judith, Eckstein; Tabea, Eser; Philine, Falk; Manuela, Feyereisen; Volker, Fingerle; Felix, Forster; Turid, Frahnow; Jonathan, Frese; Günter, Fröschl; Christiane, Fuchs; Mercè, Garí; Otto, Geisenberger; Christof, Geldmacher; Leonard, Gilberg; Kristina, Gillig; Philipp, Girl; Elias, Golschan; Michelle, Guggenbuehl Noller Jessica; Maria, Guglielmini Elena; Pablo, Gutierrez; Anslem, Haderer; Marlene, Hannes; Lena, Hartinger; Jan, Hasenauer; Alejandra, Hernandez; Leah, Hillari; Christian, Hinske; Tim, Hofberger; Michael, Hölscher; Sacha, Horn; Kristina, Huber; Christian, Janke; Ursula, Kappl; Antonia, Keßler; Zohaib, Khan; Johanna, Kresin; Inge, Kroidl; Arne, Kroidl; Magdalena, Lang; Clemens, Lang; Silvan, Lange; Michael, Laxy; Ronan, Le Gleut; Reiner, Leidl; Leopold, Liedl; Xhovana, Lucaj; Fabian, Luppa; Sophie, Nafziger Alexandra; Petra, Mang; Alisa, Markgraf; Rebecca, Mayrhofer; Dafni, Metaxa; Hannah, Müller; Katharina, Müller; Laura, Olbrich; Ivana, Paunovic; Michael, Plank; Claire, Pleimelding; Michel, Pletschette; Michael, Pritsch; Stephan, Prückner; Kerstin, Puchinger; Peter, Pütz; Katja, Radon; Elba, Raimundéz; Jakob, Reich; Friedrich, Riess; Camilla, Rothe; Raquel, Rubio-Acero; Viktoria, Ruci; Elmar, Saathoff; Nicole, Schäfer; Yannik, Schälte; Benedikt, Schluse; Lara, Schneider; Mirjam, Schunk; Lars, Schwettmann; Alba, Soler; Peter, Sothmann; Kathrin, Strobl; Jeni, Tang; Fabian, Theis; Verena, Thiel; Sophie, Thiesbrummel; Vincent, Vollmayr; Emilia, Von Lovenberg; Jonathan, Von Lovenberg; Julia, Waibel; Claudia, Wallrauch; Andreas, Wieser; Simon, Winter; Roman, Wölfel; Julia, Wolff; Tobias, Würfel; Sabine, Zange; Eleftheria, Zeggini; Anna, Zielke; Thorbjörn, Zimmer
title: Spatially resolved qualified sewage spot sampling to track SARS-CoV-2 dynamics in Munich - One year of experience
date: 2021-07-21
journal: Sci Total Environ
DOI: 10.1016/j.scitotenv.2021.149031
sha: a40483ddf92dad2ea8cc0c7a8d1454f191dbbb38
doc_id: 919762
cord_uid: u22p05u5

Wastewater-based epidemiology (WBE) is a tool now increasingly proposed to monitor the SARS-CoV-2 burden in populations without the need for individual mass testing. It is especially interesting in metropolitan areas where spread can be very fast, and proper sewage systems are available for sampling with short flow times and thus little decay of the virus. We started in March 2020 to set up a once-a-week qualified spot sampling protocol in six different locations in Munich carefully chosen to contain primarily wastewater of permanent residential areas, rather than industry or hospitals. We used RT-PCR and sequencing to track the spread of SARS-CoV-2 in the Munich population with temporo-spatial resolution. The study became fully operational in mid-April 2020 and has been tracking SARS-CoV-2 RNA load weekly for one year. Sequencing of the isolated viral RNA was performed to obtain information about the presence and abundance of variants of concern in the Munich area over time. We demonstrate that the evolution of SARS-CoV-2 RNA loads (between <7.5 and 3874/mL) in these different areas within Munich correlates well with official seven day incidence notification data (between 0.0 and 327 per 100,000) obtained from the authorities within the respective region. Wastewater viral loads predicted the dynamic of SARS-CoV-2 local incidence about 3 weeks in advance of data based on respiratory swab analyses. Aligning with multiple different point-mutations characteristic for certain variants of concern, we could demonstrate the gradual increase of variant of concern B.1.1.7 in the Munich population beginning in January 2021, weeks before it became apparent in sequencing results of swabs samples taken from patients living in Munich. Overall, the study highlights the potential of WBE to monitor the SARS-CoV-2 pandemic, including the introduction of variants of concern in a local population.

residential areas, rather than industry or hospitals. We used RT-PCR and sequencing to track the spread of SARS-CoV-2 in the Munich population with temporo-spatial resolution.

The study became fully operational in mid-April 2020 and has been tracking SARS-CoV-2 RNA load weekly for one year. Sequencing of the isolated viral RNA was performed to obtain information about the presence and abundance of variants of concern in the Munich area over time.

We demonstrate that the evolution of SARS-CoV-2 RNA loads (between <7.5 and 3874 per mL) in these different areas within Munich correlates well with official seven day incidence notification data (between 0.0 and 327 per 100,000) obtained from the authorities within the respective region. Wastewater viral loads predicted the dynamic of SARS-CoV-2 local incidence about 3 weeks in advance of data based on respiratory swab analyses. Aligning with multiple different point-mutations characteristic for certain variants of concern, we could demonstrate the gradual increase of variant of concern B.1.1.7 in the Munich population beginning in January 2021, weeks before it became apparent in sequencing results of swabs samples taken from patients living in Munich.

Overall, the study highlights the potential of WBE to monitor the SARS-CoV-2 pandemic, including the introduction of variants of concern in a local population.

J o u r n a l P r e -p r o o f Severe acute respiratory syndrome virus 2 (SARS-CoV-2) was first described in the city of Wuhan in December 2019 and has quickly spread around the world resulting in the current pandemic situation with more than 141,754,944 confirmed Coronavirus disease 2019 cases and 3,025,835 deaths as of April 20 th , 2021 (1) . In order to mitigate the spread of the virus and to avoid the collapse of the health care sector, governments have resorted to different policies including strict lockdowns or even curfews. The world is still struggling to adapt to the unprecedented economic, behavioural and societal changes associated with the pandemic. On a global scale, the COVID-19 pandemic caused by the SARS-CoV-2 virus can be regarded as the most serious economic crisis since World War II (2) (3) (4) . Transmission of SARS-CoV-2 occurs primarily through aerosols and droplets excreted from infected individuals while breathing, sneezing or coughing. Pre-and oligosymptomatic patients are unaware of their disease and can also be infectious, therefore the spread is very difficult to control. As viruses of the family Coronaviridae are enveloped single-strand, positive-sense RNA viruses, infected individuals can be identified by reverse transcription PCR (RT-qPCR) of naso-oropharyngeal swabs. This is considered the gold standard of laboratory-based detection of SARS-CoV-2 (5). For population surveillance however, this approach is not feasible, due to the number of swabs and RT-PCR reactions needed. In resource limited settings, the situation is even more challenging, and mass testing is neither feasible nor affordable.

Nevertheless, governments and health authorities need data to adapt their actions to the current local epidemiology. Many countries use mandatory notification of newly diagnosed cases and the number of hospital beds occupied by COVID-19 patients to get indications for disease activity. Due to the large fraction of untested oligo-/asymptomatic courses of disease and the often late onset of complications requiring hospital treatment, actions are often taken too late and outbreaks are detected only in advanced states.

J o u r n a l P r e -p r o o f Thus, efficient surveillance systems delivering unbiased information about the local disease burden are indispensable. It was reported that SARS-CoV-2 can be found in the faeces and urine of symptomatic and asymptomatic infected subjects (6) . Subsequently, RT-PCR from wastewater was successfully performed demonstrating SARS-CoV-2 RNA in sewage systems (7) (8) (9) . Even infections related to direct contact with human excretions have been described and warrants further investigation (10) . This is certainly a less prominent transmission path compared to the airborne route, and the duration of infectivity of the virus in sewage water is so far unclear. Still, wastewater-based epidemiology (WBE) could be a potentially useful complementary tool for the environmental surveillance of local SARS-CoV-2 outbreaks (11) (12) (13) (14) . Wastewater-based surveillance overcomes the need to test a large proportion of the population avoiding the biases of other epidemiological indicators, but still tracking the infectious agents in communities (11, 14, 15) . Recently, several reports have demonstrated a significant correlation between SARS-CoV-2 RNA concentration from wastewater samples and actual prevalence of SARS-CoV-2 infections over a defined period of time, complementing conventional screening and notification approaches (4, 8, 14, 16, 17) .

We started in March 2020 to investigate the feasibility of WBE-approaches for SARS-CoV-2 in Munich, Bavaria, southern Germany. We chose to perform qualified spot sampling of sewage in the morning once weekly in six different positions in the Munich sewage system, covering close to one third of the population of the city (504,807 /1,560,042 inhabitants; 32.4%). Emphasis was placed on short flow times from sink to sampling and drainage areas including primarily residential areas without larger hospitals, industrial complexes or significant foreign water influx.

To keep the protocol convenient and cost efficient, filtration with 1mm stainless steel strainer and subsequent ultracentrifugation was used to concentrate SARS-CoV-2 for RNA isolation. Viral load measurements were performed with RT-qPCR and digital droplet RT-PCR (dd J o u r n a l P r e -p r o o f RT-PCR) against two independent target genes. A subset of RNA eluates was also sequenced after PCR amplification using the ARTIC-protocol primer sets (18), to analyse viral genomic information.

Untreated wastewater (sewage) samples were collected weekly since mid-April 2020 from six different locations of the sewage system in Munich, Germany. The drainage areas were selected from geographically distinct regions within the city (figure 1). Maximum flow times from sink to sampling was selected to be five hours. The drainage area should not include industrial complexes, and common foreign water influx should be below 20%. There should also be no significant collateral sewage draining pipes. Five of the six regions (Numbers 1, 3, 4, 5, 6 in table 1 and figure 1) were chosen to be residential, without significant influx from hospitals. The sixth selected region (Number 2 in table 1 and figure 1) corresponded to an area with relatively few permanent inhabitants, but comprising the LMU university hospital at the Campus Grosshadern, the largest tertiary care facility of the region. Qualified spot sampling was performed during the morning flush surge. Sewage pH and temperature were measured in situ and the samples were consecutively transported to the laboratory. Care was taken to cool the samples to 4-6 ºC in a cooler box immediately. Water was transported in 500 ml cleaned and autoclaved amber glass bottles. Most samples were processed immediately upon arrival in the laboratory. Some samples had to be stored at -80 °C until further analysis. Comparison between fresh and -80 ºC stored samples was performed to evaluate possible RNA degradation and signal losses due to the freeze-thaw cycle.

Upon arrival in the laboratory, sewage was drained through a 1mm stainless steel strainer and collected in disposable 50 ml centrifuge tubes (Corning 430829). For direct nucleic acid J o u r n a l P r e -p r o o f extraction after the sewage collection, one 50 ml tube from each location was centrifuged with 3.000 g at 4 ºC for 20 minutes to pellet the debris, the other sewage samples were immediately stored at -80 ºC for further analysis. 38 ml of the debris-free sewage was transferred to high-speed centrifuge tubes (Nalgene 3138-0050). Samples were then centrifuged with 26.000 g at 4 ºC for 1 hour. Pellets were resuspended in 200 μl nuclease-free water (Ambion A9937/VWR 436912C). For nucleic acid extraction from frozen samples, two 50 ml centrifuge tubes from each location were placed at room temperature for about 3 hours to slowly thaw. Afterwards, sample debris was pelleted as mentioned above. 76 ml (twice 38 ml) of the debris-free sewage was transferred to high-speed centrifuge tubes for ultracentrifugation. Pellets from the same location were homogenized and re-suspended in a total of 200 μl nuclease-free water. In direct comparison, using double the volume corrected for freeze-thaw losses encountered. An internal positive control was added to the concentrate before extraction to verify the efficacy of the RNA isolation. As a negative control, clear tap water was treated in parallel throughout the process. Due to severe delays in the delivery of the AllPrep PowerViral DNA/RNA Kit (Qiagen), we also used the RNeasy PowerMicrobiome Kit (Qiagen) which contains the same buffer solutions also found in the other kit. The performance of the two kits was compared and found to be similar, when RNA isolation was performed following the manufacturer's recommendations for the AllPrep PowerViral DNA/RNA Kit for both kits. RNA was eluted in 50 μl and reloaded in the spin column to increase sample concentration.

SARS-CoV-2 RNA was detected using the nucleocapsid (N1) primer/probe combination described in the CDC protocol (19) . PCR reactions were performed using the one-step QuantiNova Multiplex RT-PCR Kit on a Roche LightCycler 480 II as described in more detail previously (20) . For quantification, standard curves were generated in multiple J o u r n a l P r e -p r o o f replicates using a commercially available standard for calibration containing the nucleocapsid gene with defined copy numbers (2019-nCoV-N-PositiveControl, IDT). The lower limit of detection (LOD) of the PCR reaction was extrapolated in a probit regression analysis using the same standard as described previously (21) . Viral loads of sewage samples are calculated as N1 copy numbers per 100ml of sewage. Additionally, samples were quantified using digital droplet PCR as described previously (20) .

Amplicon pools spanning the SARS-CoV-2 genome were prepared for a subset of samples, converted to barcoded sequencing libraries and sequenced on an Illumina HiSeq1500 following the ARTIC network nCoV-2019 sequencing protocol v2 (22) . Sequenced reads were demultiplexed and mapped to the SARS-CoV-2 reference genome (NC 045512.2) with bwa-mem (23) . The sequenced amplicons were assembled using the iVar package (24) .

Briefly, the package trims the primer sequences from the mapped reads, filters them by a base quality >20 and a minimal read length of 30 nt. Pileup files were generated using samtools and used for consensus sequence generation within the iVar package setting a minimum frequency threshold of 90% (-t 0.9) and a minimal read depth of 20 (-m 20). SNPs and Indels were called from the mapped reads with freebayes (25) using a ploid of 1 (-p 1). To address potential issues of contamination, negative controls (PCR-grade water) were included in all sequencing runs. The consensus sequences and metadata for the samples were uploaded to the GISAID repository.

For analyses and visualization, we used the software R, week 10 (Figure 4 B) . It can be easily appreciated that the sewage sequencing was able to predict the subsequent increase of B.1.1.7 in the population, similarly to what was observed with case numbers in our study and others (27, 28) .

Munich received a sewage system elaborate for the time, by the end of the 19th century, vast parts of which are still in use today. This system includes many collateral connections which ensure functionality in case of blockage or regional overload. Further, the sewage layout is based on gravity flow from south to north of the city due to a height difference of about 90 meters, without the general use of pumps. This is associated with several challenges for the detection of SARS-CoV-2 in the wastewater. First, it is difficult to assess the drainage area of a certain sampling point if collaterals exist, thus, we chose drainage areas which can clearly be characterized. Second, gravity flow based sewage is less homogenized and also slower regarding flow rates, due to the general absence of pumps. Third, not the whole sewage system is separated in surface water carrying pipes and fecal/sewage systems, leading to dilution artefacts and changes in pH and composition, potentially altering the concentration and decay characteristics of the virus in the samples taken.

One of the main limitations of the study is the use of qualified spot sampling. This leads to larger variation in the data. One of the main reasons thereby is the lack of homogenization prior to sampling, and the fact that the sample is drawn only during a short time period in the J o u r n a l P r e -p r o o f morning once a week and there may be different factors influencing one specific sample. On the other hand, inhabitants are also more likely to be home during the night and morning and go to work thereafter despite increasing rates of home office work recently due to the pandemic. Therefore, sampling in the morning might also give an advantage over integrated sampling over the whole day. The latter applied to residential areas might bias the data towards those inhabitants quarantined for known disease or being first degree contacts as well as white collar workers able to work from home. The practical influence of these factors is currently unknown and thus cannot be controlled for, they are also vastly ignored in most works on the topic currently. We have also omitted to try and calculate back the number of infected persons in each area, as the variation is too large using the spot sampling protocol applied here.

Another effect of the sample variation and the population size in a respective drainage area is a varying limit of detection for SARS-CoV-2 RNA. As the sewage is not sampled over the whole day and homogenized, averaging effects are achieved either by larger numbers of patients in the drainage area, or more measurements by PCR. We were also restricted to six sampling sites within the city due to capacity limitations and the layout of There is also a notification bias, which is caused by the dark field of unreported cases in the population. This might obscure the correlation between notification numbers and virus RNA load in sewage. Due to the seroprevalence and incidence studies performed in parallel in the department, we were able to estimate the dark field as well (29, 30) . Overall, we expect the real case numbers to be about 4 times higher than the reported incidences at the beginning of the spring of 2020 wave, decreasing over time. It is as low as roughly a factor 2 in the winter 2020/21 wave due to increased screening efforts. These dynamic effects cannot be fully taken into account in the analysis; however, wastewater concentrations are not affected by the notification dark field.

This study is one of the first and longest studies to follow SARS-CoV-2 RNA loads in wastewater over time in Germany (31, 32) . With a start in April 2020, the last part of the first SARS-CoV-2 wave of 2020 is covered as is the winter wave 20/21. The data demonstrates that even a simple qualified spot sample taken once weekly was able to foreshadow the SARS-CoV-2 incidence derived from notification data by 3 weeks. This study also demonstrates that sequencing of RNA isolated from raw sewage is a reliable tool to detect the introduction of SARS-CoV-2 variants of concern in the local population weeks prior to their residents. To the right of the 

Wolrd Health Organization. Coronavirus disease (COVID-19) pandemic 2021

Effects of the COVID-19 pandemic on mental well-being amongst individuals in society-A letter to the editor on "The socioeconomic implications of the coronavirus and COVID-19 pandemic: A review

The socioeconomic implications of the coronavirus pandemic (COVID-19): A review

The impacts of the novel SARS-CoV-2 outbreak on surgical oncology -A letter to the editor on "The socio-economic implications of the coronavirus and COVID-19 pandemic: A review

SARS-CoV-2 and the COVID-19

Benchmarking virus concentration methods for quantification of SARS-CoV-2 in raw wastewater

CoV-2 in wastewater: State of the knowledge and research needs

Efficient detection of SARS

CoV-2 RNA in the solid fraction of wastewater

Tracking COVID-19 with wastewater

Potential secondary transmission of SARS-CoV-2 via wastewater. The Science of the total environment

Wastewater-Based Epidemiology (WBE) and Viral Detection in Polluted Surface Water: A Valuable Tool for COVID-19 Surveillance-A Brief Review

Genomic characterization of the 2019 novel human-pathogenic coronavirus isolated from a patient with atypical pneumonia after visiting Wuhan. Emerging microbes & infections

Southeastern Virginia using wastewater-based epidemiology

Detection of SARS-CoV-2 in

Different Types of Clinical Specimens

The challenges of containing SARS-CoV-2 via test-trace-and-isolate

Measurement of SARS-CoV-2 RNA in wastewater tracks community infection dynamics

The first detection of SARS-CoV-2 RNA in the wastewater of Tehran, Iran. Environmental science and pollution research international. 2021:1-8. 18. ARTIC network. SARS-CoV-2 2020

Centers for Disease Control and Prevention. Information for Laboratories about Coronavirus (COVID-19) 2021

Multicentre comparison of quantitative PCR-based assays to detect SARS-CoV-2

Increasing Cytomegalovirus Detection Rate from Respiratory Tract Specimens by a New Laboratory-Developed Automated Molecular Diagnostic

nCoV-2019 sequencing protocol v2 (GunIt) V

Aligning sequence reads, clone sequences and assembly contigs with BWA-MEM2013

An amplicon-based sequencing framework for accurately measuring intrahost virus diversity using PrimalSeq and iVar

Haplotype-based variant detection from short-read

SARS-CoV-2 infection of the oral cavity and saliva

High-Throughput Wastewater SARS-CoV-2 Detection Enables Forecasting of Community Infection Dynamics in San Diego County. mSystems

SARS-CoV-2 titers in wastewater foreshadow dynamics and clinical presentation of new COVID-19 cases. medRxiv : the preprint server for health sciences

Prevalence and Risk Factors of Infection in the Representative COVID-19 Cohort Munich. International journal of environmental research and public health

Protocol of a population-based prospective COVID-19 cohort study Munich, Germany (KoCo19)

Long-term monitoring of SARS-CoV-2 RNA in wastewater of the Frankfurt metropolitan area in Southern Germany

Detection of SARS-CoV-2 in raw and treated wastewater in Germany -Suitability for COVID-19 surveillance and potential transmission risks

Percentages (starting at 0) represent the fraction of the indicated single nucleotide polymorphisms detected at the respective time point. Grey fields (NA) depict no coverage (less than 20 mapped reads) at the respective genome position in the sample

Signature mutations for P1 and B.1.351 were not detected over the study period;. B: Baggtitr chart of reported percentage of B.1.1.7 in sequenced SARS-CoV-2 swabs between calendar weeks 1 and 9 of 2021 in Munich. Dots represent frequencies of B.1.1.7 mutations detected in sewage as plotted in A. The solid line represents the LOESS (locally estimated scatterplot smoothing or local regression) modelling the different signature mutations. The grey region represents the 95% CI of the LOESS estimate

Red represents other variants than the three aforementioned variants of concern; green bar is identification of S1 mutants by hybridization assays without definitive confirmation by sequencing

We thank Heike Fensterseifer, Simone Lehn, Angelika Thomschke, Susanne Eva Maria Thieme and Sylvia Mallok as well as the technical assistants of the virological routine diagnostics at the Max von Pettenkofer Institut for their excellent technical support. We would also like to thank Durdica V. Marosevic (LGL) for her assistance in retrieving variants of concern reporting data. We thank the workers of the Munich Metropolitan Sewer Authority for relentless support and sample taking, as well as the team of the Department of Health of the City of Munich for their support especially with the notification data. Of great help was the fire department / disaster control team of the city of Munich who also supported the study with great expertise.

Medical Biodefense Research Program of the Bundeswehr Medical Service, BMBF initiative "NaFoUniMedCovid19" (01KX2021), subproject B-FAST. Euroimmun, Roche Diagnostics, Mikrogen, Viramed provided kits and machines for analyses at discounted rates.